Shape Selective Alkane Hydroxylation by Metalloporphyrin Catalysts+

A series of manganese and iron porphyrins with sterically protected pockets are shown to be shape selective alkane hydroxylation catalysts. With iodosobenzene as oxidant, good regioselectivity is observed for hydroxylation of alkanes at the least hindered methyl group by using the very sterically hindered (5,10,15,20-tetrakis(2’,4’,6’-triphenylphenyl)porphyrinato)manganese(III) acetate (MnTTPPP(0Ac)) as catalyst; the moderately hindered (5, I O , ] 5,20-tetrakis(2’,4’,6’trimethoxyphenyl)porphyrinato)manganese(III) acetate shows little selectivity toward terminal CH, hydroxylation but does show enhancement for the adjacent, w 1, CH, site. Primary selectivity is dependent on the size and shape of the alkane substrate, with more bulky substituents giving greater primary selectivity. Substituting pentafluoroiodosobenzene or mchloroperbenzoic acid as oxidants yields similar selectivity, thus conclusively demonstrating metal based oxidation via a common intermediate for these three systems. In contrast, tert-butyl hydroperoxide or 2,2,2-trifluoroethanoI solubilized pentafluoroiodosobenzene show no primary carbon selectivity, and reaction product ratios are independent of the metalloporphyrin catalyst; this demonstrates that the site of oxidation with these oxidants is not metal based. The iron porphyrin derivatives also show good primary selectivity, although to a lesser degree than with the Mn derivatives, proving that these oxidations too are metal based. The regioselectivities for alkane hydroxylation shown by TTPPP derivatives are comparable to or better than those found for some isozymes of cytochrome P-450 which are responsible for primary alcohol biosynthesis from steroids, fatty acids, and alkanes. The unequalled ability of enzymes to exhibit molecular recognition and to catalyze regioselective reactions has led to diverse efforts to design synthetic systems with similar capabilities. Such regioselectivity often originates from discrimination based on the size and shape of substrate molecules, Le., shape selectivity. Two criteria apply: the first is the synthetic challenge in the design of a chemical species capable of molecular shape recognition; the second criterion requires that this synthetic host be capable of catalyzing a chemical transformation on the guest substrate while held in a specific orientation. Many guest-host systems, such as the cavitands,’ calixarenes,* crown ether^,^ lacunar macrocycles,* etc., have been synthesized, and some exhibit remarkable substrate recognition. None, however, are capable of catalyzing chemical reactions with the kind of regioselectivity typical of enzymatic reactions. In microporous extended solids, most notably zeolites, shape selective catalysis of hydrocarbon reforming has been successful and formed the basis for important industrial proces~es.~ Nonetheless, the creation of homogeneous catalysts for shape selective transformations has remained unrealized. Metalloporphyrins have been used as catalysts in a variety of oxidation reactions,6 including olefin epoxidation and alkane hydroxylation, in attempts to mimic the enzymatic activity of cytochrome P-450. MnTPP(X) (where TPP is 5,10,15,20tetraphenylporphyrinate(2-), and X can be CI-, Br-, I-, N), etc.) catalyzes the hydroxylation and halogenation of cyclohexane with iodosobenzene as oxidant.’q8 A high spin d2, Mn(V)-oxo complex has been suggested as one possible intermediate. Hill has isolated and characterized two other species formed during these reactions: both of which are capable of oxidizing alkanes in solution: (XMnrVTPP(OIPh)),O (X = CI-, Br-) and (N,MnwTPP)20. The product distributions of tert-butylbenzene and norcarnane hydroxylations, as determined by Hill7a and Groves,7b respectively, are consistent with hydrogen atom abstraction and recombination, involving a radical, rather than a carbocation intermediate. We9 and others’O have used metalloporphyrins with steric protection of the porphyrin faces as catalysts for the hydroxylation of cyclic and straight chain alkanes. In all cases involving modestly hindered porphyrin^^-'^ (such as (5,10,15,20-tetrakis(2’,4’,6’-trimethoxyphenyl)porphyrinato)manganese(III) ace ta te , MnTTMPP(OAc), as shown in Figure l ) , only very modest shape selectivity has proved possible: slight regioselectivity for secondary carbon vs. secondary carbon hydroxylation was observed, and no ‘Taken in part from the Ph.D. Dissertation of B.R.C. significant production of primary alcohols was seen. We report here that the uery sterically crowded manganese(II1) and iron(II1) complexes of 5,10,15,20-tetrakis(2’,4’,6’-triphenylphenyl)porphyrinate (the bis-pocket porphyrin, TTPPP, as in Figure 1 ) show good regioselectivity for primary hydroxylation of n-alkanes and remarkable shape selective hydroxylation of branched alkanes. This is reminiscent of some isozymes of cytochrome P-450 and of the nonheme-iron w-hydroxylases, which are responsible for primary alcohol synthesis from the terminal hydroxylation of alkyl