Triadsorbed species

The facile isomerization of strained bridged molecules was inconsistent with the involvement of triadsorbed species 15. Instead, a a-alkyl adsorbed species was proposed. Its further transformation leads to a it-complex-like species (17) with a three-center, two-electron bonding with its simultaneous attachment to the surface152,157 (McKervey-Rooney-Samman mechanism) ... 

Different approaches to the C5 cyclic mechanism have been put forward. Comparative studies of the isomerization of 2,2,3-trimethylpentane, 2,2,4-trimethylpen-tane, and 2,2,4,4-tetramethylpentane indicated161 that the latter did not display any appreciable isomerization on palladium below 360°C. Since this compound cannot undergo dehydrogenation without rearrangement to form an alkene, the cyclic mechanism on palladium is suggested to occur via the 1,2,5 triadsorbed species (20) bonded to a single surface atom. 

The favored cleavage of bonds P to a tertiary carbon on platinum (e.g., in 2-methylpentane to yield isobutane and ethane) is accounted for by a,a,Y-triadsorbed species 37 ... 

Anderson and his co-workers examined the reactions of small alkanes mainly on platinum and palladium 45-48). Isobutane was isomerized to n-butane on platinum and on palladium, neopentane isomerized to isopentane on platinum, whereas other metals (including palladium) caused hydrogenolysis predominantly or exclusively. It was proposed that the slow step in the isomerization was the formation of a bridged intermediate (C) from an aory-triadsorbed species (A, B) (Fig. 11).1 Hiickel MO calculations based on this proposal suggested,... 

To provide adequate background for the work to be described next, some further findings by Anderson and Avery may be mentioned. The selectivity for isomerization versus hydrogenolysis (St = 77/77 ) of isobutane on evaporated films of platinum claimed to expose (111) faces predominantly was found to be enhanced by a factor of 5 relative to unoriented films this enhancement was not observed for n-butane (Table VI). Anderson and Avery (47) proposed that a symmetrical triadsorbed species (Diagram 1) is the preferred reaction intermediate for isobutane, such an intermediate not being possible for n-butane. This intermediate fits the triplets of metal atoms on the (111) plane of platinum, suggesting, they believed, a basis for the enhanced efficiency of the (111) plane for the isomerization of isobutane. We note that inspection of rates of isomerization given in the paper of Anderson and Avery shows a factor of only... 

Recently, Rooney and co-workers (23,58,59) have questioned the view that triadsorption by loss of 3 hydrogen atoms from the alkane is the minimum requirement for bond-shift reactions. They studied the isomerization of a series of caged hydrocarbons in excess hydrogen on palladium and platinum catalysts. The compounds were chosen in order to render difficult or totally exclude a mechanism involving aory-triadsorbed species. Thus, l,7,7-trimethyl[2,2,l]-heptane interconverts with its endo- and exo-2,3,3-trimethyl isomers, bicyclo-[3,2,2] octane changes to bicyclo[3,3,l] nonane, and protoadamantane to... 

In order to explain the isomerization of neopentane to isopentane on platinum films, Anderson and Avery [34) proposed a mechanism involving, as precursor, an a,a,y-triadsorbed species, and, in the transition state, a 71 complex of the Dewar type, attached to the surface by two carbon-metal bonds. By simplified Hiickel molecular orbital (MO) calculations, they showed that hyperconjugative effect and partial charge transfer to the metal could account for the relative isomerization rates of the various molecules studied (neopentane > isobutane > n-butane) (Scheme 19). 

The first approach to the cyclic mechanism of isomerization was the finding that the interconversion of n-hexane and methylpentanes takes place under the conditions where the nonselective mechanism of hydrogenolysis (Mechanism A) is the only one operating that is, on 0.2% Pt/AljOj (32). The identical product distributions in isomerization of hexanes and hydrogenolysis of methylcyclopentane suggested that both reactions involve a common intermediate with a methylcyclopentane structure. It was then proposed that the species responsible for dehydrocyclization of hexanes are a,j8, -triadsorbed species involving a single metal atom (55) (Scheme 40). 

This was an attempt at a more precise description of the same type of 1,2,6-triadsorbed species invoked by Herington and Rideal (96) as key intermediates in aromatization of alkanes by 1-6 ring closure (Scheme 42). 

On platinum, the a, -dicarbene mechanism which accounts for the hydrogenolysis of cycloalkanes (Scheme 34) is no longer predominant in the hydrocracking of acyclic alkanes. It has already been emphasized that the internal fission of isopentane and n-pentane is related to the metallocyclobutane bond shift mechanism of isomerization (see Section III, Scheme 29), and that in more complex molecules, the favored rupture of the C-C bonds in a p position to a tertiary carbon atom is best explained by the rupture of an a,a,y-triadsorbed species (see Section III, Scheme 30). The latter scheme can account for the mechanism of hydrocracking of methylpentanes on platinum. Finally, the easy rupture of quaternary-quaternary C-C bonds in... 

While the first process is likely in the case of iridium, nickel, and cobalt, it should not be so easy on platinum, because of its competition with carbene-olefin isomerization (see Section III, Scheme 29). We believe that the only way of explaining why 1,2-dicarbenes may account for the hydrogenolysis of cyclic hydrocarbons (Scheme 34), but only for a minor part for the hydrocracking of acyclic hydrocarbons, is the competition, for the latter, between carbene-dicarbene formation and carbene-olefin isomerization. Carbene-olefin interconversions are unlikely in the case of cyclic hydrocarbons, since a dicarbene species cannot transform into a 1,1,2,3-tetraadsorbed species (l-carbene-2,3-olefin) and further into a 1,1,3-triadsorbed species without C-C rupturing. 

These deuterium distributions were interpreted by assuming that in 11,12, and 13, rapid interconversions between a,y-diadsorbed and a,a,y-triadsorbed species, attached to the metal at the two gemdimethyl carbon atoms, were followed by metallocyclobutane dismutation. The resulting adsorbed methylene-cycloalkane may be deuterated via either n-alkyl or tertiary alkyl species. In the former case, rapid desorption would occur, with the formation of a d4 molecule, while in the latter case, a number of interconversions between cr-alkyl and 7r-olefinic cyclic adsorbed species would yield the perdeutero compound. In the case of demethylation at a tertiary center. 

The mechanism proposed by Shephard and Rooney (95) involved a (T-7i-l,2,5-triadsorbed species for the dehydrocyclization of o-ethyltoluene and a n-allylic species for the dehydrocyclization of propylbenzene (Scheme 41). We suggest, instead, a carbene-benzene addition mechanism, in better agreement with the general picture we have given for 1-5 and 1-6 ring closure (Scheme 81). 

The Anderson-Avery mechanism151155 assumes the involvement of the 15 a,a,y triadsorbed surface species ... 

---