The synthesis and study of main-group-element analogues of alkenes and alkynes containing genuine (p À p)p bonds involving p-block elements is a central theme of inorganic chemistry. [1,2] The prospect to "copy" the predictable and sophisticated reaction chemistry of C=C and CC bonds utilizing functional inorganic systems is particularly enticing. However, in many instances, the investigation of multiple bonds of heavy elements leads to fascinating, albeit unexpected, outcomes that reinforce the fundamental differences between the first and subsequent periods.Inspired by the intriguing analogy between P=C and C=C bonds in molecular chemistry, [3] we developed the addition polymerization of phosphaalkenes as a route to new functional phosphorus-containing polymers (Scheme 1). [4] Although the synthesis of phosphorus-containing macromolecules is of widespread interest because of their attractive properties and potential applications, [5] the study of the addition polymerization of P = C bonds remains in its infancy. Our studies showed that MesP=CPh 2 (1) and related monomers polymerize in the presence of radical or anionic initiators to afford poly(methylenephosphine) (Scheme 1). [6] The living anionic polymerization of 1 permits the formation of functional phosphine-containing block copolymers, [7,8] and the radical-initiated copolymerization of 1 with styrene affords random copolymers. [9] In order to explore the mechanism of radical addition to P = C bonds during polymerization, we investigated the reactions of monomer 1 with TEMPO-derived radical sources. Herein, we report the discovery of a fascinating isomerization polymerization of phosphaalkene 1 in the presence of radical alkoxyamine initiators. These striking results led to a revision of the proposed microstructure for poly(methylenephosphine) that was produced by a radical reaction.In an effort to understand the initiation step in the radical polymerization of 1, we investigated its reaction with TEMPO (1-2 equiv). 31 P NMR spectroscopic analysis of the reaction mixtures suggested the formation of multiple products, including radical species, which were detected by EPR spectroscopy. To date, none of these products have been successfully isolated or unambiguously identified. In contrast, employing the complex 1·AuCl [10,11] instead of 1 affords a single product with TEMPO. Specifically, treatment of a solution of 1·AuCl in toluene with TEMPO (1 equiv) resulted in 50 % conversion of 1·AuCl (d = 167.1) to a new species, as determined by 31 P NMR spectroscopy. Addition of more TEMPO (1 equiv, that is, 2 equiv total) resulted in complete conversion of 1·AuCl to two products in a ratio of approximately 1:3, which displayed 31 P signals at 135.6 ppm (d, J PH = 18 Hz) and 130.5 ppm (d, J PH = 18 Hz). The magnitudes of the 31 P-1 H coupling constants are not consistent with the expected product Mes(TEMPO)P(AuCl)-C-(TEMPO)Ph 2 .Colorless crystals were obtained by slow diffusion of hexanes into the reaction mixture at À30 8C. Analysis of the crystals by X-ray crys...