Correlation of spin states and spin delocalization with the dioxygen reactivity of catecholatoiron(III) complexes

Article abstract of DOI:10.1021/ic051173y

A series of catecholatoiron(III) complexes, [Fe^IIIL(4Cl-cat)] BPh₄ (L = (4-MeO)₂TPA (1), TPA (2), (4-Cl)₂TPA (3), (4-NO₂)TPA (4), (4-NO₂)₂TPA (5); TPA = tris(pyridin-2-ylmethyl)amine; 4Cl-cat = 4-chlorocatecholate), have been characterized by magnetic susceptibility measurements and EPR, ¹H NMR, and UV-vis-NIR spectroscopies to clarify the correlation of the spin delocalization on the catecholate ligand with the O₂ reactivity as well as the spin-state dependence of the O₂ reactivity. EPR spectra in frozen CH₃CN at 123 K clearly showed that introduction of electron-withdrawing groups effectively shifts the spin equilibrium from a high-spin to a low-spin state. The effective magnetic moments determined by the Evans method in a CH₃CN solution showed that 5 contains 36% of low-spin species at 243 K, while 1-4 are predominantly in a high-spin state. Evaluation of spin delocalization on the 4Cl-cat ligand by paramagnetic ¹H NMR shifts revealed that the semiquinonatoiron(II) character is more significant in the low-spin species than in the high-spin species. The logarithm of the reaction rate constant is linearly correlated with the energy gap between the catecholatoiron(III) and semiquinonatoiron(II) states for the high-spin complexes 1-3, although complexes 4 and 5 deviate negatively from linearity. The lower reactivity of the low-spin complex, despite its higher spin density on the catecholate ligand compared with the high-spin analogues, suggests the involvement of the iron(III) center, rather than the catecholate ligand, in the reaction with O₂.

Full text of DOI:10.1021/ic051173y

Inorg. Chem. 2005, 44, 8810−8821
Correlation of Spin States and Spin Delocalization with the Dioxygen
Reactivity of Catecholatoiron(III) Complexes
Masakazu Higuchi,† Yutaka Hitomi,*,† Hisataka Minami,† Tsunehiro Tanaka,† and Takuzo Funabiki*,‡
Department of Molecular Engineering, Graduate School of Engineering, Kyoto UniVersity,
Katsura, Nishikyo-ku, Kyoto 615-8510, Japan, and Biomimetics Research Center, Doshisha
UniVersity, Kyo-Tanabe, Kyoto 610-0321, Japan
Received July 13, 2005
A series of catecholatoiron(III) complexes, [FeIIIL(4Cl-cat)]BPh4 (L
(4-NO2)TPA (4), (4-NO2)2TPA (5); TPA tris(pyridin-2-ylmethyl)amine; 4Cl-cat
characterized by magnetic susceptibility measurements and EPR, 1H NMR, and UV
)
(4-MeO)2TPA (1), TPA (2), (4-Cl)2TPA (3),
4-chlorocatecholate), have been
vis NIR spectroscopies to
)
)
−
−
clarify the correlation of the spin delocalization on the catecholate ligand with the O2 reactivity as well as the
spin-state dependence of the O2 reactivity. EPR spectra in frozen CH3CN at 123 K clearly showed that introduction
of electron-withdrawing groups effectively shifts the spin equilibrium from a high-spin to a low-spin state. The
effective magnetic moments determined by the Evans method in a CH3CN solution showed that 5 contains 36%
of low-spin species at 243 K, while 1−4 are predominantly in a high-spin state. Evaluation of spin delocalization
on the 4Cl-cat ligand by paramagnetic 1H NMR shifts revealed that the semiquinonatoiron(II) character is more
significant in the low-spin species than in the high-spin species. The logarithm of the reaction rate constant is
linearly correlated with the energy gap between the catecholatoiron(III) and semiquinonatoiron(II) states for the
high-spin complexes 1−3, although complexes 4 and 5 deviate negatively from linearity. The lower reactivity of the
low-spin complex, despite its higher spin density on the catecholate ligand compared with the high-spin analogues,
suggests the involvement of the iron(III) center, rather than the catecholate ligand, in the reaction with O2.
Introduction
are thought to play a role in lowering the redox potential to
prevent the reduction of the ferric center under biological
conditions.10 When catechol coordinates to the FeIII center
in a chelate fashion, one tyrosine residue is released to form
a distorted octahedral geometry bearing a vacant site. This
intermediate species reacts with O2 to give an intradiol-
cleaving product, cis,cis-muconic acid. It is noteworthy that
the active site iron retains the high-spin ferric oxidation state
during the catalytic cycle.10-17
Catechol dioxygenases are mononuclear nonheme enzymes
playing a key role in the metabolism of aromatic compounds
such as lignin and halogenated benzene in soil bacteria.1-4
They catalyze the C-C bond cleavage of the catechol ring
and are subclassified into two classes: intradiol and extradiol
catechol dioxygenases. Intradiol catechol dioxygenases utilize
an FeIII ion as an active center, whereas extradiol catechol
dioxygenases require an FeII ion or rarely a MnII ion. In
intradiol oxygenases, the FeIII ion is ligated by two tyrosine
and two histidine residues and water (OH-) in a distorted
trigonal bipyramidal geometry.5-9 The two tyrosine residues
Functional model complexes for catechol oxygenases have
been developed for evaluation of the reaction mechanisms
and development of catalytic systems for catechol oxygena-
(5) Ohlendorf, D. H.; Lipscomb, J. D.; Weber, P. C. Nature 1988, 336,
403-405.
* To whom correspondence should be addressed. E-mail: hitomi@
moleng.kyoto-u.ac.jp.
(6) Ohlendorf, D. H.; Orville, A. M.; Lipscomb, J. D. J. Mol. Biol. 1994,
244, 586-608.
† Kyoto University.
‡ BMRC, Doshisha University.
(7) Vetting, M. W.; Ohlendorf, D. H. Structure (Cambridge, MA) 2000,
8, 429-440.
(1) Funabiki, T. Dioxygenases; Funabiki, T., Ed.; Kluwer Academic
Publishers: Dordrecht, The Netherlands, 1997; pp 19-104.
(2) Bugg, T. D. H.; Winfield, C. J. Nat. Prod. Rep. 1998, 15, 513-530.
(3) Solomon, E. I.; Brunold, T. C.; Davis, M. I.; Kemsley, J. N.; Lee, S.
K.; Lehnert, N.; Neese, F.; Skulan, A. J.; Yang, Y. S.; Zhou, J. Chem.
ReV. 2000, 100, 235-349.
(8) Vetting, M. W.; D’Argenio, D. A.; Ornston, L. N.; Ohlendorf, D. H.
Biochemistry 2000, 39, 7943-7955.
(9) Ferraroni, M.; Solyanikova, I. P.; Kolomytseva, M. P.; Scozzafava,
A.; Golovleva, L.; Briganti, F. J. Biol. Chem. 2004, 279, 27646-
27655.
(4) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que, L., Jr. Chem. ReV. 2004,
104, 939-986.
(10) Pyrz, J. W.; Roe, A. L.; Stern, L. J.; Que, L., Jr. J. Am. Chem. Soc.
1985, 107, 614-620.
8810 Inorganic Chemistry, Vol. 44, No. 24, 2005
10.1021/ic051173y CCC: $30.25
© 2005 American Chemical Society
Published on Web 10/13/2005

Catecholatoiron(III) Complexes
tions.4,18-22 In the reaction of catecholatoiron(III) complexes
with O2, the importance of participation of the semiquinona-
toiron(II) form has been proposed19,23,24 and supported by
various experimental25-31 and theoretical results.32,33 It was
suggested that the ground state of the catecholatoiron(III)
complexes is represented as a quantum mechanical admixture
of FeIII-Cat and FeII-SQ states, rather than an equilibrium
between them.25,27,28,34 On the basis of this FeII-SQ character
of the FeIII-Cat complex, two different paths have been
proposed for the reaction with O2 (Scheme 1); path a, the
substrate actiVation path, is the one in which O2 reacts first
with the catechol moiety to give an FeII-R-OO• form,27,35-38
and path b, the oxygen actiVation path, is the one in which
Scheme 1. Two Possible Reaction Mechanisms for Intradiol Catechol
Dioxygenasesa
•
O2 reacts first with an iron moiety to give an OO-FeIII-
SQ form.33,39 Path a has been favored, but data supporting
path b has recently been obtained.26,40,41
a Paths a and b show the substrate activation mechanism and the oxygen
activation mechanism, respectively.
The iron(III) center of intradiol catechol dioxygenases is
in a high-spin state, and the functional models so far
developed also contain a high-spin ferric center at ambient
temperatures.22 However, the reactivity of low-spin catecho-
latoiron(III) complexes should provide valuable information
on the reaction mechanism.29 Recently, it has been shown
that the spin state of catecholatoiron(III) complexes changes
from high-spin to low-spin in the solid state with lowering
temperatures25,42,43 and that the substituents of the catecholate
ligand greatly affect the spin-crossover transition.25 The spin-
crossover behavior has been well documented for ferric
complexes with an N4O2 donor set in the solid state,44-49
but it has been little documented in solution.50,51
We have synthesized a series of catecholatoiron(III)
complexes, [FeIIIL(4Cl-cat)]BPh4 (L ) (4-MeO)2TPA (1),
TPA (2), (4-Cl)2TPA (3), (4-NO2)TPA (4), (4-NO2)2TPA
(5)),52 to investigate the substituent effects of the supporting
ligand on spin delocalization and the O2 reactivity.53 It was
found that the introduction of electron-withdrawing groups
on the TPA ligand not only induces spin delocalization on
(11) Que, L., Jr.; Lipscomb, J. D.; Zimmermann, R.; Muenck, E.; Orme-
Johnson, N. R.; Orme-Johnson, W. H. Biochim. Biophys. Acta 1976,
452, 320-334.
(12) Whittaker, J. W.; Lipscomb, J. D.; Kent, T. A.; Muenck, E. J. Biol.
Chem. 1984, 259, 4466-4475.
(13) Kent, T. A.; Mu¨nck, E.; Pyrz, J. W.; Widom, J.; Que, L. Inorg. Chem.
1987, 26, 1402-1408.
(14) Que, L., Jr.; Heistand, R. H., II. J. Am. Chem. Soc. 1979, 101, 1,
2219-2221.
(15) Walsh, T. A.; Ballou, D. P.; Mayer, R.; Que, L., Jr. J. Biol. Chem.
1983, 258, 14422-14427.
(16) Bull, C.; Ballou, D. P.; Otsuka, S. J. Biol. Chem. 1981, 256, 2681-
2686.
(17) Felton, R. H.; Cheung, L. D.; Phillips, R. S.; May, S. W. Biochem.
Biophys. Res. Commun. 1978, 85, 844-850.
(18) Funabiki, T.; Sakamoto, H.; Yoshida, S.; Tarama, K. J. Chem. Soc.,
Chem. Commun. 1979, 754-755.
(19) Funabiki, T.; Mizoguchi, A.; Sugimoto, T.; Tada, S.; Tsuji, M.;
Sakamoto, H.; Yoshida, S. J. Am. Chem. Soc. 1986, 108, 2921-2932.
(20) Funabiki, T. Iron Model Studies on Dioxygenases; Funabiki, T., Ed.;
Kluwer Academic Publishers: Dordrecht, The Netherlands, 1997; pp
105-156.
(21) Funabiki, T. Functional Model Oxygenations by Nonheme Iron
Complexes; Simandi, L. I., Ed.; Kluwer: Dordrecht, The Netherlands,
2003; pp 157-226.
(22) Yamahara, R.; Ogo, S.; Masuda, H.; Watanabe, Y. J. Inorg. Biochem.
2002, 88, 284-294 and references therein.
(23) Funabiki, T.; Tada, S.; Yoshioka, T.; Takano, M.; Yoshida, S. J. Chem.
Soc., Chem. Commun. 1986, 1699-1700.
(38) Fujii, S.; Ohyanishiguchi, H.; Hirota, N.; Nishinaga, A. Bull. Chem.
Soc. Jpn. 1993, 66, 1408-1419.
(24) Funabiki, T.; Konishi, T.; Kobayashi, S.; Mizoguchi, A.; Takano, M.;
Yoshida, S. Chem. Lett. 1987, 719-722.
(39) Funabiki, T.; Kojima, H.; Kaneko, M.; Inoue, T.; Yoshioka, T.; Tanaka,
T.; Yoshida, S. Chem. Lett. 1991, 2143-2146.
(25) Funabiki, T.; Fukui, A.; Hitomi, Y.; Higuchi, M.; Yamamoto, T.;
Tanaka, T.; Tani, F.; Naruta, Y. J. Inorg. Biochem. 2002, 91, 151-
158.
(40) Jo, D.-H.; Que, L., Jr. Angew. Chem., Int. Ed. 2000, 39, 4284-4287.
(41) Hitomi, Y.; Tase, Y.; Higuchi, M.; Tanaka, T.; Funabiki, T. Chem.
Lett. 2004, 33, 316-317.
(26) Hitomi, Y.; Yoshida, M.; Higuchi, M.; Minami, H.; Tanaka, T.;
Funabiki, T. J. Inorg. Biochem. 2005, 99, 755-763.
(27) Cox, D. D.; Que, L., Jr. J. Am. Chem. Soc. 1988, 110, 8085-8092.
(28) Jang, H. G.; Cox, D. D.; Que, L., Jr. J. Am. Chem. Soc. 1991, 113,
9200-9204.
(42) Simaan, A. J.; Boillot, M.-L.; Rivie`re, E.; Boussac, A.; Girerd, J.-J.
Angew. Chem., Int. Ed. 2000, 39, 196-198.
(43) Floquet, S.; Simaan, A. J.; Rivie`re, E.; Nierlich, M.; Thue´ry, P.;
Ensling, J.; Gu¨tlich, P.; Girerd, J. J.; Boillot, M. L. Dalton Trans.
2005, 1734-1742.
(29) Mialane, P.; Tchertanov, L.; Banse, F.; Sainton, J.; Girerd, J. J. Inorg.
Chem. 2000, 39, 2440-2444.
(44) Haddad, M. S.; Lynch, M. W.; Federer, W. D.; Hendrickson, D. N.
Inorg. Chem. 1981, 20, 123-131.
(30) Viswanathan, R.; Palaniandavar, M.; Balasubramanian, T.; Muthiah,
T. P. Inorg. Chem. 1998, 37, 2943-2951.
(45) Ohshio, H.; Maeda, Y.; Takashima, Y. Inorg. Chem. 1983, 22, 2684-
2689.
(31) Pascaly, M.; Duda, M.; Schweppe, F.; Zurlinden, K.; Muller, F. K.;
Krebs, B. Dalton Trans. 2001, 828-837.
(46) Timken, M. D.; Hendrickson, D. N.; Sinn, E. Inorg. Chem. 1985, 24,
3947-3955.
(32) Funabiki, T.; Inoue, T.; Kojima, H.; Konishi, T.; Tanaka, T.; Yoshida,
S. J. Mol. Catal. 1990, 59, 367-371.
(47) Timken, M. D.; Strouse, C. E.; Soltis, S. M.; Daverio, S. A.;
Hendrickson, D. N.; Abdelmawgoud, A. M.; Wilson, S. R. J. Am.
Chem. Soc. 1986, 108, 395-402.
(33) Funabiki, T.; Yamazaki, T. J. Mol. Catal. A 1999, 150, 37-47.
(34) Simaan, A. J.; Boillot, M.-L.; Carrasco, R.; Cano, J.; Girerd, J.-J.;
Mattioli, T. A.; Ensling, J.; Spiering, H.; Gu¨tlich, P. Chem.sEur. J.
2005, 11, 1779-1793.
(48) Oshio, H.; Toriumi, K.; Maeda, Y.; Takashima, Y. Inorg. Chem. 1991,
30, 4252-4260.
(35) Que, L., Jr.; Lipscomb, J. D.; Muenck, E.; Wood, J. M. Biochim.
Biophys. Acta 1977, 485, 60-74.
(49) Boca, R.; Fukuda, Y.; Gembicky, M.; Herchel, R.; Jarosciak, R.; Linert,
W.; Renz, F.; Yuzurihara, J. Chem. Phys. Lett. 2000, 325, 411-419.
(50) Wilson, L. J.; Tweedle, M. F. J. Am. Chem. Soc. 1976, 98, 4824-
4833.
(36) Que, L.; Kolanczyk, R. C.; White, L. S. J. Am. Chem. Soc. 1987,
109, 5373-5380.
(37) Que, L.; Lauffer, R. B.; Lynch, J. B.; Murch, B. P.; Pyrz, J. W. J.
Am. Chem. Soc. 1987, 109, 5381-5385.
(51) Petty, R. H.; Dose, E. V.; Tweedle, M. F.; Wilson, L. J. Inorg. Chem.
1978, 17, 1064-1070.
Inorganic Chemistry, Vol. 44, No. 24, 2005 8811

Higuchi et al.
and 2-chloromethyl-4-nitropyridine (1:1) in a manner analogous
to that used to prepare (4-MeO)2TPA; it was isolated as light yellow
oil in a 41% yield. H NMR (400 MHz, CDCl3): δ 3.94 (6H, s),
4.06 (2H, s), 7.16 (2H, dd, J ) 7.3 Hz, J′ ) 4.9 Hz), 7.54 (2H, d,
J ) 7.3 Hz), 7.68 (2H, dd, J ) 7.3 Hz, J′ ) 7.3 Hz), 7.84 (1H, dd,
J ) 5.4 Hz, J′ ) 2.4 Hz), 8.35 (1H, d, J ) 2.4 Hz), 8.55 (2H, d,
J ) 4.9 Hz,), 8.79 (1H, d, J ) 5.4 Hz). 13C NMR (100 MHz,
CDCl3): δ 59.7, 60.5, 114.3, 115.3, 122.1, 123.0, 136.5, 149.1,
151.0, 154.3, 158.5, 163.7.
the catecholate ligand but also shifts the spin equilibrium to
a low-spin state. The logarithm of the reaction rate constant
with O2 is linearly correlated with the energy gap between
catecholatoiron(III) and semiquinonatoiron(II) states for the
high-spin complexes 1-3, although 5, which contains 36%
of low-spin species at 243 K, deviates negatively from
linearity. In this report, we elucidated the spin delocalization
on the catecholate ligand of high-spin and low-spin catecho-
latoiron(III) species by means of 1H NMR, EPR, and UV-
vis-NIR spectroscopies to clarify the correlation of spin
states and spin delocalization with the O2 reactivity of
catecholatoiron(III) complexes.
1
Bis(4-nitroypyridin-2-ylmethyl)(pyridin-2-ylmethyl)amine, (4-
NO2)2TPA. (4-NO2)2TPA was prepared using 4-chloro-2-nitro-
methylpyridine in a manner analogous to that used to preprare (4-
1
MeO)2TPA; it was isolated as light yellow oil in a 66% yield. H
NMR (400 MHz, CDCl3): δ 3.98 (2H, s), 4.12 (4H, s), 7.12 (1H,
ddd, J ) 7.3 Hz, J′ ) 4.9 Hz, J′′ ) 1.0 Hz), 7.49 (1H, dd, J ) 7.3
Hz, J′ ) 1.0 Hz), 7.69 (1H, ddd, J ) 7.3 Hz, J′ ) 7.3 Hz, J′′ )
1.0 Hz), 7.88 (2H, dd, J ) 5.4 Hz, J′ ) 2.0 Hz), 8.31 (2H, d, J )
2.0 Hz), 8.57 (1H, dd, J ) 4.9 Hz, J′ ) 1.0 Hz), 8.84 (2H, d, J )
5.4 Hz). 13C NMR (100 MHz, CDCl3): δ 59.9, 60.6, 114.6, 115.5,
122.4, 123.1, 136.6, 149.3, 151.3, 154.3, 157.9, 162.8.
General Procedure for the Synthesis of Catecholatoiron(III)
Complexes. Substituted TPA (0.30 mmol) in methanol (2 mL) and
4Cl-catH2 (0.043 g, 0.30 mmol) in methanol (2 mL) under N2 were
added to a stirred solution of FeCl3 (0.049 g, 0.30 mmol) in
methanol (2 mL). After the addition of Et3N (83.6 µL, 0.60 mmol),
the reaction mixture was stirred for 10 min. NaBPh4 (0.10 g, 0.30
mmol) was added to the resulting dark purple blue solution, and
the mixture was stirred for 2 h, causing the immediate precipitation
of a powder, which was washed with methanol several times and
dried in vacuo.
[FeIII((4-MeO)2TPA)(4Cl-cat)]BPh4 (1). Complex 1 was iso-
lated as a gray powder in an 83% yield. ESI-MS: m/z 548.1
([FeIII((4-MeO)2TPA)(4Cl-cat)]+). Anal. Calcd for C50H45BCl-
FeN4O4: C, 69.18; H, 5.23; N, 6.45; Cl, 4.08. Found: C, 68.93;
H, 5.20; N. 6.45; Cl, 4.07. UV-vis-NIR (CH3CN, λmax, nm (ꢀ,
M-1cm-1)): 474 (2006), 764 (3038).
[FeIII(TPA)(4Cl-cat)]BPh4 (2). Complex 2 was isolated as a dark
blue powder in a 98% yield. ESI-MS: m/z 488.1 ([FeIII(TPA)(4Cl-
cat)]+). Anal. Calcd for C48H41BClFeN4O4: C, 71.35; H, 5.11; N,
6.93; Cl, 4.39. Found: C, 71.21; H, 4.94; N. 6.92; Cl, 4.15. UV-
vis-NIR (CH3CN, λmax, nm (ꢀ, M-1cm-1)): 483 (1930), 787
(2974).
Experimental Section
Materials and Preparation of Ligands and Complexes.
Reagents and solvents used were commercially available unless
otherwise stated. 4Cl-catH2 was purified by sublimation in vacuo.
The TPA ligand, 2-chloromethyl-4-methoxypyridine, 4-chloro-2-
chloromethylpyridine, and 2-chloromethyl-4-nitropyridine were
prepared using the method reported previously.54,55
Synthesis of TPA Derivatives. Bis(4-methoxypyridin-2-yl-
methyl)(pyridin-2-ylmethyl)amine, (4-MeO)2TPA. 2-Chloro-
methyl-4-methoxypyridine (0.58 g, 3.7 mmol) and 10 N NaOH (aq)
(0.8 mL, 8.0 mmol) were added to a stirred solution of 2-picolyl-
amine (0.19 g, 1.8 mmol) in water (1 mL). The reaction mixture
was stirred at room temperature for 3 days. The organic layer was
separated, and the aqueous layer was extracted with CH2Cl2 several
times. The combined organic layers were dried over Na2SO4, and
the volatile components were removed. Purification of the crude
product on an alumina column (3:1 AcOEt/CH2Cl2) gave 0.39 g
(60%) of yellow oil. 1H NMR (400 MHz, CDCl3): δ 3.85 (6H, s),
3.86 (4H, s), 3.90 (2H, s), 6.67 (2H, dd, J ) 5.9 Hz, J′ ) 2.4 Hz),
7.14 (1H, dd, J ) 7.8 Hz, J′ ) 4.9 Hz), 7.20 (2H, d, J ) 2.4 Hz),
7.56 (1H, d, J ) 7.8 Hz), 7.64 (1H, ddd, J ) 7.8 Hz, J′ ) 7.8 Hz
J′′ ) 2.0 Hz), 8.34 (2H, d, J ) 5.9 Hz), 8.53 (1H, dd, J ) 4.9 Hz,
J′ ) 2.0 Hz). 13C NMR (100 MHz, CDCl3): δ 55.1, 60.2, 60.3,
108.3, 108.6, 121.9, 122.9, 136.3, 149.0, 150.2, 159.2, 161.1, 166.1.
Bis(4-chloropyridin-2-ylmethyl)(2-pyridylmethyl)amine, (4-
Cl)2TPA. (4-Cl)2TPA was prepared using 4-chloro-2-chlorometh-
ylpyridine in a manner analogous to that used to prepare (4-
1
MeO)2TPA; it was isolated as light yellow oil in a 66% yield. H
[FeIII((4-Cl)2TPA)(4Cl-cat)]BPh4 (3). Complex 3 was isolated
as a dark blue powder in a 75% yield. ESI-MS: m/z 556.2 ([FeIII((4-
Cl)2TPA)(4Cl-cat)]+). Anal. Calcd for C50H45BClFeN4O4: C, 65.75;
H, 4.48; N, 6.39; Cl, 12.13. Found: C, 65.51; H, 4.31; N. 6.41; Cl,
11.87. UV-vis-NIR (CH3CN, λmax, nm (ꢀ, M-1cm-1)): 496
(1916), 805 (2953).
NMR (400 MHz, CDCl3): δ 3.89 (4H, s), 3.91 (2H, s), 7.18-7.15
(3H), 7.51 (1H, d, J ) 7.8 Hz), 7.56 (2H, d, J ) 2.0 Hz), 7.68
(1H, ddd, J ) 7.8 Hz, J′ ) 7.8 Hz, J′′ ) 2.0 Hz), 8.43 (2H, d, J
) 5.4 Hz), 8.56 (1H, d, J ) 4.4 Hz). 13C NMR (100 MHz,
CDCl3): δ 59.9, 60.4, 122.2, 122.5, 123.0, 123.2, 136.5, 144.5,
149.1, 149.6, 158.5, 160.9.
[FeIII((4-NO2)TPA)(4Cl-cat)]BPh4 (4). Complex 4 was isolated
as a purple powder in a 63% yield. ESI-MS: m/z 533.3 ([FeIII((4-
NO2)TPA)(4Cl-cat)]+). Anal. Calcd for C48H42BClFeN5O5 (4‚
H2O): C, 65.85; H, 4.86; N, 8.04; Cl, 4.07. Found: C, 65.84; H,
Bis(pyridin-2-ylmethyl)(4-nitroypyridin-2-ylmethyl)amine, (4-
NO2)TPA. (4-NO2)TPA was prepared using 2,2′-dipicolylamine
(52) Abbreviations used in this paper are DTBC, 3,5-di-tert-butylcatecho-
late; 4Cl-cat, 4-chlorocatecholate; cat, pyrocatecholate; Cat, catecho-
late; SQ, semiquinonate; LMCT, ligand-to-metal charge transfer; (4-
MeO)2TPA, bis(4-methoxypyridin-2-ylmethyl)(pyridin-2-ylmethyl)amine;
TPA, tris(pyridin-2-ylmethyl)amine; (4-Cl)2TPA, bis(4-chloropyridin-
2-ylmethyl)(pyridin-2-ylmethyl)amine; (4-NO2)TPA, (4-nitropyridin-
2-ylmethyl)bis(pyridin-2-ylmethyl)amine; (4-NO2)2TPA, bis(4-nitro-
pyridin-2-ylmethyl)(pyridin-2-ylmethyl)amine; BLPA, bis((6-methyl-
2-pyridyl)methyl)(2-pyridylmethyl)amine; BBA, bis(benzimidazolyl-
2-methyl)amine.
4.78; N. 8.11; Cl, 4.28. UV-vis-NIR (CH3CN, λmax, nm (ꢀ, M-1
-
cm-1)): 498 (2038), 815 (2666).
[FeIII((4-NO2)2TPA)(4Cl-cat)]BPh4 (5). Complex 5 was isolated
as a purple-blue powder in a 77% yield. ESI-MS: m/z 577.9
([FeIII((4-NO2)2TPA)(4Cl-cat)]+). Anal. Calcd for C49H43BClFeN6O7
(5‚MeOH): C, 63.28; H, 4.66; N, 9.04; Cl, 3.81. Found: C, 63.29;
H, 4.55; N. 9.31; Cl, 3.77. UV-vis-NIR (CH3CN, λmax, nm (ꢀ,
M-1cm-1)): 509 (2408), 856 (1821).
(53) Hitomi, Y.; Higuchi, M.; Minami, H.; Tanaka, T.; Funabiki, T. Chem.
Commun. 2005, 1758-1760.
Crystallographic Data Collection and Refinement of the
Structure. All measurements were performed on a Rigaku Mercury
charge-coupled device with graphite-monochromated Mo KR (λ
(54) Tamura, M.; Urano, Y.; Kikuchi, K.; Higuchi, T.; Hirobe, M.; Nagano,
T. Chem. Pharm. Bull. 2000, 48, 1514-1518.
(55) Gafford, B. G.; Holwerda, R. A. Inorg. Chem. 1989, 28, 60-66.
8812 Inorganic Chemistry, Vol. 44, No. 24, 2005

Catecholatoiron(III) Complexes
Table 1. Crystallographic and Refinement Data for 2
emprical formula
fw
color
cryst syst
space group
a (Å)
C48H41BClFeN4O2
807.99
purple
monoclinic
P21/c (No. 14)
14.007(2)
9.994(2)
28.419(4)
90
97.396(4)
90
4
3945(1)
1.360
4.96
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
Z
V (Å3)
D
calcd (g cm-3
)
µ (cm-1
temp (K)
)
243.2
λ(Mo KR) (Å)
no. of reflns
no. of params
reflns/params ratio
GOF
0.71069
5046
514
9.82
1.141
Figure 1. ORTEP plot of the [FeIII(TPA)(4Cl-cat)]+ cation in 2. Ellipsoids
are drawn at the 50% probability level. Hydrogens have been omitted for
clarity.
Rint
R, Rw
0.038
0.058, 0.090
b
a I > 3.0σ(I). b R ) ∑||Fo|-|Fc||/∑|Fo|. Rw ) [∑w(|Fo|-|Fc|)2/
∑w|Fo| ]1/2
.
2
Table 2. Selected Bond Lengths (Å) and Angles (deg) for
[FeIII(TPA)(4Cl-cat)]BPh4 (2)
) 0.71069 Å) radiation at 243 K. Crystallographic data are given
in Table 1. The structure was solved by standard direct methods
(the Crystal-Structure crystallographic software package of the
Molecular Structure Corp. and Rigaku). Full-matrix least-squares
refinements (SHELXL-97)56 were carried out with anisotropic
thermal parameters for all non-hydrogen atoms.
Fe1-O1
Fe1-O2
Fe1-N1
Fe1-N2
1.917(2)
1.925(2)
2.195(2)
2.114(2)
Fe1-N3
Fe1-N4
C1-O1
C2-O2
2.138(3)
2.139(2)
1.352(3)
1.351(3)
N1-Fe1-N2
N1-Fe1-N3
N1-Fe1-N4
N1-Fe1-O2
O1-Fe1-N2
O1-Fe1-N3
O1-Fe1-N4
80.21(8)
76.68(9)
76.35(8)
100.50(8)
93.67(8)
100.18(10)
106.46(9)
O1-Fe1-O2
O2-Fe1-N3
O2-Fe1-N4
N1-Fe1-O1
N2-Fe1-O2
N3-Fe1-N4
85.49(8)
90.26(9)
96.17(9)
173.17(9)
177.51(9)
152.98(9)
Physical Methods. Elemental analyses were performed at the
Microanalytical Center of Kyoto University. Electrospray ionization
mass spectrometry (ESI-MS) spectra were recorded on a PE SCIEX
API 2000 mass spectrometer. 1H and 13C NMR spectra of organic
compounds and 1H NMR spectra of ferric complexes were recorded
on JEOL EX400 and AL300NMR spectrometers, respectively.
X-band EPR spectra were recorded at 123 K on a JEOL JES-
SRE2X spectrometer. UV-vis-NIR spectra were recorded on
either a Hitachi U-3500 or a Photal MCPD-1000 spectrophotometer.
Electrochemical studies were performed in a glovebox by using a
BAS CV-50W voltammetric analyzer. The solvent, supporting
electrolyte, and electrodes are acetonitrile, 0.1 M tetrabutylammo-
nium tetrafluoroborate, and a standard three-electrode system with
glassy carbon working electrode, platinum-wire counter electrode,
and Ag/Ag+/CH3CN reference electrode. Magnetic susceptibilities
were recorded on a Quantum Design SQUID susceptometer
interfaced with a HP Vectra computer system over the temperature
range of 5-340 K at the magnetic field of 0.4 T. Since the magnetic
behavior in the solid state is generally influenced by not only the
substituent effects but also other factors, such as defect, grain
boundaries, vacancies, and other imperfections,57 complexes 1-5
were ground with a mortar and pestle to obtain samples under
conditions as close to the same as possible. The data were corrected
for diamagnetism using Pascal’s constants.58
organic products were extracted with CH2Cl2, dried over anhydrous
Na2SO4, and concentrated in vacuo. The products were quantita-
tively analyzed by H NMR spectroscopy by monitoring specific
1H NMR peaks of products using anthracene as an internal
reference.
Kinetic Measurements. The solutions containing complexes
1-5 (1 mM) were prepared under N2 in a glovebox. The
oxygenation reaction was started by adding the complex solution
(0.2 mL) to CH3CN (3.8 mL), equilibrated at the atmospheric
pressure of O2 at 243 K. The reaction exhibited pseudo-first-order
kinetics for all the complex because of the excess of O2. The
pseudo-first-order rate constants were estimated by fitting a decay
in the intensity of the lower-energy LMCT band.
1
Results
X-ray Crystal Structure of [FeIII(TPA)(4Cl-cat)]+.
Crystals suitable for X-ray analysis were obtained only with
2. Figure 1 and Table 2 show an ORTEP plot and selected
bond lengths (Å) and dihedral angles (deg) for [FeIII(TPA)-
(4Cl-cat)]+, respectively. Compared with the DTBC ana-
logue,28 the distortion from octahedral geometry is smaller.
As for the bond lengths of ligands, the C1-O1/C2-O2, Fe-
O1/O2, and Fe-N1/N2 (N1 is aliphatic and N2 is trans to a
catecholate oxygen) distances are similar, longer, and shorter,
respectively, compared to those of the DTBC complexes
(1.349(4)/1.347(4), 1.898(2)/1.917(3), and 2.221(3)/2.145-
(3) Å). This indicates that the catecholate dianion coordinates
Oxygenations of Catecholatoiron(III) Complexes for Product
Analysis. The solutions containing complexes 1-5 (5 mM) in CH3-
CN (5 mL) were stirred under 1 atm of O2 at 25 °C for 5 h. Aliquots
of the reaction solution (2 mL) were diluted with CH2Cl2 (2 mL)
and washed with 2N HCl to decompose the iron complex. The
(56) Sheldrick, G. M. SHELXL-97, Program for X-ray Crystal Structure
Refinement; University of Go¨ttingen: Go¨ttingen, Germany, 1997
(57) Haddad, M. S.; Federer, W. D.; Lynch, M. W.; Hendrickson, D. N. J.
Am. Chem. Soc. 1980, 102, 1468-1470.
(58) Kahn, O. Molecular Magnetism; VCH: New York, 1993.
Inorganic Chemistry, Vol. 44, No. 24, 2005 8813

Higuchi et al.
were ca. 5.5 µB in this temperature range, which are smaller
than the 5.9 µB value expected for S ) 5/2. On the other
hand, the µeff values of 4 and 5 decreased from 5.5 to 5.3 µB
and from 5.0 to 4.6 µB, respectively, with lowering temper-
ature, indicating the spin-crossover transition from S ) 5/2
to S ) 1/2. The µeff value of the spin-crossover ferric complex
between high-spin and low-spin states is represented by eq
159
µeff2 ) [{(1 + 2k)2/3 + 8(2 + k)2/9x} exp(x) +
{10(1 - k)2/3 - 8(2 + k)2/9x} exp(-x/2) +
hs 2
3Pµ exp(-E/kBT)][exp(x) + 2 exp(-x/2) +
eff
3P exp(-E/kBT)]-1 (1)
where x ) ú′/kBT, and kB denotes the Boltzmann constant, k
the orbital reduction factor, ú′ the one-electron spin-orbit
Figure 2. Temperature dependence of the effective magnetic moments
for complexes 1-5 from SQUID susceptibility measurements.
coupling constant of the complex, P the ratio of the molecular
hs
eff
partition functions of the high- and low-spin states, µ the
effective magnetic moment of the high-spin species, and E
the energy separation between zero-point levels of the high-
spin and low-spin states. In addition, ú′ is associated with k
in the relations ú′ ) η2ú and k ) 1/2 + η2/2,59 where ú and
η denote the one-electron spin-orbit coupling constant for
a free ferric ion and the coefficient of the wave function of
the d orbitals, respectively; ú ) 460 cm-1. In case of 5, the
parameters of spin crossover, P ) 90.1 ( 3.7 and E ) 646
( 7.6 cm-1, were obtained by fitting the experimental µeff
values to eq 1 using ú′ ) 110 cm-1, k ) 0.61, which was
hs
eff
estimated using EPR analysis (vide infra), and µ ) 5.50
µB, which is the µeff value of high-spin ferric complexes 1-3.
The fitting curve is shown as a dotted line in Figure 3. Using
eq 2, these parameters lead to the Boltzmann distribution of
a 64% high-spin state (f6 ) and a 36% low-spin state (16%
Figure 3. Temperature dependence of the effective magnetic moments
for complexes 1-5 in CH3CN from 243 to 303 K. The dotted curve is
obtained by the fitting for 5.
A
2E(f 2 ) and 20% A(f 2A), where E and A are the excited
2
2
2
E
and ground-state Kramers doublets, respectively).
more weakly to the ferric center in the 4Cl-cat complex than
in the DTBC complex, while the two TPA nitrogens
coordinate more strongly in the former.
f6A) [3P exp(-E/kBT)][exp(x) + 2 exp(-x/2) +
3P exp(-E/kBT)]-1
Magnetic Susceptibility in the Solid State. Figure 2
shows the temperature dependence of the effective magnetic
moment µeff of 1-5 in the solid state. Complexes 1-4
exhibited µeff values of ca. 5.7 µB at around 300 K, indicating
the high-spin ferric state. The µeff values decreased to 5.3
µB (1), 4.2 µB (2), 4.8 µB (3), and 4.4 µB (4) at 15 K,
indicating the spin-crossover transition from S ) 5/2 to S )
1/2. On the other hand, the µeff values of 5 were substantially
smaller than those of 1-4 and varied from 3.3 µB to 2.5 µB
upon lowering the temperature from 340 to 15 K. The results
indicate that 5 contains a large proportion of low-spin species
within the temperature range observed. For all complexes,
the steep changes in µeff at 5-15 K may be the result of the
zero-field splitting of the high-spin ferric center.
f2E) [2 exp(-x/2)][exp(x) + 2 exp(-x/2) +
3P exp(-E/kBT)]-1
f2A) [exp(x)][exp(x) + 2 exp(-x/2) + 3P exp(-E/k T)]-1
B
(2)
In the case of 4, sufficient data for parameters were not
available to estimate the spin equilibrium. However, it is clear
that 4 has only a small contribution of the low-spin species
even at 243 K.
1
Temperature Effects on the Paramagnetic H NMR
Shifts. Figure 4 shows the 1H NMR spectra of 1-5 in CD3-
CN at 298 K. The wide and broad spectra indicate that all
complexes predominantly have a high-spin ferric cen-
ter.27,28,60,61 The proton signals of the TPA and 4Cl-cat ligands
Magnetic Susceptibility in Solution. The magnetic
susceptibility of the complexes in CH3CN was estimated by
the Evans method. Figure 3 shows the µeff values of 1-5 at
243-303 K. In this temperature range, the magnetic behavior
is similar to that in the solid state. The µeff values of 1-3
(59) Golding, R. M. Applied WaVe Mechanics; Van Nostrand Co.: New
York, 1969.
(60) Lim, J. H.; Lee, H. J.; Lee, K. B.; Jang, H. G. Bull. Korean Chem.
Soc. 1997, 18, 1166-1172.
8814 Inorganic Chemistry, Vol. 44, No. 24, 2005

Catecholatoiron(III) Complexes
in frozen CH3CN at 123 K. This spectral change demon-
strates that the proportion of low-spin species increases as
the substituents on the TPA ligands become more electron-
withdrawing.
EPR parameters (g value and rhombicity (E/D)) for the
high-spin species were estimated by the reported method.62
As shown in Table 3, two sets of g and E/D values were
observed with all complexes. Two different E/D values of
0.07-0.08 (major) and 0.13-0.14 (minor) indicate the
presence of two different components, probably structural
isomers.
The peaks assignable to low-spin species were analyzed
by applying the method for axial EPR spectra59,63,64 to
evaluate the spin delocalization between FeIII and ligands.
When Hamiltonian (3) is employed for the one-electron
orbitals for the t2g5 configuration, the Zeemann Hamiltonian
operators which can be applied to solve a set of three secular
equations for the ground-state Kramers doublet are repre-
sented by eq 4
2
H ) úls - δ(lz2 - 2/3) - (ꢀ/2)(l+2 + l- )
(3)
2
x
gx ) 2[2AC - B + k 2B(C - A)]
2
Figure 4. 1H NMR spectra of complexes 1-5 in CD3CN at 298 K: (a)
1, (b) 2, (c) 3, (d) 4, and (e) 5. Assignment of H signals are as follows: I,
CH2(R-py); II, CH2(py); III, â,â′-H (py); IV, â,â′-H (R-py); and V-VII,
3 Hs of 4Cl-cat.
x
gy ) -2[2AC + B + k 2B(C + A)]
gz ) -2[k(A2 - C2) + A2 - B2 + C2]
A2 + B2 + C2 ) 1
(4)
appeared down- and upfield, respectively. With the increasing
electron-withdrawing effect of the substituents on the TPA
ligand, most of the proton signals were shifted upfield: CH2
(py) 130 (1) f 103 ppm (5), â,â′-H (py) 95.5 (1) f 90.1
ppm (5), â,â′-H (R-py) 85.4 (1) f 77.4 ppm (5), and CH
(4Cl-cat) (-5.5, -36.6, -40.4) (1) f (-9.7, -45.5, -63.5)
ppm (5) (Table S1).
where ú, δ, and ꢀ, are the one-electron spin-orbit coupling
constant, the axial distortion, and the lower symmetry
distortion, respectively, A, B, and C are the coefficients
characterizing the two spinors of the ground-state Kramers
doublet, and k is the orbital reduction factor. Assuming that
C ) 0 for axial EPR spectra, the parameters are represented
by eq 5
Participation of the low-spin species of 4 and 5 in solution
was also studied by variable-temperature 1H NMR measure-
ments. Figure 5 shows the Curie plots of some proton signals
of 1-5 at 243-303 K. As shown in Figure 5A-C, the proton
signals of the TPA ligands of 1-3 gave linear plots. The
intercept values at T-1 ) 0, which are shown in parentheses
in Figure 5, are close to the chemical shifts of free TPA
ligands, indicating that the spin-crossover transition does not
take place in this temperature range. Complex 4 also gave
linear plots, but the intercept values at T-1 ) 0 for â-H
deviated greatly from the diamagnetic region (e.g., 19.0, 19.8
ppm), which may result from the spin-crossover transition.
Complex 5 showed nonlinear lines, clearly reflecting the
spin-crossover transition of 5. On the other hand, as shown
in Figure 5D-F, the temperature dependence of the proton
signals of the 4Cl-cat ligands showed linear features but in
a different manner than the Curie law. It is very interesting
that even 4 and 5 exhibited linear features similar to those
of 1-3, despite their spin-crossover transition.
2
x
g ) -2[sin R + 2k sin R cos R]
g| ) 2[sin2 R - (1 + k)cos2 R]
δ/ú ) 2-1/2(tan R - cot R + 2-1/2
ꢀ/ú ) 0
)
(5)
where A and B are replaced with cos R and sin R,
respectively, on the basis of A2 + B2 ) 1. In this analysis,
two sign sets of the g values (i.e., (g , g|) ) (-, +) and (g ,
g|) ) (-, -)) are probable.64,65 The fitting of the former set
to eq 5 leads to a k value of 0.61 and a very small A value
(0.093-0.107), a very large B value (0.994-0.996), and a
large δ/ú (7.01-8.01) value. On the other hand, the latter
set of g values leads to a k value of 1.06, an A value of
0.801-0.804, a B value of 0.595-0.599, and a δ/ú value of
0.07-0.08. The k value close to 0.6 for the former set
(62) Wickman, H. H.; Klein, M. P.; Shirley, D. A. J. Chem. Phys. 1965,
42, 2113-2117.
(63) Griffith, J. S. The Theory of Transition-Metal Ions; University Press,
Cambridge, U.K., 1961.
EPR Parameters and Electronic Geometry of Com-
plexes in Solution. Figure 6 shows the EPR spectra of 1-5
(61) Yoon, S.; Lee, H. J.; Lee, K. B.; Jang, H. G. Bull. Korean Chem. Soc.
2000, 21, 923-928.
(64) DeSimone, R. E. J. Am. Chem. Soc. 1973, 95, 6238-6244.
(65) McGarvey, B. R. Coord. Chem. ReV. 1998, 170, 75-92.
Inorganic Chemistry, Vol. 44, No. 24, 2005 8815

Higuchi et al.
Figure 5. Curie plots for some signals of 1-5: (A) â-H (py), (B) â-H (R-py), (C) γ-H (py), and (D)-(F) 3Hs of the 4Cl-cat ligand. The apparent
intercepts at 1000T-1 ) 0 are given in parentheses on the left.
theoretically leads to the spin delocalization of 20% low-
spin FeIII and 80% ligands in the ground-state Kramers
doublet,59,66 which suggests the low-spin semiquinonatoiron-
(II) complex as the ground state. Thus, the former set may
be excluded because the EPR spectra around g ) 2 are
attributed to low-spin ferric complexes. We however favor
the former set because low-spin iron(III)-TPA complexes
are expected to have a large δ/ú because of the t2g orbitals
degenerated by interactions between three nitrogen pπ orbitals
of the TPA ligand and only one dπ orbital. The k value
smaller than 0.75, which is the lower limit expected for the
low-spin FeIII-Cat state, is probably obtained because the
analysis method used here has been developed for simpler
metal complexes, such as porphyrinate and dithiocarbamate
complexes.67 It is still noteworthy, however, that the small
k value affords reasonable support for the significant spin
delocalization on the 4Cl-cat ligand.
varied from -903 (1) to -641 mV (5), whereas the Epa of
SQ•-/Cat2- was not sensitive to the substituents in the range
of 198-228 mV. This result suggests that the substituents
of the TPA ligand influence the d-orbital energy of the ferric
center but not the orbital energy of the 4Cl-cat ligand.
LMCT Bands in UV-vis-NIR Spectroscopy. As shown
in Table 4 and Figure S1, complexes 1-5 in CH3CN
exhibited two intense LMCT bands (λ1 and λ2) characteristic
of catecholatoiron(III) complexes.19,68,69 The electron-
withdrawing substituents on the TPA ligand brought about
red shifts of both λ1 and λ2. As for the peak intensity, 5
exhibited unique spectra, ꢀ1 > ꢀ2, whereas ꢀ1 < ꢀ2 for 1-4.
Because this characteristic feature was thought to be caused
by the low-spin fraction, the temperature effect on the spectra
was studied by lowering the temperature from 303 to 243
K. Complexes 1-3 exhibited no effect on the intensities of
the λ1 and λ2 bands but showed slight blue shifts with
isosbestic points (Figure 7A for 2, Figure S2 for 1 and 3).
On the other hand, 5 exhibited spectral changes with an
increase in the intensity of λ1 and with a decrease in the
intensity and blue shift of the λ2 band. In addition, the spectral
Electrochemical Properties. In the cyclic voltammogram
measurements at room temperature, quasi-reversible waves
showed cathodic FeIII f FeII (Epc) and anodic Cat2- f SQ•-
(Epa) wave peaks. As shown in Table 4, the Epc of FeIII/FeII
was dependent on the substituents on the TPA ligand and
(68) Cox, D. D.; Benkovic, S. J.; Bloom, L. M.; Bradley, F. C.; Nelson,
M. J.; Que, L., Jr.; Wallick, D. E. J. Am. Chem. Soc. 1988, 110, 2026-
2032.
(66) Stevens, K. W. H. Proc. R. Soc. London A 1953, 219, 542-555.
(67) Rieger, P. H. Coord. Chem. ReV. 1994, 135, 203-286.
(69) Dei, A.; Gatteschi, D.; Pardi, L. Inorg. Chem. 1993, 32, 1389-1395.
8816 Inorganic Chemistry, Vol. 44, No. 24, 2005

Catecholatoiron(III) Complexes
Figure 6. X-band EPR spectra recorded in CH3CN at 123 K. The EPR
conditions are microwave frequency, 9.06 kHz; microwave power, 20 mW;
modulation amplitude, 1 mT; and modulation frequency, 100 kHz: (a) 1,
(b) 2, (c) 3, (d) 4, and (e) 5.
change in 5 was accompanied by the growth of new peaks
at around 1100 nm. Complex 4 exhibited a mixture of the
above two types of spectral changes, with smaller peaks at
around 1100 nm than those of 5. The lowest-energy peaks
could be assigned to the LMCT bands of low-spin species.34
Reactivity of the Complexes with Molecular Oxygen.
The reaction of 1-5 with O2 in CH3CN afforded exclusively
cis-dienelactone, which is produced by the intradiol oxy-
genation of the 4Cl-cat ligand followed by dechlorination
of the muconic anhydride.70
Figure 7. Variable-temperature UV-vis-NIR spectra for (A) 2 and (B)
5 at 298, 293, 283, 273, 263, 253, and 243 K. The arrows represent changes
on going to lower temperature.
reaction mechanisms. However, there have been discussions
for a long time on how the high-spin catecholatoiron(III)
complex reacts with O2. To provide insight into the reaction
mechanism, we prepared an admixture of high- and low-
spin catecholatoiron(III) complexes in solution. Although the
low-spin complexes are not regarded as models for high-
spin catechol dioxygenases, the reactivity of the low-spin
catecholatoiron(III) species with O2 is expected to provide
important information about the reaction mechanisms of the
high-spin species. Scheme 2 summarizes the relationship
among spin crossover, spin delocalization, and the reaction
of the complexes with O2.
To elucidate the effect of the ligand substituents on the
reactivity with O2, we estimated the reaction rates by
monitoring the decay of the lower-energy LMCT band (λ2)
(Figure S3).70 The oxygenation rate constant, kobs, was
obtained under the conditions of pseudo-first-order kinetics.
As shown in Table 4, kobs increased in the order of 1 < 2 <
3 < 4 < 5.
Admixture of High- and Low-Spin Catecholatoiron-
1
(III) Complexes, (i). The H NMR spectra at 298 K show
the substituent effects of the TPA ligand. Both the TPA and
4Cl-cat protons are shifted upfield as the substituents become
more electron-withdrawing. The upfield shifts of the TPA
protons indicate that the interactions of the TPA ligand with
eg orbitals of the ferric ion become less effective because
the TPA protons appear in the downfield region as a result
of a σ-spin delocalization mechanism via the eg orbitals. On
the other hand, because the 4Cl-cat protons appear in the
upfield region as a result of a π-spin delocalization mech-
anism via the t2g orbitals, the upfield shifts of the 4Cl-cat
Discussion
In the developments in bioinorganic chemistry involving
the catechol dioxygenases and their model systems, the
formation of the high-spin catecholatoiron(III) complex as
an intermediate and the participation of the semiquinona-
toiron(II) form have become well accepted to explain the
(70) Funabiki, T.; Yamazaki, T.; Fukui, A.; Tanaka, T.; Yoshida, S. Angew.
Chem., Int. Ed. 1998, 37, 513-515.
Inorganic Chemistry, Vol. 44, No. 24, 2005 8817

Higuchi et al.
Table 3. Experimental EPR Parameters for High-Spin and Low-Spin Species of 1-5 at 123 K
high-spin
low-spin
g values (E/D)
g
value
g| value
∑g2
cos R
sin R
k
δ/ú
a
b
b
b
b
b
1
2
8.5 (0.13); 7.7, 4.1 (0.08)
8.6 (0.14); 7.6, 4.3 (0.07)
-2.160
-2.167
-2.167
-2.160
-2.160
-2.153
-2.153
-2.141
-2.141
-
-
-
-
-
-
1.940
-1.940
1.941
-1.941
1.945
-1.945
1.955
-1.955
13.16
13.16
13.10
13.10
13.05
13.05
12.99
12.99
0.107
0.801
0.106
0.802
0.103
0.802
0.093
0.804
0.994
0.599
0.994
0.598
0.995
0.597
0.996
0.595
0.63
1.07
0.61
1.07
0.60
1.06
0.61
1.06
7.01
0.08
7.04
0.08
7.27
0.08
8.01
0.07
3
4
5
8.6 (0.14); 7.3, 4.2 (0.07)
8.5 (0.13); 7.3, 4.2 (0.07)
8.5 (0.13); 7.3, 4.2 (0.07)
a Not determined because of a broad EPR spectrum for 1. b Not calculated because there was no observation of g|.
Table 4. Spectroscopic and Kinetics Data for 1-5 in CH3CN
potentialc (mV)
complex
µeffa (µB)
λ1, λ2b (nm) (ꢀ1, ꢀ2 (M-1cm-1))
Epc(FeIII/FeII)
Epa (SQ•-/Cat2-
)
kobsa (× 10-4 s-1
)
1
2
3
4
5
5.5
5.5
5.5
5.3
4.6
474 (2010), 767 (3040)
483 (1930), 787 (2970)
496 (1920), 805 (2950)
498 (2040), 815 (2670)
509 (2410), 839 (1820)d
-903
-828
-748
-731
-641
228
198
203
198
228
0.18
0.67
1.6
2.1
2.9
a Measured at 243 K. b Measured at 298 K. c Measured at room temperature. All the E values are referenced to the Fc/Fc+ couple. Conditions were a scan
rate of 100 mV/s, under ambient temperature and N2. Epc and Epa show cathodic and anodic potential, respectively. d Other peaks observed, λ (ꢀ ): 1090
(640), 1160 (400).
Scheme 2
Evans method at 243-303 K, are 5.5 µB, indicating that
complexes 1-3 do not show spin crossover behavior in this
temperature range and are in the high-spin state. This result
is also consistent with the chemical shifts in the diamagnetic
region at T-1 ) 0 in the linear Curie plots of 1H NMR signals
of the TPA ligand (Figure 5A-C). In the case of complex
5, the µeff value varies from 4.6 µB (243 K) to 5.0 µB (303
protons indicate that the interactions of the 4Cl-cat ligand
K). The proportion of the high-spin species is estimated to
with t2g orbitals of the ferric ion become more effective.
be 64% at 243 K by simulation of the observed data using
Weak coordination of the TPA ligand with electron-
withdrawing substituents should stabilize both eg and t2g
orbitals, which take into account the respective upfield shifts
of the TPA and 4Cl-cat protons. The effective induction of
low-spin species with electron-withdrawing substituents
indicates greater stabilization of the t2g orbital than the eg
orbital because it leads to a larger crystal-field splitting.
The admixture of two spin states in the solid state was
clearly shown by measurement of the magnetic susceptibility.
As shown in Figure 2, complexes 1-4 are in the high-spin
state with µeff ) ca. 5.7 µB at around room temperature, and
they exhibit spin-crossover behavior with decreasing tem-
perature. This µeff value smaller than 5.9 µB, which is
expected from ferric complexes in the fully S ) 5/2 ground
state, may be the result of the spin delocalization on the
ligand, the existence of intermolecular magnetic interactions,
or both.51,57,71 Remarkably, complex 5 is in the admixture
state even at room temperature (µeff ) 3.3 µB), indicating
the strong substituent effect of the two NO2 groups on the
crystal field parameter.
hs
eff
µ
) 5.5 µB. The nonlinear Curie plots of the TPA
protons, as shown in Figure 5A-C, can be accounted for
by the spin admixture in 5. The 1H NMR shifts of the metal
complex in spin equilibrium are represented by a weighted
average of the signals for high- and low-spin species in eq
7
hs
iso
ls
iso
δobs ) δdia + (1 - x)δ + xδ
(7)
where x is the portion of the low-spin species and δobs, δdia,
hs
iso
ls
iso
δ , and δ are the observed, diamagnetic, and isotropic
shifts of the high-spin and low-spin complexes, respectively.
These nonlinear Curie plots (i.e., the negative deviation of
δobs of the TPA protons from the expected linear plots at
the lower temperatures, despite the increase in the x value)
can be attributed to δ < δ for the TPA protons. The
large downfield shifts of the TPA protons are mainly the
result of the σ-spin delocalization;72 therefore, the relation
ls
iso
hs
iso
ls
iso
hs
iso
of δ < δ reflects the smaller spin number in the eg
orbitals of the low-spin ferric complex. Examples are found
for the decrease in isotropic shift caused by the decrease in
the spin number in eg orbitals (e.g., iron(III) dithiocarbamates
The admixture in solution was shown by measurements
1
of the magnetic susceptibility, H NMR, EPR, and UV-
vis-NIR spectra. The µeff values of 1-3, estimated by the
(72) Que, L., Jr., Ed. Physical Methods in Bioinorganic Chemistry:
Spectroscopy and Magnetism; University Science Books: Sausalito,
CA, 2000.
(71) Heistand, R. H., II; Lauffer, R. B.; Fikrig, E.; Que, L., Jr. J. Am. Chem.
Soc. 1982, 104, 2789-2796.
8818 Inorganic Chemistry, Vol. 44, No. 24, 2005

Catecholatoiron(III) Complexes
in the equilibrium between S ) 5/2 and 1/273,74 and iron(III)
porphyrinates in the equilibrium between S ) 3/2 and 1/2).75
EPR spectra provide a clear support for the admixture caused
by the substituent effects. Figure 6 shows that the proportion
of low-spin species increases in the order of 1 < 2 ≈ 3 < 4
< 5 at 123 K. UV-vis-NIR spectra also offer support for
the admixture at 243-303 K. All complexes exhibit two
characteristic peaks of the LMCT band, λ1 and λ2 (λ1 > λ2
in energy). The intensities of the bands of high-spin
complexes 1-4 are ꢀ1 < ꢀ2, but that of 5 is uniquely ꢀ1 >
ꢀ2. As shown in Figure 7B, the intensities of the bands of 5
decrease with decreasing temperature, in harmony with the
other spectroscopic supports for the increase in low-spin
species at lower temperatures. In addition, new peaks at
around 1100 nm grow with decreasing temperature, which
may be responsible for progression coupled to the LMCT
for low-spin species, as observed for low-spin [Fe(TPA)-
(cat)]BPh4 in the solid state.34
complexes at T-1 ) 0 are observed in the nondiamagnetic
region of -4.9 to -24.4 ppm. A similar non-Curie-law
temperature dependence has been observed for the 4-H and
6-H proton signals of the DTBC ligand of [FeIII(BLPA)-
DTBC]BPh4,69 [FeIII(BBA)DTBC]BPh4,61 and [FeIII(TPA)-
DTBC]BPh4.76 The explanation for iron(III) meso-substituted
porphyrinates77,78 may be applied to this non-Curie-law
behavior (i.e., the participation of thermally accessible plural
excited states in the catecholate-to-iron charge transfer).
Semiquinone Character in the Low-Spin Complex:
Spin Delocalization on the 4Cl-cat Ligand, (iils). It is very
interesting and important to determine the spin delocalization
in the low-spin catecholatoiron(III) complexes. The EPR
spectra in Figure 6 indicate the increasing tendency of the
low-spin complexes, 1 < 2 ≈ 3 < 4 < 5. Analysis of the
low-spin peaks of complexes 2-5 by the method for axial
EPR spectra results in a small value of the orbital reduction
factor k (0.61), which represents the significant spin delo-
calization on the ligands in the low-spin catecholatoiron-
(III) complex. As for the 4Cl-cat ligand, this was supported
by the observation of LMCT bands in the significant low-
energy region (i.e., at around 1100 nm).34
Semiquinone Character in the High-Spin Complex:
Spin Delocalization on the 4Cl-cat Ligand, (iihs). The
ground state of catecholatoiron(III) complexes is represented
as the quantum mechanical admixture of FeIII-Cat and FeII-
SQ states. This means that the ground state Ψ- and the
excited state Ψ+ of the catecholatoiron(III) complex are
represented by
As shown in Figure 5D-F, the 4Cl-cat protons of spin-
crossover complexes 4 and 5 exhibit a non-Curie-law plot
similar to that of high-spin complexes 1-3. The linear
feature, despite the increasing proportion of the low-spin
species at lower temperatures, indicates that the absolute
value of δiso for the low-spin complex is at least not smaller
Ψ- ) ΨFeIII-Cat + RΨFeII-SQ
(8)
(9)
Ψ+ ) ΨFeII-SQ - RΨFeIII
-Cat
ls
iso
hs
iso
than that of the high-spin analogues: |δ | g |δ | in eq 7,
R ) Hin/∆E
(10)
ls
iso
hs
iso
although δ < δ for the TPA protons. Thus, the change
in the d-orbital spin configuration from eg t2g3 to eg t2g5 may
be effective in increasing the dπ-pπ overlap interaction
between the ferric center and the 4Cl-cat ligand, resulting
in an increase in Hin. In addition, the observation of a LMCT
band in the low-spin species lower than that in the high-
spin analogues shows that the ∆E of the low-spin species is
smaller than that of the high-spin analogues. Thus, both the
increase in Hin and the decrease in ∆E of the low-spin species
relative to the high-spin analogues should result in an
increase in the R value (i.e., the stronger contribution of the
FeII-SQ state in the low-spin species).
2
0
II
III-Cat
where ΨFe
and ΨFe -SQ are the wave functions of the
pure FeIII-Cat and FeII-SQ states, respectively. Hin and ∆E
denote the charge-transfer integral and the energy difference
between the two pure states, respectively (Figure S4). On
the other hand, the chemical shift of paramagnetic metal
complexes is represented by eq 11
δobs ) δdia + δiso
(11)
where δiso are the isotropic shifts. From a comparison of the
relative terms of eq 8 with eq 11, it is clear that δiso of the
II
4Cl-cat protons corresponds to RΨFe -SQ and is an index of
Reactivity of Catecholatoiron(III) Complexes, (iiihs) and
(iiils). The oxygenation of catechols catalyzed by iron-
containing catechol dioxygenases and model complexes
proceeds as follows: (1) the catechol binds to iron to form
a catecholatoiron(III) complex, (2) oxygen binds to the
catecholatoiron(III) complex, (3) oxygen inserts into the C-C
bond of the aromatic ring, (4) the monooxygenated product
is hydrolyzed, and 5) the oxygenated product is eliminated
to regenerate the iron active species. For step 1, many
catecholatoiron(III) complexes which are reactive with O2
have been developed. For step 2, two processes have been
proposed as shown in Scheme 1, but no direct evidence for
the semiquinone character of a catecholate ligand in the
ground state.
As shown in Figure 4, the upfield shifts of the 4Cl-cat
protons are ascribed to the π-spin delocalization on the 4Cl-
cat ligand via LMCT but not to the σ-spin delocalization.72
For the high-spin ferric complexes 1-3, the increase in the
magnitude of δiso of the 4Cl-cat protons in the order of 1 <
2 < 3 indicates the greater semiquinone character of the
catecholate ligand in this order. In addition, as shown in
Figure 5D-F, although complexes 1-3 remain in the high-
spin state, the intercept values of the 4Cl-cat protons of these
(73) Golding, R. M.; Tennant, W. C.; Kanekar, C. R.; Martin, R. L.; White,
A. H. J. Chem. Phys. 1966, 45, 2688-2693.
(76) Higuchi, M.; Hitomi, Y.; Tanaka, T.; Funabiki, T. Unpublished data.
(77) Shokhirev, N. V.; Walker, F. A. J. Phys. Chem. 1995, 99, 17795-
17804.
(78) Banci, L.; Bertini, I.; Luchinat, C.; Pierattelli, R.; Shokhirev, N. V.;
Walker, F. A. J. Am. Chem. Soc. 1998, 120, 8472-8479.
(74) Golding, R. M.; Tennant, W. C.; Bailey, J. P. M.; Hudson, A. J. Chem.
Phys. 1968, 48, 764-771.
(75) Ikeue, T.; Ohgo, Y.; Ongayi, O.; Vicente, M. G. H.; Nakamura, M.
Inorg. Chem. 2003, 42, 5560-5571.
Inorganic Chemistry, Vol. 44, No. 24, 2005 8819

Higuchi et al.
the structure of the O2 adduct has been obtained.41,61 For step
3, a reaction via a Criegee79 or epoxide-type intermediate33
has been proposed, but details have not been clarified. Present
results are valuable for discussions on step 2.
Many catecholatoiron(III) complexes have been synthe-
sized as model complexes for catechol dioxygenases to
understand the mechanisms of the oxygenation reaction.4,22,80
The importance of the semiquinonatoiron(II) character in the
catecholatoiron(III) complex has been proposed from the
early stage of the model studies.19,23,39,68,81 In efforts to clarify
the mechanism of O2-binding to the catecholatoiron(III)
complex, correlations have been studied between the reac-
tivities and physical properties of these model complexes,
such as the energy of the LMCT band,27-31 the redox
potential,30,31 and the NMR shift of the catecholate proton.27-29
In these studies, the reactivity was compared with the rate
constants of the reactions of the complexes with O2, although
the logarithm of the rate constants is thought to be appropriate
for discussions on the basis of the kinetic energies. Recently,
we have reported a linear correlation between the logarithm
of the rate constants and the energy of the lower-energy
LMCT band in the oxygenation of a series of [FeIII(TPA)-
(R-Cat)]BPh4 complexes with substituted Cat ligands.26
In this study, the logarithm of the rate constants was
plotted against the energy of the lower-energy LMCT band
assignable to the high-spin species (Figure 8A). As expected,
high-spin complexes 1-3 exhibit good linear correlations,
supporting the belief that the reactivity of the complexes with
O2 is directly correlated with the FeII-SQ state.26 Complexes
4 and 5 show a negative deviation from the linear line,
indicating that the low-spin species have lower reactivity than
the high-spin analogues. The rate constant for the high-spin
species of 5 can be estimated to khs[O2] ) 4.30 × 10-4 s-1
on the basis of the extrapolation of the linear line from the
high-spin LMCT energy of 5 in Figure 8A because the
negative deviation from the linear correlation is the result
of the admixture of the low-spin species. Assuming that both
the high- and low-spin species react with O2, the rate constant
of the low-spin species can be estimated to be kls[O2] ) 0.41
× 10-4 s-1 using the values of kobs[O2] ) 2.90 × 10-4 s-1,
khs[O2] ) 4.30 × 10-4 s-1, which is calculated by the linear
correlation in Figure 8A, and the equilibrium constant K )
0.56 at 243 K. Thus, this result indicates that the reactivity
of the low-spin species is much lower than that of the high-
spin analogues.
Figure 8. Correlation between ln kobs and the lower-LMCT energy at 243
K (A) and between ln kobs and the NMR shift (|δiso
|
-0.5) of the 4Cl-cat
proton at 243 K (B). The dotted line represents the linear relationship
between 1-3.
where a stable O2 adduct can be formed). Although the spin-
state-dependent reactivity has been known in other iron
systems,83-86 it is still unclear why the low-spin catechola-
toiron(III) species has a lower O2 reactivity than the high-
spin species.
The isotropic shift controls mainly the contact shift when
the protons are sufficiently distant from the metal center.
Thus, as shown in eq 12, δiso for high-spin ferric complexes
is directly proportional to the unpaired-electron spin density,
FC, in the pπ orbital of the carbon to which the proton is
attached
We have suggested that O2 attacks the ferric center of the
catecholatoiron(III) complex,26 although the interaction of
an iron(III) center with O2 has been little known, in contrast
to the well-known O2 binding to an iron(II) center. Recently,
however, Collins and co-workers demonstrated that the Fe-
(III)-TAML complex can react with O2 to yield a µ-oxo-
diiron(IV) complex.82 Thus, a direct O2 attack on the ferric
center may be possible in certain cases (e.g., in the case
QCHFCh γe
S(S + 1)
δiso ) -
(12)
( )
2S
γN 3kBT
where S (5/2), h, and QCH denote the effective spin for the
high-spin ferric ion, the Plank constant, and the McConnel
(82) Ghosh, A.; de Oliveira, F. T.; Yano, T.; Nishioka, T.; Beach, E. S.;
Kinoshita, I.; Mu¨nck, E.; Ryabov, A. D.; Horwitz, C. P.; Collins, T.
J. J. Am. Chem. Soc. 2005, 127, 2505-2513.
(83) Jensen, K. P.; Ryde, U. J. Biol. Chem. 2004, 279, 14561-14569.
(84) Shiota, Y.; Yoshizawa, K. J. Chem. Phys. 2003, 118, 5872-5879.
(85) Franzen, S. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 16754-16759.
(79) Mayer, R. J.; Que, L., Jr. J. Biol. Chem. 1984, 259, 13056-13060.
(80) Funabiki, T. Oxygenases and Model Systems; Kluwer Academic
Publishers: Dordrecht, The Netherlands, 1997; Vol. 19.
(81) Funabiki, T.; Mizoguchi, A.; Sugimoto, T.; Yoshida, S. Chem. Lett.
1983, 917-920.
8820 Inorganic Chemistry, Vol. 44, No. 24, 2005

Catecholatoiron(III) Complexes
constant, respectively. Because the spin density, FC, is
represented by FC R2 ) (Hin/∆E)2,77 |δiso|-0.5 is proportional
to ∆E in the case where Hin is constant for the high-spin
complexes. Because ∆E can be regarded as the LMCT
energy,87 the plot of ln kobs against |δiso|-0.5 is also expected
to exhibit a linear correlation similar to the plot of ln kobs
against the LMCT energy. Indeed, as shown in Figure 8B,
the linear correlation was found for high-spin complexes
1-3. Again, this plot also indicates that low-spin complexes
have lower reactivity than their high-spin analogues.
(III) complexes with O2 for the high-spin species, the
complexes containing low-spin species negatively deviate
from the above linear correlation line, despite their higher
semiquinonate character. This finding suggests a crucial role
of the iron(III) center in the reaction of the catecholatoiron-
(III) complexes with O2. The detailed reaction mechanism
is under investigation in our laboratory.
Acknowledgment. We are grateful to Prof. S. Kitagawa
and Dr H.-C. Chang for their assistance with the X-ray
crystallographic analysis. Prof. M. Kodera provided useful
advice. This work was financially supported by a Grant-in-
Aid for Scientific Research (No. 14750679) from the
Ministry of Education, Science, Sports, and Culture, Japan.
Y.H. is grateful to the Mizuho Foundation for the Promotion
of Sciences for financial support.
Conclusion
We evaluated the spin delocalization on the catecholate
ligands by the isotropic shifts of the catecholate protons and
found that the semiquinonatoiron(II) character is more
significant in the low-spin species than in the high-spin
species. The higher semiquinonatoiron(II) character in the
low-spin species is also supported by the LMCT band
observed at around 1100 nm for the low-spin species.
Although a linear relationship between the spin delocalization
and the logarithm of the reaction rate of the catecholatoiron-
Supporting Information Available: The NMR shifts of
complexes 1-5 at 298 K (Table S1), UV-vis-NIR spectra for
1-5 at 298 K (Figure S1), variable-temperature UV-vis-NIR
spectra for 1, 3, and 4 at 243-303 K (Figure S2), progress of the
oxygenation reaction of 2 at 243 K (Figure S3), potential surface
for the charge-transfer complex (Figure S4), and X-ray crystal-
lographic data for 2 in CIF format. This material is available free
of charge via the Internet at http://pubs.acs.org.
(86) Bassan, A.; Blomberg, M. R. A.; Siegbahn, P. E. M.; Que, L. J. Am.
Chem. Soc. 2002, 124, 11056-11063.
(87) Foster, R. Organic Charge-Transfer Complexes; Academic Press: New
York, 1969.
IC051173Y
Inorganic Chemistry, Vol. 44, No. 24, 2005 8821