W-based catalysts

Si-P and Si-AI are not effective, either. The basic catalyst such as Si-K is very active for decomposition of pyruvic acid, but citraconic anhydride is not produced. The best results are obtained with W and W-based catalyst. The one-pass yield of citraconic anhydride reaches about 60 mol% at a pyruvic acid conversion of about 92%. The combination of P to W decreases the activity markedly. The combination of Mo or K decreases the selectivity to citraconic anhydride. The combination of Ti or Sn is scarcely effective, when the amount is less than 10 atomic %. When the amount of Sn or Ti is high, the selectivity falls. 

Upon discovery of this mechanism, new catalysts have been developed, now presenting alkylidene ligands in the metal coordination sphere, such as [(=SiO) Ta(=CH Bu)Np2 and [(=SiO)Mo(=NAr)(=CH Bu)Np] [43, 88]. Table 11.4 presents results obtained with several catalysts prepared by SOMC. Although [(=SiO) Ta(CH3)3Cp (=SiOSi=)] is not active in alkane metathesis (the tantalum site would not be as electrophilic as required) [18], results obtained with [(=SiO)Mo(=NAr) (=CH Bu)Np] show that ancillary ligands are not always detrimental to catalytic activity this species is as good a catalyst as tantalum hydrides. Tungsten hydrides supported on alumina or siHca-alumina are the best systems reported so far for alkane metathesis. The major difference among Ta, Mo and W catalysts is the selectivity to methane, which is 0.1% for Mo and less than 3% for W-based catalysts supported on alumina, whereas it is at least 9.5% for tantalum catalysts. This... 

Alkene metathesis, a remarkable reaction catalyzed by transition metal catalysts, can be traced back to Ziegler-Natta chemistry as its origin [11], In 1964, Natta et al. reported a new type polymerization of cyclopentene using Mo- or W-based catalyst, without knowing the mechanism. This was the first example of ring-opening metathesis polymerization (ROMP eq. 1.9) [12],... 

Some of these W-based catalysts could be transformed into insoluble salts for example, the phosphotungstate-pyridinium salt in a toluene-tributyl phosphate (4 3) solvent mixture [328]. In this solvent the pre-catalyst is insoluble, but upon reaction with H2O2 it gives the catalytically active W-peroxo complex, which is soluble. The catalytic action is then performed under homogeneous conditions and, at the end of the reaction, H2O2 being completely consumed, the precatalyst precipitates and can be easily filtered off and recovered. The method is smart, but from an industrial point of view the separation and recovery is costly. 

The second step of this process would be the cleavage of 1,2-cyclohexandiol 37 this reaction occurs with W-based catalysts and H P under homogeneous and heterogeneous conditions [37a, b[. For example, the diol (1 1 mixture of cis and trans) was deaved to an 88% yield of AA with aqueous H P and with an heteropolyacid tris (cetylpyridinium)-12-tungstphosphate catalyst, at a reflux temperature of t-BuOH in 24h reaction time, under two-phase conditions [37c[. 

Among the large variety of catalytic methods available for the epoxidation of alkenes, those based on tungstate catalysts hold a prominent place when it comes to industrial relevance. Several positive features of W-based catalyst systems explain their popularity in industry ... 

Tungstate is readily available, inexpensive, and not very toxic. Active W-based catalysts are usually devoid of any expensive organic ligands, and merely assembled from simple inorganic components plus a phase-transfer catalyst when required. 

We have carried out an experimental comparison of the epoxidation activity of various W-based catalysts in order to assess the performance of the [WZn3 (ZnW9034)2] sandwich POM relative to other W-based epoxidation catalysts [10]. The 11 catalyst systems used for this experimental study are listed below. The catalysts were either added as an isolated tungsten compound or formed without isolation in situ ... 

FIGURE 16.2 Conversion versus time for the epoxidation of cyclooctene catalyzed by various W-based catalyst systems (0.1 mol% W) at 60 °C in toluene in the presence of 1.5 equiv 50% H2O2. Reprinted with permission from [10]. Copyright 2004 American Chemical Society. 

Apart Ifom the molybdenum carbene complexes already listed in Tables 2.1 and 2.2 Mo-based catalysts are of three main types (i) other Mo complexes, activated by a suitable cocatalyst (ii) M0CI5, also activated by a cocatalyst and (iii) supported oxides, generated in various ways. For the metathesis of terminal olefins higher than propene. Mo-based catalysts are generally more effective than the corresponding W-based catalysts. 

There are perhaps more known W-based catalysts for olefin metathesis than all others together. Many studies have been made on the catalyst systems themselves in an attempt to elucidate the nature of the active species and its mode of formation. They are especially effective for internal and cyclic olefins. Many examples will be found throughout this book. Here we will illustrate some of the main types. 

W-based catalysts for the metathesis of terminal olefins are comparatively few in number. However, this is partly an illusion because systems such as WClg/EtAlCla/ EtOH, although not effective in the sense of yielding ethene and an internal olefin, cause rapid non-productive metathesis in which the products can only be distinguished from the reactants by isotopic labelling see Ch. 5. 

Table 2.4 Examples of W-based catalyst systems" for productive olefin metathesis (see...

An example of this type of forward reaction has already been given in Seheme 2.4. Further examples are provided by (i) the use of norbomene epoxide with W-based catalysts to enhance the ROMP of norbomene derivatives (Devine 1982) (ii) the use of oxygen with norbomene in Ru-based systems to generate such an epoxide (Ivin 1981b) (iii) the use of cyclopropane derivatives with Re207/Al203 (Anisimov 1991) see Ch. 2. 

Although, on the whole. Mo-based catalysts tend to be more stereoselective than W-based catalysts, it should nevertheless be noted that some of the highest stereoselectivities are observed with W-based catalysts. 

The ROMP of 212 proceeds best with WCl6/Et3Al (1/4) as catalyst in PhCl at 60°C (34% yield). The NMR spectrum of the polymer in CDCI3 shows only one line for each carbon, which indicates that the polymer is certainly all-HT with one dominant double-bond configuration (Cho 1985). The ROMP of 213 using a W-based catalyst at 70°C gives an 11% yield of soluble polymer, A/ = 5240. If the benzyl group in the monomer is replaced by methyl it fails to polymerize (Watkins 1994). 

Early attempts at making block copolymers using first-generation catalysts met with limited success. Thus, when cyclooctatetraene (M2) is added to a freshly polymerized solution of cyclooctene (MO, made with a W-based catalyst, the UV spectrum of the product indicates that at least two or three molecules of M2 have added to the chains (Korshak 1985). 

Sulfur uptake and exchange data 1 referred in Sec. 2.3 indicated substantial differences of alumina supported Pd and Pt, in comparison with the molybdena based catalysts. The mobile sulfur amounts were lower [0.25 Smob/Pd (or /(Pt)] than those experienced with the Mo and W based catalysts. Near to all sulfur, accommodated on the noble metal was mobile and — different from the Mo- and W-based catalysts — participated in the reaction the sulfur mobility was substantially higher due to the much lower S-Pd and S-Pt bond strengths. This indicates that the HDS mechanism, detailed above for Mo based catalysts cannot be valid for the noble metals. Again, the number of vacancies for bonding sulfur-organic compounds is not limiting, as in the case with monometallic Co and Ni. 

Using reference substances, either comercially available or synthesized and purified by known methods like 5-ethinyl-2-norbornene (ET-NBE, Scheme 3), and adding these separately to the solution of 1 in highly purified DCPD, we found that ET-NBE activates the system. This was rather unexpected, since acetylenes are known as catalyst poisons in the RIM of DCPD with Mo- and W-based catalyst and are therefore carefully removed during the purification of DCPD. If the monomer contained very low levels of ethinyl-norbornene, a very bad curing behaviour with 1 as catalyst was observed. Best results were obtained with approximately equimolar amounts of ET-NBE and 1. It turned out that crude DCPD (e.g. 94% from Shell) contained - by chance -approximately the appropriate amount of this acetylene and therefore was nicely cured with 1. 

---