Application of PyrBOX Triphenyl Allyl Complexes 62a-d in Asymmetric Allylic Substitutions

The triphenyl allyl substrate belongs to the class of unsymmetrically substituted allyl systems, thus, can serve as an example for this type of mechanism for enantiodiscrimination (see Chapter 5.4.3). As already mentioned in the previous section, the obtained complexes provided in solution only one isomer, which facilitates mechanistic considerations for the enantiocontrol.

In comparison to many other unsymmetric allylic substrates, the employed substrate offers an exclusive regioselectivity for the nucleophilic attack. Furthermore, it exemplifies the application of allylic acetates, which can easily be derived from the commercially available β-phenylcinnamaldehyde and related aldehydes. Subsequently, the substitution products can (after asymmetric allylic substitution) be transformed further into enantiomerically enriched succinic acids,^[160,169] lactones,^[169] amino acids and derivatives (Scheme 7.23).[160,203,204,205]

Unfortunately, these substrates often exhibit a lower or no reactivity, thus making the synthesis of catalysts that can perform such transformations a desirable goal.^[203]

Scheme 7.23: Potential downstream products for 64 starting from various aldehydes.

7.3.4.1 Substrate Synthesis and General Remarks

Although the synthesis of allylic acetates normally is quite straightforward, it proved to be more difficult than expected in case of the triphenyl allyl system. In Scheme 7.24 the synthetic strategy by Bosnich et al. is depicted.^[160] Starting from the tertiary alcohol 60, the tertiary acetate 63 should be obtained after deprotonation and addition of acetyl chloride.

The subsequent rearrangement to the more stable secondary acetate 64 under acidic conditions (traces of acid in chloroform, silica and alumina columns) or thermal induction (60° C) was also reported. Aside from this, in alcohol solution the formation of an allyl ether was observed.^[160] Thus, the approach starting from the tertiary alcohol 60, would to provide synthetic access to both a prochiral 63 and racemic chiral substrate 64.

Scheme 7.24: Envisioned synthesis of 63 and 64 based on reference [160].

Thus, the reaction was carried out analogously to the reported procedure using acetic anhydride instead of acetyl chloride, due to availability. Unfortunately, after work-up only a product mixture could be isolated without any indication of the desired product 63. Instead the secondary acetate 64, as well as another allylic species could be identified. Based on integration, the other species was assigned to a dimeric allylic ether 65, resulting from a nucleophilic attack of a deprotonated alcohol to 63 or 64. This product mixture might be explained by the reported instability of the tertiary acetate 63.

Scheme 7.25: Synthetic route towards 64.

Based on previous findings, it was assumed, that under acidic conditions in presence of an acetyl transfer agent, the product mixture of 64 and 65 should be selectively transformed into the secondary acetate 64 (Scheme 7.25). Thus, using acetic anhydride in combination with acetic acid yielded the pure secondary acetate 64 after aqueous work up.

Further attempts to isolate the tertiary acetate 63 using acetyl chloride or different conditions were not undertaken, as the rearrangement seems to be a common problem.^[206]

Therefore, only the reactivity of the racemic chiral substrate 64 was studied in the course of this thesis.

7.3.4.2 Asymmetric Allylic Alkylation using Dimethyl Malonate Initially, similar reaction conditions as for the diphenyl allyl substrate 55 were applied.

However, at room temperature neither the BSA-method nor the traditional sodium malonate procedure yielded any detectable amounts of the target product 66 (Scheme 7.26). This is in agreement with the described lower activity for these substrates.[169,183,207]

Scheme 7.26: General reaction scheme for asymmetric allylic alkylation of 64 with dimethyl malonate.

Therefore, the reaction was performed under harsher conditions, applying a catalyst loading of 5 mol% using sodium malonate as nucleophile in refluxing THF. After 8 days, in case of precatalyst 46c = [L³Pd2(C4H7)2]BF4, a yield of 13% of the target product 66 could be isolated, while for the analogous ligand 3a bearing phenyl substituents on the oxazoline ring, a conversion of around 2% could be estimated from the ¹H-NMR spectrum of the reaction mixture. Unfortunately, only a low stereoselectivity (ee = 12%) for the expected (R)-enantiomer^[208] could be observed for the most likely (S)-configured 62c. This selectivity does not come up to one’s expectations, considering the preferred formation of only one isomers and a subsequent nucleophilic attack under inversion of the stereochemistry.

Possible explanations might be a certain degree of retention due to an initial attack at the metal center or competing uncatalyzed processes yielding racemic 66.

The model proposed in Scheme 7.31 might offer an explanation not only for the observed stereochemistry, but also for the low activity of the PyrBOX ligands. Taking 62a as example, nucleophilic attack trans to the oxazoline ring yields the (R)-enantiomer, while iso-propyl substituted oxazolines should yield the (S)-enantiomer. During the rotation to the palladium(0)-olefin complex, severe interactions of the allyl substituents with the ligand

scaffold can be anticipated. As the twist of the allyl-moiety in solid state already indicated such repulsive forces, high activation barriers are very likely.

Scheme 7.27: Enantiocontrol provided by dinuclear pyrazole-bridged Pd catalyst 62a.

Comparable results have been previously reported for triphenyl allyl systems, where in solution usually only one allyl intermediate and no indication for fluxional behavior could be detected by NMR spectroscopy.[136b,152,202]

Even more interesting is the fact that in two cases the observed and assigned isomers^[152,202] do not match the observed chiral products.^[152,201]

So far only the chiral phosphine complexes of Bosnich^[160] were able to provide reasonable results for these substrates.

In a nutshell, although the dinuclear triphenyl allyl palladium complexes were unable to efficiently catalyze transformation of the corresponding substrates, their isolation helped to gain insight into the steric restriction within the chiral pocket provided by the PyrBOX ligands.