The synthesis is described of the first stable pentaalkylphosphorane, 6, and of some phenyl-substituted analogs (Scheme I). These phosphoranes are stable because strain in the precursory tetravalent phosphonium salt is relieved when a fifth substituent attaches to phosphorus. An alternative explanation, that the methyl protons in 8, for example, are less acidic than those in tetramethylphosphonium iodide, is excluded by the observation that in 0.27 TVNaOD at 26.8°the methyl protons in 8 exchange with deuterons 44 times faster than those in tetramethylphosphonium iodide. The latter explanation also does not account for the reactions shown in Scheme II, of phosphine oxide 9 with phenyllithium and HBr giving diphenylphosphonium salt 2 and of 10 with methyllithium giving 8, while the former explanation does. Alkylphosphoranes in general should be preparable if two of the substituents are constrained in a ring that is sufficiently small. Whether a cyclic phosphonium salt reacts with an organolithium reagent to give an ylide or a pentasubstituted phosphorane should depend on the ring angle at the phosphorus, and, if so, the angle below which pentasubstituted phosphoranes form and above which ylides form is bracketed between that in the bicyclo[4.2.l]nonane and the bicyclo[2.2.1]heptane ring systems. Thus 24a with phenyllithium gives the ylide 25, while 20, which differs from 24a by only one bond, gives phosphorane 21 instead. In accord with these results is the failure to prepare the diphenylphosphonium salt 24a from phos-Journal of the American Chemical Society / 95:13 / June 27, 1973 (12) J. Mayerle, Columbia University. We thank Dr. Mayerle for his results.

No abstract

No abstract

No abstract

No abstract