iron(o) 6['.'] reveals an unusual trigonal bipyramidal type structure with the q '-bonded phosphaalkene ligand occupying an axial position. In view of its spatial requirements, this ligand should occupy an equatorial (bond angle 120") rather than an axial position (bond angle 90°) as found for the recently described carbonyliron complex with a phosphaalkene ligandI6l and for a tetracarbonyliron compound with the very bulky diphosphene ligand"].

Radical reactions involving two or more intermediates and many mutual reaction channels may lead to the specific formation of one cross-reaction product if one species is rather persistent and if transient and persistent species are produced with equal rates. A previous kinetic analysis of the phenomenon revealed that the concentrations of the intermediates and the selectivities of product formation depend critically on the relative initiation rates. The present experimental ESR studies on systems involving simultaneously generated phenoxyl radicals as persistent and alkyl radicals as transient species confirm the predicted behavior. They also lead to absolute rate constants for reactions of alkyl with phenoxyl radicals and for hydrogen atom transfer from phenols to alkyl radicals.

Rate constants of various simultaneous reactions of t-butoxicarbonyl and t-butyl radicals generated by photolysis of t-butylpivalate in n-heptane are directly determined by kinetic electron spin resonance. The temperature dependence of the decarboxylation reaction t-BuOCO + t-Bu + CO, obeys log(k/s-') = 13.8-49.0/8 where 8 = 2.303 RT/ kJ . mol-'. The self-and cross-termination of the radicals are diffusion limited.

Kinetic electron spin resonance spectroscopy with intermittent radical generation has been used to obtain the rate constants of various simultaneous reactions in systems containing 2,6-di-t-butylphenoxyl or 2,6di-t-butyl-4-methylphenoxyl radicals, the latter terminating reversibly. The radicals were generated in various solvents by photolysis of di-t-butyl peroxide in presence of the parent phenols. A data-analysis procedure is presented for reaction systems also involving species not detectable by e.s.r. The results show that self-termination rate constants of the phenoxyl radicals are up to two orders of magnitude lower than expected for diffusion control. This is usually attributed to steric hindrance due to the bulky tbutyl groups, but the activation energies found for most solvents contradict this interpretation. An explanation is presented which involves an intermediate in the termination reaction and an extra stabilization of the free radicals by dipole-induced-dipole interactions with the solvents, the latter being characterized by their molar polarization.

FeMo-Cofaktor aus Azobacter vinelandii e n t s p r i~h t~~. ~~ (vgl. auch [' I). Die Synthese von 1 ist somit wahrscheinlich ein wesentlicher Schritt in Richtung einer Verbindung, die in diesem Zusammenhang als Modellsubstanz bezeichnet werden kann. Arbeitsvorschr.@Alle Arbeiten werden unter Stickstoff in wasser-und sauerstofffreien uisungsmitteln ausgefiihrt. 0.06 g (1.07 mmol) Fe, 0.13 g (0.52 mmol) E 4 N I und 0.07 g (0.27 mmol) I2 werden in 60 mL CHICll 24 h bei Raumtemperatur geriihrt. Nach Zugabe von 1.0 g (0.52 mmol) (Et4N)2FebSbIb wird die Suspension mehrmals im geschlossenen Kolben (unter Druck) auf ca. 60°C erhitzt. Der entstehende amorphe Niederschlag wandelt sich bei Raumtemperatur innerhalb mehrerer Wochen zum Teil in kristallines 1 um. Die gut ausgebildeten Kristalle lassen sich in ca. 40proz. Ausbeute leicht von dem amorphen Rest mechanisch trennen. Die lange Kristallisationsdauer ist durch die geringe Loslichkeit von l in CH2C12 bedingt. In stirker polaren Lbsungsmitteln konnte 1 bisher jedoch nicht synthetisiert werden.