Source: Dissertation Abstracts International, Volume: 57-11, Section: B, page: 6939.

Director: John Philip Caradonna.

The focus of the research presented here is to study the spectroscopic and reactivity properties of binuclear non-heme iron complexes as models for metalloenzymes (methane monooxygenase, ribonucleotide reductase, purple acid phosphatase, H chain ferritin and others) containing non-heme diiron cores in their active sites. Our approach has been to synthesize electronic models for these cores with special emphasis on systematically studying the reactivity properties that can be supported by these complexes.

The remarkable ability of the diferrous non-heme iron compound, $\rm \lbrack Fe\sb 2\sp{2+}(H\sb2Hbamb)\sb2(N$-MeIm)$\sb2$), 10, $\rm (H\sb4Hbamb,$ 2,3-bis(2-hydroxybenzamido)2,3-dimethylbutane), characterized as a good electronic model for MMO, to catalyze the OIPh oxidation of a variety of organic substrates (cyclohexane, adamantane, PhSMe, toluene, norbornane, decalin, hexamethyl ethane, cyclohexene, stilbenes) is reported along with the reactivity properties of its mixed-valence $\rm \lbrack Fe\sp{2+},Fe\sp{3+}$), 11, and diferric $\rm \lbrack Fe\sp{3+},Fe\sp{3+}\rbrack ,$ 12, states. While complexes 10 and 11 behave as good oxidation catalysts, the diferric complex 12 is inactive as a catalyst in all these reactions. Both 10 and 11 induce nearly complete $(\ge$90%) heterolytic cleavage of the peracyl O-O bond of phenylperacetic acid (PPAA), while 12 catalyzes the decomposition of PPAA following a homolytic pathway $(\ge$80%). Stereospecificity and stereoselectivity studies with diagnostic substrates suggest a discrete oxygen atom transfer in a non (oxygen based) free radical manner for 10 and 11. Parallels between the catalytic chemistry of non-heme 10 and 11 and that reported for cytochrome P-450 and methane monooxygenase suggest the involvement of higher valent metal-oxo species as the oxidizing intermediate in the catalytic cycles. The higher reactivity of these complexes in comparison to the analogous $\rm Fe\sp{2+}$ complex, $\rm \lbrack Fe\sb2(H\sb2Hbab)\sb2(N$-$\rm MeIm)\sb2\rbrack ,$ 1, indicates the effect of redox potential on catalysis.

Furthermore, the effect of d-electron count on catalysis is demonstrated by the studies on $\rm \lbrack Co\sp{m+},Co\sp{n+}\rbrack $ complexes, 7-9. The binuclear complex $\rm \lbrack Co\sb2\sp{2+}(H\sb2Hbab)\sb2(N$-MeIm)$\sb2\rbrack ,$ 7 (which is isostructural to the $\rm Fe\sp{2+}$ complex 1), and its oxidized analogues, 8 and 9, react with olefin substrates (norbornene, cyclooctene, styrene, cyclohexene) to yield mostly epoxidation products. The mechanistic implications of the significantly different product distributions obtained with the $\rm Co\sp{2+}$ (d$\sp7$) dimer, 7, versus the isostructural $\rm Fe\sp{2+}$ (d$\sp6$) dimer, 1, are presented. These data strongly infer the operation of two independent mechanisms for the $\rm Co\sp{2+}$ and $\rm Fe\sp{2+}$ catalyzed reactions. It is proposed that the cobalt complexes 7-9, incapable of forming higher valent metal-oxo species, behave as Lewis acid catalysts and epoxidize alkenes through a metal-OIPh adduct intermediate.