Pairwise Mechanisms

This reaction is invariably catalyzed by transition metal compounds and its mechanism is of special interest. The first explanation for this transformation is based on the so-called pairwise mechanism , in which two olefins coordinate to a transition metal center to form a transient cyclobutane-like intermediate [2], However, this idea was later replaced... 

A major objection to these mechanisms was raised by Herisson and Chauvin,64 who found that cross-metathesis between a cycloalkene and an unsymmetric alkene resulted in a statistical distribution of cross-products even at very low conversion, whereas a simple pairwise mechanism would lead to a single product. It is also important to point out that cyclobutanes are not isolated as intermediates and are unreactive under metathesis conditions.30... 

Scheme 12.14 The quasi-cyclobutane (upper) and metallocyclopentane (lower) pathways for metathesis both are pairwise mechanisms (see the text for full details).

Consider a telomer being formed from a cyclopentenyl polymer growing under the pairwise mechanism (Scheme 12.14) with growth being curtailed by cross-metathesis under two extreme conditions (i) with only pent-2-ene present (C4 C5 C6 = 0 100 0) and (ii) with a fully equilibrated mixture of acyclic monoalkenes (C4 C5 C6 = 1 2 1). Under condition (i), one would expect the formation of only hierarchical telomers (n = 1,2,3,4,5, etc.) of the type (C2)-[(cyc-C5) ]-(C3) as the pent-2-ene is split into a C2 and a C3 unit across the growing cyclo polyene. In contrast, under condition (ii), one would expect each hierarchical telomer to be formed in a 1 2 1 ratio of (C2)-[(cyc-C5)n]-(C2) (C2)-[(cyc-C5) ]-(C3) (C3)-[(cyc-Q)n]-(C3)> depending on whether there is cross-metathesis with C4, C5 or C6 (ratio = 1 2 1). The outcome will thus depend on how quickly the pent-2-ene is equilibrated by homo-metathesis to yield the C4, C5 and C6 mixture. Analysis of the rate of pent-2-ene homo-metathesis showed that it was not fast. Indeed, it proceeded at approximately the same rate as the telomerisation reaction. One would thus expect the telomer product early in the reaction to be essentially pure (C2)-[(cyc-C5) ]-(C3) species. Then, as C4 and C6 increase in concentration relative to C5, formation of the (C2)-[(cyc-C5) ]-(C2) and (C3)-[(cyc-C5) ]-(C3) telomers should increase proportionally. This was not found to be the case. 

In fact, the (C2)-[(C5)n]-(C2) and (C3)-[(C5) ]-(C3) telomers were produced inparallel with the C4 and Cs alkenes, with the constant statistical telomer ratio of 1 2 1. The result was inconsistent with the pairwise mechanism and suggests that the two ends of the telomer arise from two different alkene units thus, even in the early stages of the reaction, the pent-2-ene can generate all three telomer forms. In other words, a sequential rather than pairwise mechanism seems to have been operative. It was proposed, therefore, that the catalyst acts as a carrier for one half of the alkene and thereby allows sequential reactions of one alkene with... 

There thenfollowed reports by Katz [13] and Grubbs [14] and their co-workers on studies that aimed to simplify and confirm the analysis. The key remaining issue was whether a modified pairwise mechanism, in which another alkene can coordinate to the metal and equilibrate with the product prior to product displacement, would also explain the appearance of the anomalous cross-over products early in the reaction evolution. However, a statistical kinetic analysis showed that for a 1 1 mixture of equally reactive alkenes, the kinetic ratio of cross-metathesis should be 1 1.6 1 for the pairwise mechanism and 1 2 1 for the Chauvin mechanism. Any equilibration (substrate or product) would, of course, cause an approach towards a statistical distribution (1 2 1) and thus allow no distinction between the mechanisms. 

A pairwise mechanism involving a quasi-cyclobutane structure was suggested first, Eq. (3) [18], but cross metathesis between cyclopentene and 2-pentene produced a statistical ratio of cross-products 1-3 (1 2 3 = 1 2 1), Eq. (4) [19]... 

The mechanism of the alkene metathesis reaction is now very well understood and is shown in Scheme 1. The initial mechanistic proposal of a pairwise reaction (the pairwise mechanism) of two alkenes at a transition metal center in a pseudocyclobutane transition metal complex has been discarded in favor of the carbene mechanism (the Chauvin Mechanism) of Scheme 1. ... 

Evidence for the carbene mechanism is now so overwhelming (as discussed below) that the pairwise mechanism is only mentioned in this review for historical reasons. All alkene metathesis reactions are catalyzed by a metal carbene complex of some description, and the widely variable compositions used as catalysts are necessary to generate an active metal carbene group. 

In 1970, Chauvin proposed a mechanism in which the catalytically active species was a transition metal carbene complex. The carbene mechanism was proposed to account for the products observed when cyclopentene and pent-2-ene were metathesized in the presence of the catalysts WOCLi/Bu4Sn or WOCl4/Et2AlCl (equation 4). The products were observed in a 1 2 1 ratio, even as the initial products. A simple pairwise mechanism, however, would have predicted that compound (1) will be the only product initially formed. 

The mechanism described in Scheme 1 (the Chauvin mechanism) is the accepted mechanism of alkene metathesis, and its validity has been demonstrated in two ways. First, classical kinetic studies, including isotopic labeling and crossover experiments performed using poorly defined catalysts, conclusively demonstrated that the carbene mechanism was consistent with the experiments, while the pairwise mechanism was not. More recently, the synthesis of isolable carbene complexes that catalyze the reaction has allowed a more direct observation of the reaction. Each individual step in the Chauvin mechanism has now been observed spectroscopically for several of the well-defined catalyst systems. 

The suggestion that a metal carbene was the active metal-containing species involved in the reaction inspired an impressive number of elegant experiments, designed to test the vahdity of this mechanism. The results of double crossover experiments as well as isotopic labeling experiments showed conclusively that a pairwise mechanism could not account for the observed data. The reaction shown in equation (5) shows the possible products of the metathesis of two isotopically labeled dienes, and these products include a cyclic alkene derived from the closing of the diene as well as a series of deuterated ethylenes. At very low conversions, the observed ratio of ethylenes was 1 2 1 dQ d.2.d ). A detailed analysis of these results demonstrated that the pairwise mechanism could not possibly account for this result, while the carbene mechanism could. 

The mechanism of the reaction remained mysterious for many years. Several early papers, the first by Chauvin, suggested the correct solution, but this was not generally accepted until much later. The question was whether the two alkenes bound to the metal and underwent rearrangement (called the pairwise mechanism), or whether the alkenes reacted one at a time (the nonpairwise mechanism). The Chauvin Mechanism, equation (41), is now the accepted pathway, and was a particularly imaginative suggestion at a time when both the required metalacyclobutane formation and fragmentation reactions and nonhetroatom substituted carbenes were unknown. 

In 1971, Herisson and Chauvin12 performed an experiment described in equation 11.7. They isolated three major products, 2 (C10), 3 (C9), and 4 (Cn), upon tungsten-catalyzed CM of cyclopentene and 2-pentene. Product 2 is expected as the direct result of pairwise metathesis of the two starting materials the other two products could result from subsequent reactions of 2 with 2-pentene. At equilibrium, all products would be present in nearly statistical distribution according to the pairwise mechanism. What troubled Herisson and Chauvin, however, was the observation that, upon quenching the reaction well before equilibrium could be achieved, a statistical distribution of the three products was already present. 

Mechanism 3 shows a pathway that was strongly influenced by the results of Herisson and Chauvin and is outlined in Scheme 11.2. Two key intermediates in this pathway are an alkene-metal carbene complex (5) and a metallacyclobu-tane (6), formed through concerted cycloaddition of the M=C and C=C bonds. A highly significant feature of the mechanism, caused by the unsymmetrical structure of 6, is its explanation of randomization early in the course of reaction. The Herisson-Chauvin mechanism does not require a specific pair of alkenes to interact directly for metathesis to occur, hence the name non-pairwise mechanism. 

J. L. Herisson and Y. Chauvin, Makromol. Chern., 1971, 141, 161. This paper does not propose the discrete intermediates, 5 and 6, but it does suggest that carbene complexes could interact with alkenes separately in a non-pairwise manner such that a new alkene and a new carbene complex could form after a bond reorganization. The mechanism shown in Scheme 11.2 is the non-pairwise mechanism that was elucidated after much work by other chemists, and it is discussed later in this section and in Section 11-1-2. 

The early development of Mechanism 3 was bold for its day because Fischer carbene complexes had just been discovered a few years earlier, and alkylidenes were not yet known. The carbene complexes prepared before 1971 also did not catalyze olefin metathesis. With the discovery of Schrock carbene complexes and the demonstration that some alkylidenes could promote metathesis, the non-pairwise mechanism became more plausible (Section 11-1-2). It was, however, the elegant work of Katz and co-workers that provided early substantial support for the Herisson-Chauvin mechanism. 

Despite this strong evidence supporting the non-pairwise mechanism, others objected that the pairwise mechanism could explain the results of Katz s experiment if another step in the pairwise mechanism were rate determining. Consider Scheme 11.4, in which initial single cross metathesis is rapid to form the C12 alkene (7). If 7 sticks to the metal and 4-octene attacks in a rate-determining step... 

Since the definitive experiments of Katz and Grubbs appeared in the literature, the non-pairwise mechanism has become the accepted pathway for metathesis. Subsequent investigations have supported this pathway.1718 Key intermediates in the non-pairwise mechanism are metal-carbene and metallacyclo-butane complexes. Both have been prepared and shown to catalyze metathesis, and more recently both species have actually been observed in the same reaction mixture and shown to interconvert during it, thus offering additional support for Mechanism 3.19 The next section will cover the discovery of discrete metal-carbene complexes that do serve as metathesis catalysts. 

The metal carbene/metallacyclobutane mechanism of olefin metathesis, as outlined in Section 1.3, was first proposed by Herisson and Chauvin in 1971. By 1975 the evidence in its favour had become so compelling that the earlier pairwise mechanism had been totally discarded. From 1980 onwards well-defined carbene complexes of Ta, Mo, W, Re, and Ru were discovered which would act as initiators without the need for activation by heat, light, or cocatalyst. This in turn led to the spectroscopic detection of the propagating metal-carbene complexes in many systems, to the detection of the intermediate metallacyclobutane complexes in a few cases, and in one case to the detection of the metal-carbene-olefin complex that precedes the formation of the metallacyclobutane complex. In no individual case have all three intermediates been detected at most two have been observed, sometimes one, more often none. After 1980 metallacyclobutane complexes of Ti and Ta were found which would act as initiators at 60°C, but where the intermediate metal carbene complexes could not be detected. 

The first important conclusion from these experiments is that the unsymmetrical telomer (C14) is formed from the start, which would not be possible on the pairwise mechanism. The results may be interpreted in terms of the metal carbene mechanism shown in Scheme 15.3. In order to obtain useful expressions from this scheme, it is necessary to make an assumption similar to that of eqn. (7) for Scheme 15.2, namely that the rate constant ratios are related by eqn. (19). One then obtains eqns. (20) and (21) from a steady-state treatment of Scheme 15.3. Multiplying the two together gives eqn. (22). The product of the two intercepts in Fig. 15.7 is 4.5 0.6. A more accurate value of 4.1 0.1 is obtained by extrapolating the quantity [C 4] /[Ci2][Ci6] to zero conversion, confirming the prediction of eqn. (22) and justifying eqn. (19) a posteriori. 

---