Electronic configuration assignment and the importance of low-lying excited states in high-spin imidazole-ligated iron(II) porphyrinates

Article abstract of DOI:10.1021/ja044077p

The synthesis and characterization of six new high-spin deoxymyoglobin models (imidazole-(tetraarylporphyrinato)iron(II)) are described. These have been intensively studied by temperature-dependent Moessbauer spectroscopy from 295 to 4.2 K. All complexes

Full text of DOI:10.1021/ja044077p

Published on Web 03/25/2005
Electronic Configuration Assignment and the Importance of
Low-Lying Excited States in High-Spin Imidazole-Ligated
Iron(II) Porphyrinates
Chuanjiang Hu,† Arne Roth,† Mary K. Ellison,† Jin An,† Christina M. Ellis,‡
Charles E. Schulz,*,‡ and W. Robert Scheidt*,†
Contribution from the Department of Chemistry and Biochemistry, UniVersity of Notre Dame,
Notre Dame, Indiana 46556, and Department of Physics, Knox College,
Galesburg, Illinois 61401
Received September 28, 2004; E-mail: scheidt.1@nd.edu
Abstract: The synthesis and characterization of six new high-spin deoxymyoglobin models (imidazole-
(tetraarylporphyrinato)iron(II)) are described. These have been intensively studied by temperature-dependent
Mo¨ssbauer spectroscopy from 295 to 4.2 K. All complexes show a strong temperature dependence for the
quadrupole splitting consistent with low-lying excited states of the same or lower multiplicity. An analysis
of the data obtained in applied magnetic fields leads to the assignment of the sign of the quadrupole splitting.
All model compounds as well as those of deoxymyoglobin and deoxyhemoglobin, previously studied, have
a negative sign for the quadrupole splitting. Although not previously predicted, this experimental observation
leads to the assignment of the ground-state electronic configuration for all high-spin imidazole-ligated iron-
2
1
1
1
1
2
2
2
(II) porphyrinates as (dxz) (dyz) (dxy) (dz ) (dx -y ) . This is a distinctly different ground-state electronic
configuration from other high-spin iron(II) porphyrinates; differences in structural details for the two classes
of high-spin complexes are also discussed. The apparent anomaly of differing signs for the zero-field splitting
constant between previously studied model complexes and the heme proteins is addressed; the difference
appears to result from the fact that the assumptions used in the spin Hamiltonian approach that has been
applied to these complexes are not adequately satisfied. Structures of four of the new five-coordinate species
have been determined. Core conformations in these derivatives show variation, but these and previously
studied compounds reveal a limited number of conformational patterns. The bond lengths and other
geometrical parameters such as porphyrin core size and iron out-of-plane displacement support a high-
spin state assignment for the iron(II).
Introduction
pressure. The best known systems are the O2-binding proteins
myoglobin and hemoglobin, which shuttle between deoxy- and
Iron porphyrinate complexes (hemes) are the active sites in
many biologically important systems. Knowledge of both the
electronic and geometric structure at iron appears to be the key
to understanding the many diverse and complicated functions
of these hemoproteins. Indeed, the electronic ground states of
models of these iron(II) species have been intensively investi-
gated both experimentally and theoretically and yet are far from
being determined with certainty.
One widely spread class of heme-based systems are those
that bind the diatomic molecules O2, NO, or CO. The primary
purposes of binding diatomic molecules include transport,
storage and sensing. Sensing proteins include the FixL system,1
which detects the presence of O2, and the guanylate cyclase
system,2 which detects NO and functions in the control of blood
oxy-states. The iron of the hemes remains in the 2+ formal
oxidation state. In the deoxy state, the iron(II) is five-coordinate
and bound to the protein by the proximal histidine ligand. The
hemes become six-coordinate upon binding of an O2 molecule.
In an earlysdare we say heroicsstudy, Pauling and Coryell3
measured the magnetic moments for a number of hemoprotein
systems including deoxyhemoglobin and oxyhemglobin. The
oxy state is found to be diamagnetic, whereas the deoxy state
is high spin with four unpaired electrons.3a The structural
changes that accompany the spin state change are essential to
the mechanism of cooperative O2-binding by hemoglobin and
are the basis of the T f R transition of hemoglobin.4-6 In brief,
the five-coordinate deoxy heme has a structure with a significant
displacement of the iron atom out of the heme plane whereas
the six-coordinate oxyheme has the iron in the plane. The
† University of Notre Dame.
‡ Knox College.
(3) (a) Pauling, L.; Coryell, C. D. Proc. Natl. Acad. Sci. U.S.A. 1936, 22, 159.
(b) Pauling, L.; Coryell, C. D. Proc. Natl. Acad. Sci. U.S.A. 1936, 22, 211.
(c) Coryell, C. D.; Stitt, F.; Pauling, L. J. Am. Chem. Soc. 1937, 59, 633.
(4) Perutz, M. F. Nature 1970, 228, 226.
(5) Perutz, M. F. Nature 1972, 237, 495.
(6) Perutz, M. F.; Fermi, G.; Luisi, B.; Shaanan, B.; Liddington, R. C. Acc.
Chem. Res. 1987, 20, 309.
(1) (a) Rodgers, K. R.; Lukat-Rodgers, G. S.; Barron, J. A. Biochemistry 1996,
35, 9539. (b) Miyatake, H.; Mukai, M.; Park, S. Y.; Adachi, S.; Tamura,
K.; Nakamura, H.; Tsuchiya, T.; Lizuka, T.; Shiro, Y. J. Mol. Biol. 2000,
301, 415.
(2) (a) Murad, F. J. Clin. InVest. 1986, 78, 1. (b) Waldman, S. A.; Murad, F.
Pharmacol. ReV. 1987, 39, 163.
9
10.1021/ja044077p CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 5675-5688
5675

A R T I C L E S
Hu et al.
signaling of the binding state between the four hemes of
tetrameric hemoglobin as dioxygen is bound forms the basis of
the cooperativity and is strongly coupled to the structure of the
five-coordinate iron(II) porphyrin sites.
ences. Although the two structures show many common features
that are expected for high-spin iron(II) species (large iron atom
displacement, long Fe-Np bonds), there are also some interest-
ing differences. The earlier [Fe(TPP)(2-MeHIm)] structure9 has
a domed core with a substantial out-of-plane displacement of
the iron13 whereas the later structure12 had a much less domed
core with a smaller out-of-plane displacement of the iron.
Indeed, the core conformation showed a stepped (or wave)
conformation with an apparent asymmetry in the equatorial
bonds related to the orientation of the axial imidazole ligand.
As part of a more general program of characterization of high-
spin iron(II) porphyrinates, we have now determined the
molecular structure of four new high-spin imidazole-ligated iron-
(II) porphyrinates to determine if there are general structural
features for this class of derivative.
We have also been concerned with the electronic structure
of this class of heme species, which can be considered to be
model compounds for deoxymyoglobin and deoxyhemoglobin.
The electronic structure of iron(II) hemes is quite challenging
to study since most spectroscopic probes provide little or no
information about the states of the d6 metal ion. Iron(II) is a
non-Kramers system and, except in fortuitous circumstances is
EPR silent. Fortunately, Mo¨ssbauer spectroscopy has proven
to be an extremely useful spectroscopic probe for the electronic
structure of iron(II), and we report detailed results in applied
magnetic field for all new derivatives. The Mo¨ssbauer data
measured for the four new high-spin iron(II) samples in both
zero and applied magnetic field, along with previously studied
porphyrin derivatives, provide a consistent picture of the
electronic structure. Imidazole-ligated iron(II) porphyrinates
form a distinctly different set of electronic structures than iron-
(II) porphyrinates with other axial ligands.
The difficulty of understanding the electronic structure of the
five-coordinate imidazole-ligated hemes is reflected in both the
relatively large amount published on the issue and some
changing interpretations for both experimental and theoretical
studies. A number of recent density functional theory (DFT)
studies for five-coordinate imidazole-ligated hemes have
appeared.14-17 However, the conclusions reached about the
electronic ground state are by no means uniform. Three of the
four DFT calculations found that a triplet state was lower in
energy than the experimentally observed quintet state. In all of
these cases, a quintet state is predicted to be lowest in energy
when the iron(II) atom is constrained to be further out of the
porphyrin plane than the value obtained for the triplet state. In
other words, DFT says large values of ∆Fe stabilize higher spin
multiplicities. The fourth study16 does predict a quintet ground
state.
The structural features of the five-coordinate heme considered
to be of prime functional significance are the out-of-plane
displacement of the iron with respect to the porphyrin plane,
the porphyrin core conformation, which is usually considered
to have features in accord with C4V doming, the axial bond
distance, and possible changes in orientation and off-axis tilts
of the proximal imidazole ligand. Despite these interesting and
important features, relatively little structural information is
available on five-coordinate imidazole-ligated iron(II) species.
There are two practical reasons: (i) the compounds are easily
oxidized to iron(III) complexes, and (ii) the equilibrium of
binding only one imidazole to an iron(II) porphyrinate is quite
unfavorable. Thus, the synthesis of six-coordinate bis-ligated
complexes is much easier than that of five-coordinate species.
A synthetic strategy for preparing five-coordinate imidazole-
ligated high-spin iron(II) derivatives was reported by Reed and
Collman in 19737 and used the sterically hindered 2-methylimi-
dazole ligand. This ligand is expected to lead to a significantly
distorted molecule only if a six-coordinate species was formed.
However, stereochemical issues concerning five-coordinate
species remain. The crystalline complex of [Fe(TPP)(2-Me-
HIm)]8 prepared by Reed and Collman was structurally char-
acterized in the laboratory of the late Prof. J. L. Hoard. A
preliminary report of the structure was given at an American
Chemical Society meeting,9 and results were additionally cited
and used by Hoard and Scheidt,10 but complete structural details
were never published. One crystallographic feature that marred
the metrical usefulness of the structure was the presence of
crystallographically required two-fold disorder normal to the
porphyrin plane. This two-fold axis leads to positional disorder
in the axial imidazole and significantly limits the accuracy of
some features involving the axial ligand and possibly that of
the porphyrin ligand as well. A related species, [Fe(TpivPP)-
(2-MeHIm)], also displayed this type of disorder11 and suffers
the same limitations. We recently reported the structure of a
new, more ordered variant of the five-coordinate species [Fe-
(TPP)(2-MeHIm)]. As noted below this new structure reveals
a number of stereochemically important features for iron(II)
porphyrinates that are possibly functionally significant.12
Quite surprisingly, the two crystalline forms of [Fe(TPP)(2-
MeHIm)] show both geometric and electronic structure differ-
(7) Collman, J. P.; Reed, C. A. J. Am. Chem. Soc. 1973, 95, 2048.
(8) The following abbreviations are used in this paper: Porph, a generalized
porphyrin dianion; Tp-OCH3PP, dianion of meso-tetra-p-methoxyphe-
nylporphyrin; TPP, dianion of meso-tetraphenylporphyrin; TTP, dianion
of meso-tetratolylporphyrin, TpivPP, dianion of R,R,R,R-tetrakis( o-
pivalamidophenyl)porphyrin; Piv2C8P, dianion of R,R,5,15-[2,2′-(octanedia-
mido)diphenyl]-R,R,10-20-bis(o-pivalamidophenyl)porphyrin; Im, gener-
alized imidazole; RIm, generalized hindered imidazole; HIm, imidazole;
1-MeIm, 1-methylimidazole; 2-MeHIm, 2-methylimidazole; 1,2-Me2Im, 1,2-
dimethylimidazole; Np, porphyrinato nitrogen; Ct, the center of four
porphyrinato nitrogen atoms; EPR, electron paramagnetic resonance; NRVS,
nuclear resonance vibrational spectroscopy.
A key issue for completely defining the quintet state of the
imidazole-ligated model compounds can be simply stated as
“which of the five occupied d-orbitals is the doubly populated
1
1
2
1
1
2
2
2
orbital?” The possibilities include (i) (dxz) (dyz) (dxy) (dz ) (dx -y ) ,
1
1
1
2
1
2
1
1
1
1
2
2
2
2
2
2
(ii) (dxz) (dyz) (dxy) (dz ) (dx -y ) , (iii) (dxz) (dyz) (dxy) (dz ) (dx -y ) ,
(9) (a) Collman, J. P.; Kim, N.; Hoard, J. L.; Lang, G.; Radonovich, L. J.;
Reed, C. A. Abstracts of Papers; 167th National Meeting of the American
Chemical Society; Los Angeles, CA, April 1974; American Chemical
Society: Washington, DC, 1974; INOR 29. (b) Hoard, J. L., personal
communication to WRS. In particular, Prof. Hoard provided a set of atomic
coordinates for the molecule.
(13) One convenient measure of core doming is given by the difference in the
displacement of the iron from the mean plane of four nitrogen atoms
compared to the displacement from the from the 24-atom mean plane.
(14) Rovira, C.; Kunc, K.; Hutter, J.; Ballone, P.; Parrinello, M. J. Phys. Chem.
A 1997, 101, 8914.
(15) Kozlowski, P. M.; Spiro, T. G.; Zgierski, M. Z. J. Phys. Chem. B 2000,
104, 10659.
(10) Hoard, J. L.; Scheidt, W. R. Proc. Natl. Acad. Sci. U.S.A. 1973, 70, 3919.
(11) Jameson, G. B.; Molinaro, F. S.; Ibers, J. A.; Collman, J. P.; Brauman, J.
I.; Rose, E.; Suslick, K. S. J. Am. Chem. Soc. 1980, 102, 3224.
(16) Liao, M.-S.; Scheiner, S. J. Chem. Phys. 2002, 116, 3635.
(17) Ugalde, J. M.; Dunietz, B.; Dreuw, A.; Head-Gordon, M.; Boyd, R. J. J.
Phys. Chem. A 2004, 108, 4653.
(12) Ellison, M. K.; Schulz, C. E.; Scheidt, W. R. Inorg. Chem. 2002, 41, 2173.
9
5676 J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005

High-Spin Imidazole-Ligated Iron(II) Porphyrinates
A R T I C L E S
1
2
1
1
1
2
2
2
then with water until the aqueous layer was neutral, then dried with
MgSO4, and then distilled twice over P2O5. 2-Methylimidazole and 1,2-
dimethylimidazole were purchased from Aldrich, recrystallized from
toluene, and dried under vacuum. The free-base porphyrin ligands meso-
tetra-p-methoxyphenylporphyrin (H2Tp-OCH3PP), meso-tetratolylpor-
phyrin (H2TTP), and meso-tetraphenylporphyrin (H2TPP) were prepared
according to Adler et al.24 The metalations of the free-base porphyrins
to give [Fe(Porph)Cl] were done as previously described.25 [Fe(Tp-
OCH3PP)]2O, [Fe(TTP)]2O, and [Fe(TPP)]2O were prepared according
to a modified Fleischer preparation.26
or (iv) (dxz) (dyz) (dxy) (dz ) (dx -y ) . The latter two states are
nominally degenerate but may differ in energy, owing to the
relative orientation of the axial imidazole ligand with respect
to the porphyrin core. The DFT calculations fail to provide a
clear picture to this question. The experimental Mo¨ssbauer data,
however, clearly rule out some orbital population possibilities
and suggests a clear, most probable electronic configuration.
Moreover, an examination of all Mo¨ssbauer data for high-spin
iron(II) with a wide variety of axial ligands suggests the division
of high-spin iron(II) porphyrinates into two classes of electronic
configuration.
The following reactions were carried out under strict anaerobic
conditions. All solvents were degassed prior to use by three freeze/
pump/thaw cycles. The four-coordinate species [Fe(II)(Tp-OCH3PP)]
was prepared by reduction of [Fe(Tp-OCH3PP)]2O (0.04 mmol) in
benzene (10 mL) with excess ethanethiol (1.0 mL, >200-fold) according
to Stolzenberg et al.27 The benzene solution was stirred overnight
followed by solvent removal under vacuum. [Fe(II)(TTP)] was prepared
from [Fe(TTP)]2O as above. [Fe(II)(TPP)] was prepared from [Fe-
(TPP)]2O as above, except that toluene was used as the solvent. Solid
[Fe(II)(Porph)] samples were never exposed to air to avoid the easily
formed [Fe(Porph)]2O. UV-vis spectra were recorded on a Perkin-
Elmer Lambda 19 UV/vis/near-IR spectrometer, and IR spectra were
recorded on a Perkin-Elmer Paragon 10000 or Nicolet Nexus 670 FT-
IR spectrometer as KBr pellets. Mo¨ssbauer measurements were
performed on a constant acceleration spectrometer from 4.2 to 300 K
with optional small field and in a 9-T superconducting magnet system
(Knox College). Samples for Mo¨ssbauer spectroscopy were prepared
by immobilization of the crystalline material in Apiezon M grease.
Synthesis of [Fe(Tp-OCH3PP)(2-MeHIm)] and [Fe(Tp-OCH3PP)-
(1,2-Me2Im)]. The complexes were prepared by dissolving the pre-
cipitated [Fe(Tp-OCH3PP)] (0.08 mmol) in chlorobenzene, adding ∼1
mL ethanethiol by pipet and a C6H5Cl solution of imidazole (2-fold
excess) by cannula. The reaction mixture was then stirred for 24 h.
UV-vis in C6H5Cl: λmax, nm; for [Fe(Tp-OCH3PP)(2-MeHIm)], 370,
438, 540, 571, 612; for [Fe(Tp-OCH3PP)(1,2-Me2Im)], 371, 439, 541,
568, 613.
Synthesis of [Fe(TPP)(1,2-Me2Im)]. In this case, after reduction
with ethanethiol, the toluene was not removed under vacuum, but a
toluene solution of 1,2-dimethylimidazole (2-fold excess) was added
by cannula directly to the solution of [Fe(TPP)] (0.08 mmol). UV-vis
in C6H5Cl: λmax, nm; 369, 435, 538, 563, 608.
Synthesis of [Fe(TTP)(2-MeHIm)]. This complex was prepared by
adding excess ligands (2-methylimidazole) (0.4 mmol) in chlorobenzene
(15 mL) by cannula to the solid [Fe(II)(TTP)] and stirred for 1 h. UV-
vis in C6H5Cl: λmax, nm; 368, 439, 538, 567, 613.
For all imidazole complexes, X-ray quality crystals were obtained
in 8 mm × 250 mm sealed glass tubes by liquid diffusion using hexanes
as non-solvent after 10 days (in the case of [Fe(TPP)(1,2-Me2Im)] and
[Fe(TTP)(2-MeHIm)]) and three weeks (in the case of [Fe(Tp-OCH3-
PP)(2-MeHIm)] and [Fe(Tp-OCH3PP)(1,2-Me2Im)]). Microcrystalline
solids for Mo¨ssbauer measurements were obtained by liquid diffusion
in Schlenk tubes using hexanes as the nonsolvent. The solids were
isolated in an inert-atmosphere box and immobilized in Nylon sample
holders.
Finally, we examine the issues of the differing crystal field
parameters that have been reported for the proteins deoxymyo-
globin and -hemoglobin and the imidazole-ligated iron(II) por-
phyrinates, the interpretation of which has changed significantly
over time. The crystal field analyses are based on the most recent
Mo¨ssbauer12,18 and magnetic susceptibility studies.19-21
A
distinct difference, seemingly unexpected, is that the proteins
and the models show differing signs of the zero-field splitting
parameter, D, and the rhombicity E/D of the crystal field among
the studied compounds. A synopsis of the analyses for imida-
zole-ligated iron(II) porphyrinates derivatives is that all of the
model complexes analyzed to date12,18,22 have negative zero-
field splitting constants, while the heme proteins myoglobin and
hemoglobin18 have positiVe values of the zero-field splitting
constant. While most of the assignments are based on Mo¨ssbauer
spectroscopy, the pattern of differing signs is also supported
by an integer spin EPR study.23
Since the difference in the sign of D requires major differ-
ences in the relative orbital energies of the dxz, dyz, and dxy
orbitals, as described in the Discussion, achieving an under-
standing of the origin of the apparent difference between the
proteins and the model complexes was desirable. We have used
Mo¨ssbauer spectroscopy in applied magnetic field to determine
the sign of the zero-field splitting constant for several new high-
spin imidazole-ligated iron porphyrinates. While a completely
satisfactory solution to the understanding of the question of the
differences in the sign of D has not been found, further insight
and a path for additional study has been accomplished.
The analysis in this contribution allows us to give the best
experimental electronic configuration to date for all known
imidazole-ligated high-spin iron(II) porphyrinates: (dxz)2(dyz)1(dxy)1-
1
1
2
2
2
(dz ) (dx -y ) . This assignment is equally valid for the analogous
heme proteins. Moreover, in hemes, this electronic configuration
appears to be found only for high-spin imidazole-ligated iron-
(II) species; all other high-spin five-coordinate iron(II) deriva-
tives have a different electronic configuration.
Experimental Section
General Information. All reactions and manipulations for the
preparation of the iron(II) porphyrin derivatives (see below), were car-
ried out under argon using a double-manifold vacuum line, Schlenkware
and cannula techniques. Benzene and hexanes were distilled over
sodium benzophenone ketyl. Ethanethiol (Aldrich) was used as received.
Chlorobenzene was purified by washing with concentrated sulfuric acid,
X-ray Structure Determinations. Single-crystal experiments were
carried out on a Bruker Apex system with graphite monochromated
Mo K radiation (λ ) 0.71073 Å). The crystalline samples were placed
(24) Adler, A. D.; Longo, F. R.; Finarelli, J. D.; Goldmacher, J.; Assour, J.;
Korsakoff, L. J. Org. Chem. 1967, 32, 476.
(25) (a) Adler, A. D.; Longo, F. R.; Kampus, F.; Kim, J. J. Inorg. Nucl. Chem.
1970, 32, 2443. (b) Buchler, J. W. In Porphyrins and Metalloporphyrins;
Smith, K. M., Ed.; Elsevier Scientific Publishing: Amsterdam, The
Netherlands, 1975; Chapter 5.
(26) (a) Fleischer, E. B.; Srivastava, T. S. J. Am. Chem. Soc. 1969, 91, 2403.
(b) Hoffman, A. B.; Collins, D. M.; Day, V. W.; Fleischer, E. B.; Srivastava,
T. S.; Hoard, J. L. J. Am. Chem. Soc. 1972, 94, 3620.
(27) Stolzenberg, A. M.; Strauss, S. H.; Holm, R. H. J. Am. Chem. Soc. 1981,
103, 4763.
(18) Kent, T. A.; Spartalian, K.; Lang, G. J. Chem. Phys. 1979, 71, 4899.
(19) Nakano, N.; Otsuka, J.; Tasaki, A. Biochim. Biophys. Acta 1971, 236, 222.
(20) Nakano, N.; Otsuka, J.; Tasaki, A. Biochim. Biophys. Acta 1972, 278, 355.
(21) Alpert, A.; Banerjee, R. Biochim. Biophys. Acta 1975, 405, 114.
(22) Kent, T. A.; Spartalian, K.; Lang, G.; Yonetani, T.; Reed, C. A.; Collman,
J. P. Biochim. Biophys. Acta 1979, 580, 245.
(23) Hendrich, M. P.; Debrunner P. G. J. Magn. Reson. 1988, 78, 133-141.
9
J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005 5677

A R T I C L E S
Hu et al.
Table 1. Crystallographic Details for the Four Structures
[Fe(Tp-OCH3PP)(2-MeHIm)]
[Fe(Tp-OCH3PP)(1,2-Me2Im)]
[Fe(TPP)(1,2-Me2Im)]
[Fe(TTP)(2-MeHIm)]
formula
C52H42FeN6O4‚2.3C6H5Cl
‚0.2C4H6N2
1146.05
C53H44FeN6O4‚C6H5Cl
‚0.5C6H14
1032.91
C49H36FeN6‚C7H8
C52H42FeN6‚C6H5Cl
‚0.5C6H14
962.40
FW
856.82
a, Å
b, Å
c, Å
16.2258(6)
17.1365(7)
20.7837(8)
9.7135(11)
10.7474(13)
25.525(3)
85.980(2)
87.730(2)
74.323(3)
2558.6(5)
2
13.0504(8)
9.4303(6)
35.931(2)
13.9178(3)
22.8372(4)
15.4107(3)
R
â, deg
104.053(1)
98.468(1)
96.999(1)
γ
V, Å3
5606.0(4)
4
4373.8(5)
4
4861.7(2)
4
Z
space group
Dc, g/cm3
F(000)
P21/n
1.358
2385
0.436
P1h
P21/n
1.301
1792
0.391
P21/c
1.315
2020
0.413
1.341
1081
0.403
0.20 × 0.12 × 0.03
µ, mm-1
crystal dimens, mm3
absorption correction
radiation, Mo KR, λh
T, K
0.3 × 0.3 × 0.2
0.5 × 0.4 × 0.2
0.41 × 0.15 × 0.14
SADABS
0.71073 Å
130(2)
100(2)
100(2)
100(2)
total data collected
unique data
61136
13917
(Rint ) 0.043)
10323
23820
10478
(Rint ) 0.088)
4470
47270
10903
(Rint ) 0.023)
9879
66800
17880
(Rint ) 0.035)
12349
unique obsd data [I > 2 σ(I)]
refinement method
on F2 (SHELXL)
final R indices [I > 2 σ(I)]
R1 ) 0.0496,
wR2 ) 0.1258
R1 ) 0.0714,
wR2 ) 0.1416
R1 ) 0.0668,
R1 ) 0.0396,
wR2 ) 0.1077
R1 ) 0.0434,
wR2 ) 0.1123
R1 ) 0.0488,
wR2 ) 0.1283
R1 ) 0.0815,
wR2 ) 0.1454
wR2 ) 0.1346
R1 ) 0.1643,
wR2 ) 0.1622
final R indices [for all data]
in inert oil, mounted on a glass pin, and transferred to the cold gas
stream of the diffractometer. Crystal data were collected at 100 K (in
case of [Fe(TPP)(1,2-Me2Im)], [Fe(TTP)(2-MeHIm)], and [Fe(Tp-
OCH3PP)(1,2-Me2Im)]) or 130 K (in case of [Fe(Tp-OCH3PP)(2-
MeHIm)]). The structures were solved by direct methods using
SHELXS-9728 and refined against F2 using SHELXL-97;29,30 subsequent
difference Fourier syntheses led to the location of most of the remaining
non-hydrogen atoms. For the structure refinement all data were used
including negative intensities. All non-hydrogen atoms were refined
anisotropically if not remarked otherwise below. Hydrogen atoms were
idealized with the standard SHELXL-97 idealization methods. The
program SADABS31 was applied for the absorption correction. Brief
crystallographic details are given for each structure in Table 1. Complete
crystallographic details, atomic coordinates, anisotropic thermal pa-
rameters, and fixed hydrogen atom coordinates are given in the
Supporting Information for all four structures.
[Fe(Tp-OCH3PP)(2-MeHIm)]‚2.3C6H5Cl‚0.2C4H6N2. A dark-
purple crystal with the dimensions 0.3 × 0.3 × 0.2 mm3 was used for
the structure determination. The 2-methylimidazole was found to be
disordered over two positions, a major and a minor position. The
anisotropic displacement parameters of the atoms of the minor
2-methylimidazole N(5b), N(6b), C(4b), and C(3b) were each con-
strained to be equal to the anisotropic displacement parameters of the
corresponding atoms N(5a), N(6a), C(4a), and C(3a) of the major
imidazole orientation. After the final refinement the occupancy of the
major imidazole orientation was found to be 85%. The unit cell contains
two fully occupied chlorobenzene molecules. A third solvent molecule
(30% chlorobenzene and 20% 2-methylimidazole) was found to be
disordered around an inversion center. It was refined as two rigid
groups, the non-hydrogen atoms were isotropically refined. The
hydrogen atoms of the major imidazole ligand were located directly
from a difference Fourier and subsequently idealized while allowing
the torsion angles to refine.
[Fe(Tp-OCH3PP)(1,2-Me2Im)]‚C6H5Cl‚0.5C6H14. A dark-purple
crystal with the dimensions 0.20 × 0.12 × 0.03 mm3 was used for the
structure determination. The axial ligand was found to have a single
orientation. The asymmetric unit contains one porphyrin complex, one-
half of a hexane molecule and two half molecules of chlorobenzene.
Both of the chlorobenzene half molecules were found to be disordered
over two positions related by an inversion center. The first chloroben-
zene molecule and Cl(3S) of the second one were refined anisotropi-
cally; other atoms of the second chlorobenzene molecule and hexane
molecule were refined isotropically. The chlorobenzene half-molecules
were refined with the carbon atoms fixed as idealized six-membered
rings that were allowed to ride on their respective chlorine atoms. One
of the methoxyphenyl groups (C42‚‚‚C47) had large temperature factors
and was found to be disordered over two positions. However, these
positions are too close to be refined independently. For hydrogen atoms
of methyl C(4) and C(5), a difference electron density synthesis was
calculated around the circle which represents the loci of possible
hydrogen positions (for a fixed C-H distance and X-C-H angle).
The maximum electron density (in the case of a methyl group after
local 3-fold averaging) was taken as the starting hydrogen position.
Then the hydrogen coordinates were re-idealized and rode on the atoms
to which they were attached with fixed thermal parameters (uij ) 1.5Uij-
(eq)). Other hydrogen atoms were placed at idealized positions, and a
riding model (uij ) 1.2Uij(eq)) was used for subsequent refinements.
[Fe(TPP)(1,2-Me2Im)]‚C7H8. A dark-purple crystal with the dimen-
sions 0.5 × 0.4 × 0.2 mm3 was used for the structure determination.
The 1,2-dimethylimidazole was found to be disordered over two
positions, a major and a minor position. The anisotropic displacement
parameters of the atoms of the minor 1,2-dimethylimidazole N(5b),
N(6b), C(5b), C(4b), and C(3b) were each constrained to be equal to
the anisotropic displacement parameters of the corresponding atoms
N(5a), N(6a), C(5a), C(4a), and C(3a) of the major imidazole ligand.
After the final refinement the occupancy of the major imidazole ligand
was found to be 71%. The unit cell contains one fully occupied toluene
molecule. It was refined as a rigid group, the non-hydrogen atoms were
(28) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467.
(29) Sheldrick, G. M. Program for the Refinement of Crystal Structures.
Universita¨t Go¨ttingen, Germany, 1997.
(30) R1 ) ∑||Fo| - |Fc||/∑|Fo| and wR2 ) {∑[w(Fo2 - Fc2)2]/∑[wFo4]}1/2. The
conventional R-factors R1 are based on F, with F set to zero for negative
F2. The criterion of F2 > 2σ(F2) was used only for calculating R1. R-factors
based on F2 (wR2) are statistically about twice as large as those based on
F, and R-factors based on ALL data will be even larger.
(31) Sheldrick, G. M. Program for Empirical Absorption Correction of Area
Detector Data; Universita¨t Go¨ttingen, Germany, 1996.
9
5678 J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005

High-Spin Imidazole-Ligated Iron(II) Porphyrinates
A R T I C L E S
Scheme 1
Figure 1. ORTEP diagram of [Fe(Tp-OCH3PP)(2-MeHIm)]. The major
orientation (85%) of the imidazole ligand is shown. The hydrogen atoms
of the porphyrin ligand have been omitted for clarity; the hydrogen atoms
of the imidazole ligand are shown; 50 probability ellipsoids are depicted.
isotropically refined. The hydrogen atoms of the major imidazole ligand
were located directly by using a difference Fourier and subsequently
idealized while allowing the torsion angles to refine.
[Fe(TTP)(2-MeHIm)]‚C6H5Cl‚0.5C6H14. A red crystal with the
dimensions 0.41 × 0.15 × 0.14 mm3 was used for the structure
determination. The asymmetric unit contains one porphyrin complex
and one fully occupied chlorobenzene molecule and a half molecule
of hexane. The chlorobenzene molecule is disordered over two positions
and was refined as a rigid group. For the methyl hydrogen atoms of
the imidazole group, a difference electron density synthesis was
calculated around the circle which represents the loci of possible
hydrogen positions (for a fixed C-H distance and X-C-H angle).
The maximum electron density (in the case of a methyl group after
local 3-fold averaging) was taken as the starting hydrogen position.
Then the hydrogen coordinates were re-idealized and rode on the atoms
to which they were attached with fixed thermal parameters (uij ) 1.5Uij
(eq)).
dihedral angle between the major imidazole ligand and the 24-
atom mean porphyrin plane is 81.4°. Other selected geometric
parameters involving the iron and imidazole ligand are given
in Table 2 for this and the remaining structures.
[Fe(Tp-OCH3PP)(1,2-Me2Im)] is illustrated in the ORTEP
diagram of Figure 2. The 1,2-dimethylimidazole ligand has a
single orientation. The orientation of the hydrogen atoms on
the hindering methyl, as defined by the X-ray diffraction result,
is illustrated in this figure. Similar orientations are seen for all
of the other structures as well. The dihedral angle φ between
the imidazole plane and the plane defined by N(4), Fe, N(5) is
20.7°. The dihedral angle between the imidazole ligand and the
24-atom mean porphyrin plane is 88.6°.
An ORTEP diagram of [Fe(TPP)(1,2-Me2Im)] is given in
Figure 3. The 1,2-dimethylimidazole ligand is disordered over
two positions with the major orientation (71% occupancy)
shown. The dihedral angle between the two axial ligand planes
is 2.6°. The dihedral angle φ between the imidazole plane and
the plane defined by N(1), Fe, N(5a) is 20.9°. The dihedral angle
between the major imidazole ligand and the 24-atom mean
porphyrin plane is 84.6°. An ORTEP diagram showing both
orientations of the axial ligand is provided in the Supporting
Information (Figure S2).
The final structure is that of [Fe(TTP)(2-MeHIm)] which is
illustrated in Figure 4. The dihedral angle φ between the
imidazole plane and the plane defined by N(1), Fe, N(5) is 35.8°.
The dihedral angle between the imidazole ligand and the 24-
atom mean porphyrin plane is 88.7°.
Illustrated in Figure 5 are formal diagrams of the porphyrin
cores of the four new iron(II) imidazole structures. Given are
the displacements of each atom from the mean plane of the
24-atom porphyrin core in units of 0.1 Å. The orientation of
the imidazole ligand with respect to the core atoms are shown
by the line with the circle representing the methyl group bound
at the 2-carbon atom position. An analogous diagram showing
atomic displacements from the mean plane of the four nitrogen
atoms is given in the Supporting Information (Figure S3).
Also included on the diagrams of Figure 5 are the individual
Fe-Np bond lengths. The average Fe-Np bond lengths are
2.087(7) Å for [Fe(Tp-OCH3PP)(2-MeHIm)], 2.076(8) Å for
[Fe(Tp-OCH3PP)(1,2-Me2Im)], 2.079(8) Å for [Fe(TPP)(1,2-
Me2Im)], and 2.076(3) Å in [Fe(TTP)(2-MeHIm)]. Other
selected bond distances and angles are given in Table 2. A
Results
The five-coordinate, imidazole-ligated, iron(II) porphyrin
complexes were synthesized starting from the corresponding
iron(III) µ-oxo derivatives ([Fe(Porph)]2O). Ethane thiol reduc-
tion to give the iron(II) species [Fe(Porph)] was followed by
reaction with a hindered imidazole ligand as shown in Scheme
1. The iron(II) species, especially four-coordinate [Fe(Porph)],
are air sensitive and must be handled under rigorous inert-
atmosphere conditions. Crystalline samples suitable for X-ray
structure determinations were obtained by liquid diffusion of a
nonsolvent layered above the porphyrin solution in sealed glass
tubes. The structures of four derivatives have been obtained:
[Fe(Tp-OCH3PP)(2-MeHIm)], [Fe(Tp-OCH3PP)(1,2-Me2Im)],
[Fe(TPP)(1,2-Me2Im)], and [Fe(TTP)(2-MeHIm)]. Although two
of the structures have disordered imidazole ligands, the second
orientation of the imidazole has only a minor occupation, and
the major orientation of the ligand dominates the structural
features of the molecule.
An ORTEP diagram of [Fe(Tp-OCH3PP)(2-MeHIm)] is given
in Figure 1. The axial ligand is disordered over two positions
with a major and a minor orientation. The diagram shows the
major orientation (85% occupancy) of the imidazole ligand. An
ORTEP diagram showing both orientations of the axial ligand
is provided in the Supporting Information (Figure S1). The
labeling scheme is consistent with all of the diagrams and tables.
The two orientations of the imidazole ligand are essentially
coplanar; the dihedral angle between the two axial ligand planes
is 0.7°. The dihedral angle (φ) between the major imidazole
ligand and the plane defined by N(1), Fe, N(5a) is 44.7°. The
9
J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005 5679

A R T I C L E S
Hu et al.
ref
Table 2. Selected Bond Distances (Å) and Angles (deg) for [Fe(Porph)(Im)] and Related Speciesa
complexb
Fe−Np
Fe−Nim
∆N4
∆
Ct‚‚‚Nd
Fe−N−
Cg,h
Fe−N−
Cg,i
θg,j
φg,k
c,d
d
d,e
d,f
iron(II)
0.42
0.39
0.51
0.38
0.55
0.38
[Fe(TPP)(1,2-Me2Im)]
[Fe(TTP)(2-MeHIm)]
2.079(8)
2.076(3)
2.087(7)
2.077(6)
2.086(8)
2.073(9)
2.080(6)
2.072(6)
2.075(20)
2.11(2)
2.108(15)
2.107(2)
2.114(2)
2.076(20)
2.096(4)
1.997(4)
2.057(4)
2.158(2)l
2.144(1)
2.155(2)l
2.137(4)
2.161(5)
2.127(3)l
2.147(13)
2.095(6)
2.13(2)
2.002(15)
2.301(2)n
2.034(3)o
1.937(4)o
2.370(3)p
2.360(2)p
2.014(5)
2.351(3)s
0.36
0.32
0.39
0.35
0.42
0.32
2.048
2.050
2.049
2.046
2.044
2.049
2.048(2)
2.033
2.051
2.045
2.040
2.033
2.037
2.033
2.030
1.997
2.057
129.3(2)
132.8(1)
130.4(2)
131.9(3)
131.4(4)
131.1(2)
131.2(10)
132.1(8)
126.5
NRm
124.9(2)
121.4(1)
123.4(2)
122.7(3)
122.6(4)
122.9(2)
123.0(11)
126.3(7)
120.4
NR
11.4
6.6
8.6
6.1
10.3
8.3
20.9
35.8
44.5
20.7
6.5
tw
tw
tw
tw
9
[Fe(Tp-OCH3PP)(2-MeHIm)]
[Fe(Tp-OCH3PP)(1,2-Me2Im)]
[Fe(TPP)(2-MeHIm)](two-fold)
[Fe(TPP)(2-MeHIm)]‚1.5C6H5Cl
average of the six
24.0
12
0.36(4)
0.40
0.31
0.52
0.53
0.55
0.56
0.42
0.52
0.0r
0.44(7)
8.6(20)
[Fe(TpivPP)(2-MeHIm)]
[Fe(Piv2C8P)(1-MeIm)]
[Fe(TpivPP)(2-MeIm-)]-
[Fe(TpivPP)Cl]-
0.43
0.34
0.65
0.59
0.64
0.62
NR
9.6
5.0
5.1
22.8
34.1
14.7
-
11
32
33
34
35
35
34
36
37
38
-
-
-
[Fe(TpivPP)(O2CCH3)]-
[Fe(TpivPP)(OC6H5)]-
[Fe(TpivPP)(SC6HF4)]-
[Fe(TPP)(SC2H5)]-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.62
-
-
-
[Fe(TPP)(1-MeIm)2]q
0.0r
128.2
-
128.3
-
1.1
-
14.7
-
[Fe(TPP)(THF)2]
0.0r
0.0r
iron(III)
0.36
0.0r
[Fe(OEP)(2-MeHIm)]+
[Fe(OEP)(2-MeHIm)2]+
[Fe(TPP)(2-MeHIm)2]+ q
[Fe(TMP)(1,2-Me2Im)2]+ q
[Fe(OEP)(1-MeIm)2]+ q
2.038(6)
2.041(11)
1.970(4)
1.937(12)
2.004(2)
2.068(4)
2.275(1)
2.012(4)
2.004(5)
1.975(2)
0.35
0.0r
0.0
2.008
2.041
1.970
1.937
2.004
131.7(3)
134.15(10)
132.8
134.6
NR
122.1(3)
120.76(9)
120.6
119.8
NR
5.0
3.2
3.9
22.2
33,32
45,45
20
39
40
41
42
43
0.0
∼4,4
0.01
0.01
NR
NR
0.0r
0.0r
a Estimated standard deviations are given in parentheses. b Unless noted otherwise, complex is high spin. c Averaged value. d In Å. e Displacement of iron
from the mean plane of the four pyrrole nitrogen atoms. f Displacement of iron from the 24-atom mean plane of the porphyrin core. g Value in degrees.
h 2-carbon, sometimes methyl substituted. i Imidazole 4-carbon. j Off-axis tilt (deg) of the Fe-NIm bond from the normal to the porphyrin plane. k Dihedral
angle between the plane defined by the closest Np -Fe-Nim and the imidazole plane in deg. l Major imidazole orientation. m Not reported. n Chloride.
o Anionic oxygen donor. p Thiolate. q Low spin. r Six-coordinate; required to be zero by symmetry. s Neutral oxygen donor.
Figure 2. ORTEP diagram of [Fe(Tp-OCH3PP)(1,2-Me2Im)]. The single
orientation of the (ordered) imidazole ligand is shown. The hydrogen atoms
of the porphyrin ligand have been omitted for clarity; 50% probability
ellipsoids are depicted.
Figure 3. ORTEP diagram of [Fe(TPP)(1,2-Me2Im)]. The major orientation
of the imidazole ligand is shown. The hydrogen atoms of the porphyrin
ligand have been omitted for clarity; 50% probability ellipsoids are depicted.
complete listing of bond distances and angles for all four
structures is given in the Supporting Information.
rupole splitting and isomer shift values show a strong temper-
ature dependence. The quadrupole splitting for the crystalline
species [Fe(Tp-OCH3PP)(2-MeHIm)] at 4.2 K is 2.18 mm/s and
at 298 K is 1.67 mm/s. The isomer shifts at these respective
temperatures are 0.94 and 0.81 mm/s. At 4.2 K the quadrupole
splitting for [Fe(Tp-OCH3PP)(1,2-Me2Im)] is 2.44 mm/s and
the isomer shift is 0.95 mm/s.These values at 298 K are 1.77
and 0.81 mm/s respectively. For [Fe(TPP)(1,2-Me2Im)], ∆EQ
at 4.2K is 1.97 mm/s and δFe is 0.92 mm/s. At 295 K these
values are 1.61 and 0.82 mm/s respectively. The quadrupole
splitting for the crystalline species [Fe(TTP)(2-MeHIm)] at 15
K is 2.03 and 1.40 mm/s at room temperature. The isomer shifts
at these temperatures are 0.92 and 0.81 mm/s, respectively.
These and the Mo¨ssbauer parameters for other iron(II)tet-
The Fe-N(imidazole) bond lengths for the four structures
range from 2.14 to 2.16 Å, and in all four structures the
imidazole ligand Fe-N bond vector is tilted off of the normal
to the 24-atom porphyrin mean plane. The value of the tilt angle,
given by θ in Table 2, is 8.6° for [Fe(Tp-OCH3PP)(2-MeHIm)],
6.1° for [Fe(Tp-OCH3PP)(1,2-Me2Im)], 11.4° for [Fe(TPP)(1,2-
Me2Im)], and 6.6° for [Fe(TTP)(2-MeHIm)]. Two other im-
portant angles associated with the imidazole ligands (Fe-N-
C) are given in Table 2. The range of iron atom displacements
out of the porphyrin plane is 0.38-0.51 Å. The radii of the
porphyrin cores, given by Ct‚‚‚N in Table 2, are nearly identical
at 2.046-2.050 Å.
These crystalline species as well as others were studied with
variable temperature Mo¨ssbauer spectroscopy. Both the quad-
9
5680 J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005

High-Spin Imidazole-Ligated Iron(II) Porphyrinates
A R T I C L E S
Figure 4. ORTEP diagram of [Fe(TTP)(2-MeHIm)]. The single orientation
of the (ordered) imidazole ligand is shown. The hydrogen atoms of the
porphyrin ligand have been omitted for clarity; 50% probability ellipsoids
are depicted.
raarylporphyrins at various temperatures are given in Table S25
of the Supporting Information.
Discussion
Mo1ssbauer /Electronic Structure. Mo¨ssbauer isomer shift
(δ) and quadrupole splitting (∆Eq) values for six new iron(II)
tetraarylporphyrinates have been measured from room temper-
ature to 4.2 K. The values of ∆Eq that were observed at 4.2 K
ranged from 1.93 to 2.44 mm/s in magnitude. These are within
the range of previously reported high-spin iron(II) species; the
large value of the isomer shift (∼0.9 mm/s) is consistent with
high-spin iron(II).44 Although the quadrupole splitting is in
principle directly correlated with the detailed structure of the
complex,45 a comparison of the ∆Eq values with the respective
structure does not suggest an obvious correlation.
A potential difficulty in attempting such a correlation is our
observation of the appearance of polymorphic forms. The
examples of [Fe(TPP)(2-MeHIm)] and [Fe(TTP)(2-MeHIm)]
represent the best explored system for the issue of polymorphs.
Prior to this study, the crystal system and unit cell constants
for [Fe(TPP)(2-MeHIm)] had been determined for about 15
crystals for the original structural investigation12 and subse-
quently for NRVS vibrational studies.46 The oriented crystal
(32) Momenteau, M.; Scheidt, W. R.; Eigenbrot, C. W.; Reed, C. A. J. Am.
Chem. Soc. 1988, 110, 1207.
(33) Mandon, D.; Ott-Woelfel, F.; Fischer, J.; Weiss, R.; Bill, E.; Trautwein,
A. X. Inorg. Chem. 1990, 29, 2442.
(34) Schappacher, M.; Ricard, L.; Weiss, R.; Montiel-Montoya, R.; Gonser, U.;
Bill, E.; Trautwein, A. X. Inorg. Chim. Acta 1983, 78, L9.
(35) Nasri, H.; Fischer, J.; Weiss, R.; Bill, E.; Trautwein, A. X. J. Am. Chem.
Soc. 1987, 109, 2549.
(36) Caron, C.; Mitschler, A.; Riviere, G.; Schappacher, M.; Weiss, R. J. Am.
Chem. Soc. 1979, 101, 7401.
(37) Steffen, W. L.; Chun, H. K.; Hoard, J. L.; Reed, C. A. Abstracts of Papers;
175th National Meeting of the American Chemical Society; Anaheim, CA,
March, 1978; American Chemical Society: Washington, DC, 1978; INOR
15. (b) Hoard, J. L., personal communication.
(38) Reed, C. A.; Mashiko, T.; Scheidt, W. R.; Spartalian, K.; Lang, G. J. Am.
Chem. Soc. 1980, 102, 2302.
(39) Scheidt, W. R.; Geiger, D. K.; Lee, Y. J.; Reed, C. A.; Lang, G. J. Am.
Chem. Soc. 1985, 107, 5693.
(40) Geiger, D. K.; Lee, Y. J.; Scheidt, W. R. J. Am. Chem. Soc. 1984, 106,
6339.
Figure 5. Formal diagrams of the porphyrinato cores of (top to bottom)
(a) [Fe(Tp-OCH3PP)(2-MeHIm)], (b) [Fe(Tp-OCH3PP)(1,2-Me2Im)], (c)
[Fe(TPP)(1,2-Me2Im)], and (d) [Fe(TTP)(2-MeHIm)]. Illustrated are the
displacements of each atom from the mean plane of the 24-atom core in
units of 0.01 Å. Positive values of displacement are toward the imidazole
ligand. The diagrams also show the orientation of the imidazole ligand with
respect to the atoms of the porphyrin core. The location of the methyl group
at the 2-carbon position is represented by the circle.
(41) Scheidt, W. R.; Kirner, J. F.; Hoard, J. L.; Reed, C. A. J. Am. Chem. Soc.
1987, 109, 1963.
(42) Munro, O. Q.; Serth-Guzzo, J. A.; Turowska-Tyrk, I.; Mohanrao, K.;
Shokhireva, T. Kh.; Walker, F. A.; Debrunner, P. G.; Scheidt, W. R. J.
Am. Chem. Soc. 1999, 121, 11144.
(43) Mylrajan, M.; Andersson, L. A.; Sun, J.; Loehr, T. M.; Thomas, C. S.;
Sullivan, E. P., Jr.; Thomson, M. A.; Long, K. M.; Anderson, O. P.; Strauss,
S. H. Inorg. Chem. 1995, 34, 3953.
(44) Debrunner, P. G. In Iron Porphyrins Part 3; Lever, A. B. P., Gray, H. B.,
Eds.; VCH Publishers Inc.: New York, 1983; Chapter 2.
(45) Zhang, Y.; Mao, J.; Oldfield, E. J. Am. Chem. Soc. 2002, 124, 7829.
(46) Rai, B. K.; Durbin, S. M.; Prohofsky, E. W.; Sage, J. T.; Ellison, M. K.;
Scheidt, W. R.; Sturhahn, W.; Alp, E. E. Phys. ReV. E 2002, 66, 051904-
1-12.
NRVS studies utilized several large crystals with final specimens
enriched to 95% 57Fe, all crystallographically characterized and
identical. These crystals were obtained from several different,
9
J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005 5681

A R T I C L E S
Hu et al.
Figure 6. Plot of the temperature variation of the quadrupole splitting value as a function of temperature for the six new imidazole-ligated iron(II) porphyrinates
and the heme proteins deoxymyoglobin and deoxyhemoglobin. Data for deoxymyoglobin are taken from ref 22, while those for deoxyhemoglobin are from
ref 53.
independent preparations and an even larger number of crystal-
growing tubes. When we subsequently made a new preparation
of [Fe(TPP)(2-MeHIm)] for a complete temperature-dependent
Mo¨ssbauer study, a new spectrum was obtained, not identical
to that previously observed.12 Cell constant determinations for
single crystals of this new preparation revealed that a new phase
had been obtained.47 One crystal-growing tube containing an
57Fe sample prepared sometime earlier was available; quite
surprisingly, this sample was identical to the later sample both
with respect to cell constants and Mo¨ssbauer spectrum. How-
ever, four separate preparations of the complex [Fe(TTP)(2-
MeHIm)] yielded samples that had equivalent Mo¨ssbauer
spectra. Limitations of experimental time did not allow such
detailed comparisons on the remaining species. We did observe
that two independent samples [Fe(Tp-OCH3PP)(1,2-Me2Im)]
were slightly different in Mo¨ssbauer values, two samples of [Fe-
(TPP)(2-MeHIm)] were the same, while two samples of [Fe-
(TTP)(1,2-Me2Im)] were different. Complete details of all
Mo¨ssbauer observations are given in Table S25.
There is a substantial temperature variation of the quadrupole
splitting value; complete values of δ and ∆Eq vs T are tabulated
in Table S25. A plot of the data is given in Figure 6. The value
of ∆Eq decreases significantly as the temperature is increased;
the temperature variation for these samples is similar over the
temperature range examined. Although we have not explored
the issue in detail, the maximum value of ∆Eq appears to
typically be at a temperature somewhat above 4.2 K. The
explanation for this temperature variation is almost certainly
the presence of a number of close-lying excited states. The
excited states could have the same or differing spin multiplicity
relative to the ground state.
previously seen (over a smaller temperature range) for deoxy-
myoglobin and -hemoglobin samples.18,22,48-53 The observations
are compared in Figure 6. It is almost certain that all of these
five-coordinate imidazole-ligated species share many common
features of electronic structure. A detailed study of the tem-
perature dependence observed for deoxy-Hb was used to attempt
to distinguish which of the quintet states is the ground state.53
1
2
1
1
1
2
2
2
The ground-state thus assigned was (dz ) (dxy) (dxz) (dyz) (dx -y
)
with the other quintet states very close in energy. This
assignment of the ground state is unlikely to be correct as is
discussed below.
The application of applied magnetic field Mo¨ssbauer spec-
troscopy provides more rigorous information concerning the
electronic ground states. Lang and co-workers pioneered in the
use of applied field measurements to deoxyMb and -Hb as well
as two five-coordinate model compounds, [Fe(TPP)(2-MeHIm)]-
(two-fold) and [Fe(TPP)(1,2-Me2Im)] (structure unknown). An
analysis of the spectra showed the largest component of the
electric field gradient, Vzz, to have a negative value and hence
the sign of the quadrupole splitting value is also negative. For
an iron compound with nonionic ligands, it is reasonable to
assume that the electric field gradient (EFG) at the nucleus is
dominated by contributions from the d-electrons. Given that the
quadrupole splitting is negative, this clearly shows that the
2
doubly occupied orbital must be chosen between dyz, dxz, or dz ,
although Lang et al. did not specify which. All of the systems
of Lang et al. were found to have an asymmetry parameter η )
(Vxx - Vyy)/Vzz > 0.6. As shown below, we find that the sign of
(48) Kent, T. A.; Spartalian, K.; Lang, G.; Yonetani, T. Biochim. Biophys. Acta
1977, 490, 331.
(49) Bade, D.; Parak, F. Biophys. Struct. Mechanism 1976, 2, 219.
(50) Eicher, H.; Bade, D.; Parak, F. J. Chem. Phys. 1976, 64, 1446.
(51) Huynh, B. H.; Paraefthymiou, G. C.; Yen, C. S.; Wu, C. S. J. Chem. Phys.
1974, 61, 3750.
The observed temperature dependence of ∆Eq for these iron-
(II) tetraarylporphyrinate derivatives is similar to the variation
(52) Eicher, H.; Trautwein, A. J. Chem. Phys. 1970, 52, 932.
(53) Eicher, H.; Trautwein, A. J. Chem. Phys. 1969, 50, 2540.
(47) Hu, C.; Scheidt, W. R. Work in progress.
9
5682 J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005

High-Spin Imidazole-Ligated Iron(II) Porphyrinates
A R T I C L E S
Table 3. Mo¨ssbauer Parameters for Five-Coordinate, High-Spin
Imidazole-Ligated Iron(II) Tetraarylporphyrinates and Hb and Mb
a
a
complex
∆EQ
δFe
Γb
T, K
ref
[Fe(Tp-OCH3PP)(1,2-Me2Im)]
[Fe(Tp-OCH3PP)(2-MeHIm)]
[Fe(TPP)(1,2-Me2Im)]
[Fe(TPP)(2-MeHIm)]
[Fe(TTP)(2-MeHIm)]
[Fe(TTP)(1,2-Me2Im)]
[Fe(TPP)(2-MeHIm)]
[Fe(TPP)(2-MeHIm)(two-fold)
[Fe(TPP)(1,2-Me2Im)]
[Fe((Piv2C8P)(1-MeIm)]
Hb
-2.44 0.95 0.46
-2.18 0.94 0.58
-1.93 0.92 0.44
-1.96 0.86 0.55
-1.95 0.85 0.42
-2.06 0.86 0.43
-2.40 0.92 0.50
-2.28 0.93 0.31
-2.16 0.92 0.25
4.2 twc
4.2 tw
4.2 tw
4.2 tw
4.2 tw
4.2 tw
4.2 12
4.2 22
4.2 22
4.2 32
4.2 22
4.2 22
4.2 57
4.2 58
4.2 59
4.2 35
4.2 35
-2.3c
0.88 0.40
Figure 7. Comparison of the typical stereochemical features of the
imidazole-ligated high-spin iron(II) complexes with a (dxz)2 ground con-
figuration with the anionic high-spin iron(II) complexes with a (dxy)2 ground
configuration. Data were taken from Table 2.
-2.40 0.92 0.30
-2.22 0.92 0.34
+2.18 0.83 0.30
+4.01 1.03
+4.25 1.05 0.30
+3.67d 1.03 0.40
+3.90d 1.06 0.38
+3.51d 0.97
Mb
[Fe(TpivPP)(SC2H5]-
[Fe(OC6H5)(TPP)]-
[Fe(O2CCH3)(TpivPP)]-
[Fe(OCH3)(TpivPP)]-
[Fe(OC6H5)(TpivPP)]-
[Fe(TpivPP)(2-MeIm)]-
Thus, high-spin iron(II) porphyrinates can be divided into
two different electronic configurations, those with a doubly
occupied dxz orbital and those with a doubly occupied dxy orbital.
As can be seen from a comparison of structural data (Table 2)
with the signed quadrupole derivatives of Table 3, this division
into two groups is also manifested in the structures. The
77
33
4.2 57
4.2 57
[Fe(TpivPP)(SC6HF4)][NaC12H24O6] +2.38d 0.84 0.28
[Fe(TpivPP)(SC6HF4)][Na⊂222]
+2.38d 0.83 0.32
[Fe(TpivPP)Cl]-
+4.36d 1.01 0.31 77
34
a mm/s. b Line width, fwhm. c Sign not determined experimentally,
presumed negative. d Sign not determined experimentally, presumed posi-
tive.
1
1
2
1
1
2
2
2
complexes with known (dxz) (dyz) (dxy) (dz ) (dx -y ) configu-
ration have significantly larger iron atom displacements (∼ 0.18
Å larger) and equatorial Fe-Np bonds (∼ 0.04 Å longer). To
date, only the oxygen-carrying hemoproteins and their model
the quadrupole splitting value for all new imidazole-ligated
derivatives is also negative. Moreover, all complexes are of very
low symmetry. This clearly rules out dz as the doubly occupied
orbital. The remaining pair of possible orbitals are expected to
be nearly degenerate with the degeneracy lifted owing to the
coordination of the planar axial imidazole ligand which will
interact to differing degrees with the pair. The low symmetry
deduced clearly shows a lifting of the degeneracy. We can thus
conclude that the ground-state configuration for all of the
2
1
1
1
1
2
2
2
complexes are known to have the (dxz) (dyz) (dxy) (dz ) (dx -y
)
2
configuration. It is interesting to speculate that the electronic
configuration differences are related to functional or ligand
binding dynamical differences. The structural differences in the
two electron configuration types is schematically shown in
Figure 7.
The sign of the quadrupole splitting has thus yielded, with
high probability, the assignment that the ground electronic
2
1
1
1
2
imidazole-ligated iron complexes is the (dxz) (dyz) (dxy) (dz ) -
2
1
1
1
1
2
2
2
configuration is (dxz) (dyz) (dxy) (dz ) (dx -y ) . However, other
features of the spectral analysis and the fits in magnetic field
by Lang et al.18 were regarded as being much less definitive. A
particularly surprising feature was that the Mo¨ssbauer spectral
analysis led to a zero-field splitting constant (D) of -5.0 cm-1
in the fit for the complex [Fe(TPP)(2-MeHIm)](two-fold), while
on the other hand, the value of D in deoxy-Mb was a positive
4.8 cm-1. This positive value was based on both the Mo¨ssbauer
and magnetic susceptibility measurements. Spectra for the other
model complex, [Fe(TPP)(1,2-Me2Im)], were not fit in detail,
but also suggest a negative value of D. Moreover, the value of
D for deoxy-Hb was positive. This difference in the sign of D
between the protein samples and the model compounds was
unexpected since all show strong similarities in the Mo¨ssbauer
at ambient and near-ambient temperatures. In our subsequent
analysis of the magnetic Mo¨ssbauer spectrum of the second
crystalline form of [Fe(TPP)(2-MeHIm)],12 we have also found
a negative value for the quadrupole splitting and a negative value
of D, continuing the unanticipated distinction between the
proteins and the model compounds.
1
2
2
(dx -y ) configuration, and where we have chosen the dxz orbital
to have slightly lower energy than dyz.
2
1
1
1
1
2
2
2
A similar (dxz) (dyz) (dxy) (dz ) (dx -y ) ground configuration
probably pertains to the six-coordinate, high-spin complex [Fe-
(TPP)(THF)2],54 known to have a negative quadrupole splitting
(-2.75 mm/s at 4.2 K). This is consistent with the experimental
X-ray electron density assignment of Lecomte et al.55
This newly assigned electronic configuration appears con-
sistent with the NMR studies of Goff and LaMar on five-
coordinate high-spin iron porphyrinates,56 although they assigned
the doubly occupied orbital as the dxy orbital. However, this
configuration ((dxy)2) must lead to a positive sign for the
quadrupole splitting value. Interestingly, there are a number of
five-coordinate high-spin iron(II) porphyrinate derivatives defi-
nitely known57-59 to have a positive value of the quadrupole
splitting, while a number of others must be regarded as highly
probable. These derivatives are listed in Table 3 with a “+”
sign. As shown in the table, all of these complexes have an
anionic species as the axial ligand.
(54) (a) Reed, C. A.; Mashiko, T.; Scheidt, W. R.; Spartalian, K.; Lang, G. J.
Am. Chem. Soc. 1980, 102, 2302. (b) Boso, B.; Lang, G.; Reed, C. A. J.
Chem. Phys. 1983, 78, 2561.
(55) Lecomte, C.; Blessing, R. H.; Coppens, P.; Tabard, A. J. Am. Chem. Soc.
1986, 108, 6942.
(56) Goff, H.; La Mar, G. N. J. Am. Chem. Soc. 1977, 99, 6599.
(57) Schappacher, M.; Ricard, L.; Fisher, J.; Weiss, R.; Montiel-Montoya, R.;
Bill, E. Inorg. Chem. 1989, 28, 4639.
(58) Shaevitz, B. A.; Lang, G.; Reed, C. A. Inorg. Chem. 1988, 27, 4607.
(59) Bominaar, E. L.; Ding, X.; Gismelseed, A.; Bill, E.; Winkler, H.; Trautwein
A. X.; Nasri, H.; Fisher, J.; Weiss, R. Inorg. Chem. 1992, 31, 1845.
What is the significance of the sign of D for these species?
A difference in sign would appear to imply differences in the
d-orbital energy levels of the complexes in question. The sign
issue can be approached by the spin Hamiltonian approximation.
The spin Hamiltonian approximation attempts to describe the
ground electronic state of the systems, which for high-spin
ferrous is an orbital singlet and a spin quintet, in terms of
9
J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005 5683

A R T I C L E S
Hu et al.
equivalent spin operators, commonly written as60
1
3
H ) D S2z - S(S + 1) + E(Sx2 - Sy2) + µBHB‚g˜‚BS. (1)
[
]
The first two terms (zero-field splitting) describe the shifts in
energy among the spin states due to spin/orbit coupling mixing
in excited spin/orbital states, while the third term describes the
Zeeman interaction between the electron spin and orbital
magnetic moments and an external applied field H.
This Hamiltonian will only accurately describe the ground
state if the energies of the excited states are large relative to
the spin orbit coupling which mixes them into the ground state.
The spin orbit coupling parameter λ is normally assumed to be
about 100 cm-1 for high-spin ferrous iron. If we assume that
the doubly occupied d-orbital is a dxz state, then second-order
perturbation theory gives the following expressions for the
parameters D and E:
3
D ) λ2
2
1
1
1
2
+
+
-
(2)
(3)
[
[
]
3Exy 3Ex2-y2 Ez2 3Eyz
1
3
1
1
E ) λ2
+
-
Figure 8. Figure illustrating the fits to the Mo¨ssbauer data obtained at 9
T. Spectrum (a) is that obtained for [Fe(Tp-OCH3PP)(1,2-Me2Im)], spectrum
(b) for [Fe(TPP)(1,2-Me2Im)], and spectrum (c) is for [Fe(TTP)(1,2-Me2-
Im)].
]
2
Ez2 Ex2-y2 Exy
where Eii is the energy of the relevant d orbital. Again, these
equations assume (i) large splittings to excited states, (ii) no
configurations of lower spin multiplicity have low enough
energy to mix in, and (iii) one can ignore possible equivalent
spin operators of fourth power in the quadratic form above (eq
1).
by the components (Vii) of the electric field gradient. Simulta-
neous fits to spectra at all fields were obtained by least-squares
minimization using a downhill simplex method.61
To obtain the best simultaneous fits to the applied-field spectra
at 3, 6 and 9 T, we had to allow a nonaxial magnetic hyperfine
tensor (A/xx * Ay/y) and also allow the principal axes of the
quadrupole tensor to rotate with respect to the axes of the g
and A˜* tensors. The parameters that were varied for the
simultaneous fits were D, E, the components of A˜*, η, and the
Euler angles relating the principal axes of the quadrupole and
D tensors. Fits to the applied field Mo¨ssbauer data were obtained
for the six new complexes (top six complexes of Table 3). The
sign of D for four complexes, [Fe(Tp-OCH3PP)(2-MeHIm)],
[Fe(TTP)(2-MeHIm)], [Fe(TPP)(2-MeHIm)], and [Fe(TTP)(1,2-
Me2Im)], were clearly found to be positive in the fitting
procedure, similar to the earlier results for deoxymyoglobin and
deoxyhemoglobin. Even when the fits were constrained so that
D < 0, the E values so obtained were found to be defined in an
improper coordinate system. Transforming E back to a proper
coordinate system always led to the condition that D was
positive. The sign determinations of D for [Fe(TPP)(1,2-Me2-
Im)] and [Fe(Tp-OCH3PP)(2-MeHIm)] were found to be
ambiguous, presumably because the magnetic splitting in applied
magnetic field is small and thus not sensitive to D. A
reexamination of the data previously obtained for one form of
[Fe(TPP)(2-MeHIm)]12 still gave fits that D was negative.
Typical fits of the Mo¨ssbauer data in applied field for the
six new derivatives are shown in Figures S4-S9 of the
Supporting Information. Figure 8 shows the 9-T data for (a)
[Fe(Tp-OCH3PP)(1,2-Me2Im)], (b) [Fe(TPP)(1,2-Me2Im)], and
(c) [Fe(TTP)(1,2-Me2Im)]. The significantly differing and
increasing magnetic splittings are clear in the progression. It is
2
2
2
Since the dz and dx -y orbitals are high enough in energy to
be negligible, eq 2 simplifies to the following inequality for D
> 0:
Eyz > 2Exy
(4)
Equation 4 clearly shows that the sign of D depends on the
relative energies of the excited dyz and dxy orbitals. A reasonable
assumption would be that the energy of the dxy orbital is well
above that of the dyz orbital given that dxz is expected to be the
ground state. It is thus apparent that one should expect all of
the imidazole-ligated complexes to have a negative value of D.
We have measured the Mo¨ssbauer spectra in applied magnetic
field, typically at 3, 6, and 9 T, to further examine the nature
of these deoxymyoglobin and -hemoglobin models. The Mo¨ss-
bauer data were fit using the spin Hamiltonian model used
previously by us12 and by Kent et al.:22
1
3
H ) D S2z - S(S + 1) + E(Sx2 - Sy2) + HB‚g˜‚BS + HQ -
[
]
g/NâNHB‚BI + BS‚A˜*‚BI (5)
where D and E are again the axial and rhombic zero-field
splitting parameters, A˜* is the magnetic hyperfine tensor and
HQ gives the nuclear quadrupole interaction:
eQVzz
HQ )
[3I2z - I(I + 1) + η(Ix2 - I2y)]
(6)
12
Q is the quadrupole moment of the 57Fe nucleus and η is given
(61) Press, W. H.; Teukolsky, S. A.; Vetterling, W. T.; Flannery, B. P. Numerical
Recipes in C, The Art of Scientific Computing, 2nd ed.; Cambridge
University Press: 1992; p 408.
(60) Abragam A.; Bleaney, B. Electron Paramagnetic Resonance of Transition
Ions; Oxford University Press: Oxford 1970.
9
5684 J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005

High-Spin Imidazole-Ligated Iron(II) Porphyrinates
A R T I C L E S
to be emphasized that we required simultaneous fits to the data
obtained at three different applied fields, 3, 6, and 9 T. The
simultaneous fitting of spectra in 3, 6, and 9 T applied field
did not completely constrain all of the parameters used in the
fit. We found that adequate simulation of the observed spectra
required rotation of the principal axes of the quadrupole tensor,
and thus three more parameters than the minimal model of eq
5. However, for a wide range of choices of the axial ZFS
parameter D the fitting algorithm was able to readjust the other
parameters and achieve equally good ø-squared values for the
fits. Thus, although we can be confident of the sign of D based
on the requirement that D and E from the fits correspond to a
proper coordinate system, we are unable to claim the determi-
nation of specific values of D for these samples. At best we
can say the ø-squared had a broad minimum for D in the range
all, five-coordinate, imidazole-ligated, high-spin species that
have been studied to date by either Mo¨ssbauer or by integer-
spin EPR.
Further insight into this problem might be gained through
Mo¨ssbauer data taken in higher fields than 9 T. It is possible
that complete understanding of the Mo¨ssbauer results would
require careful application of a crystal field model which
includes all five of the d orbitals and low-symmetry crystal field
potentials. Importantly, we do find among the model compounds
examples that agree with the results on deoxyhemoglobin and
deoxymyoglobin in having a positive sign for D.
Given the clear complexity of the electronic state of these
imidazole-ligated iron(II) species, the difficulties of the theoreti-
cal calculations14-17 in predicting the observed quintet ground
state is explicable. The low symmetry of the electronic state
that we infer strongly suggests that further theoretical calcula-
tions should not use any symmetry simplifications in the
calculations.
of 5 cm-1 to as large as 30 cm-1
.
There are some aspects of the fit using the spin Hamiltonian
(eq 1) that should be noted. First, the anisotropic part of the
magnetic hyperfine tensor A should be proportional to the D
tensor. If one assumes a reasonable value for the spin/orbit
coupling parameter (λ ≈ -100 cm-1), the calculated values for
D and E from the fit components of the A tensor are too large
by factors as large as 3. Calculated D and E values can be
adjusted to match the fit values only with a very small λ (-30
to -50 cm-1), less than half of what is normally used for the
free-ion value. This is an indication of strong delocalization, as
λ should also be roughly proportional to the expectation value
of r-3 , assuming the d electrons feel an effective potential
proportional to 1/r. Second, the isotropic component of the A
tensor, which measures the Fermi contact contribution to the
magnetic hyperfine interaction, is quite small for most of the
fits. As this component is also proportional to the expectation
value r-3 , this is again consistent with strong delocalization
of the d-wave functions. Finally, to get acceptable ø-squared
values for the fits to high-field data, we found it necessary to
allow the orientation of the EFG tensor to be rotated relative to
the D and A tensors. This indicates that the effective symmetry
at the iron site is even lower than the rhombic symmetry
assumed in eq 1.
Structural Results. Another goal of this study was to further
investigate the structural details of the five-coordinate deoxy-
heme model compounds [Fe(Porph)(Im)]. There are a number
of overall structural features expected for such high-spin iron-
(II) complexes.62 These include an expanded porphinato core,
large equatorial Fe-Np bond distances and a significant out-
of-plane displacement of the iron(II) atom. Such features were
observed in a recently characterized, new crystalline form of
[Fe(TPP)(2-MeHIm)]12 and an earlier, partly described crystal-
line form.9 However, we found distinctly different structural
features between these two nominally equivalent species. In the
newer crystalline form we observed a previously unrecognized
porphyrin core distortion that we ascribed to steric interaction
between the axial 2-methylimidazole ligand and the porphyrin
core. An even more interesting response to the steric effects of
the axial ligand was found in the unsymmetrical Fe-Np bond
lengths where two short bonds result from the porphyrin ring
tilting down and two long bonds result from the up-wave of
the core. This is clearly illustrated in the formal diagram of the
porphyrin core given in ref 12. This was in distinct contrast to
the strongly domed features of the first crystalline form of
[Fe(TPP)(2-MeHIm)] that also apparently led to a larger out-
of-plane displacement of the iron (0.55 Å vs 0.38 Å from the
24-atom plane).
Because we thought that the structural differences between
these two derivatives might relate to functional significance,
we have prepared and structurally characterized several ad-
ditional examples in order to determine what structural pattern,
if any, is dominant. We have again employed the sterically
hindered imidazole ligand strategy to make four new five-
coordinate, high-spin species [Fe(Porph)(RIm)], with either
2-methylimidazole or 1,2-dimethylimidazole. The steric interac-
tion of the methyl group with the porphyrin core when using
the ligands 2-methylimidazole and 1,2-dimethylimidazole greatly
reduces K2 without significantly affecting K1.63 All of the
porphyrin derivatives were meso-substituted tetraarylporphyrins.
A summary of the structural parameters for the four new
examples is given as the first four entries of Table 2. Although
disorder in the position of the axial imidazole ligand has been
considered a major problem for this class of derivative, two of
While the spin Hamiltonian approximation (eq 5) led to
acceptable fits to the magnetic Mo¨ssbauer spectra of the models
for fields up to 9 T, the positive sign of D inferred from the fits
is at odds with the negative sign of the quadrupole splitting, to
the extent they are both explainable by a crystal field model
with an isolated ground orbital. Given that λ seems to be quite
small for these models, and that D (proportional to λ2/∆, where
∆ is an excited-state energy) is likely not to be exceptionally
small, we conclude that there must be low-lying excited states
which mix strongly with the ground dxz orbital. We note that
this conclusion is also supported by the strong temperature
dependence of the quadrupole splitting.
In summary, we can obtain good simultaneous fits to the field
dependence of the Mo¨ssbauer spectra for these models, but the
spin Hamiltonian parameters resulting are not consistent with
a simple crystal field model, and lead us to doubt the
applicability of the spin-Hamiltonian approximation. Proper
modeling of the iron state would thus require a larger crystal
field model which takes the excited states into account directly,
rather than through perturbation theory. The inadequacy of the
spin Hamiltonian treatment appears to be true for most, if not
(62) Scheidt, W. R.; Reed, C. A. Chem. ReV. 1981, 81, 543.
(63) Rougee, M.; Brault, D. Biochem. Biophys. Res. Commun. 1974, 57, 654.
9
J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005 5685

A R T I C L E S
Hu et al.
the derivatives have a single, ordered axial ligand and the other
two have a major orientation of the ligand. As expected, the
general structural features that are common to five-coordinate
iron(II) high-spin species are seen for these four structures.
These include an expanded porphyrin core where Ct‚‚‚N is 2.05
Å, long equatorial Fe-Np bond lengths where the averages range
from 2.076(3) Å to 2.087(7) Å, and moderately large iron atom
out-of-plane displacements (∆N4 > 0.32 Å). In addition, the
large size of the iron(II) ion has been thought to lead to a
significant C4V doming of the core. A measure of the doming is
given by the difference between the displacement of the iron
atom from the mean plane of the four nitrogen atoms (∆N4)
and the mean plane of the 24-atom core (∆). Also listed in
Table 2 are the values previously reported for the two forms of
[Fe(TPP)(2-MeHIm)].9,12 Averaged values for these six deriva-
tives are given on the next line. Values for other species are
also given in Table 2 for further discussion.
There are several necessary structural responses to binding a
hindered imidazole to form the five-coordinate iron porphyri-
nates. The overall effect required is to maximize the attractive
interaction between iron and its ligands while minimizing the
repulsive interaction between the 2-methyl group of the imi-
dazole ligand and the porphyrin plane. This is accomplished
by adjustments in the tilting of the Fe-N(Im) bond off of the
normal to the heme plane, the Fe-N(Im)-C(Im) angles, the
iron out-of-plane displacement, the Fe-N(Im) bond length, and
the orientation of imidazole plane with respect to the porphyrin
core. However, there is still considerable variation among these
geometric parameters.
Although these structural features are a result of the large
size of the high-spin iron(II) ion and the asymmetric steric bulk
of the axial ligand, an examination of the values in Table 2
shows that there is significant variation within the set of six
imidazole-ligated high-spin five-coordinate iron(II) derivatives.
While the radius of the central hole (Ct‚‚‚N) remains constant
at 2.048(2) Å, the value of the average Fe-Np bond distances
in the complexes varies from 2.073 Å up to 2.087 Å. The
displacement of the iron atom from the four nitrogen atom mean
plane shows a range of 0.32 to 0.42 Å while the iron
displacement from the 24-atom plane shows a larger range of
variation; the difference between the two displacements shows
a still larger variation. The difference in displacement between
the two planes range from as low as 0.03 Å ([Fe(Tp-OCH3-
PP)(1,2-Me2Im)]) to 0.13 Å ([Fe(TPP)(2-MeHIm)](two-fold)).
To understand these variations we first examine the atomic
displacements of the four new derivatives from the 24-atom
mean plane. This information is given in Figure 5. The core
conformation of the four derivatives are clearly very different.
They range from a “domed” conformation for [Fe(Tp-OCH3-
PP)(2-MeHIm)] to a “tented” conformation for [Fe(TTP)(2-
MeHIm)] to one pyrrole ring showing large displacement from
the mean plane ([Fe(TPP)(1,2-Me2Im)]) to a close-to-planar
conformation for [Fe(Tp-OCH3PP)(1,2-Me2Im)]. The orientation
of the imidazole ligands are also shown in Figure 5; there is no
apparent correlation between the core conformation and the
ligand orientation.
Figure 9. Diagrams illustrating the core conformation and iron displacement
for seven known imidazole-ligated high-spin iron(II) porphyrinates. The
displacement of the iron and the atoms of the porphyrin core from the mean
plane defined by the four pyrrole nitrogen atoms is given. The position of
the imidazole ligand with respect to directions defined by the Fe-Np
directions is shown.
derivatives along with a hybrid porphyrin system are illustrated
in Figure 9. Displayed in linear fashion are the displacements
of the iron atom and the 24 atoms of the porphyrin core from
the four pyrrole nitrogen atom mean plane. The structures are
illustrated with the derivative having the largest displacements
of the atoms from the plane at the top and then in the order of
decreasing displacements; the hybrid porphyrin structure is
shown last. The orientation of the imidazole ligand with respect
to the porphyrin core is also shown: the vertical line represents
the position of the imidazole shown with the pyrrole closest to
The four new derivatives have different core conformations,
but there are some similarities to the core conformations
previously observed for both crystalline forms of [Fe(TPP)(2-
MeHIm)]. Core conformations of the six tetraarylporphyrin
9
5686 J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005

High-Spin Imidazole-Ligated Iron(II) Porphyrinates
A R T I C L E S
the 2-methyl group. When there is a minor orientation this is
illustrated with a second vertical line; the major orientation is
always shown to the left. With the exception of the [Fe(TPP)-
(2-MeHIm)](two-fold) structure, which has two equivalent
orientations of the axial ligand, it is expected that the major
ligand orientation dominates the observed structure.
The results summarized in Figure 9 make evident that there
is not a single preferred conformation for imidazole-ligated high-
spin iron(II) complexes. Although doming of the porphyrin core,
which leads to a substantial displacement of the iron from the
N4 plane and a larger displacement from the 24-atom plane,
might be expected to best accommodate the large high-spin iron-
(II) center, a number of other conformations are also observed.
Variations in core conformations and iron atom displacement
notwithstanding, all derivatives have porphyrin cores in which
the radial expansion of the central hole is effectively constant
at 2.048 Å. These variations do lead a modest range of Fe-Np
bond distances from 2.073 to 2.087 Å. The low electronic
symmetry and close-lying states of iron may be reflected in the
high variability of the core conformations.
The strong similarity in the core conformations between the
“domed” conformation of [Fe(TPP)(2-MeHIm)](two-fold)9 and
that of [Fe(Tp-OCH3PP)(2-MeHIm)] is clearly seen in the top
two structures illustrated in Figure 9. These two structures have
a number of similarities in addition to core conformation. The
iron displacements from the N4 and 24-atom planes are the
largest for these two derivatives (0.42 and 0.55 Å for [Fe(TPP)-
(2-MeHIm)](two-fold), and 0.39 and 0.51 Å for [Fe(Tp-OCH3-
PP)(2-MeHIm)]). Concomitant with the large displacement of
the iron, the average Fe-Np distances (2.086 and 2.087 Å) are
also the largest for these two derivatives.
The ORTEP drawings of Figures 1-4 clearly show that the
axial Fe-NIm bond vector is tilted from the heme normal. The
tilt angle ranges from 6.1 to 11.4°. This tilting is the partial
result of minimizing the interaction between the bulky imidazole
methyl group and the porphyrin core; unequal Fe-N-C angles
in the imidazole also serve to minimize the steric interactions.
As seen in Table 2, these angles have an average value of 131.2°
on the methyl side and 123.0° on the other side. However, these
structural features are also seen in [Fe(Piv2C8P)(1-MeIm)], albeit
with smaller values, suggesting that the steric interactions are
not the sole factor in the asymmetric binding of the imidazole
ligand. The small differences in ligand tilting appears not to be
correlated with any other structural feature, showing no apparent
relation to varying core conformation or the orientation of the
axial ligand plane with respect to Fe-Np directions.
The Fe-N(imidazole) bond lengths for the four structures
show a small but significant variation. The small variation is
no doubt an accommodation to the hindering methyl group and
differences in relative orientation of the axial ligand and differing
core conformations. All Fe-N(imidazole) values are nominally
longer than the one value for an unhindered imidazole bound
to an iron(II) ([Fe(Piv2C8P)(1-MeIm)]), but the differences are
statistically not significant. The rather short distance observed
in [Fe(TpivPP)(2-MeHIm)] may be artifactual; the rigid group
refinement parameters used to describe the disordered imidazole
had an unusually small C-N-C angle.
The core conformations of the second form of [Fe(TPP)(2-
MeHIm)] and [Fe(TTP)(2-MeHIm)] also display similarities as
can be seen from the third and fourth entries of Figure 9. The
unusual feature first noted in [Fe(TPP)(2-MeHIm)]12 that the
core conformation asymmetry apparently induced by the axial
ligand leads to unequal Fe-Np bond distances is also seen in
the tetratolyl derivative. The Fe-Np bond distances appear to
fall into two distinct categories. In both derivatives, the two
pyrrole rings that bracket the 2-methyl group are tilted down-
ward. The Fe-Np bonds to these two pyrrole rings appear to
be shortened while the opposite two are lengthened. The pattern
of distances can be seen in Figure 5d. In Fe(TPP)(2-MeHIm)]
the two pairs have values of 2.066(4) and 2.081(3) Å, a
statistically significant difference, whereas the pair values of
2.072(3) and 2.080(4) Å in [Fe(TTP)(2-MeHIm)] are somewhat
less significant. These two derivatives have the smallest iron
atom displacements from the N4 plane (0.32 Å) and very modest
amounts of core doming. The overall effect of these various
conformational features lead to overall average Fe-Np bond
distances that are shorter than those seen for the domed
conformers.
The last two new compounds, both 1,2-dimethylimidazole
derivatives, have porphyrin cores that have only one pyrrole
ring that is significantly tilted from the N4 mean plane with the
remaining three rings nearly exactly planar. These are displayed
as the fifth and sixth compounds in Figure 9. The tilted pyrrole
ring is, not unexpectedly, the ring closest to the bulky methyl
group of the imidazole. The magnitude of the iron displacement
from the N4 plane has increased slightly from those of the
previous set (average ∆N4 value 0.35 Å).
The final derivative illustrated in Figure 9 is the only known
five-coordinate high-spin iron(II) derivative with an unhindered
imidazole ligand.32 To achieve this, a hybrid porphyrin with
both a strap and two pickets, all on one side of the porphyrin
plane, must be used to allow access to only one face of the
porphyrin. This is the only known high-spin iron(II) complex
with a strongly ruffled core. It is most likely that the trans
superstructure strap leads to the core ruffling. Although the N4
and 24-atom planes are nearly coincident and core ruffling
usually leads to shortened M-Np bonds, the iron atom dis-
placement and Fe-Np distances appear to be unaffected and
are not outside the range of the other derivatives described
herein.
Summary. The extensive Mo¨ssbauer studies undertaken in
this investigation clearly show that all imidazole-ligated high-
spin iron(II) porphyrinates have a negative value for the
quadrupole splitting and the most likely ground electronic
2
1
1
1
1
2
2
2
configuration of all is the (dxz) (dyz) (dxy) (dz ) (dx - y ) config-
uration. Although we failed to achieved absolute conclusions
concerning the sign of the zero-field splitting constant D, we
have shown that an apparent dichotomy in the sign between
the model compounds and the heme protein systems is illusory.
The electronic state of all these derivatives is of extremely low
symmetry. For all species there are close-lying excited states
of unknown spin multiplicity. High-spin iron(II) porphyrinates
can be divided into two classes of electronic configuration with
significantly different geometric structures. The importance of
these features for heme protein functionality remains unknown
but intriguing. The structural characterization of four deoxy-
hemoglobin models has been reported. All derivatives have an
iron atom displacement of ∼0.36 Å from the four nitrogen atom
plane with larger displacements and larger variation in displace-
ment from the 24-atom plane reflecting core conformation
9
J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005 5687

A R T I C L E S
Hu et al.
differences. Although there are a number of differing core
conformations observed, there are some strong similarities with
all falling into perhaps three classes of conformation.
derived crystal field parameters obtained; Figure S1 showing
the two disordered axial imidazoles in [Fe(Tp-OCH3PP)(2-
MeHIm)]; Figure S2 giving the same information for [Fe(TPP)-
(1,2-Me2Im)]; Figure S3 showing the displacement of the core
atoms from the four nitrogen atom plane; Figures S4-S9 giving
illustrations of the Mo¨ssbauer spectra and fits in 3-, 6-, and
9-T applied magnetic field for [Fe(TPP)(2-MeHIm)], [Fe(TPP)-
(1,2-Me2Im)], [Fe(TTP)(2-MeHIm)], [Fe(TTP)(1,2-Me2Im)],
[Fe(Tp-OCH3PP)(2-MeHIm)], and [Fe(Tp-OCH3PP)(1,2-Me2-
Im)], respectively (PDF). An X-ray crystallographic file, in CIF
format, is available. This material is available free of charge
via the Internet at http://pubs.acs.org.
Acknowledgment. We thank the National Institutes of Health
for support of this research under Grant GM-38401. We thank
Dr. Bruce Noll and A. Beatty for assistance with X-ray data
collection.
Supporting Information Available: Tables S1-S24, giving
complete crystallographic details, atomic coordinates, bond
distances and angles, anisotropic temperature factors, and fixed
hydrogen atom positions for all four structures; Table S25,
giving complete values of Mo¨ssbauer quadrupole splitting and
isomer shift as a function of temperature; Table S26 giving the
JA044077P
9
5688 J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005