Ruthenacyclopentadiene Intermediates

In this homocoupling, two orientations are possible, head-to-head as in cycle A and head-to-tail as in cycle B. Indeed, exposing the propargyl alcohol 17 to the trisacetoni-trile complex 16 [21], whose ease of dissociation of this ligand makes it a functional equivalent of a highly coordinatively unsaturated ruthenium, effects coupling even 

21 (Equation 1.25), which served as a synthetic intermediate for the synthesis of kainic add, a neuroexcitatory amine acid. 

A most useful application of this process is the invention of a highly atom economic synthesis of pyridines, wherein the only stoichiometric by-product is water. Cycloisomerization of diyne 22 followed by reaction with hydroxylamine provides the tricyclic pyridine 23 with only water as the stoichiometric by-produd (Equation 1.26) [24]. The Ru-catalyzed cycloisomerization of propargyl alcohols can also generate six membered rings, which then form tetrahydroisoquinolines as shown in Equation 1.27. 

An intermolecular cross-coupling requires that the simple alkyne be a better binder to Ru than the propargylic alcohol. co-Cyanoalkynes meet that requirement. Thus, 6-cyano-l-hexyne generates the crossed coupled produd in 75% yield (Equation 1.28) [25]. 

In substrates bearing suitable functionality that could lead to internal nucleophilic attack by an oxygen, even a carbonyl oxygen, excellent regioselectivity may occur. Equation 1.31 illustrates an example of a ketone playing such a role [27]. Such an effect can even lead to hydration at the more hindered alkyne as shown in 

---

The assumed mechanism proceeds via ruthenacyclopentadiene intermediates of type 160 or structure 161 as possible intermediates. Subsequently, the common bicyclic intermediate 162 is formed through insertion of the C=N double bond into the C-Ru bond, and reductive elimination of the ruthenium fragment gives rise to the desired bicyclic pyridones 163. 

The catalytic cyclocarbonylations of diynes proceed efficiently to afford fused cyclohexadienes via trapping of the ruthenacyclopentadiene intermediate by an alkene component <2000JA4310>. Thus, the ruthenium-catalyzed cyclo-co-trimerization of 1,6-heptadiyne derivatives possessing a heteroatom at the 4-position affords heterotricycles in good yields (Equation 110). 

With strained bicycloalkenes such as norbornene derivatives a ruthenium-catalyzed tandem cyclopropanation occurred together with common [2+2+2] cy-clotrimerization, showing a biscarbenoid hybride structure for the ruthenacyclopentadiene intermediate [92] (Eq. 72). 

Cyclotrimerization of alkynes mediated by the cationic complex [(77-Cp)Ru(acetonitrile)3](PF6) was shown by the DFT methods to proceed via the ruthenacyclopentadiene intermediates in accord with experimental findings <2003JOM(682)204>. One illustration for the transformation of such an intermediate into the final product is illustrated... 

Reactions between two olefins or two alkynes are also possible, although less common. 1,6-Diynes are fascinating substrates they are often paired with a third alkyne moiety or activated olefin to produce a seven-membered ruthenacycle, as the ruthenacyclopentadiene intermediate cannot undergo /3-hydride elimination because there is no available hydrogen. This second intermediate likely undergoes direct reductive elimination to afford the observed product, a six-membered ring Yamamoto and co-workers have reported regioselective synthesis of benzene derivatives from this synthetic pathway (81). 

In 2003, Saa and coworkers performed a comprehensive study on cationic [Cp Ru(CH3CN)3]PF6 complex-catalyzed [2 -I- 2 -I- 2] cycloaddition of 1,6-diynes to a,CT-dinitriles or electron-deficient nitriles (Scheme 3.13) [33]. The reaction with asymmetric electron-deficient alkynes could give the corresponding 2,3,6-trisubstituted pyridines in good yield. Based on their studies, they propose that the reactions with dinitriles seem likely to proceed via ruthenacyclopentadiene intermediates and the reactions with electron-poor nitriles via azaruthenacyclopentadienes. 

Although ruthenacyclopentadiene intermediates are able to coordinate a third molecule of alkyne to afford benzene derivatives, catal3Aic intermolecular [2+2+2] cycloadditions involving three nontethered a]k3mes are seldom because of their low selectivity. For example, substituted o-phtalates could be chemoselectively obtained by reaction of two equivalents of terminal alkynes with dimethylacetylenedicarboxylate (DMAD) [11] or by cycloaddition of an internal alkyne, a terminal alkyne and DMAD [12] [Eq. (8)] in the presence of the Cp RuCl(cod) complex. 

The reaction of 1,6-heptadiynes with alkenes led to a [2+2+2] cyclotrimer-ization in the case of cyclic or linear alkenes possessing heteroatoms at the al-lylic position. Bicyclic cyclohexadienes were thus produced in good yields with RuCl(COD)C5Me5 [92,93] (Eq. 71). A ruthenacyclopentadiene is invoked as an intermediate in the mechanism. Insertion of the alkene becomes possible by a heteroatom-assisted reaction. 

Ruthenacyclopentane 331 has been postulated as an intermediate in the ruthenium-catalyzed cycloisomerization of lactones <2003TL2157>. Cycloisomerization of phenylsulfonylallenes to the cyclohexane derivatives catalyzed by a ruthenium benzylidene complex might proceed through the ruthenacyclopentane intermediate 332 <2006TL3971>. The [( 7 -Cp")RuCl( -COD)]-catalyzed cyclotrimerization of 1-octyne with dimethyl acetylenedi-carboxylate proceeds via a ruthenacyclopentadiene <2004JMO(209)35>. 

The strong forward donation-back donation of electrons (i.e., the Chatt model) between alkynes and ruthenium makes such a bond a very good ligand for Ru. Hence it is not surprising that reactions involving ruthenacyclopentadienes as intermediates, notably in the trimerization of alkynes to benzenes, occur readily. Intercepting the ruthenacyclopentadiene prior to its reaction with an additional alkyne, however, is rather rare. A unique juxtaposition of functionality occurs when a propargyl alcohol is the alkyne partner which allows such a diversion as shown in Scheme 1.3. 

---