Titanium silicates catalytic sites

In some reports, the presence of the two forms of titanium has been described as arising from two types of sites identified as isolated titanium framework sites and Ti02 particles, and the reactions have been attributed to the catalytic activity of one or the other phase (Huybrechts et al., 1992). Since it is possible to obtain pure phase titanium silicates, it seems preferable to identify the Ti02 phase as an impurity. 

The catalytic activity of titanium silicates must be ascribed Tilv sites, because pure crystalline silicas are totally inactive. As was discussed in Section III, Tilv is present in the crystalline structure at random. Very likely, the random distribution that is obtained in the precursor reagents is maintained in the solid. Being dilute, each Tilv is expected to be surrounded by OSiIV groups and isolated from other Tiiv ions by long O—Si—O—Si—O sequences. It has been... 

The incorporation of Ti into various framework zeolite structures has been a very active research area, particularly during the last 6 years, because it leads to potentially useful catalysts in the oxidation of various organic substrates with diluted hydrogen peroxide [1-7]. The zeolite structures, where Ti incorporation has been achieved are ZSM-5 (TS-1) [1], ZSM-11 (TS-2) [2] ZSM-48 [3] and beta [4]. Recently, mesoporous titanium silicates Ti-MCM-41 and Ti-HMS have also been reported [5]. TS-1 and TS-2 were found to be highly active and selective catalysts in various oxidation reactions [6,7]. All other Ti-modified zeolites and molecular sieves had limited but interesting catalytic activities. For example, Ti-ZSM-48 was found to be inactive in the hydroxylation of phenol [8]. Ti-MCM-41 and Ti-HMS catalyzed the oxidation of very bulky substrates like 2,6-di-tert-butylphenol, norbomylene and a-terpineol [5], but they were found to be inactive in the oxidation of alkanes [9a], primary amines [9b] and the ammoximation of carbonyl compounds [9a]. As for Ti-P, it was found to be active in the epoxidation of alkenes and the oxidation of alkanes and alcohols [10], even though the conversion of alkanes was very low. Davis et al. [11,12] also reported that Ti-P had limited oxidation and epoxidation activities. In a recent investigation, we found that Ti-P had a turnover number in the oxidation of propyl amine equal to one third that of TS-1 and TS-2 [9b]. As seen, often the difference in catalytic behaviors is not attributable to Ti sites accessibility. 

The catalysts used in the aforementioned studies were always titanium silicates of MFI structure prepared by hydrothermal synthesis. Ti can, however, be inserted in the silica lattice by post-synthesis treatment of a dealuminated H-ZSM-5 with TiCl4 vapor [11]. Titanium silicalite-2 (TS-2), with the MEL structure of ZSM-11, was prepared shortly after the first synthesis of TS-1 [15]. Both catalysts have been used for the hydroxylation of phenol. Kraushaar-Czarnetzki and van Hooff showed that no major catalytic differences resulted from the method of synthesis of TS-1 [11]. The slow rate of reaction they observed was probably the result of large crystal size and low titanium content [7]. Tuel and Ben Taarit demonstrated there was no perceptible difference between the catalytic activity of TS-2 and TS-1 [8]. This was predictable, because of the close similarity of the Ti-site structure, chemical composition, and pore dimensions of the two titanium silicates. 

In addition to the characteristic XRD patterns and photoluminescence, UV-visible and X-ray absorption spectra, another fingerprint thought to indicate lattice substitution of titanium sites was the vibrational band at 960 cm-1, which has been recorded by infrared and Raman spectroscopy (33,34). Although there is some controversy about the origin of this band, its presence is usually characteristic of a good TS-1 catalyst, although it turned out to be experimentally extremely difficult to establish quantitative correlations between the intensity of the 960 cm-1 band and the Ti content of a Ti silicate and/or its catalytic activity. 

Epoxidation of alkeneic reactants is faster on titanium-grafted silicates (such as A, B and C) than on the coprecipitated titanosilicates (such as D and E). This difference was attributed to the fact that on extra-framework titanium-grafted silicates, the catalytically active sites are virtually all exposed and accessible, whereas on the coprecipitated material some of them may be buried within the silicate walls and, thus, cannot adsorb reactant molecules. 

The vanadium silicalites (with MFI and MEL stmcture) are active oxidation catalyst in gas and liquid phase reactions [180]. As for the titanium silicalites, only the ftamework associated vandium exhibits redox properties [181]. For example, in the hydroxylation of phenol, silicalite impregnated with vanadium compounds is catalytically inactive [182]. The catalytically active vanadium species is speculated to be located in non-tetrahedral positions, most probably chemically bound to the framework. Vanadium bound in that way is not extractable from the lattice [ 183]. A proposed stmcture of the vanadium site is schematically shown in Scheme 21. Note that the Si-O-V bonds are longer than the Si-O-Ti bonds and that V seems to be more exposed. The redox properties are affiliated with the changes in the oxidation state of vanadium between +IV and +V. Vanadium silicates with SiA ratios ranging from 40 to 160 have been reported and these high values suggest (in accordance with V MAS-NMR measurements) that the V sites are isolated in the lattice. 

The turnover numbers (TON) of phenol are 62.4 and 105.4 with a H2O2 efficiency of 28.9 wt.% and 48.9 wt.% for the Zr-Sil-2 samples A and B, respectively. A significant difference in the product distribution between these two runs is also observed. The catechol (CAT) to hydroquinone (HQ) ratios are 0.9 and 1.7 for Zr-Sil-2 (A) and (B) samples, respectively. A CAT/HQ ratio of 0.9 to 1.3 has been reported for titanium and vanadium silicate molecular sieves (TS-2 and VS-2) [13]. The samples synthesized using Zr(acac)4 show a nearly two fold activity in the reaction probably due to the smaller particle size. These results indicate that in the case of Zr-Sil-2 samples synthesized using ZrCU, the Zr " ions are well dispersed within the channels of the MEL structure while in the samples synthesized using Zr(acac)4, the hydroxylation occurs at the external surface as well, where a part of Zr species may be located. For small submicron crystals (<1 pm), external surface sites could be a significant fraction of the total surface area. If the external surface sites are catalytically either the same or more active than the intracrystalline active sites, then the shape selectivity of a zeolite could... 

Inclusion of these cations does impart new catalytic activities, but in many cases the active site results from a metal ion that has left the framework and entered the pore space upon heating, especially in the presence of water vapour. This is thought to be the case for zinc- and gallium-containing solids used in the dehydrocyclisation of butane and propane to aromatics in the Cyclar process (Chapter 9). Boron, iron, chromium and vanadium all appear to leave the framework under harsh conditions. The incorporation of titanium and more recently tin into framework sites within silicates have become very important substitutions, because both titanosilicates and stannosilicates have been shown to contain stable Lewis acid sites of importance in selective oxidation catalysis. The metal atom can coordinate additional water molecules in the as-prepared material, but these can be removed by heating. In the synthesis of titanosilicates, titanium is usually added to the gel as the alkoxide, and synthesis performed in the absence of sodium hydroxide to avoid precipitation of sodium titanate or nanoparticulate titanium oxides. 

---