Periodic Phases 31 MCM41 P6mm

MCM-41 is the simplest and most extensively studied meso-porous silica [4]. In a typical example of synthesis, the alkyltrimethylammonium surfactants, C„H2b+1-N(CH3)+X-(X = Cl or Br, n = 8-10, 12, 14, 16), one of so-called anionic surfactants described as S+I-, were used. According to the simplest picture of the LCT mechanism, these surfactant molecules form micellar rods (Fig. 1a), which further self-assemble into a hexagonal array (Fig. 1b). Silicate ions deposit on the surface of the micellar rods to form an inorganic framework. When the surfactant molecules are removed, a hexagonal array of unidirectional mesopores appears eventually (Fig. 1c).

Direct structural information can be obtained from the TEM images as shown in Figure 2 [28]. When we view down the pore axis, we can see the hexagonally arranged white disks that correspond to the positions of the pores. If we look along a direction perpendicular to the pore axis, a group of fringes is seen, indicating straight pores. The highest possible space group is P6mm. The material has in fact a two-dimensional structure, since its c axis is infinite.

Figure 1. Schematic drawing of (a) a micellar rod, (b) hexagonal array of micellar rods, and (c) MCM-41 mesoporous silica, showing three important steps in the liquid crystal mechanism of the formation of MCM-41.

Figure 2. Two principal TEM images of MCM-41, viewing down the (a) channel axis and (b) a direction perpendicular to the channel axis. The arrows in (b) indicate the diffraction spots from the ordering of the mesopores. Reprinted with permission from [28], W. Zhou et al., Progr. Natural Sci. (Chinese Ed.) 8, 528 (1998). © 1998, National Natural Science Foundation of China.

Figure 2. Two principal TEM images of MCM-41, viewing down the (a) channel axis and (b) a direction perpendicular to the channel axis. The arrows in (b) indicate the diffraction spots from the ordering of the mesopores. Reprinted with permission from [28], W. Zhou et al., Progr. Natural Sci. (Chinese Ed.) 8, 528 (1998). © 1998, National Natural Science Foundation of China.

Since MCM-41 silica lacks strict crystallographic order at the atomic level and the one-dimensional pore system results in a two-dimensional (2D) hexagonal symmetry, the main structural variations in MCM-41 occur in the pore diameter and the wall thickness. The pore size is affected by several factors. In a highly acidic synthetic system with reactants of tetraethyl orthosilicate (TEOS), HCl, and cetyltrimethylam-monium bromide (CTMA+M-), the silica-containing liquid crystals have a combination of CTMA+-M--SiO+. Therefore, the size of M-, which could be Cl- or NO- or Br-, etc., will certainly affect the pore size of the final products. Hydrothermal treatment at different times can also result in different pore sizes [29, 30] (Fig. 3). Obviously, the largest effect on the pore size is from the chain length of the surfactant molecules and addition of auxiliary organics [4, 31]. Although a typical MCM-41 has a pore diameter of about 2 nm and the minimum wall thickness of <1 nm with specific surface area of over 1000 m2/g according to the Brunauer-Emmett-Teller (BET) calculation [4], the range of pore size of MCM-41 can be changed from 1.5 to 10 nm corresponding to the different chain lengths of surfactant molecules.

The shape of the pores in MCM-41 can be either cylindrical or hexagonal. The cylindrical pore structure was proposed based on a classical molecular dynamics simulation approach [32]. The hexagonal pore structure was proposed by Behrens et al. [33]. In the third model, proposed by Garces [34], the pores can be regarded as spherical cages connected by slightly narrower necks. Each model has some experimental evidence to support. However, none of them have been completely proved. On the other hand, due to

20 (degrees)

Figure 3. Typical XRD patterns of calcined purely siliceous MCM-41 (a) prepared at different temperatures for 48 h; (b) prepared for different reaction times at 165 °C Reprinted with permission from [29], C. F. Cheng et al., Chem. Phys. Lett. 263, 247 (1996). © 1996, Elsevier Science.

20 (degrees)

Figure 3. Typical XRD patterns of calcined purely siliceous MCM-41 (a) prepared at different temperatures for 48 h; (b) prepared for different reaction times at 165 °C Reprinted with permission from [29], C. F. Cheng et al., Chem. Phys. Lett. 263, 247 (1996). © 1996, Elsevier Science.

the high flexibility of the amorphous wall of MCM-41, the shape of the pores may be variable with different synthetic conditions.

In the last decade, control of particle size and morphology of MCM-41 has been extensively investigated. In addition, derived from the MCM-41 structure, some surface tubular forms with diameters around 30 to 100 nm may appear during a postsynthesis ammonia hydrothermal treatment [35]. Smaller tubular form such as bundle of nanotubes and even single-channel nanotubes with a few nanometer in diameter were also observed on the surface of MCM-41, resulting in a surface paintbrush pattern [36].

Similar to zeolites, pure silica materials have no acidic sites and are of limited use for catalysis and metal incorporation. In order to produce acidic mesostructured materials for the application in catalysis, aluminum has been successfully introduced into the framework of MCM-41 [37, 38] and can be detected by EDX commonly equipped in electron microscopes. However, it is very difficult to tell whether Al cations are located inside or outside the wall by using EDX. 27Al MAS NMR confirmed that all Al cations were tetracoordinated (i.e., in the framework). XRD and TEM characterizations indicate that the degree of order of the mesopores in MCM-41 is significantly reduced by incorporation of even a few percentage of aluminum into the wall [27].

Incorporation of many other cations into MCM-41 has been investigated in the last decade. Those guest elements include Ti [39, 40], V [41], Ga [42], B [43], Mn [44, 45], Fe [46, 47], and Sn [48]. The doped materials often have a lower degree of order in comparison with the parent pure MCM-41 silica.

Another method for creating active sites in MCM-41 is introduction of catalytically active metal clusters on the inner surface of the channels. If the metal particles are very small (e.g., in the form of individual atoms) and are evenly distributed in the channels, high resolution TEM (HRTEM) may not give enough image contrast to see them. But the presence of the metal can be confirmed by EDX elemental mapping [49]. When larger clusters (>1 nm) form, they can be detected by HRTEM. It has been shown that direct images can easily differentiate between clusters inside the channels and those on the outer surface of the mesoporous MCM-41 [50]. In the case of Ag3Ru10, anchored on the inner surface of the channels, the clusters can be imaged as some dark dots even though they are not ordered [51]. Ordered or partially ordered metal clusters inside the meso-pores result in a so-called rosary pattern of image contrast (Fig. 4) [52, 53]. In this case, an average distance of the cluster separation can be calculated. The incorporation of metals inside the MCM-41 channels usually has no significant effect on the original silica framework.

It has been found that MCM-41 templated by the alkyltrimethylammonium surfactants is not the only P6mm phase. For example, using decaoxyethylene cetyl ether [C16(EO)10] with <30 wt% in the reaction system, the product CMI-1 (P6mm) has the same structure as MCM-41 with the pore diameter about 4 nm [54]. SBA-3 (P6mm) synthesized in an acidic condition using C16TMA+ surfactant is also of a MCM-41 type [17]. HMM-1 has a 2D hexagonal structure with a hybrid ethane-silica framework [(CH3O)3Si-CH2-CH2-Si(OCH3)3], instead of pure inorganic composition [55, 56]. It is very interesting to find that

Figure 4. TEM image of MCM-41 loaded with Ru6 clusters showing a rosary pattern of the image contrast along the pore axis, with its Fourier transform (top inset). The bottom inset shows one-dimensional brightness scanning along the line marked in the image. Reprinted with permission from [28], W. Zhou et al., Progr. Natural Sci. (Chinese Ed.) 8, 528 (1998). © 1998, National Natural Science Foundation of China.

Figure 4. TEM image of MCM-41 loaded with Ru6 clusters showing a rosary pattern of the image contrast along the pore axis, with its Fourier transform (top inset). The bottom inset shows one-dimensional brightness scanning along the line marked in the image. Reprinted with permission from [28], W. Zhou et al., Progr. Natural Sci. (Chinese Ed.) 8, 528 (1998). © 1998, National Natural Science Foundation of China.

these organic/inorganic hybrid mesoporous materials have better ordered mesopore arrangements. The following sections demonstrate more 2D hexagonal phases that are similar to MCM-41.

Using bolaform surfactant (CH3)3-N-CnH2n-O-C6H4-C6H4-O-CnH2n-N-(CH3) 3(n = 4, 6, 8, 10, 12), the product SBA-8 has a two-dimensional C-centered rectangular structure with typical unit cell parameters a = 6 and b = 3.96 nm and space group cmm when the surfactant with n = 12 was used [57]. A typical SBA-8 structure has pore diameters of 2.65 to 2.91 nm and average wall thickness of about 2 nm with the BET surface area of about 1000 m2/g. This structure has a close relation with MCM-41 since the perfect hexagonal cell of MCM-41 is equivalent to a C-centered rectangular cell with a/b = V3, while the a/b value in SBA-8 is slightly smaller than \/3. In other words, SBA-8 consists of arrays of one-dimensional mesopores with a distorted hexagonal symmetry. Although SBA-8 is a thermally stable phase, its uncompleted condensation of the siloxane framework is probably the main reason for the lattice distortion and, indeed, SBA-8 can be transformed into MCM-41 through postsynthesis treatment in water at about 100 ° C.

It is interesting to see that the bolaform surfactants include two hydrophilic headgroups connected by a hydro-phobic chain (i.e., they are amphilphiles), which are different from the alkyltrimethylammonium surfactants normally used for producing MCM-41. They are more similar to gemini surfactants which usually form globular micelles instead of cylindrical rods, resulting in a three-dimesional hexagonal phase SBA-2. It is obvious that such surfactants are not essential to the cmm phase formation. A most recent work by Haskouri et al. reported that SBA-8 could also be synthesized by using cetyltrimethylammonium bromide instead of the rigid bolaform amphiphiles [58].

Quite different from the formation mechanisms of MCM-41 and SBA-8 mentioned above, FSM-16 is derived from a layered polysilicate without forming any liquid crystals. It was first synthesized by reaction of single layered polysilicate (NaHSi2O5 • 3H2O) with alkyltrimethylammonium chloride [C16H33-(CH3)3N+Cl- ] [59] and attracted a large attention after the appearance of the M41S series in 1992 [16, 60-62]. A typical FSM-16 has a 2D hexagonal cell with a about 4 nm, pore diameter slightly smaller than 3 nm, and BET surface area over 1000 m2/g. These parameters can also be varied by changing the chain length in the surfactants. For example, the use of C10H21-(CH3)3N+Br- leads to the product FSM-10 with smaller pore size and unit cell dimension [62]. Nevertheless, the structure of FSM-16 is almost the same as MCM-41, both having similar pore sizes, surface areas, and hexagonal arrangement of the one-dimensional straight pores.

Although the formation of FSM-16 is due to a structural rearrangement of layered silicates, our knowledge about the detailed procedure of this transformation is still very limited. The originally proposed "folded sheets mechanism"

(FSM) may be an inaccurate description since we do not know whether the sheets of silicates "fold" to form meso-pores or some other intermediate phases are involved. According to a HRTEM study, a domain structure was observed during the phase transformation from the layered compound to FSM-16 [63].

On the other hand, if a layered alkyltrimethylammonium-kanemite (NaHSi2O5 • 3H2O) complex was synthesized first followed by a mild acid treatment, novel rectangular arrangements of square or lozenge channels can be formed. The material was denoted as KSW-2 [64], which has flat silicate walls, BET surface area of 1100 m2/g, and an average pore diameter of about 2.1 nm. The formation of KSW-2 strongly suggests that the "folded sheets mechanism" is correct at least for this particular case.

Another important 2D hexagonal mesoporous silica, SBA-15, was synthesized by using triblock copolymers [poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide)], for example, EO20PO70EO20, as the structure-directing agents [65, 66]. The structure seems to be similar to that of MCM-41, having a 2D hexagonal unit cell in symmetry of P6mm as detected by XRD and TEM. Its BET surface area can be from 690 to 1040 m2/g. One of the significant differences between SBA-15 and MCM-41 is that the pore size (5 to 30 nm) in the former is much larger due to the large surfactant molecules used. The second remarkable feature of SBA-15 is that the wall is much thicker, 3 to 6 nm, which results in a higher hydrothermal stability.

The most extraordinary feature of SBA-15 is that the mesopores are connected by micropores or smaller mesopores through the wall, which has been indirectly proved by TEM observation of metal negative replicas of SBA-15. These replicated structures are consequently three-dimensional. Metal nanorods produced inside the MCM-41 channels, on the other hand, are separated [67]. The micropore volume in SBA-15 can be around 0.1 cm3/g or larger. The formation of these micropores has been proposed. During the aggregation of triblock copolymer surfactant (PEO)„-(PPO)m-(PEO)„ and silicate species, the hydrophilic heads (PEO) are deeply occluded within the wall, forming some "corona" regions with relatively low density of silica. These regions become microporous upon calcination [68]. The intrawall micropores can also be detected by the adsorption and desorption method [69]. The existence of nanoscale pores in the wall of SBA-15 is also possible as directly observed by Fan et al. from the TEM images viewed down the directions both along and perpendicular to the principal channel axis [70]. These interchannel micropores and mesopores are randomly located along the [001] direction of SBA-15 and, therefore, could not be detected by XRD. To date, structural control of these pores seems to be much more difficult than controlling the main group of mesopores. Nevertheless, Newalkar and Komar-neni reported their experiments of tuning the intrawall micropores by means of salt addition under microwave-hydrothermal conditions [71].

As for MCM-41, investigation of doped SBA-15 has been extensively performed in the last few years, for example, alu-minaion [72, 73], incorporations of Ti [74], V [75], Zr [76], sulfonic acid groups [77], etc. In comparison with MCM-41, the chemical doping in SBA-15 has less effect on the meso-pore structures due to the thicker wall [72, 73].

In addition to loading catalytically active nanoparticles and coating on the inner surface of SBA-15 [78, 79] in order to modify the properties of the material, another important application of the SBA-15 structure due to its relatively large pore size, high stability of the wall, and the existence of microporous connection between mesopores is to use this material as template for fabrication of porous materials of other compositions, such as silica-free Pd [80], Pt [81], carbon [82, 83] frameworks, etc. To achieve 3D porous bulk materials, the microporous bridges in SBA-15 must be filled. Otherwise, separated nanowires would be produced [84]. Most of these 3D porous replicas of SBA-15 mentioned above are either amorphous or polycrystalline. However, 3D porous single crystals (PSCs) of metals and ceramics have been proved to be possible using SBA-15 as template [85, 86]. Figure 5 shows a TEM image of 3D PSCs of Cr2O3 prepared in SBA-15. When the silica framework is completely removed by using 10% HF solution, small bridges connecting the Cr2O3 nanorods are clearly revealed. The state of the single crystal is confirmed by selected area electron diffraction (SAED) from an area covering many nano-rods. Because the structures of these new porous materials are exactly negative replicas of SBA-15, it can be expected that their low-angle XRD patterns should be similar to that of SBA-15 [87].

0 0

Post a comment