Pt-Catalysed Alkylation of Secondary Phosphines

This reaction is only plausible if phosphide complex 73 is the most nucleophilic species present in the mixture. Therefore, the base must not be too strong to deprotonate 72, otherwise the generated phosphide anion would quiekly react with the benzyl halide to give racemic 74. Sodium trimethylsilanoate appeared to be a good choice. When benzyl bromide was used as the electrophile in the presence of 5% of 73, the tertiary phosphine 74 was obtained in quantitative yield and with 77% ee. Interestingly, with benzyl chloride the reaction proceeds equally well but is extremely slow it took one month to reach completion compared to 3 h with benzyl bromide. These remarkable results were extended to a variety of benzylic substrates and secondary phosphines (Table 6.12). In these reactions instead of using the rather sensitive complexes 73 or analogues, the air-stable precursor Pt((7, 7 )-Me-DuPhos)(Q)(Ph)] was employed. 

A NMR monitoring method was developed to evaluate the rate and diastereoselectivity of the reaction quickly and efficiently. This is possible because in the products the NMR corresponding to C2-76 and meso-16 isomers are well separated. 

The reaction of 75b with benzyl bromide was used to optimise the catalyst. After screening Pt complexes containing several typical chiral diphosphines, the best catalysts were again found to be Pt(DuPhos)(Cl)(Ph)]. A variety of benzylic, naphthylmethyl and anthracenylmethyl halides were screened as alkylating agents. Low to moderate diastereoselectivies (0 0% de to C2-76) were observed and in one case (75b, Ar = o-trifluoromethylbenzyl) both enantiomers of compound C2-I6 were isolated enantiomerically pure on a gram scale. 

Some trends were observed in the catalysis. The reactivity of 75 follows the order ob d a, while the diastereoselectivity towards C2-76 is similar, Ob d a. No direct structure/selectivity relationship is evident among the benzyl halides, although as expected bromides react faster than chlorides. 

A requirement for the NMR monitoring method is that all the species must be soluble. That is not the case in the reaction of anthracenyl methyl chloride with 75b. In this case an anomalously high diastereoselectivity (97% de towards C2-I6) was found, which could not be reproduced when the reaction (30% yield, 70% ee) was scaled up. This can be explained by the insolubility of meso-76, which precipitates during the catalysis. 

---

Scheme 6.38 Proposed mechanism of Pt-catalysed alkylation of secondary phosphines.

---