was predicted that this cavity would promote substrate binding and trap reactive intermediates
adjacent to the metal center, although a deeper, more well defined cavity is required to fully
realize these beneficial properties. This "micro-reactor" environment would thus improve
catalyst activity and might also influence selectivity. We also predicted that highly substituted
porphyrins such as our iron dodecaphenylporphyrin catalysts would have improved stability
relative to traditional planar porphyrin catalysts because there is considerable steric hindrance to
bifacial approach of two porphyrin molecules, thus inhibiting bimolecular catalyst destruction.
Electron withdrawing groups were substituted on the phenyl rings of iron dodecaphenyl-
porphyrin to create a series of catalysts, FeFRDPPCl where x =0, 20, 28, 36, with a range of
overall electron depletion at the metal center. This catalyst series is unique because the bulky
phenyl substituents create a nonplanar distortion leading to the formation of a cavity, as
discussed above. In addition, these catalysts maintain this same shape across the series, even
with the addition of fluorine substituents. In most other investigations of the effect of electron
withdrawing substituents on metalloporphyrin catalyst activity, the addition of electron
withdrawing groups to the porphyrin macrocycle has been accompanied by a change in the
degree of porphyrin nonplanarity. Our unique catalyst series allowed us to study the effect of
increased electron depletion of the metal center isolated from significant structural variation.
The FeFXDPPCl catalyst series was tested in the oxidation of isopentane by molecular
oxygen. For comparison, we also tested the commercial planar catalyst FeF20TPPCl. This
reaction was very selective for the production of alcohols. We observed the predicted trend --
catalytic activity increased with the degree of fluorination for the FeFDPPCl series. However,
the overall activity of the FeFDPP catalysts was much lower than that of the planar catalyst,
FeF20TPPCl, despite the built-in cavity of the DPP catalysts. Furthermore, we observed that the
porphyrins tested as oxidation catalysts degraded completely after several hours, even when the
reactor was charged with a large amount of catalyst. Such rapid catalyst deactivation is in
conflict with literature reports which indicate that metalloporphyrin catalysts in this type of
reaction are stable for much longer time periods, some even for days (Ellis, P.E., Jr., Lyons, J.E.,
Catal. Lett. 1989, 3, 389 and Lyons, J.E., Ellis, P.E., Jr., Catal. Lett. 1991, 8, 45). Our catalysts
were stable at the temperatures and pressures used in catalyst testing. Catalyst degradation was
only observed in the presence of a reactive alkane substrate. This indicates that some species
formed in the course of the oxidation reaction is responsible for the catalyst degradation. The
short life of our catalysts, a problem in itself, also prevented us from making adequate
comparisons for our designed catalysts. Although the amount of alcohol produced by
FeF20DPPCl was substantially less than the amount produced by FeF20TPPCl under identical
experimental conditions, we did not know if this was an indication that the DPP catalyst is less
active, less stable, or a combination of both. One would expect similar activities and stabilities
for the two catalysts based on the degree of electron depletion of the porphyrin by the
substituents. (Charge depletion can be estimated from the sum of the Hammett a constants of
the substituents, and FeF20DPPCl is predicted to be more active and stable or less active and
stable depending on whether meta or para constants are used, respectively.)
Conclusions which can be drawn from our previous testing with the FeFXDPP series are
as follows: The importance of adding electron withdrawing substituents to alter the redox
potential of the metal center was validated by the trend of increasing activity with increased
degree of fluorination observed for this structurally homologous series. The catalysts were less
stable than expected. However, if the reaction proceeds via a radical autooxidation mechanism,