Metallocyclobutane

The catalysts are metal-carbene complexes that react with the alkene to form a metal-locyclobutane intermediate.290 If the metallocyclobutane breaks down in the alternative path from its formation, an exchange of the double-bond components occurs. 

A.J. Carty, Guelph-Waterloo Centre It might be expected that the facility with which metallocyclobutanes undergo a- or B-eliminations would be rather dependent on the proximity of hydrogen atoms on the a- or B-carbon atoms to the metal. In other words, there might be a structural dependence. Is there any evidence that such structural features contribute substantially to the preferred reaction pathway ... 

Second, several investigations have implied that the formation of cyclopropanes is related to olefin metathesis (5, 12, 13). Extrication of the metal from the metallocyclobutane intermediate has been suggested (5) as a route to the three-membered ring ... 

However, the generation of cyclopropane derivatives was also shown to implicate an alternate route not requiring a metallocyclobutane transition state (14). In addition, a metathesis catalyst successfully converted certain cyclopropanes to metathesis-related olefins by way of a carbene elimination process (15-17), according to Eq. (3). 

In comparison with other catalysts, (diphenylcarbene)tungstenpenta-carbonyl is surprisingly sluggish (26). This observation is significant as it relates to diverse views regarding the carbene-to metallocyclobutane interconversion. Whereas Casey emphasizes a need to accommodate the incoming olefin within the coordination sphere of the metal prior to rearrangement to a metallocycle [Eq. (5)], Katz (28) has described the process essentially as a dipolar attack of a polarized carbene-metal (R2C+—M-) on the olefin [Eq. (6)]. The latter does not specify a need for w-complexation of the olefin as a precondition to metathesis. 

Green demonstrated (48) that a 7r-allyl complexed to W or Mo can undergo a nucleophilic attack by H- on the central carbon of the tj3 allylic group, forming a stable metallocyclobutane. [See Eq. (20).] Cyclopropane and propylene were evolved (49) on heating the metallocyclic... 

Osborn and Green s elegant results are instructive, but their relevance to metathesis must be qualified. Until actual catalytic activity with the respective complexes is demonstrated, it remains uncertain whether this chemistry indeed relates to olefin metathesis. With this qualification in mind, their work in concert is pioneering as it provides the initial experimental backing for a basic reaction wherein an olefin and a metal exclusively may produce the initiating carbene-metal complex by a simple sequence of 7r-complexation followed by a hydride shift, thus forming a 77-allyl-metal hydride entity which then rearranges into a metallocyclobutane via a nucleophilic attack of the hydride on the central atom of the 7r-allyl species ... 

As opinions regarding the metathesis reaction pathway have shifted in recent years to favor a nonpairwise carbene-to-metallocyclobutane transformation, increasing attention has been given to the mechanistic significance of cyclopropane olefin interconversions. This interconversion process seems to occur primarily when certain relatively inefficient catalysts are employed, which in itself raises questions. Under ideal conditions, "good metathesis catalysts are remarkably efficient promoters of transalkylidenation, and consequently are well suited for olefin and polymer syntheses. Thus, most early studies focused primarily on applications. When side reactions did occur, they were usually ignored or presumed to be trivial cationic processes. 

Several studies now exist which add credence to Casey s attractive proposal that metallocyclobutanes intervene in metathesis and cyclopro-panation reactions. Additionally, metallocyclobutanes have been observed to interconvert to propylene derivatives by way of /3-hydrogen transfer reactions. It is not well established, however, whether precoordination of the reacting olefin is required in all these processes. The proposed interrelation of these reactions may be formally presented as follows ... 

The fact that Schrock s proposed metallocyclobutanes decomposed to propylene derivatives rather than cyclopropanes was fortunate in that further information resulted regarding the stereochemistry of the olefin reaction with the carbene carbon, as now the /3-carbon from the metal-locycle precursor retained its identity. The reaction course was consistent with nucleophilic attack of the carbene carbon on the complexed olefin, despite potential steric hindrance from the bulky carbene. Decomposition via pathways f-h in Eq. (26) was clearly confirmed in studies utilizing deuterated olefins (67). 

Reaction pathways apparently analogous to d and f of Eq. (26) yield a mixture of propylene and cyclopropane. Only when photochemical activation was employed were the major products olefins derived from metathesis-decomposition of the metallocycle. The failure to form metathesis olefins under moderate conditions is significant. It may be that either unimolecular dissociation of the olefin from the complex (in the absence of excess olefin to restabilize the carbene) is energetically unfavored, or the metallocyclobutane structure in the equilibrium given by steps a and b in Eq. (26) is highly stabilized and favored. These results... 

Metallocyclobutanes from cyclopropanes have been frequently invoked in transition metal-catalyzed rearrangements of strained ring hydrocarbons, and this body of chemistry is quite rich and diverse, as evidenced in the excellent review by Bishop (72). Because of this diversity, the significance of isolated observations should not be overstated nevertheless, certain reactions outlined by Bishop are closely related to the carbene retroadditions reported by Gassman and co-workers using metathesis catalysts. 

Soluble metathesis catalysts yield trans products in reactions with // / v-2-pentene, but generally are not very stereospecific with c/.v-2-pen-tene. In the latter case, the initially formed butenes and hexenes are typically about 60 and 50% cis, respectively. Basset noted (19) that widely diverse catalyst systems behaved similarily, and so it was suggested that the ligand composition about the transition metal was not a significant factor in the steric course of these reactions. Subsequently, various schemes to portray the stereochemistry have been proposed which deal only with interactions involving alkyl substituents on the reacting olefin or on the presumed metallocyclobutane intermediate. 

This scheme now appears to be of very limited predictive value, because the results described previously with c / -4-methyl-2-pentene, which selectively gave pure trans-Ce as one of the products (76), would be entirely unexpected. Furthermore, if the commonly observed lack of stereospecificity in reactions of c/.v-2-pentene were to be attributed to an ionic process involving the metallocyclobutane, one would not expect uniformly similar steric results with diverse catalysts such as those summed up by Basset (81), as varying degrees of cationic isomerization would have been expected with these diverse catalysts. 

Therefore, one cannot presume a fixed amount of puckering to occur in various substituted metallocyclobutanes. 

Alternatively, if regenerative metathesis involved preferentially a methyl-substituted carbene, the prefered metallocyclobutane would favor isomerization of the starting olefin. 

To explain the observed rupture of sterically-hindered substituted tertiarysecondary C-C bond on these low dispersed Pt catalysts, a third mechanism was proposed that competes with the dicarbene mechanism.55 This third mechanism involves a metallocyclobutane species as an intermediate. In fact, both McVicker et al.11 and Coq et al.56 have cited the metallocyclobutane mechanism as responsible for the observed hydrogenolysis of 1,2,4-trimethylcyclohexane and 2,2,3,3-tetra-methylbutane, respectively. 

Fig. 19 Ring opening modes dicarbene (a), 7t-adsorbed olefin (b), metallocyclobutane (c). Adapted from ref. 49.

Initiation involves coordination of the double bond of monomer with the transition metal (imino and OR ligands not shown), cleavage of the 7t-bond with formation of a 4-membered metallocyclobutane intermediate, followed by rearrangement to form a metal-carbene propagating center ... 

Fig. 8. 3C complexes in isomerization. Various pathways, as suggested in the literature (see text), for the "one-site (small ensemble) conversions of metallocyclobutane rings. 

Figure 20.11 Mechanistic scheme for n-heptane isomerization according to a bond-shift mechanism with a metallocyclobutane intermediate.

Scheme 2. Schematic relationships between C2 hydrocarbon adsorbates and the surface species that could be derived from them. Possible relationships are indicated between surface species that involve not more than one addition/subtraction of H and/or M atoms. Strong bonding of these species to metal atoms could give rise to structures approximating to metallocyclopropanes or metallocyclopropenes, respectively. These structures with C=C or C = C groupings frequently occur on surfaces with additional n bonds to further metal atoms, M in other cases, two additional CM bonds may replace a CC double bond. JOn surfaces the metal atoms are usually bonded to each other so that the analogous molecular cluster compounds would be metallocyclopropanes or metallocyclobutanes, etc. 1 See footnote to Scheme I. The dashed rectangles indicate surface species that involve no CH bond breaking on adsorption of the parent hydrocarbon. These are most likely to be present under low-temperature adsorption conditions.

Two resonance-contributing structures (3a and 3b), in the formalism of ylide structures, can be used to describe metal carbene intermediates. The highly electrophilic character of those derived from Cu and Rh catalysts suggests that the contribution from the metal-stabilized carbocation 3b is important in the overall evaluation of the reactivities and selectivities of these metal carbene intermediates. Emphasis on the metal carbene structure 3a has led to the subsequently discounted proposal that cyclopropane formation from reactions with alkenes occurs through the intervention of a metallocyclobutane intermediate [18]. The metal-stabilized carbocation structure 3b is consistent with the cyclopropanation mechanism in which LnM dissociates from the carbene as bond-formation occurs between the carbene and the reacting alkene (Eq. 5.4) [7,15]. 

The extensive data accumulated by Nakamura and Otsuka, although interpreted by them as being due to the intervention of metal carbene and metallocyclobutane intermediates, can also be rationalized by an alternative mechanism in which coordination of the chiral Co(II) catalyst with the alkene activates the alkene for electrophilic addition to the diazo compound (Scheme 5.4). Subsequent ring closure can be envisioned to occur via a diazonium ion intermediate, without involving at any stage a metal carbene intermediate. 

Akeypart in the analysis and proposal was electron accountancy - on the basis of the usual propensity for the adoption of a noble gas outer shell configuration - and all of the proposed intermediates have 16- or 18-electron configurations except the metallocyclobutane (26) in pathway II which has 14. On this basis, one might expect that pathway I, the non-dissociative pathway, would predominate over pathway II. In the presence of a large excess of Cy3P, which was used only to simplify the kinetic analysis, this is clearly the case. However, under the... 

Because transition metals other than Pt (as far as they have been examined) have not such a propensity to form a metallocyclobutane directly from a gem-dimethyl group the way to ready isomerization of neopentane is barred. However, an alkane such as isobutane may have an indirect route to the ay-diadsorbed species provided that 1,2-migration of hydrogen is relatively easy (60d), as follows. 

---