Electrochemical and electronic absorption spectroscopic studies of substituent effects in iron(IV) and manganese(IV) corroles. Do the compounds feature high-valent metal centers or noninnocent corrole ligands? Implications for peroxidase compound I and II intermediates

Article abstract of DOI:10.1021/jp012037r

We report here an electrochemical and optical spectroscopic study of new Fe(IV) and Mn(IV) meso-triarylcorrole complexes. The complexes studied are three Fe(IV)Cl, three Mn(IV)Cl, and three dimeric Fe(IV)-OFe(IV) meso-tris(p-X-phenyl)corrole complexes, where X = CH₃, H, and CF₃. The first oxidation potentials of the Fe(IV)Cl and Mn(IV)Cl corrole complexes are considerably higher than those of the corresponding Fe(IV) corrole ??-oxo dimers, suggesting that the corrole ligands in the chloride complexes are already oxidized to a radical-like state. This is consistent with the suggestion by Walker and co-workers (ref 12) that the iron center in an (octaalkylcorrolato)Fe^IVCl complex is best described as intermediate spin (S = 3/2) and that it is antiferromagnetically coupled to a corrole ??-radical. We have attempted to clarify the nature of this antiferromagnetic coupling by means of DFT calculations and propose that it results from an metal(d_z2)-corrole( b₁ ) orbital interaction. In contrast, the corrole ligand in the Fe(IV) corrole ??-oxo dimers does not seem to have radical character. The optical spectra of the Fe(IV)Cl and Mn(IV)Cl corrole derivatives exhibit distinctive split Soret bands, one arm of which is strongly substituent sensitive. This behavior contrasts with that of free-base corroles and porphyrins and of typical metalloporphyrins whose optical spectra are relatively substituent-insensitive. We qualitatively assign this substituent-sensitive feature to a transition with significant ligand-to-metal charge-transfer character.

Full text of DOI:10.1021/jp012037r

11406
J. Phys. Chem. B 2001, 105, 11406-11413
ARTICLES
Electrochemical and Electronic Absorption Spectroscopic Studies of Substituent Effects in
Iron(IV) and Manganese(IV) Corroles. Do the Compounds Feature High-Valent Metal
Centers or Noninnocent Corrole Ligands? Implications for Peroxidase Compound I and II
Intermediates
Erik Steene, Tebikie Wondimagegn, and Abhik Ghosh*
Institute of Chemistry, Faculty of Science, UniVersity of Tromsø, N-9037 Tromsø, Norway
ReceiVed: May 30, 2001; In Final Form: September 13, 2001
We report here an electrochemical and optical spectroscopic study of new Fe(IV) and Mn(IV) meso-
triarylcorrole complexes. The complexes studied are three Fe(IV)Cl, three Mn(IV)Cl, and three dimeric Fe(IV)-
OFe(IV) meso-tris(p-X-phenyl)corrole complexes, where X ) CH3, H, and CF3. The first oxidation potentials
of the Fe(IV)Cl and Mn(IV)Cl corrole complexes are considerably higher than those of the corresponding
Fe(IV) corrole µ-oxo dimers, suggesting that the corrole ligands in the chloride complexes are already oxidized
to a radical-like state. This is consistent with the suggestion by Walker and co-workers (ref 12) that the iron
center in an (octaalkylcorrolato)FeIVCl complex is best described as intermediate spin (S ) 3/2) and that it
is antiferromagnetically coupled to a corrole π-radical. We have attempted to clarify the nature of this
antiferromagnetic coupling by means of DFT calculations and propose that it results from an metal(dz2)-
corrole(“b1”) orbital interaction. In contrast, the corrole ligand in the Fe(IV) corrole µ-oxo dimers does not
seem to have radical character. The optical spectra of the Fe(IV)Cl and Mn(IV)Cl corrole derivatives exhibit
distinctive split Soret bands, one arm of which is strongly substituent sensitive. This behavior contrasts with
that of free-base corroles and porphyrins and of typical metalloporphyrins whose optical spectra are relatively
substituent-insensitive. We qualitatively assign this substituent-sensitive feature to a transition with significant
ligand-to-metal charge-transfer character.
Introduction
on the meso phenyl groups (Figure 1). In addition, we report
preliminary DFT calculations on unsubstituted model com-
pounds (corrolato)FeIVX and (corrolato)MnIVX (X ) Cl, OLi).
The combined electrochemical and theoretical results afford
significant insights into the nature of the formally high-valent
iron and manganese centers, the issue of ligand noninnocence,
and substituent effects on the electronic character of these high-
valent complexes.6
Iron(IV) complexes have been detected or proposed as
reactive intermediates for various heme and nonheme iron
enzymes, as well as for a number of synthetic biomimetic model
systems.1 A wide variety of spectroscopic techniques have been
used to probe the geometric and electronic structures of
enzymatic and synthetic iron(IV) intermediates. A key point of
interest in these studies has involved the question of where the
oxidation equivalents reside in these high-valent molecules, i.e.,
whether these species are oxidized in a metal- or ligand-centered
fashion, relative to analogous lower-valent complexes. Recent
theoretical studies2 have contributed to a greatly enhanced
understanding of the factors that result in this rich electronic-
structural diversity among compound I intermediates. Develop-
ments in corrole chemistry suggest that a similar diversity of
electronic structures might emerge for iron3,4 and manganese4,5
corroles. From a practical standpoint, high-valent metallocor-
roles3,4 are also of particular interest because of their relative
stability and ease of handling and characterization.
Methods. Free-base meso-triarylcorroles with the para-
substituents CH3, H, and CF3 were synthesized from pyrrole
and the relevant para-substituted benzaldehydes by a straight-
forward application of the procedure reported by Gross and co-
workers for tris(pentafluorophenyl)corrole, H3TPFPC.7 Although
the procedure described in ref 7 specifically refers to electron-
deficient aromatic aldehydes, we found that it worked for a
broad range of aldehydes. The free-base corroles were exhaus-
tively characterized by spectroscopic methods and elemental
analysis.8 Iron was inserted into these ligands via the interaction
of the free-base corroles with Fe2(CO)9 in refluxing toluene,
exactly as described by Vogel et.al. for octaethylcorrole.3 After
aerial oxidation, this reaction gave the Fe(IV) µ-oxo dimer
derivatives of the corroles. Monomeric Fe(IV) chloride com-
plexes were prepared from the µ-oxo dimers by treatment with
dilute aqueous HCl, as described by Vogel et al.3 Manganese
was inserted into the corroles by reaction with Mn(OAc)2•4H2O
We report here the preparation, optical spectra, and an
electrochemical study of new iron(IV) and manganese(IV) meso-
triarylcorroles with different para-substituents (CH3, H, and CF3)
* To whom correspondence should be addressed. E-mail: abhik@
chem.uit.no. Fax: +47 7764 4072.
10.1021/jp012037r CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/25/2001

Study of Iron(IV) and Manganese(IV) Corroles
J. Phys. Chem. B, Vol. 105, No. 46, 2001 11407
TABLE 1: Half-Wave Potentials (V vs SCE) for
[FeIVT(p-X-P)C]Cl, [FeIVT(p-X-P)C]2O, and
[MnIVT(p-X-P)C]Cl (X ) CH3, H, CF3) in CH2Cl2
Containing 0.1 M TBAPa
corrole
X
3σ
E1/2ox
E1/2red
FeIV[T(p-X-P)C]Cl
CH3
H
CF3
CH3
H
CF3
CH3
H
-0.51
0
1.62
-0.51
0
1.62
-0.51
0
1.015
1.068
1.178
0.593
0.635
0.830
0.970
1.032
1.150
0.033
0.050
0.190
-0.350
-0.305
-0.115
0.072
[FeIVT(p-X-P)C]2O
MnIV[T(p-X-P)C]Cl
0.093
0.230
CF3
1.62
a A scan rate of 0.1 Vs-1 was used. See also Figures 2, 3, and 4 for
additional redox potential data.
Figure 1. Compounds studied in this work; X ) CH3, H, and CF3.
rolic), 7.09 (d, J ) 6.9 Hz, 2H, phenyl), 7.03 (d, J ) 4.5 Hz,
4H, phenyl and/or â-pyrrolic), 6.84 (d, J ) 4.5 Hz, 4H, phenyl
and/or â-pyrrolic), 2.51 (s, 12H, p-CH3), 2.46 (s, 6H, p-CH3).
λmax (log ꢀ/(M-1cm-1)): 389 nm (4.94).
in refluxing DMF. Elemental analyses and MALDI-TOF mass
spectra of all compounds prepared were as expected on the basis
of the molecular formulas. Proton NMR (CDCl3, room tem-
perature) data for the iron compounds, along with proposed
assignments, and UV-vis absorption spectroscopic data (in
dichloromethane) are as follows. The NMR spectra and a
discussion of the proposed assignments are given in the
Supporting Information. The Mn(IV) Cl corrole complexes gave
rise to broad, paramagnetically shifted peaks in the proton NMR
spectra, which have not yet been assigned.
(T(p-CF3-P)C)MnIVCl. λmax (log ꢀ/(M-1cm-1)): 316 (4.32),
364 (4.40), 423 nm (4.62).
(TPC)MnIVCl. λmax (log ꢀ/(M-1cm-1)): 314 (4.37), 358
(4.44), 433 nm (4.66).
(T(p-CH3-P)C)MnIVCl. λmax (log ꢀ/(M-1cm-1)): 319 (4.40),
363 (4.47), 443 nm (4.64).
Cyclic voltammetry was carried out using an EG&G Model
263A Potensiostat with a three-electrode system consisting of
a glassy carbon working electrode, a platinum wire counter-
electrode and a saturated calomel reference electrode (SCE).
Tetra(n-butyl)ammonium perchlorate (TBAP) recrystallized
from ethanol and dried in a vacuum for at least one week was
used as supporting electrolyte. Dichloromethane, distilled and
kept over molecular sieves, was used as solvent. The reference
electrode was separated from bulk solution by a fritted-glass
bridge filled with the solvent/supporting electrolyte mixture, and
all potentials were referenced to the SCE. Pure nitrogen was
bubbled though solutions containing the metallocorroles for at
least 2 min prior to running the experiments and the solutions
were also protected from air by a nitrogen blanket during the
experiment.
DFT calculations on (corrolato)MX complexes (M ) Fe, Mn;
X ) Cl, OLi) were carried out using Slater-type valence triple-ú
plus polarization basis sets, the VWN local functional, the
Perdew-Wang 1991 gradient corrections, a spin-unrestricted
formalism, a fine mesh for numerical integration of matrix
elements, full geometry optimizations with Cs symmetry
constraints, and the ADF10 program system.
Electrochemical Results. Table 1 presents the oxidation and
reduction potential data for the iron and manganese corroles
prepared. The corresponding cyclic voltammograms are shown
in Figure 2, 3, and 4. Figure 5, 6, and 7 shows Hammett plots
for the redox potentials of the Fe(IV) corrole chloride, the Fe(IV)
corrole µ-oxo dimer, and the Mn(IV) corrole chloride series,
respectively.
The oxidation potentials of the Fe(IV) corrole chloride
complexes are considerably higher than those of the corre-
sponding Fe(IV) corrole µ-oxo dimers. This suggests that the
nature of the one-electron oxidation is different in the two
families of compounds. The oxidation potentials of the Fe(IV)
corrole µ-oxo dimers are comparable to those observed for the
corresponding Cu(III) corroles (e.g., E1/2ox (CuIIITPC) ) 0.76
V vs SCE).11 Because the electrochemical generation of Cu(IV)
and Fe(V) corroles is unlikely, we propose that the Fe(IV)
1
(T(p-CF3-P)C)FeIVCl. H NMR: 24.0 (s, 2H, â-pyrrolic),
22.8 (s, 2H, â-pyrrolic), 22.0 (s, 2H, â-pyrrolic), 5.55 (s, 2 H,
â-pyrrolic or o-phenyl), -1.75 (s, 2H, m-phenyl), -1.95 (s, 2H
m-phenyl), -2.77 (s, 1H m-phenyl), -2.87 (s, 1H m-phenyl),
-6.29 (s, 2H, o-phenyl), -6.78 (s, 2H, o-phenyl), -39.8 (s, 2
H, â-pyrrolic or o-phenyl). λmax (log ꢀ/(M-1cm-1)): 366 nm
(4.60), 402 nm (4.72).
1
(TPC)FeIVCl. H NMR: 25.2 (s, 2H, â-pyrrolic), 24.0 (s,
2H, â-pyrrolic), 23.1 (s, 2H, â-pyrrolic), 19.6 (s, 2H, p-phenyl),
17.2 (s, 1H, p-phenyl), 6.01 (s, 2H, â-pyrrolic or o-phenyl),
-2.56 (s, 2H, m-phenyl), -2.85 (s, 2H, m-phenyl), -3.71 (s,
1H, m-phenyl), -3.91 (s, 1H, m-penyl), -5.79 (s, 2H, o-phenyl),
-6.94 (s, 2H, o-phenyl), -41.1 (s, 2H, â-pyrrolic or o-phenyl?).
λmax (log ꢀ/(M-1cm-1)): 359 nm (4.53), 411 nm (4.63).
1
(T(p-CH3-P)C)FeIVCl. H NMR. 26.0 (s, 2H, â-pyrrolic),
24.7 (s, 2H, â-pyrrolic), 23.9 (s, 2H, â-pyrrolic), 5.75 (s, 2H,
â-pyrrolic or o-phenyl), -3.23 (s, 2H, m-phenyl), -3.38 (s, 2H,
m-phenyl), -4.51 (s, 1H, m-phenyl), -4.67 (s, 1H, m-phenyl),
-5.41 (s, 2H, o-phenyl), -7.25 (s, 2H, o-phenyl), -9.61 (s,
3H, p-CH3), -11.9 (s, 6H, p-CH3), -41.4 (s, 2H, â-pyrrolic or
o-phenyl). λmax (log ꢀ/(M-1cm-1)): 361 nm (4.60), 421 nm
(4.61).
[(T(p-CF3-P)C)FeIV]2O. 1H NMR. 7.71-7.77 (m, 10H,
phenyl), 7.64 (d, J ) 6.7 Hz, 4H, phenyl), 7.59 (d, J ) 7.3 Hz,
2H, phenyl), 7.52 (d, J ) 7.3, 6H,9 phenyl), 7.35 (d, J ) 4.3
Hz, 4H, â-pyrrolic), 7.24 (d, J ) 4.9 Hz, 4H, â-pyrrolic), 7.11
(d, J ) 7.9 Hz, 2H, phenyl), 6.94 (d, J ) 4.3 Hz, 4H, â-pyrrolic),
6.76 (d, J ) 4.9 Hz, 4H, â-pyrrolic). λmax (log ꢀ/(M-1cm-1)):
384 nm (5.03).
1
[(TPC)FeIV]2O. H NMR. 7.61 (s, 4H, phenyl), 7.37-7.54
(m, 22H, phenyl and â-pyrrolic), 7.29 (m, 8H, phenyl and/or
â-pyrrolic), 7.25 (d, J ) 7.3 Hz, 2H, phenyl), 7.11 (d, J ) 7.3
Hz, 2H, phenyl), 7.00 (d, J ) 4.9 Hz, 4H, â-pyrrolic), 6.83 (d,
J ) 4.9 Hz, 4H, â-pyrrolic). λmax (log ꢀ/(M-1cm-1)): 385 nm
(5.09).
1
[(T(p-CH3-P)C)FeIV]2O. H NMR. 7.29-7.49 (m, 26H,
phenyl and/or â-pyrrolic), 7.21 (m, 4H, phenyl and/or â-pyr-

11408 J. Phys. Chem. B, Vol. 105, No. 46, 2001
Steene et al.
Figure 2. Cyclic voltammograms of FeIV[T(p-X-P)C]Cl (X ) CH3,
H, CF3) in CH2Cl2 with TBAP as the supporting electrolyte. The various
redox potentials are indicated.
Figure 3. Cyclic voltammograms of [FeIVT(p-X-P)C]2O (X ) CH3,
H, CF3) in CH2Cl2 with TBAP as the supporting electrolyte. The various
redox potentials are indicated.
corrole µ-oxo dimers undergo ligand-centered oxidation. The
significantly higher oxidation potentials for the Fe(IV) corrole
chloride complexes suggests that the corrole ligand in these
complexes is already oxidized and that the electrochemical one-
electron oxidation is at least significantly metal-centered. This
is consistent with the proposal by Walker et.al. that FeIV(OMC)-
Cl (OMC ) â-octamethylcorrole) is better described as inter-
mediate-spin Fe(III) antiferromagnetically coupled to an OMC
“cation” radical.12 However, based on spectroelectrochemical
measurements, Gross and co-workers have suggested that one-
electron oxidation and reduction of FeIV(TPFPC)Cl are corrole-
and metal-centered, respectively.4c Thus, Gross and co-workers
have favored a clean Fe(IV) description for FeIV(TPFPC)Cl.
The electrochemical results show that substituents at the para-
positions of the meso-phenyl groups exert significant effects
on the oxidation and reduction potentials. The observed F-values
(F ) 1/3 dE1/2/dσ) are 74 mV for the first oxidation of the Fe(IV)
corrole chloride series, 77 mV for the first reduction of this
series, 114 mV for the first oxidation of the Fe(IV) corrole µ-oxo
dimer series, and 109 mV for the first reduction of this series.
These F-values are comparable to those found for corresponding
metallotetraphenylporphyrins13 where (F ) 1/4 dE1/2/dσ). The
relatively high F-values for oxidation, comparable to the ones
found for metallotetraphenylporphyrins, are consistent with an
oxidation from the a2u-like b1 HOMO that has considerable
electron density at the meso-positions (see Figure 8).14
The F-values for oxidation and reduction in the Fe(IV) corrole
µ-oxo dimer series are significantly higher than for the corre-
sponding Fe(IV) corrole chloride series. This suggests that the
oxidation and reduction equivalents in the former series of
compounds are distributed in such a manner that they sense
the substituent effects of more than three meso-aryl substituents.
DFT Calculations. Spin-unrestricted DFT calculations on
(corrolato)FeIVCl and (corrolato)MnIVCl reveal rather complex
spin density profiles (Figure 9). Thus, for (corrolato)FeIVCl, the
Fe and Cl atoms carry 2.00 and 0.22 R spins and the â-carbons
carry a total of 0.10 R spins. In contrast, the meso carbons carry
a total of 0.28 â spins and the nitrogens carry 0.14 â spins. For
(corrolato)MnIVCl, the Mn and Cl atoms carry 3.22 and 0.14 R
spins and the â-carbons carry a total of 0.087 R spins. In
contrast, the meso carbons carry a total of 0.30 â spins and the
nitrogens carry 0.29 â spins. This spatial distribution of the â
spins in these two M(IV)Cl complexes coincides precisely with
the shape of the b1 HOMO of the corrole ligand, which crudely
resembles the a2u HOMO of porphyrins. The overall spin density
profiles of these molecules, in particular, the significant spatial
separation of the R and â spins, are highly characteristic of
antiferromagnetically coupled spin systems.6d In this case, the

Study of Iron(IV) and Manganese(IV) Corroles
J. Phys. Chem. B, Vol. 105, No. 46, 2001 11409
Figure 5. Hammett plots for the first oxidation and reduction of the
FeIV[T(p-X-P)C]Cl series (X ) CH3, H, CF3).
Figure 4. Cyclic voltammograms of MnIV[T(p-X-P)C]Cl (X ) CH3,
H, CF3) in CH2Cl2 with TBAP as the supporting electrolyte. The various
redox potentials are indicated.
antiferromagnetic coupling seems to involve the metal center
on the one hand and the corrole b1 HOMO on the other hand.
An examination of the high-lying occupied MOs of each of these
molecules revealed an MO with an metal(dz2)-corrole(“b1”)
overlap. This overlap is clearly facilitated by the significant
displacement of 0.4-0.5 Å of the metal atom from the N4 plane
of the corrole in the optimized geometry of (corrolato)MIVCl,
which is in good agreement with a displacement of 0.422 Å in
the X-ray crystal structure of (OEC)FeIVCl.3 The Fe(dz2)-corrole-
(“b1”) overlap revealed by our calculations is consistent with
the suggestion by Walker and co-workers that the iron center
is best described as intermediate spin (S ) 3/2) and that it is
12
Figure 6. Hammett plots for the first oxidation and reduction of the
[FeIVT(p-X-P)C]2O series (X ) CH3, H, CF3).
antiferromagnetically coupled to a corrole π-radical. In the
same way, we propose that the electronic structures of (corro-
lato)MnIVCl has some significant S ) 2 Mn(III) character, with
antiferromagnetic coupling to a corrole b1 radical.
An analogous electronic-structural scenario has been proposed
for five-coordinate high-spin iron(III) porphyrin π-cation radicals
such as [(OEP)FeCl]+ and [(OEP)FeBr]+.15 Structural, Mo¨ss-
bauer,15 and theoretical6d studies on these and related compounds
indicate that the dz2 electron of the high-spin iron(III) center is
antiferromagnetically coupled to a porphyrin a2u radical, leading
to an overall S ) 2 spin state.
Obviously, the spatial separation of the R and â spins does
not correspond to any real spin density profile but is an artifact
of the spin-unrestricted DFT method. A proper description of
antiferromagnetic coupling is better achieved via traditional ab
initio methods such as multiconfigurational SCF calculations,
but we have not yet accomplished such calculations on the large
molecules of interest in this study, although a modest beginning
has been made.6d

11410 J. Phys. Chem. B, Vol. 105, No. 46, 2001
Steene et al.
which is strongly substituent sensitive, as shown in Table 2.
This strong sensitivity may be contrasted with the relative
invariance of the absorption spectra of free-base triarylcorroles
as well as of free-base and metallotetraarylporphyrins as a
function of substituents in the para-position of the phenyl
rings.13 We propose that the substituent sensitive band in the
absorption spectra of the Fe(IV) and Mn(IV) corrole chloride
complexes have significant ligand-to-metal charge-transfer
character. Interestingly, the optical spectra of the Fe(IV) corrole
µ-oxo dimers studied do not exhibit similarly large substituent
effects. The reason for this difference in behavior of the Fe(IV)
corrole chlorides and the Fe(IV) corrole µ-oxo dimers is not
clear at present.
Note also that the optical spectra of the Fe(IV) and Mn(IV)
corrole chloride complexes are considerably red-shifted and
exhibit “hyperporphyrin” character, relative to those of the
Fe(IV) corrole µ-oxo dimers. This is somewhat of a novelty
compared to Fe(IV)-oxo porphyrins such as peroxidase com-
pound II intermediates and their synthetic models. However,
chloroperoxidase compound II (CPO-II) also exhibits a hyper
spectrum, with a split Soret band with maxima at 373 and 436
nm; in addition, according to Egawa et al., “Resonance
enhancement of the FeIVdO stretching (νFeO) Raman band was
found for CPO-I16c when Raman scattering was excited at
wavelengths within both transition bands around 365 and 415
nm, whereas the νFeO Raman band was not identified for CPO-
II at any of the excitation wavelengths examined”.17 On the
basis of our own results on FeIVCl and FeIV-oxo corroles and
those of Egawa et al.,16 it may even be resonable to totally
reconsider the molecular and electronic structure of CPO-II:
It is possible that CPO-II may not be a high-valent iron-oxo
intermediate! Instead, the putative oxo group in CPO-II may
actually be protonated, i.e., an OH group. An OH group, like
the axial Cl ligand in our FeIVCl corroles, would be less able to
stabilize a high-valent iron center than a terminal oxo ligand,
which would also be consistent with the hyperporphyrin
character of CPO-II.16
Resonance Raman Spectra (note added in proof): Since
the initial submission of this paper, we have succeeded in
obtaining the resonance Raman (RR) spectra of the Mn(IV) and
Fe(IV) corroles reported in this paper (Figure 13) using a Spex
Jobin-Yvon double monochromator equipped with a liquid
nitrogen cooled Spex Jobin-Yvon CCD 2000 detector and the
457.9 nm laser line of a Coherent Innova 70 Series mixed Ar-
Kr laser. We had hoped that the RR spectra would provide
valuable insights into differences in the electronic character of
the corrole ligand among the different compounds. Unfortu-
Figure 7. Hammett plots for the first oxidation and reduction of the
MnIV[T(p-X-P)C]Cl series (X ) CH3, H, CF3).
In contrast to the rather complex spin density profile of
(corrolato)FeIVCl, a DFT calculation on (corrolato)FeIVOLi
shows that the unpaired spin density is entirely localized on
the Fe-O unit, with the atomic spin populations being 1.2 for
Fe and 0.82 for O. This result shows that with a strongly basic
axial ligand such as oxide, the corrole ligand does not have
radical character. This parallels a similar lack of unpaired spin
density on the porphyrin ring of peroxidase compound II and
synthetic Fe(IV)-oxo model systems. These calculations provide
an explanation for the considerably lower oxidation potentials
of the Fe(IV) corrole µ-oxo dimers relative to the corresponding
Fe(IV) corrole chloride complexes. These calculated electronic
structural descriptions are also consistent with the significantly
lower Mo¨ssbauer isomer shift for [(OEC)Fe]2O (0.02 mm/s)
relative to (OEC)FeCl (0.19 mm/s).3
Optical Spectra. The optical spectra of the Fe(IV) and
Mn(IV) corrole chloride derivatives (Figures 10 and 12 respec-
tively) exhibit highly distinctive split Soret bands, one arm of
Figure 8. The b1 and a2 HOMOs of Ga(III) corrole, chosen as a representative closed-shell metallocorrole.

Study of Iron(IV) and Manganese(IV) Corroles
J. Phys. Chem. B, Vol. 105, No. 46, 2001 11411
Figure 12. UV-vis absorption spectra of MnIV[T(p-X-P)C]Cl.
TABLE 2: “Soret” Band Maxima (nm) in CH2Cl2 for the
Fe(IV) and Mn(IV) Triarylcorroles Studied
para-substituent (X)
corrole
CH3
H
CF3
FeIV[T(p-X-P)C]Cl
361, 421
389
359, 411
385
366, 402
384
[FeIVT(p-X-P)C]2O
MnIV[T(p-X-P)C]Cl 319, 363, 443 314, 358, 433 316, 364, 423
a The highly substituent-sensitive bands are marked in bold.
Figure 9. Calculated gross atomic spin populations obtained from spin-
unrestricted DFT calculations.
Further, certain peaks in the Fe(IV)Cl corrole complexes appear
to be modestly upshifted in the corresponding Fe(IV)-O-Fe-
(IV) derivatives. For instance, the bands at 1533 and 1494 cm-1
for FeIV[T(p-CF3-P)C]Cl appear to upshift to 1537 and 1499
cm-1, respectively, for {FeIV[T(p-CF3-P)C]}2O. Similarly, the
bands at 1527, 1485, and 1499 cm-1 for FeIV[TPC]Cl appear
to upshift to 1534, 1489, and 1503 cm-1, respectively, for {FeIV-
[TPC]}2O, and the bands at 1529 and 1489 cm-1 for FeIV[T(p-
CH3-P)C]Cl appear to upshift to 1532 and 1495 cm-1, respec-
tively, for {FeIV[T(p-CH3-P)C]}2O. These upshifts may be
indicative of a systematic difference in electronic character of
the corrole ligand between the Fe(IV)Cl versus Fe(IV)-O-
Fe(IV) derivatives, as discussed elsewhere in the paper.
Ongoing experimental work in our laboratory include tem-
perature-dependent NMR studies of the paramagnetic Fe(IV)
corrole chloride complexes as well as Mo¨ssbauer spectroscopy
of all the iron(IV) corroles reported herein.
Figure 10. UV-vis absorption spectra of FeIV[T(p-X-P)C]Cl.
Conclusions
The following are some of the key conclusions of this study.
1.The observed variations in redox potentials among the
different Fe(IV) and Mn(IV) complexes studied indicate that
meso-aryl substituents can strongly modulate the electronic
nature of metallocorrole complexes.
2.The first oxidation potentials of the Fe(IV) corrole chloride
complexes are considerably higher than those of the corre-
sponding Fe(IV) corrole µ-oxo dimers, suggesting that the
corrole ligands in the former complexes are already oxidized
to a radical-like state. This is consistent with the suggestion by
Walker and co-workers that the iron center is best described as
intermediate spin (S ) 3/2) and that it is antiferromagnetically
coupled to a corrole π-radical. DFT calculations have clarified
the nature of this antiferromagnetic coupling and attribute it to
an Fe(dz2)-corrole(“b1”) orbital interaction. In contrast, the
corrole ligand in the Fe(IV) corrole µ-oxo dimers does not seem
to have radical character. The RR spectra, although not analyzed
in detail, may also be consistent with the conclusions.
3.The first oxidation potentials of the Mn(IV) corrole chloride
complexes are approximately the same as those for the
analogous Fe(IV)Cl complexes.
Figure 11. UV-vis absorption spectra of FeIV[T(p-X-P)C]2O.
nately, the spectra do not exhibit many readily interpretable
trends. Nevertheless, a few preliminary observations may be
made, albeit at a rather descriptive level. For example, the band
at 1533 cm-1 in the RR spectrum of FeIV[T(p-CF3-P)C]Cl
downshifts to 1527 and 1529 cm-1 for FeIV[T(p-CH3-P)C]Cl
and FeIV[TPC]Cl, respectively. Similarly, the band at 1528 cm-1
for MnIV[T(p-CF3-P)C]Cl downshifts to 1517 and 1521 cm-1
for MnIV[T(p-CH3-P)C]Cl and MnIV[TPC]Cl, respectively.
These results may indicate that the electronic character of the
corrole ligand in the MIV[T(p-CF3-P)C]Cl (M ) Fe, Mn)
derivatives may be significantly different, perhaps with respect
to radical character, relative to the analogous derivatives
involving the more electron-rich TPC and T(p-CH3-P)C ligands.

11412 J. Phys. Chem. B, Vol. 105, No. 46, 2001
Steene et al.
Figure 13. Resonance Raman spectra (λex ) 457.9 nm) of the Fe(IV) and Mn(IV) corroles studied.
4.The optical spectra of the Fe(IV) and Mn(IV) corrole
chloride derivatives exhibit distinctive split Soret bands, one
arm of which is strongly substituent sensitive. This behavior
contrasts with that of free-base corroles and porphyrins and of

Study of Iron(IV) and Manganese(IV) Corroles
J. Phys. Chem. B, Vol. 105, No. 46, 2001 11413
Z. Chem. Comm. 2000, 1957. (c) Golubkov, G.; Bendix, J.; Gray, H. B.;
Mahammed, A.; Goldberg, I.; DiBilio, A. J.; Gross, Z. Angew. Chem. Int.
Ed. Engl. 2001, 40, 2132.
(6) (a) Vangberg, T.; Ghosh, A. J. Am. Chem. Soc. 1999, 121, 12 154.
(b) Ghosh, A.; Wondimagegn, T.; Gonzalez, E.; Halvorsen, I. J. Inorg.
Biochem. 2000, 78, 79. (c) Ghosh, A.; Gonzalez, E. Israel J. Chem. 2000,
40, 1. (d) Ghosh, A.; Gonzalez, E.; Vangberg, T.; Taylor, P. J. Porphyrins
Phthalocyanines 2001, 5, 345.
(7) Gross, Z.; Galili, N.; Simkhovich, L.; Saltsman, I.; Botoshansky,
M.; Bla¨ser, D.; Boese, R.; Goldberg, I. Org. Lett 1999, 1, 599.
(8) Halvorsen, I.; Wondimagegn, T.; Ghosh, A. J. Am. Chem. Soc.,
Submitted for publications.
typical metalloporphyrins. We qualitatively assign this substitu-
ent sensitive transition to a ligand-to-metal charge transfer.
5.The optical spectra of the Fe(IV) and Mn(IV) corrole
chloride complexes are considerably red-shifted and exhibit
“hyperporphyrin” character, relative to those of the Fe(IV)
corrole µ-oxo dimers. This is somewhat of a novelty compared
to Fe(IV)-oxo porphyrins such as peroxidase compound II
analogues and their synthetic models, but is a feature shared
17
by chloroperoxidase compound II.
On the basis of these
developments, an interesting possibility to consider is that CPO-
II may not be a high-valent oxo intermediate at all but rather
an iron(IV)-hydroxy species.
(9) Two doublets on top of each other, based on COSY.
(10) The ADF program system was obtained from Scientific Computing
and Modeling, Department of Theoretical Chemistry, Vrije Universiteit,
1081 HV Amsterdam, The Netherlands. For details of basis sets, grids for
numerical integration, etc., the reader is referred to the program manual
obtainable from this source.
Acknowledgment. We acknowledge financial support from
the Norwegian Research Council and the VISTA program of
Statoil, Norway. A.G. acknowledges a Senior Fellowship from
the San Diego Supercomputer Center.
(11) Halvorsen, I.; Ghosh, A., Unpublished results.
(12) Cai, S.; Walker, F. A.; Licoccia, S. Inorg. Chem. 2000, 39, 3466.
(13) Halvorsen, I.; Steene, E.; Nilsen, H. J.; Kadish, K. M.; Ghosh, A.,
Unpublished results.
(14) (a) Ghosh, A.; Wondimagegn, T.; Parusel, A. B. J. J. Am. Chem.
Soc. 2000, 122, 5100. (b) Bendix, J.; Dmochowski, I. J.; Gray, H. B.;
Mahammed, A.; Simkhovich, L.; Gross, Z. Angew. Chem. Ind. Ed. 2000,
39, 4048.
(15) (a) Gans, P.; Buisson, G.; Due´e, E.; Machon, J.-C.; Erler, B. S.;
Scholz, W. F.; Reed, C. A. J. Am. Chem. Soc. 1986, 108, 1223. (b) Erler,
B. S.; Scholz, W. F.; Lee, Y. J., Scheidt, W. R.; Reed, C. A. J. Am. Chem.
Soc. 1987, 109, 2644. (c) Schulz, C. E.; Song, H.; Mislankar, A.; Orosz, R.
D.; Reed, C. A. Inorg. Chem. 1997, 36, 406.
(16) (a) Dawson, J. H.; Sono, M.; Eble, K. S.; Hager, L. P. ReV. Port.
Quim. 1985, 27, 205. (b) Nakajima, R.; Yamazaki, I.; Griffith, B. W.
Biochem. Biophys. Res. Commun. 1985, 128, 1. (c) Egawa, T.; Proshlyakov,
D. A.; Miki, H.; Makino, R.; Ogura, T.; Kitagawa, T.; Ishimura, Y. J. Biol.
Inorg. Chem. 2001, 6, 46-54.
(17) The location of the radical in CPO-I is a controversial issue. Thus,
Terner et al. provided resonance Raman evidence that CPO-I may be best
described as an Aiu-type porphyrin π-cation radical (Hosten, C. M.; Sullivan,
a. M.; Palaniappan, V.; Fitzgerald, M. M.; Terner, J. J. Biol. Chem. 1994,
269, 13966). Green’s2b recent DFT calculations on CPO-I revealed a radical
that was substantially centered on the axial thiolate ligand. Even more
recently, Shaik and coworkers (Ogliaro, F.; de Visser, S. P.; Groves, J. T.;
Shaik, S. Angew. Chem. Int. Ed. 2001, 40, 2874) have shown that when
solvent effects are accounted for in the DFT calculations in his modeling
of catalase compound I2d (CAT-I), one obtains a porphyrin π-cation radical
description for CAT-I only when the proximal charge relay system consisting
of hydrogen bonding interactions involving the axial tyrosinate ligand is
accounted for in the calculations. Otherwise, one obtains a description
involving an iron(IV)-oxo porphyrin with an axial tyrosinate radical.
Supporting Information Available: The NMR spectra and
a discussion of the proposed assignments. This material is
available free of charge via the Internet at http://pubs.acs.org.
References and Notes
(1) Groves, J. T.; Shalyaev, K.; Lee, J. In The Porphyrin Handbook;
Kadish, K. M., Smith, K. M., Guilard, R., Eds.; Academic: San Diego,
2000, Vol. 4, Ch. 27, 17.
(2) (a) Green, M. T. J. Am. Chem. Soc. 2000, 122, 9495. (b) Green,
M. T. J. Am. Chem. Soc. 1999, 121, 7939. (c) Green, M. T. J. Am. Chem.
Soc. 1998, 120, 10 772. (d) Green, M. T. J. Am. Chem. Soc. 2001, 123,
9218.
(3) Vogel, E.; Will, S.; Schulze Tilling, A.; Neumann, L.; Lex, J.; Bill,
E.; Trautwein, A. X.; Wieghardt, K. Angew. Chem., Int. Ed. Engl. 1994,
33, 731. Vogel, E.; Will, S.; Schulze Tilling, A.; Neumann, L.; Lex, J.;
Bill, E.; Trautwein, A. X.; Wieghardt, K. Angew. Chem. 1994, 106, 771.
(4) For recent reviews on the chemistry of corroles, see: (a) Paolesse,
R., In The Porphyrin Handbook; Kadish, K. M., Smith, K. M., Guilard, R.,
Eds.; Academic: San Diego, 2000, Vol. 2, Ch. 11, 201. (b) Erben, C.; Will,
S.; Kadish, K. M., In The Porphyrin Handbook; Kadish, K. M., Smith, K.
M., Guilard, R., Eds.; Academic: San Diego, 2000, Vol. 2, Ch. 12, 233.
(c) Gross, Z. J. Biol. Inorg. Chem. 2001, 6, 733. (d) Ghosh, A.; Steene, E.
J. Biol. Inorg. Chem. 2001, 6, 739.
(5) (a) Gross, Z.; Golubkov, G.; Simkhovich, L. Angew. Chem., Int.
Ed. Engl. 2000, 39, 4045. (b) Bendix, J.; Gray, H. B.; Golubkov, G.; Gross,