Achiral or chiral phosphines are widely used in two main domains: ligands in organometallic catalysis and organocatalysis. For this reason, the obtention of optically pure phosphine has always been challenging in the development of asymmetric catalysis. The simplest method to obtain phosphines is the reduction of phosphine oxides. The essential difficulty is the strength of the P=O bond which involves new procedures to maintain a high chemio- and stereoselectivity. The reduction can occur with retention or inversion of the stereogenic phosphorus atom depending on the nature of the reducing agent and the presence of additives. In fact, the reactivity of the phosphine oxides and the mechanism of the reduction are not always well understood. Since the first work in the 1950's, numerous studies have been realised in order to develop methodologies with different reagents or to understand the mechanism of the reaction. In the last decade, efficient stereospecific methodologies have been developed to obtain optically pure tertiary phosphines from P-stereogenic phosphine oxides. In this review, we intend to provide a comprehensive and critical overview of these methodologies.

Well-defined optically pure Transition Metal-complexes bearing C1-and C2-symmetric N-Heterocyclic Carbene (NHC) ligands were prepared from prochiral NHC precursors. As predicted by DFT calculations, our strategy capitalizes on the formation of a metal-carbene bond which induces an axis of chirality. Configurationally stable atropisomers of various NHC containing TM-complexes were isolated by preparative HPLC on a chiral stationary phase in good yields and excellent optical purities (up to 99.5% ee). The carbene transfer from an optically pure Cu-complex to gold or palladium center reveals, for the first time, a full stereoretentivity, supporting the hypothesis of an associative mechanism for the transmetalation. The potential of these new chiral TM-complexes was illustrated in asymmetric catalysis with up to 98% ee.