Oxaphosphetanes cycloaddition

Olefination Reactions Involving Phosphonium Ylides. The synthetic potential of phosphonium ylides was developed initially by G. Wittig and his associates at the University of Heidelberg. The reaction of a phosphonium ylide with an aldehyde or ketone introduces a carbon-carbon double bond in place of the carbonyl bond. The mechanism originally proposed involves an addition of the nucleophilic ylide carbon to the carbonyl group to form a dipolar intermediate (a betaine), followed by elimination of a phosphine oxide. The elimination is presumed to occur after formation of a four-membered oxaphosphetane intermediate. An alternative mechanism proposes direct formation of the oxaphosphetane by a cycloaddition reaction.236 There have been several computational studies that find the oxaphosphetane structure to be an intermediate.237 Oxaphosphetane intermediates have been observed by NMR studies at low temperature.238 Betaine intermediates have been observed only under special conditions that retard the cyclization and elimination steps.239... 

The generally quite stable methylenephosphorane nevertheless resembles the short-lived highly reactive methyleneoxophosphorane. The oxaphosphetane intermediate 25 formed by -I- n2]-cycloaddition, which can only be isolated in exceptional cases21, is to be seen against the stable oxaphosphetanes of type 26, which can be photolyzed if suitably substituted or thermolyzed under drastic... 

The same reaction sequence performed in methanol affords a mixture of diastereo-mers of the phosphorylated phosphinic ester 48b, of which one pure isomer can be isolated32 . In the presence of piperidine, reductive elimination of nitrogen 28,29) from 45 to give bis(diphenylphosphoryI)methane competes with the prevailing formation of the phosphinic piperidide 48c32). Expected trapping of 47 by [2 + 2]-cycloaddition with benzaldehyde fails to occur in place of 1,2k5-oxaphosphetanes, products are obtained which arise mainly by way of the benzoyl radical32,33). 

According to Section 11.1.3, P-ylides and aldehydes first react in a [2+2]-cycloaddition to form a heterocycle, which is referred to as oxaphosphetane (Figure 4.44). 

We have a fairly detailed knowledge of the mechanism of the Wittig reaction (Figure 11.3). It starts with a one-step [2+2]-cycloaddition of the ylide to the aldehyde. This leads to a heterocycle called an oxaphosphetane. The oxaphosphetane decomposes in the second step—which is a one-step [2+2]-cycloreversion—to give triphenylphosphine oxide and an alkene. This decomposition takes place stereoselectively (cf. Figure 4.44) a cw-disubstituted oxaphosphetane reacts exclusively to give a cis-alkene, whereas a fraws-disubstituted oxaphosphetane gives only a trans-alkene. The reaction is stereospecific. 

The [2+2]-cycloaddition between P ylides and carbonyl compounds to give oxaphosphetanes can be stereogenic. It is stereogenic when the carbanionic C atom of the ylide bears—besides the P atom—two different substituents and when this holds true for the carbonyl group, too. The most important stereogenic oxaphosphetane syntheses of this type start from monosubstituted ylides PhgP —CH —X and from substituted aldehydes R— CH=0. We will therefore study this case in Figure 11.3. 

The [2+2]-cycloaddition between the mentioned reaction partners to form an oxaphosphetane is not only stereogenic hut frequently also exhibits a considerable degree of stereoselectivity. The latter is more precisely called simple diastereoselectivity. 

Stereogenic Wittig reactions of nonstabilized ylides of the structure Ph3P+—CH —R2 have been studied in-depth in many instances. They give the cis-configured oxaphosphetane rapidly, with the rate constant kcis, and reversibly (Figure 9.7). On the other hand, the same nonstabilized ylide produces the /ran.v-oxaphosphetane slowly, with the rate constant ktrans, and irreversibly. The primary product of the [2+2]-cycloaddition of a nonstabilized P ylide to a substituted aldehyde is therefore a cis-oxaphosphetane. Why this is so has not been ascertained despite the numerous suggestions about details of the mechanism which have been made. 

The P—C bonds of 83 are essentially equivalent, suggesting contribution of polar forms. An extremely long P—O bond in contrast to 581 indicates open chain mesomeric structures, stable due to electron delocalization (60). At 125°C a Wittig reaction occurs, and 84 is formed (26). The same type of [2 + 2] cycloaddition has been found with a series of P-(chloro)alkylidene-phosphoranes. The oxaphosphetanes 83b-d thermally eliminate hydrogen chloride to yield vinyloxophosphoranes (165a). 

Currently accepted mechanism of the Wittig reaction of aldehydes with non-stabilized ylides involves the formation of oxaphosphetanes through a [2-I-2]-cycloaddition-like reaction . The oxaphosphetanes are thermally unstable and collapse to alkene and phosphine oxide below room temperature. Under salt-free conditions there is no formation of betaine intermediates. The salt-free ylides can be prepared by the reaction of phosphines with carbenes generated in situ. Vedejs etal proposed a puckered 4-centre cyclic transition state I for sy -oxaphosphetane and planar structure J for anff-oxaphosphetane. In general, the flnfi-oxaphosphetane J is more stable than the syn-oxaphosphetane I, and under equilibrium conditions (when stabilized ylides are used) the E-alkene product is favoured (Scheme 4.24). However, kinetic control conditions, which appear to dominate when non-stabilized ylides are used, would lead to Z-alkene. 

Keglevich et al. have reported a series of papers on the mechanistic aspects of what they term inverse Wittig reactions , i.e. the preparation of phosphoranes from the [2-f2] cycloadditions of phosphine oxides and acetylenedicar-boxylates, an example of which is given in Scheme 1. A raft of spectroscopic and structural evidence, coupled with theoretical calaculations, indicate that these reactions proceed via oxaphosphetane intermediates (16). ... 

In the first step a Wittig transformation is performed. The a-position of the phosphonium salt 12 is C,H-acidic, and therefore the strong base NaHMDS abstracts a proton forming the labile ylide 38. This ylide is a carbon nucleophile, which attacks the ketone moiety of Segment B (7) forming the oxaphosphetane intermediates cw-39 and trans-39. The mode of action is thought to proceed via a [2+2]-cycloaddition. 

The mechanism proposed for the ( )-alkene selectivity involves a nonsynchro-nous cycloaddition with a relatively advanced, productlike transition state leading to the kinetic fran -oxaphosphetane intermediate. Extrusion of triphenylphosphine oxide produces the ( )-alkene. ... 

This reaction is obviously of the greatest significance. Yet the detailed mechanism of the formation of the oxaphosphetane is not agreed. No previous intermediates have been isolated and one of the two main contenders is a one-step mechanism 76. This mechanism has the advantage of a believable explanation for the stereoselectivity. A 2+2 cycloaddition must have an antarafacial component - it must be a K2S + rt2a reaction 76a. 

In a Diels-Alder-like (4 + 2) cycloaddition followed by a 1-3 H-shift the benzoxaphosphinine is formed (Scheme 55) <91BCJ713>. Originally (278) was formulated as a 1,2-oxaphosphetane <73AG(E)1010, 80CB3303). 

Stabilization of the oxaphosphete formed by the cycloaddition of the P = O group and the acetylene moiety involves formally the rupture of the P-O bond and the formation of a P = C and a C = O double-bond (Fig. 16). It is recalled that a Wittig reaction follows the opposite direction the cycloaddition of a P = C and O = C unit gives a 1,2-oxaphosphetane that is opened up to furnish a phosphine oxide and an olefin (Fig. 16). 

The short-lived phosphene (63) yields the oxaphosphetan (64) upon [27T + 27r] cycloaddition with ajS-unsaturated ketones. The reaction of sodium phenylacetylide with tellurium metal forms a 1,3-ditelluretan and not, as had previously been reported, a 1,3-ditellurole. ... 

Several mechanistic variations might be possible under either of the main options (1 or 2). For example, the four-center process might involve a direct conversion from the P=C and the C=0 reactants into the oxaphosphetane (asynchronous cycloaddition) (18,59,66,219,220). In this case, there would be no other intermediates and no energy minima between the reactants and... 

Ionic mechanisms based on betaine intermediates or TS are difficult to reconcile with the absence of solvent effects on lithium-free nonstabilized ylide reactions (Table 12) or reactivity-selectivity considerations (15). Also, there is no apparent reason why the reactants should prefer to form a high-energy intermediate such as 93 when the direct conversion to a more stable oxaphosphetane 97 is possible. Orbital symmetry should not interfere with the four-center process since phosphorus can provide 3d orbitals of appropriate symmetry for a 2s - - 2s cycloaddition. Nevertheless, the betaine mechanism has persisted in the literature because there was no direct evidence against the formation of 93 as a transient intermediate until recently (229). 

---