Nucleophilic racemic allyl acetates

Although Helmchen et al. showed that asymmetric iridium-catalyzed allylic substitution could be achieved, the scope of the reactions catalyzed by iridium complexes of the PHOX ligands was limited. Thus, they evaluated reactions catalyzed by complexes generated from [lr(COD)Cl]2 and the dimethylamine-derived phosphoramidite monophos (Scheme 8) [45,51]. Although selectivity for the branched isomer from addition of malonate nucleophiles to allylic acetates was excellent, the highest enantiomeric excess obtained was 86%. This enantiomeric excess was obtained from a reaction of racemic branched allylic acetate. The enantiomeric excess was lower when linear allylic acetates were used. This system catalyzed addition of the hthium salts of A-benzyl sulfonamides to aUylic acetates, but the product of the reaction between this reagent and an alkyl-substituted linear aUylic acetate was formed with an enantiomeric excess of 13%. 

Kinetic resolutions of allylic esters have also been conducted. As noted in Chapter 14, in most cases, catalysts that are selective for kinetic resolution of substrates containing one type of functional group are also selective for reactions of meso substrates and vice versa.The enantioselective reaction of one acetate of a meso substrate involves similar stereochemical recognition to reaction of two enantiomers of a racemic mixture of allylic esters. In one illustrative example, an allylic acetate underwent reaction with a pivalate nucleophile to form the allylic pivalate product, which is less reactive than the starting compound (Equation 20.54). Reaction of one enantiomer of the racemic allylic acetate occurs preferentially, and the pivalate product is formed enantioselec-tively. The product of the resolution in Equation 20.54 was carried forward to form (+)-cyclophellitol. 

Palladium-catalyzed nucleophilic substitutions of activated allylic alcohols have been investigated using a bicyclic phosphine as the chiral element. Reactions have been reported with racemic acyclic allylic acetates <1999TL7791> and cycloalkenyl carbonates <2001TL1297> using the phosphine 203, with good to excellent enantiomeric excesses. 

There are also a few reactions showing more or less preference for an anti SN2 reaction, including a cyclopentane-forming reaction with an enolate nucleophile,589 and a sulfur nucleophile.590 In addition, the reactions of alkyl cuprates with allylic acetates are always stereospecifically anti 5.176 > 5.177, although the formation of racemic product shows that regiocontrol has been lost. 

Allyl acetates 1/ent-l that possess identical R groups undergo aUyUc substitution via an achiral intermediate 2. Both enantiomers of starting material proceed via the same intermediate. In the absence of any controlling influence, approach of the nucleophile via pathways a and b is equally likely, and a racemic product 3/ent-3 will be formed (Scheme 1). However, the opportunity for an asymmetric catalytic reaction exists if the reaction can be channeled through one pathway selectively. Overall, the process represents a dynamic resolution, since a racemic starting material is converted into an enan-tiomericaUy emiched product. 

When a racemic allylic substrate is employed in an enantioselective substitution reaction, one of the two substrate enantiomers may react more quickly than the other. This effect is a kinetic resolution and has been noted reasonably often in enantioselective allylic substitution reactions. Several studies on kinetic resolution have been reported, - and a few highlight reactions are noted in Scheme 45. These include recovery of unreacted cy-clohexenyl acetate 92, as well as the tetraacetate 225. Kinetic resolution has also been observed in allylic snbstitution using a snlfinate nucleophile (Li02STlu) with allyl acetate... 

For an allylic acetate such as 9.125 in which the substituents at each end are the same, reaction with a source of palladium(O) will generate an achiral palladium complex 9.126, and then a racemic product 9.127, regardless of the original chirality of the starting material (Scheme 9.39). If the ligands on the palladium are chiral, then the intermediate tf-allyl complex will not be a symmetrical compound. Under the right circumstances, the nucleophile may be directed so that a non-racemic product is obtained. 

Optically active barbiturates important in pharmacy, are obtained by classical synthesis and resolution procedures [14]. A new approach to the synthesis of optically active barbiturates is enantioselective catalysis with Pd complexes. An example is shown in Scheme 10 on the right side. The synthesis of the enantiomers of a N-methyl barbiturate containing a quaternary asymmetric carbon atom is achieved by allylation of the corresponding precursor with allyl acetate. Pd/triphenylphosphine complexes are efficient catalysts for this reaction which give a racemic mixture of the product. The reaction proceeds via a jt-allyl complex of Pd as an intermediate (Scheme 10, left side). This 7t-allyl complex is attacked by the nucleophile, the anion of N-... 

In the allylic substitution of racemic 2-propenyl acetates or related substrates with the same substituents at 1 and 3 positions, the jt-allylpalladium intermediate containing a meso type 7r-allyl group is formed from both enantiomers of the allylic substrate. Two jt-allyl carbons at the 1- and 3-positions are diastereotopic on coordination of a chiral phosphine ligand to palladium. The asymmetric induction arises from preferential attack by the nucleophile on either of the two diastereotopic TT-allyl carbon atoms (Scheme 2-28). 

The control of stereoselectivity in acyclic substrates is more difficult than in cyclic substrates. As discussed previously, palladium can serve as a template to provide rigidity in an acyclic system thus favoring higher stereoselectivity. Palladium(0)-mediated substitution of the chiral nonracemic ally acetate depicted below yields only racemic material68, n-o-n Rearrangement of the intermediate 7t-allyl complex involving the unsubstituted allyl terminus is probably faster than nucleophilic attack. 

Since ethanol and acetic acid have similar ionizing powers, it is not surprising that racemization of each optically active chloride occurs at about the same rate in the two solvents. The more nucleophilic solvent, ethanol, is much more efficient than acetic acid in reacting with the allylic carbonium ion, with the result that isomerization is faster, and solvolysis slower, in acetic acid than in ethanol. That is, the partitioning of the ion pair between solvolysis products and racemic chloride is more in favor of solvolysis products in the more nucleophilic solvent. 

The stereochemical results of these reactions have been carefully studied." If a chiral acetate is employed, then the product is found to have retention of stereochemistry. Hence, the cis acetoxy ester 9.98 gives the cis product 9.100, while the trans acetoxy ester 9.101 gives the trans product 9.103 (Scheme 9.32)." Retention is a result of two inversions - inversion during formation of the t -allyl complexes, 9.99 and 9.102, and a second inversion during attack by the nucleophile. In the case of the acetoxy esters 9.98 and 9.101, a curious observation can be made when the substrate is non-racemic. It is found that the product is racemic. This is not a general observation. It occurs here, because the intermediate V-allyl complexes, 9.99 and 9.102, each have a plane of symmetry - they are meso compounds and the nucleophile is equally likely to attack either terminus. Systems without the symmetrical intermediate do not show racemization (Scheme 9.33). 

Thanks to the fundamental studies of Tsuji, Trost, and others, palladium-catalyzed allylic substitution has become a versatile, widely used process in organic synthesis [40]. The search for efficient enantioselective catalysts for this class of reactions is an important goal of current research in this field [41]. It has been shown that chiral phosphine ligands can induce substantial enantiomeric excesses in Pd-catalyzed reactions of racemic or achiral allylic substrates with nucleophiles [42]. Recently, promising results have also been obtained with chiral bidentate nitrogen ligands [43]. We have found that palladium complexes of neutral aza-semicorrin or methylene-bis(oxazoline) ligands are effective catalysts for the enantioselective allylic alkylation of l,3-diphenyl-2-propenyl acetate or related substrates with dimethyl malonate (Schemes 18 [25,30] and 19 [44]). 

---