Oxonium ylide 1,2 -Stevens rearrangement

Applications in Synthesis Oxonium Ylide [1,2]-Stevens Rearrangements... 

The addition of /i-toluenesulfmate to the silyl ether of l-hydroxybut-3-yn-4-yl(phenyl)iodonium triflate triggers a sequence of reactions that ultimately delivers 2-substituted-3-(/i-toluenesulfonyl)dihydrofuran products in variable yields (Scheme 31). A putative 1,2-group shift within an unsaturated oxonium ylide (Stevens rearrangement) accounts for the oxygen-to-carbon transfer of the ether substituent <2000JOC8659>. 

The other major reaction pathway for oxonium ylide is [l,2]-shift (Stevens rearrangement). Compared with [2,3]-sigmatropic rearrangement, which is an orbital symmetry-allowed concerted process, the [l,2]-shift has higher activation barrier, [1,2]-Shift is generally considered as stepwise process with radical pair as possible intermediates. 

Van den Berg et al. suggested an intramolecular Stevens rearrangement of the oxo-nium ylide to ethyl methyl ether to interpret carbon-carbon bond formation463 [Eq. (3.55)]. Olah and coworkers, however, provided evidence (based on isotopic labeling studies) that the oxonium ylide undergoes intermolecular methylation to ethyl dimethyloxonium ion [Eq. (3.56)] instead of Stevens rearrangement 447... 

Cyclic mixed acetals with pendant diazo ketone side-chains undergo rearrangement to ether-bridged cycloheptane ring systems on treatment with Cu(hfacac)2.118 A Stevens [1,2]-shift of an oxonium ylide gives the major product (62), in some cases accompanied by minor amounts of a product (63) resulting from a [l,2]-shift of a sulfonium ylide. 

Woerpel and coworkers interpreted the results of these mechanistic experiments as evidence that the insertion of silylene into the C-O bond occurs through a [1,2]-Stevens rearrangement of oxonium ylide 198 and that a competitive [2,3]-sigmatropic rearrangement of 198 could account for allylic transposition. 

While [2,3]-onium ylide rearrangements are proceeding most often by concerted pathways (review [388]), [l,2]-onium ylide rearrangements (Stevens rearrangements) often involve radical intermediates (recent reviews [389, 390]). A study aimed at the comparison of Rh(II) and Cu(II)-catalyzed oxonium ylide reactions of oxygenated diazoketones using 3 mol% of the former and 15 mol% of the latter showed that the copper-catalyzed process provided better yields and selectivities for [1,2]-rearrangement products (see Part 2, Sect. 6, Fig. 107) [391, 392]. 

Three issues need to be addressed in connection with oxonium ylide mechanisms. The first question concerns the existence of oxonium ylides. Whereas sulfonium, phosphonium, and ammonium ylides are well known, oxonium ylides have not been isolated. Secondly, if they exist, will they undergo Stevens rearrangement, or decompose via other routes Finally, is the zeolite conjugate base sufficiently basic to abstract a proton from oxonium ions to form an ylide ... 

It is apparent that much resourceful, imaginative experimentation has been done to resolve the question of C-C bond formation from methanol. Although the answer remains elusive, these experiments tell us at least what is probably not involved in the bond formation, particularly in the presence of zeolite catalysts. The Stevens rearrangement of oxonium ylide can be ruled out, as well as the carbocationic route invoking hypervalent carbon transition states. Not excluded are surface-bound species such as carbenoids and ylides. Again there seems to be a consensus that surface methoxyls are precursors to these reactive C- intermediates, which seems somehow to be "coming full circle", since surface methoxyls were early shown to be intermediates in the formation of DME, which is itself an intermediate in hydrocarbon formation. Finally, if the free radical character of the initiation step proves correct, the implications to zeolite catalysis will be far-reaching. 

The most popular mechanisms at present invoke oxonium ylides as intermediates. van den Berg et al. [20] proposed that DME is protonated by a Bronsted site, and the resultant ion suffers nucleophilic attack by a second molecule of DME to form TMO with release of MeOH. The TMO ion is then deprotonated by a basic site to form the dimethyloxonium methylide, which undergoes a Stevens-type rearrangement to give methylethyl-oxonium ion. MeOEt is subsequently formed upon p— elimination. No experimental evidence was offered in support of the scheme. 

At this point, three questions must be addressed. Do oxonium ylides indeed exist If so, will they undergo a Stevens-type rearrangement Finally, will a zeolite conjugate base be sufficiently basic for proton abstraction from an oxonium ion to generate the ylide ... 

A mechanism that has received a great deal of attention is the oxonium ylide mechanism.Dimethylether is methylated to trimethyloxonium, which is subsequently deprotonated to form surface associated methylene-dimethyloxoniumylide. The next step is either an intramolecular Stevens rearrangement, leading to the formation of methylethyl-ether, or an intermolecular methylation, leading to the formation of ethyl-dimethyloxoniumion. In both cases ethylene is obtained via jS-elimination. 

Scheme 5.5 First C—C bond formation by Stevens rearrangement (left) or oxonium ylide route ...

---