Metathesis reaction, Cross

For a discussion on the beneficial effect of using homodimers of one CM substrate in cross metathesis reactions, see Blackwell HE, O Leary DJ, Chatterjee AK, Washenfelder RA, Bussmann DA, Grubbs RH (2000) J Am Chem Soc 122 58... 

While alkane metathesis is noteworthy, it affords lower homologues and especially methane, which cannot be used easily as a building block for basic chemicals. The reverse reaction, however, which would incorporate methane, would be much more valuable. Nonetheless, the free energy of this reaction is positive, and it is 8.2 kj/mol at 150 °C, which corresponds to an equihbrium conversion of 13%. On the other hand, thermodynamic calculation predicts that the conversion can be increased to 98% for a methane/propane ratio of 1250. The temperature and the contact time are also important parameters (kinetic), and optimal experimental conditions for a reaction carried in a continuous flow tubiflar reactor are as follows 300 mg of [(= SiO)2Ta - H], 1250/1 methane/propane mixture. Flow =1.5 mL/min, P = 50 bars and T = 250 °C [105]. After 1000 min, the steady state is reached, and 1.88 moles of ethane are produced per mole of propane consmned, which corresponds to a selectivity of 96% selectivity in the cross-metathesis reaction (Fig. 4). The overall reaction provides a route to the direct transformation of methane into more valuable hydrocarbon materials. 

Despite those challenges, both Johnson [161] and Grela [162] performed several cross metathesis reactions with vinylhalides using phosphine free catalysts. Turnover numbers (TON) above 20 were very few, while in many cases the TON stayed below ten. The diastereoselectivity of CMs with vinylhalides is shghtly in favour of the Z product which is similar to their acrolein-counterparts. 

The synthesis and olefin metathesis activity in protic solvents of a phosphine-free ruthenium alkylidene bound to a hydrophilic solid support have been reported. This heterogeneous catalyst promotes relatively efficient ring-closing and cross-metathesis reactions in both methanol and water.200 The catalyst-catalyzed cross-metathesis of allyl alcohol in D20 gave 80% HOCH2CH=CHCH2OH. 

Scheme 6.73 Enyne ring-closing and cross-metathesis reactions.

This review focuses on the cross-metathesis reactions of functionalised alkenes catalysed by well-defined metal carbene complexes. The cross- and self-metath-esis reactions of unfunctionalised alkenes are of limited use to the synthetic organic chemist and therefore outside the scope of this review. Similarly, ill-defined multicomponent catalyst systems, which generally have very limited functional group tolerance, will only be included as a brief introduction to the subject area. 

Although the bulk of this review is concerned with well-defined metal carbene catalysts, it is important to note the contributions made to cross-metathesis chemistry by ill-defined or multicomponent catalysts. A brief discussion of the cross-metathesis reactions of functionalised alkenes using catalysts of this type will therefore be included here [1]. 

In contrast, ruthenium catalysts gave the best results for the cross-metathesis reactions of vinylsilanes with a range of unfunctionalised alkenes [8] (a typical example is shown in Eq. 4). 

Initial reports of cross-metathesis reactions using well-defined catalysts were limited to simple isolated examples the metathesis of ethyl or methyl oleate with dec-5-ene catalysed by tungsten alkylidenes [13,14] and the cross-metathesis of unsaturated ethers catalysed by a chromium carbene complex [15]. With the discovery of the well-defined molybdenum and ruthenium alkylidene catalysts 3 and 4,by Schrock [16] and Grubbs [17],respectively, the development of alkene metathesis as a tool for organic synthesis began in earnest. 

Although the application of tungsten catalyst 5 to the cross-metathesis reaction of other alkenes has not been reported, Basset has demonstrated that to-un-saturated esters [18] and glycosides [21], as well as allyl phosphines [22], are tolerated as self-metathesis substrates. 

However, not all substrates readily underwent cross-metathesis. Alkenes containing ketone groups were found to be very poor substrates for the cross-metathesis reaction, although acceptable yields were obtained by protecting them as their silyl enol ethers prior to metathesis. 

Like styrene, acrylonitrile is a non-nucleophilic alkene which can stabilise the electron-rich molybdenum-carbon bond and therefore the cross-/self-metathe-sis selectivity was similarly dependent on the nucleophilicity of the second alkene [metallacycle 10 versus 12, see Scheme 2 (replace Ar with CN)]. A notable difference between the styrene and acrylonitrile cross-metathesis reactions is the reversal in stereochemistry observed, with the cis isomer dominating (3 1— 9 1) in the nitrile products. In general, the greater the steric bulk of the alkyl-substituted alkene, the higher the trans cis ratio in the product (Eq. 11). 

The success of the cross-metathesis reactions involving styrene and acrylonitrile led to an investigation into the reactivity of other Ji-substituted terminal alkenes [27]. Vinylboranes, enones, dienes, enynes and a,p-unsaturated esters were tested, but all of these substrates failed to undergo the desired cross-metathesis reaction using the molybdenum catalyst. 

Cross-metathesis reactions with styrenes or acrylonitrile gave yields and cist trans selectivities that were comparable with the best results obtained in the previous reports (for example Eq. 12). 

As expected, there was no formation of stilbenes or a dinitrile product and, more surprisingly, in all of the reactions reported only 5-7% of the allyltrimeth-ylsilane self-metathesis product was observed. It was proposed that this lack of allylsilane self-metathesis was due to the steric bulk of the TMS group reducing the reactivity of the Me3SiCH2 substituted alkylidene. In a more recent report by Blechert and co-workers it was noted that allyltrimethylsilane and its hydrocarbon equivalent (4,4-dimethylpent-l-ene) had comparable reactivities in the cross-metathesis reaction [28], further suggesting that the selectivity arises from steric rather than electronic effects. 

From the single comparative study reported, it appears that increasing the steric bulk of the silyl group in the allyltrialkylsilane can significantly enhance the trans selectivity of the cross-metathesis reaction without adversely affecting the yield (Eq. 14). 

Barrett and Gibson also reported the application of the molybdenum catalysed cross-metathesis reaction to the elaboration of (3-lactams [31,32]. Protected allyloxy p-lactams 16 were successfully cross-metathesised with a selection of substituted styrenes to yield trans cross-metathesis products (Eq. 16). 

It is worth noting that the reactions with the unsubstituted styrene gave noticeably higher yields than the corresponding cross-metathesis reactions with the substituted styrenes. This appears to be quite a common characteristic of the molybdenum catalysed cross-metathesis reaction. 

With the development of an analogous ruthenium benzylidene catalyst 17 by Grubbs and co-workers in 1995, a ruthenium carbene catalyst suitable for the cross-metathesis reaction was in place [34]. Benzylidene 17 exhibited the same impressive tolerance of air and moisture, and the same stability towards functional groups as its predecessor 4, but benefited from easier preparation [35,36] and much improved initiation rates. 

The first published report on the use of this catalyst for the cross-metathesis of functionalised acyclic alkenes was by Blechert and co-workers towards the end of 1996 [37]. This report was also noteworthy for its use of polymer-bound alkenes in the cross-metathesis reaction. Tritylpolystyrene-bound AT-Boc N-al-lylglycinol 18 was successfully cross-metathesised with both unfunctionalised alkenes and unsaturated esters (Eq. 17) (Table 1). 

It is noticeable that cross-metathesis with the unfunctionalised alkenes occurred in significantly higher yields over shorter reaction times and required a smaller excess of the soluble alkene. This was possibly due to the unfunctionalised alkenes, which are more nucleophilic than their ester containing counterparts, complementing the less nucleophilic/more carbon-metal bond stabilising allylglycinol 18. Comparable results were obtained from cross-metathesis reactions of the polymer-bound isomeric N-Boc C-allylglycinol with the same four alkenes. 

Table 1. Cross metathesis reactions of tritylpolystyrene-bound AT-Boc AT-allylglycinol... 

As with the allylsilane cross-metathesis reactions, significant quantities of allyl stannane self-metathesis were not detected in any of the reactions and the trans isomer predominated in the cross-metathesis products. Identical reactions were carried out using allyltributyl stannane, in place of allyltriphenyl stannane, but the yields of the cross-metathesis products were consistently lower and in many cases dropped below 25%. 

Performing the cross-metathesis reactions at reflux (open to an argon atmosphere in a glove box) was designed to help remove the unwanted ethene produced by the reaction. 

Some further examples of the allyltrimethylsilane cross-metathesis reactions, originally investigated by Crowe, with more complex substrates were also in-... 

Table 3. A comparison of the cross metathesis reactions between oct-l-ene and homoallyl-,... 

We have also studied the effect that moving the double bond closer to the amino acid moiety has upon the reactivity of unsaturated a-amino acids [43]. To this end, the cross-metathesis reactions of similarly protected homoallyl-, allyl-and vinylglycine with oct-l-ene were investigated under identical conditions (Eq. 25) (Table 3). 

It appears that the molybdenum catalyst is more suited to the cross-metathesis of the sterically bulky vinylglycines. The cross-metathesis reaction of a similarly protected dehydro alanine gave only recovered starting material. 

Although the metathesis reaction with allylglycine 23 did not go to completion, a moderate yield of the desired cross-metathesis product was isolated. Very recently, Blechert has reported two similar cross-metathesis reactions of an allylglycine 25 using the ruthenium catalyst [44]. In these cases higher yields of the cross-metathesis products were isolated, presumably due to the higher reaction temperatures employed (Eq. 26). 

Applications of the cross-metathesis reaction in more diverse areas of organic chemistry are beginning to appear in the literature. For example, the use of alkene metathesis in solution-phase combinatorial synthesis was recently reported by Boger and co-workers [45]. They assembled a chemical library of 600 compounds 27 (including cisttrans isomers) in which the final reaction was the metathesis of a mixture of 24 oo-alkene carboxamides 26 (prepared from six ami-nodiacetamides, with differing amide groups, each functionalised with four to-alkene carboxylic acids) (Eq.27). 

The ring-opening cross-metathesis reaction is similar to the acyclic cross-metathesis reaction discussed above, except that one of the acyclic alkenes is replaced with a strained cyclic alkene (Scheme 5). 

Since one of the substrates is a cyclic alkene there is now the possibility of ring-opening metathesis polymerisation (ROMP) occurring which would result in the formation of polymeric products 34 (n >1). Since polymer synthesis is outside the scope of this review, only alkene cross-metathesis reactions resulting in the formation of monomeric cross-coupled products (for example 30) will be discussed here. 

The report by Basset and co-workers on the metathesis of sulphur-containing alkenes using a tungsten alkylidene complex, mentioned previously for the acyclic cross-metathesis reaction (see Sect. 2.2), also contained early examples of ring-opening cross-metathesis of functionalised alkenes [20]. Allyl methyl sulphide was reacted with norbornene in the presence of the tungsten catalyst 5, to yield the desired ring-opened diene 35 (Eq. 29). 

In 1995 the first examples of ring-opening cross-metathesis reactions for the preparation of functionalised monomeric products using the Grubbs ruthenium vinylalkylidene catalyst 4 were published by Snapper and co-workers [47]. Reaction of a variety of symmetrical cyclobutenes with simple terminal alkenes... 

Optimal yields were obtained by slow addition of the alkene substrates to a solution of the ruthenium vinylalkylidene and this allowed just two equivalents of the acyclic alkene to be used without significant formation of polymeric products. Unlike the acyclic cross-metathesis reactions, which generally favour the formation of tram products, the above ring-opening metathesis reactions yielded products in which the cis stereoisomer is predominant. Particularly noteworthy was the absence of significant amounts of products of type 31, formed from metathesis of one cyclic and two acyclic alkenes. In fact, considering the number of possible ring-opened products that could have been formed, these reactions showed remarkable selectivity (GC yields > 80%). 

A subsequent publication by Blechert and co-workers demonstrated that the molybdenum alkylidene 3 and the ruthenium benzylidene 17 were also active catalysts for ring-opening cross-metathesis reactions [50]. Norbornene and 7-oxanorbornene derivatives underwent selective ring-opening cross-metathesis with a variety of terminal acyclic alkenes including acrylonitrile, an allylsilane, an allyl stannane and allyl cyanide (for example Eq. 34). 

---