Oxidation versus dioxygenation of catechol: The iron-bispidine system

Article abstract of DOI:10.1002/ejic.201100802

The iron-bispidine-catalyzed oxidation and dioxygenation (catechol dioxygenase activity) [i.e., the oxidation of 3,5-di-tert-butylcatecholate (dbc^2-) by [Fe^II(L)X₂]ⁿ⁺ (L = 3,7-dimethyl-9-oxo-2,4-bis(2-pyridyl)-3,7-diazabicyclo[3.3.1]nonane-1, 5-dicarboxylate methyl ester)] and air (O₂) was studied experimentally and supported by the analysis of the X-ray crystal structure of [Fe(L·MeOH)(tcc)][B(Ph)₄] with the deactivated tetrachlorocatecholate tcc^2- and a DFT-based analysis. The [Fe ^II(L)X₂]ⁿ⁺/O₂/dbc^2- system catalyzes the intradiol cleavage of dbc^2- but with a relatively low activity (5 % yield); most of the substrate is oxidized in a two-electron oxidation to the benzoquinone (dbq) product (48 % yield). The crystallographic and DFT-based theoretical analyses indicate that this is due to the high oxidation potential of the Fe^III oxidant (fast and efficient electron transfer), that is, the oxidation to the benzoquinone side product is faster, and due to the bonding mode of the catecholate substrate to the Fe^III oxidant, with little spin density transferred to the catecholate substrate. The iron-bispidine-catalyzed oxidation and dioxygenation of 3,5-di-tert-butylcatecholate (dbc^2-) by [Fe^II(L)X ₂]n⁺ and air (O₂) was studied experimentally and supported by a crystal structure determination and by a DFT-based analysis. The [Fe^II(L)X₂]n⁺/O₂/dbc ^2- system catalyzes the intradiol cleavage of dbc^2- but only with relatively low activity; most of the substrate is oxidized in a two-electron oxidation to give the benzoquinone (dbq) product. Copyright

Full text of DOI:10.1002/ejic.201100802

FULL PAPER
DOI: 10.1002/ejic.201100802
Oxidation versus Dioxygenation of Catechol: The Iron–Bispidine System
Peter Comba,*[a] Hubert Wadepohl,[a] and Steffen Wunderlich[a]
Keywords: Biomimetic catalysis / Oxidation / Environmental chemistry / Coordination chemistry / Iron complexes / Density
functional calculations
The iron–bispidine-catalyzed oxidation and dioxygenation
(catechol dioxygenase activity) [i.e., the oxidation of 3,5-di-
tert-butylcatecholate (dbc2–) by [FeII(L)X2]n+ (L = 3,7-di-
(5% yield); most of the substrate is oxidized in a two-electron
oxidation to the benzoquinone (dbq) product (48% yield).
The crystallographic and DFT-based theoretical analyses
methyl-9-oxo-2,4-bis(2-pyridyl)-3,7-diazabicyclo[3.3.1]- indicate that this is due to the high oxidation potential of the
nonane-1,5-dicarboxylate methyl ester)] and air (O2) was
studied experimentally and supported by the analysis of the
X-ray crystal structure of [Fe(L·MeOH)(tcc)][B(Ph)4] with the
deactivated tetrachlorocatecholate tcc2– and a DFT-based
analysis. The [FeII(L)X2]n+/O2/dbc2– system catalyzes the in-
tradiol cleavage of dbc2– but with a relatively low activity
FeIII oxidant (fast and efficient electron transfer), that is, the
oxidation to the benzoquinone side product is faster, and due
to the bonding mode of the catecholate substrate to the FeIII
oxidant, with little spin density transferred to the catecholate
substrate.
dioxygenation, which leads to intradiol or extradiol cleav-
age (see Scheme 1; enzymes that follow the intradiol path-
Introduction
Catechol dioxygenases are mononuclear nonheme iron en- way use FeIII, the extradiol enzymes use FeII in their active
zymes, and their biochemistry, supported by experimental site). Catechol dioxygenases are isolated from soil bacteria,
and computational modeling, has been studied exten- and the most extensively studied intradiol-cleaving catechol
sively.[1–4] Three different reaction channels are possible dioxygenase is protocatechuate-3,4-dioxygenase, for which
when dioxygen reacts with a catecholate bound to a non- a vast range of structural and mechanistic data are avail-
heme iron center: the oxidation to the ortho-quinone (cate- able.[5] Various biomimetic systems have been studied, and
cholase activity) (i.e., a simple two-electron oxidation), and these have helped to elucidate the catalytic mechanism.[6–16]
Scheme 1. Proposed mechanisms for the iron-catalyzed intradiol (a) and the extradiol cleavage (b) of catechol.[1–4]
[a] Universität Heidelberg, Anorganisch-Chemisches Institut,
The biotic and abiotic natural degradation of organic
pollutants in the soil,[17,18] possible industrial applications
for the conversion of aromatic compounds to water-soluble
aliphatic products,[19] and the transformation of renewable
Im Neuenheimer Feld 270, 69120 Heidelberg, Germany
Fax: +49-6221-546617
E-mail: peter.comba@aci.uni-heidelberg.de
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201100802
5242
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 5242–5249

Oxidation versus Dioxygenation of Catechol
biomass resources to fuels and chemicals[20,21] by catechol ation of 3,5-di-tert-butylcatecholate (dbc2–) by [FeII(L)-
dioxygenases and model compounds continue to attract at- X2]n+ and air (O2), presented here, are combined with ex-
tention. An interesting recent observation is that volatile perimental and computational structural work and a DFT-
halogenated hydrocarbon compounds, which are known to based analysis of the reaction mechanism.
be involved in stratospheric chemistry, in particular in
ozone depletion,[22] are produced through abiotic natural
processes in the soil, which involve FeIII, H2O2, and humic
compounds, which have aromatic residues and specifically
also catechol groups.[23]
We have recently been able to show that the iron–bispid-
ine complex with the same ligand L as used in the present
communication leads to the first biomimetic example for
the iron-catalyzed halogenation of aliphatic hydrocarbon
Scheme 2. Structure of the ligand L and other ligands used in
compounds, and which halogenates with high chemoselec-
tivity.[24] Since a possible mechanism to produce volatile ha-
logenated hydrocarbon compounds in the soil might involve
the iron-catalyzed degradation of humic compounds (di-
oxygenation of catechol derivatives) and subsequent iron-
catalyzed halogenation of the degradation products, we
were interested to test the same catalyst for a possible cate-
chol dioxygenase activity to better understand possible
mechanisms for the formation of volatile halogenated hy-
catechol dioxygenase modeling.
Results and Discussion
Syntheses and Structural Properties
The FeII precursor [Fe(L)(OTf)2] (OTf = triflate = tri-
–
drocarbons from humic substances. The tetradentate bispi- fluoromethanesulfonate, CF3SO3 ) was prepared by a
dine ligand L, used in this study, is shown in Scheme 2, known procedure,[40] and the corresponding FeIII tetrachlo-
together with a collection of ligands used in efficient biomi- rocatecholate (tcc2–) complex was obtained by ligand ex-
metic catechol dioxygenase reactions. Bispidines are very ri- change in methanol and subsequent air oxidation. The
gid adamantane derivatives. They were first prepared by magnetic moment of the catalyst–substrate complex
Mannich,[25] exist in a large variety[26] and have been used [Fe(L·MeOH)(tcc)][B(Ph)4], μ = 5.61 BM (Bohr Magneton,
extensively to enforce specific geometries to metal ions and μB), is only slightly lower than the spin-only value of
desirable properties to their complexes.[27–33] High-valent 5.92 BM for a high-spin d5 electronic ground state, and this
iron–bispidine complexes have been found to be particularly is as expected from the enzyme protocatechuate-3,4-dioxy-
interesting due to their high redox potentials[34,35] and, con- genase.[41] A plot of the experimental structure (single-crys-
sequently, as efficient oxidants and potential oxygenation tal X-ray diffraction) of [Fe(L·MeOH)(tcc)][B(Ph)4] is
catalysts.[36–39] The experimental studies of the dioxygen- shown in Figure 1, and selected structural parameters are
Figure 1. (a) ORTEP plot of [Fe(L·MeOH)(tcc)]B(Ph)4 (ellipsoids are drawn at the 30% probability level, hydrogen atoms and counterions
are omitted for clarity). (b) Computed structure of [Fe(L)(dbc)]+ with the computed spin densities (spin densities: Fe 3.783, N3 0.044,
N7 0.038, O8 0.27, O7 0.34, C1 0.094, C2 0.175).
Eur. J. Inorg. Chem. 2011, 5242–5249
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5243

P. Comba, H. Wadepohl, S. Wunderlich
FULL PAPER
listed in Table 1. The FeIII-donor distances are between optimization. The computed structural parameters, also
1.91–2.27 Å, as expected for high-spin FeIII [42]
;
the bond to shown in Table 1, are in acceptable agreement with the ex-
N7 is, as in other bispidine complexes, significantly longer perimental data.[46] As observed experimentally, a high-spin
than the bond to the other tertiary amine, N3 (2.27 versus electronic configuration (S = 5/2) is predicted by DFT
2.18 Å).[26,30] However, the two FeIII–Ocatecholate bonds are (the low-spin state (S = 1/2) is significantly higher in
with 1.91 and 1.95 Å (trans to N3 and trans to N7, respec- energy (50.9 kJmol–1); the intermediate-spin state is at
tively), very similar in length, which is in sharp contrast to 57.3 kJmol–1), and this was expected on the basis of the
the corresponding Jahn–Teller-active CuII complex, in large and rigid ligand cavity.[26,30]
which the CuII–Ocatecholate bond trans to N7 is strongly
elongated, and for which a second isomer with mono-
dentate catecholate has also been isolated.[43] The structure
reported here supports our assumption that the generally
long bond to N7 is a ligand-enforced structural ef-
Catechol Dioxygenase Activity
Since FeIII–bispidine complexes in general are relatively
fect,[30,44,45] and this elongation obviously has little in-
fluence on the bonding to the bidentate catecholate co-
ligand. The emerging relatively symmetric bonding of cate-
cholate to the FeIII center was expected to have a significant
influence on the catecholase and dioxygenase activity (vide
infra).
unstable, high-spin [Fe(L)(OTf)2] was used as a precatalyst,
and the FeIII/catecholate species was prepared in situ. The
reactive high-spin [Fe(L)(dbc)]+ complex was obtained by
addition of an equimolar amount (stoichiometric reactions)
or an excess amount (catalytic reactions, 10 equiv.) of dbc2–
[addition of the catechol derivative and 2 equiv. of NEt3
(NEt3 = triethylamine) in dry and degassed MeCN (MeCN
= acetonitrile) as solvent] to a solution of the FeII precur-
sor; the oxidation reaction was then started by the addition
of O2 (1 atm). As usual in catechol dioxygenase studies, af-
ter completion the reaction was quenched with HCl (reac-
tion time of approx. 10 h), and the organic products were
then extracted into Et2O (Et2O = diethyl ether) and charac-
terized by GC/GC–MS analyses. Products of the dbc2– oxi-
dation and oxygenation (see Scheme 3) in dry MeCN are (a)
3,5-di-tert-butyl-benzoquinone (dbq, the “catechol oxidase“
product), (b) muconic acid anhydride (the immediate “cate-
chol dioxygenase“ product) (i.e., the muconic acid precur-
sor in dry solution), and (c) traces (at most) of methyl 3,5-
di-tert-butyl-2-oxofuran-5-carboxylate [derived from lac-
tone formation from the muconate; yields of the stoichio-
metric reactions (in %) and of the catalytic transformations
(TON, TONmax = 10) are also given in Scheme 3].[47] As
expected (redox potential, symmetric bonding of the sub-
strate), the iron–bispidine-catalyzed reactions mainly pro-
duce the benzoquinone product. On account of the high
Table 1. Structural parameters (distances in Å, angles in °) of
[Fe(L·MeOH)(tcc)][B(Ph)4] (X-ray diffraction) and [Fe(L)(X)]+
(DFT, X = tcc, dbc).
[FeIII(L)(tcc)]+
S = 1/2 3/2 5/2
[FeIII(L)(dbc)]+
S = 1/2 3/2 5/2
exp.
Fe–N3
Fe–N7
Fe–N1
Fe–N2
Fe–O8
Fe–O7
2.03
2.15
2.00
2.00
1.95
1.96
88.5
83.5
83.5
176.4
94.2
177.3
166.2
50.9
2.11
2.16
2.24
2.24
1.87
1.89
86.8
78.0
78.0
2.22 2.176(3) 2.03
2.29 2.273(4) 2.17
2.20 2.103(3) 1.99
2.20 2.131(3) 1.99
2.03 1.914(3) 1.98
2.07 1.946(3) 1.96
2.12 2.24
2.16 2.32
2.27 2.20
2.27 2.20
1.90 2.02
1.89 2.07
86.8 83.7
77.8 77.2
77.8 77.2
178.1 176.1
94.1 96.3
179.2 180.0
154.4 153.3
68.2 0.0
N3–Fe–N7
N3–Fe–N1
N3–Fe–N2
N3–Fe–O8
N3–Fe–O7
N7–Fe–O7
N1–Fe–N2
84.5 80.8(1)
77.2 77.5(1)
77.2 76.8(1)
88.3
83.6
83.6
179.4 178.2 171.8(1) 175.8
94.9 99.5 104.0(1) 93.4
178.3 176.0 175.1(1) 178.3
154.9 153.1 153.5(1) 166.2
ΔG[a] [kJmol–1
]
57.3
0.0
56.2
[a] All energies quoted refer to the lowest-lying spin state, which
was set as the origin.
The structural and electronic properties of the catechol- redox potential of the catalyst, even with O2 as oxidant, the
ate FeIII site were also evaluated by a DFT-based structure oxidation of dbc2– is efficient but primarily produces the
Scheme 3. The iron–bispidine-based oxidation of H2dbc.
5244
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 5242–5249

Oxidation versus Dioxygenation of Catechol
benzoquinone (dbq) product. The formation of dbq as the kO = 3.77ϫ10–1
m
–1 s–1 for the decay of [FeIII(L–N4Me2)-
2
main product also follows experimentally from reaction (dbc)]+ in oxygen-saturated MeCN at 25 °C[8]), the reactiv-
1
control of a catalysis experiment by H NMR spectroscopy ity of the iron–bispidine-based model system is relatively
(see Figure S1 in the Supporting Information).
low. The energy of the LMCT transitions has been corre-
In a stoichiometric reaction of the FeIII–catecholate com- lated to the Lewis acidity of the FeIII center (i.e., the ease of
plex ([FeII(L)(OTf)2]/H2dbc/NEt3, 1:1:2) with O2 (1 atm; see electron transfer of the catecholate substrate to the catalytic
Figure 2; the immediate formation of [FeIII(L)(dbc)]+ is not center has been assumed to be a measure for the catechol
shown in this figure), there is a relatively slow and clean dioxygenase activity).[7,50,51] With the lower-energy transi-
monophasic reaction (isosbestic point, pseudo-first-order tion at 610 nm, compared to that of [FeIII(L–N4Me2)-
kinetics), which leads to a decay of the relatively intense (dbc)]+ at 553 nm, the observed reactivities do not seem to
electronic transitions at 610 and 910 nm, attributed to be correlated to the Lewis acidities. The relatively sluggish
FeIII–dbc ligand-to-metal charge-transfer (LMCT) transi- reaction and the low yield of cleavage versus oxidation
tions and a concomitant increase of a band at 390 nm (ab- product (dioxygenase versus catecholase activity) might
sorption of dbq).[3] The pseudo-first-order rate constant of rather be due to efficient electron transfer (powerful oxi-
dbq formation is kobs = 3.43ϫ10–4 s–1, the half-life of the dant), and this was expected on the basis of the high redox
high-spin [FeIII(L)(dbc)]+ complex is t1/2 = 0.56 h (t1/2
=
potential and efficient outer-sphere electron transfer.[34]
0.693/kobs) and, with the known concentration of O2 in
MeCN ([O2] = 8.1 mm),[48] the second-order rate constant
is kO = 4.23ϫ10–2 –1 s–1. From Figure 2 it also emerges
m
2
DFT-Based Mechanistic Studies
that the rate of decay of the FeIII–catecholate complex is
close to identical to the rate of formation of dbq (kobs
=
DFT was used to further analyze the reaction mecha-
2.84ϫ10–4 s–1 versus 3.43ϫ10–4 s–1). On the basis of the nism of the iron–bispidine-catalyzed oxidation of catechol.
selective formation of dbq (90%, catecholase reactivity) and The experimentally observed and computed structural
muconic acid anhydride (10%, catechol dioxygenase reac- properties and computed energies of [FeIII(L)(dbc)]+ in the
tivity, see Scheme 3), it follows that the approximate various possible spin states are shown in Table 1. Based on
pseudo-first-order rate for catechol cleavage is kobs
=
the relative energies, the FeIII catalyst has a high-spin elec-
tronic configuration (S = 5/2), and the low-spin state (S
= 1/2) is significantly higher in energy (56.2 kJmol–1; the
intermediate-spin configuration is at 68.2 kJmol–1). This is
in agreement with the experimental data of the tcc-based
model system (see above), with computational work re-
ported for the enzyme,[52] and with a recent DFT-based
analysis of another low-molecular-weight model system.[53]
The computed structural and energetic parameters of
[FeIII(L)(dbc)]+ are similar to those of [FeIII(L)(tcc)]+: in
their high-spin electronic ground state, the largest deviation
in the metal–donor distances is for the relatively weak Fe–
N7 bond (0.03 Å). This indicates that the experimental
structure of [FeIII(L)(tcc)]+ with the deactivated tcc sub-
strate is a sensible model for the catalytically active
[FeIII(L)(dbc)]+ complex.
0.38ϫ10–2 –1 s–1.
m
Shown in Figure 1 (b) is a plot of the computed structure
of [FeIII(L)(dbc)]+ with a visualization of the spin densities.
Based on the experimentally observed similarity of the
bond lengths from the FeIII center to the two catecholate
oxygen atoms (O7 and O8), confirmed in the DFT-opti-
mized structures, only little spin density is expected on the
catecholate substrate, and this is shown to be distributed
symmetrically between the two catecholate carbon atoms
C1 and C2 (see Figure 1, b) (i.e., there is only little
FeII–semiquinonate (localized radical) as opposed to FeIII
–
Figure 2. (a) Observed spectral changes in the stoichiometric reac-
tion of the catalyst with dbc ([Fe(L)(OTf2)], 1ϫ10–4 m; H2dbc, 1
equiv.; NEt3, 2 equiv.; O2, 1 atm; dry MeCN; 298 K). (b) Observed
absorption versus time traces of the same experiment at 390 nm
catecholate character (see Table S3 in the Supporting Infor-
mation). The expectation therefore is that [Fe(L)(dbc)]+ is
(increasing absorption, formation of dbq), and at 610 nm (decreas- not among the most efficient intradiol–catechol dioxygen-
ing absorption, decay of the FeIII–catecholate complex) as a func-
tion of the time.
ase mimics, and this is what was shown by the experiments
(see above and Scheme 3). This is amplified by the fact that
Compared to the enzymatic reaction (kO
=
the bispidine–FeIII/II redox potential is relatively high (a re-
2
2.5ϫ105 m–1 s–1)[49] and other biomimetic reactions (e.g., sult of the large and rigid bispidine cavity[54]) [i.e., the driv-
Eur. J. Inorg. Chem. 2011, 5242–5249
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5245

P. Comba, H. Wadepohl, S. Wunderlich
FULL PAPER
ing force for catecholase activity (electron transfer) is large tally). This is in contrast to a recent DFT-based analysis of
and catechol degradation (intradiol cleavage) therefore is an efficient intradiol–catechol–dioxygenase model system,
disfavored].
in which the two spin states (quartet and sextet) of the per-
The formation of an alkyl–peroxo–iron(III) intermediate oxo intermediate are very close in energy and which both
along the reaction coordinate for the intradiol–catechol have FeIII–semiquinone character.[53] The large structural
cleavage (Schemes 1 and 3) leads to two possible isomers of difference between the two intermediates in the present case
the dioxygen adduct [Fe(L)(O2)(dbc)]+ (see Figures S4 and (FeIII–semiquinone versus FeII–benzoquinone) is the basis
S5, Table S2 in the Supporting Information). The more for the relatively large energy difference and, unfortunately,
stable isomer has Fe–OO trans to N7 and is, compared to is one of the reasons for the inefficiency with respect to
its isomer (Fe–OO trans to N3), 22.9 kJmol–1 more stable dioxygenase activity.
in its S = 5/2 spin state; the S = 3/2 state is destabilized by
The attack of triplet dioxygen at the high-spin-config-
93.2 kJmol–1 for the trans to N7 isomer and by ured enzyme–substrate complex of protocatechuate-3,4-di-
58.5 kJmol–1 for the trans-to-N3 isomer; the S = 1/2 config- oxygenase was analyzed in detail, and the spin-forbidden
uration is at 117.6 kJmol–1 for the trans-to-N7 isomer and reaction between triplet dioxygen and singlet catecholate
at 107.3 kJmol–1 for the trans-to-N3 isomer. On the basis was shown to be initiated by iron-centered redox processes:
of the more stable trans-to-N7 isomer, there are remarkable electrophilic attack by O2 is facilitated in the FeII–semi-
differences between the preferred S = 5/2 and the higher quinone intermediate (see Scheme 4).[52] This pathway is
energy S = 3/2 and S = 1/2 electronic states: (i) in the S = disfavored in our bispidine complexes with respect to a fur-
1/2 and S = 3/2 electronic configurations, the catechol C–O ther iron-based electron transfer and benzoquinone forma-
bond lengths are 1.40 and 1.39 Å, respectively, and this is tion.
typical for a C–O single bond. In the electronic ground
The simplified reaction coordinate is visualized in Fig-
state, the two C–O bond lengths are 1.28 Å, thus indicating ure 3.[55] On the basis of the transformation of the high-
a distinct carbonyl character; (ii) whereas the Fe(OO)– spin FeIII–catecholate complex to [FeIII(L)–OH]+, the for-
C
catechol distance in the S = 1/2 or S = 3/2 states is at 1.54 Å,
thus indicating a relatively strong interaction, in the S =
5/2 ground-state electronic configuration the Fe(OO)–
Ccatechol distance is significantly longer (2.37 Å) [i.e., there
is only a very weak interaction of the iron-bound (activated)
dioxygen molecule with the substrate]. As a consequence, in
the energetically preferred S = 5/2 electronic state, the spin
density on the exo-oxygen atom is +0.70 and on the sub-
strate carbon atom it is –0.07 [i.e., this intermediate (i) has
little carbon-based radical character, (ii) has little prefer-
ence for C–O bond formation, and (iii) is largely FeII–
benzoquinone in character]. Accordingly, the S = 5/2 state
does not lead to the products of the intradiol cleavage and
is close to the product of the catecholase reaction (i.e. the
Figure 3. Simplified reaction coordinate on the basis of DFT calcu-
two-electron oxidation, which is also observed experimen- lations of the oxygenation of [Fe(L)(dbc)]+.[55]
Scheme 4. Simplified MO-based scheme to show the spin-allowed pathways for the transfer of three electrons for dioxygenase and
catecholase activity (see also Scheme 1 and Scheme 3).[52]
5246
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 5242–5249

Oxidation versus Dioxygenation of Catechol
Oxidative Cleavage of 3,5-Di-tert-butylcatechol: In a typical reac-
tion, 0.1 mmol of [Fe(L)(OTf)2] was mixed with 1–10 equiv. of 3,5-
di-tert-butylcatechol in MeCN (10 mL, H2O-free, Ar, 298 K), fol-
lowed by the addition of NEt3 (2 equiv.). The solution was stirred
for a further 5 min, and dioxygen was then bubbled through the
solution for 3 min. The reaction was quenched after 9 h with 0.01 m
HCl (5 mL), then the organic products were extracted with Et2O
(5 mL three times) and dried with Na2SO4. Naphthalene was added
as the internal standard, and the mixture was analyzed by GC.
The retention for the product peaks were compared with authentic
compounds, and their identity was confirmed by GC–MS. All reac-
tions were done in duplicate; the reported data are the average of
these reactions. For kinetic studies, the reaction solution was di-
luted to 1ϫ10–4 m based on the iron catalyst.
mation of muconic acid anhydride is an exergonic process
(ΔG = –185.9 kJmol–1). On this pathway, the iron center
keeps its oxidation state (+III) over the entire process, and
the [FeIII(L)(O2)(dbc)]+ intermediate in its high energy S =
1/2 and S = 3/2 electronic configurations (172.2 kJmol–1,
trans-to-N7 isomer) leads to intradiol products. However,
the formation of benzoquinone (dbq) on the S = 5/2 spin
surface has
a
much lower-energy intermediate
(90.8 kJmol–1, trans-to-N3 isomer), and this supports the
experimental observations.[56]
Conclusion
Synthesis of [Fe(L·MeOH)(tcc)]B(Ph)4]: NEt3 (2 equiv.) was added
to a solution with equal amounts of [Fe(L)(OTf)2] (1ϫ10–2 m), tet-
rachlorocatechol (H2tcc), and Na[B(Ph)4] in MeOH under ambient
conditions. Oxidation to iron(III) was immediately observed under
air. Blue single crystals were obtained by slow evaporation of the
solvent at 0 °C. ESI-MS: calcd. for [Fe(L)(OH2)(tcc)]+: 756.00107;
The reactivity of the iron–bispidine-based system is lower
(up to 2 orders of magnitude) than that of other model
complexes, which are up to 5 orders of magnitude less reac-
tive than the natural systems. This is believed to be due
to the efficient electron transfer and relatively high redox
potential of the bispidine–iron site, which leads to a prefer-
ence for catecholase over intradiol dioxygenation activity
(approx. 90 versus 10%). This is also supported by the ob-
served relatively low energy of the LMCT electronic transi-
tions. Support for the preference for catecholase over cate-
chol dioxygenase activity comes from the DFT-based and
the experimental structural analyses, which show that there
is only little spin density transferred from the FeIII center
to the catecholate substrate (i.e. there is little FeII–semi-
quinonate-localized-radical and primarily FeIII–catecholate
and FeII–quinone character in the catalytically active form).
An important conclusion therefore is that, whereas the
iron–bispidine complexes are active halogenation catalysts
as well as strong oxidants, they are not very active in cate-
chol dioxygenase. It is not unlikely that catechol and humic
acid degradation require a different type of catalytic center
than subsequent halogenation of the degradation products.
found: 756.00111. UV/Vis (CH3CN): λmax = 531 nm (ε531
=
2190 m–1 cm–1), 805 nm (ε805 = 980 m–1 cm–1). CV: E = –0.03 V ver-
sus SCE [0.1 m nBu4N(ClO4)]. Crystal data: C54H50BCl4FeN4O8,
Mr=1091.44, 0.400.100.10 mm3, monoclinic, space group P21/n, a
=
12.1685(10) Å, b = 28.234(2) Å, c = 14.7596(12) Å, β =
94.573(2)°, 5054.8(7) Å3, Z = 4, F(000) = 2260, T = 100(2) K, θ
range 1.8 to 25.0°. Index ranges (indep. set): –14ՅhՅ14,
0ՅkՅ33, 0ՅlՅ17. Reflections measd: 84819, indep.: 8939 (Rint
= 0.0853). Final R indices [IϾ2σ(I)] R1 = 0.0635, wR2 = 0.1664,
GoF = 1.126. Intensity data were collected with a Bruker AXS
Smart 1000 CCD diffractometer (Mo-Kα radiation, graphite mono-
chromator, λ = 0.71073 Å). Data were corrected for air and detec-
tor absorption, Lorentz and polarization effects;[57] absorption by
the crystal was treated with a semiempirical multiscan method.[58,59]
The structure was solved by conventional direct methods,[60,61] and
refined by full-matrix least-squares methods based on F2 against
all unique reflections.[60,61] All non-hydrogen atoms were given an-
isotropic displacement parameters. Hydrogen atoms were input at
calculated positions and refined with a riding model.
CCDC-837082 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
Experimental Section
General: Chemicals (Aldrich, Fluka) and solvents were of highest
possible grade and used as purchased. Mass spectra were obtained
with a Bruker ApexQe hybrid 9.4 FT-ICR or Finnigan TSW 700
instrument. Elemental analyses were performed by the analytical
laboratories of the chemical institutes of the University of Heidel-
berg. Product analyses by GC were done with a Varian 3900 instru-
ment equipped with a ZB-1701 column. UV/Vis spectra and kinetic
studies were performed at 298 K with a J&M Tidas II spectropho-
tometer with a quartz cells (path length 10 mm). NMR spectra at
Computational Studies: All DFT calculations were performed with
the Jaguar 6.5 program package[62] unless otherwise specified. The
B3LYP functional[63,64] and LACVP basis set (double ζ, with a Los
Alamos effective core potential for the Fe center, and 6-31G for
the other atoms) were used.[65] All intermediates were confirmed
by frequency calculations using Gaussian 03.[66] Single-point calcu-
lations were performed on the B3LYP/LACVP-optimized geome-
tries using the LACV3P++** basis set (LanL2DZ on the Fe center
200 MHz (1H) were obtained with a Bruker ARX 200 instrument. and 6-311++G** on the other atoms). The energies reported are
The solvent was used as internal reference; δ in ppm, J in Hz. Elec-
trochemical measurements were conducted with a CH Instruments
660D workstation with a three-electrode setup that consisted of a
glassy carbon working electrode, a Pt-wire auxiliary electrode, and
an Ag/AgCl reference electrode [0.01 m Ag+, 0.1 m Bu4N(PF6)]. The
potentials were determined in MeCN and converted versus SCE by
adding 290 mV.
those calculated at the B3LYP/LACV3P++** level and include
zero-point and free-energy corrections derived from the B3LYP/
LACVP calculations. A simplified model system was used in all
calculations, in which the ester groups on the ligand backbone were
replaced by hydrogen atoms.
Supporting Information (see footnote on the first page of this arti-
cle): NMR spectroscopic control of the oxidation reaction; com-
puted structural data and spin densities. Structural parameters (dis-
tances in Å, angles in °) of [Fe(L·MeOH)(tcc)][B(Ph)4] (X-ray dif-
fraction) and [Fe(L)(X)]+ (DFT, X = tcc, dbc).
Synthesis of the Ligand and the Iron–Bispidine Complexes: The bis-
pidine ligands and the corresponding FeII complex [FeII(L)(OTf)2]
were prepared as described before.[40]
Eur. J. Inorg. Chem. 2011, 5242–5249
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5247

P. Comba, H. Wadepohl, S. Wunderlich
FULL PAPER
M. Costas, L. Que Jr, J. Am. Chem. Soc., manuscript in prepa-
ration.
Acknowledgments
[36] J. Bautz, M. Bukowski, M. Kerscher, A. Stubna, P. Comba, A.
Lienke, E. Münck, L. Que Jr, Angew. Chem. Int. Ed. 2006, 45,
5681; Angew. Chem. 2006, 118, 5810.
Generous financial support by the Deutsche Forschungsgemein-
schaft (DFG) is gratefully acknowledged.
[37] J. Bautz, P. Comba, C. Lopez de Laorden, M. Menzel, G. Raja-
raman, Angew. Chem. 2007, 119, 8213; Angew. Chem. Int. Ed.
2007, 46, 8067.
[1] J. D. Lipscomb, A. Orville, in: Metal Ions in Biological Systems,
vol. 28 (Ed.: H. Sigel, A. Sigel), Marcel Dekker, New York,
1992, p. 243.
[38] P. Comba, M. Maurer, P. Vadivelu, J. Phys. Chem. A 2008, 112,
13028.
[2] L. Que Jr, in: Iron Carriers and Iron Proteins (Ed.: T. M.
Loehr), VCH Publishers, New York, 1989, p. 467.
[3] H. J. Krüger, in: Biomimetic Oxidations Catalyzed by Transition
Metal Complexes (Ed.: E. B. Meunier), Imperial College Press,
Oxford, 2000, p. 363.
[4] E. I. Solomon, M. Y. M. Pau, R. K. Hocking, in: Computa-
tional Inorganic and Bioinorganic Chemistry (Eds.: E. I. Solo-
mon, R. A. Scott, R. B. King), Wiley, New York, 2009, p. 241.
[5] D. H. Ohlendorf, A. M. Orville, J. D. Lipscomb, J. Mol. Biol.
1994, 244, 586.
[39] P. Comba, M. Maurer, P. Vadivelu, Inorg. Chem. 2009, 48,
10389.
[40] H. Börzel, P. Comba, K. S. Hagen, M. Merz, Y. D. Lampeka,
A. Lienke, G. Linti, H. Pritzkow, L. V. Tsymbal, Inorg. Chim.
Acta 2002, 337, 407.
[41] L. Que Jr, J. D. Lipscomb, R. Zimmermann, E. Münck, N. R.
Orme-Johnson, W. H. Orme-Johnson, Biochim. Biophys. Acta
Enzymol. 1976, 452, 320.
[42] A. G. Orpen, L. Brammer, F. H. Allen, O. Kennard, D. G. Wat-
son, R. Taylor, J. Chem. Soc. Dalton Trans. 1989, 1.
[43] H. Börzel, P. Comba, H. Pritzkow, Chem. Commun. 2001, 97.
[44] C. Bleiholder, H. Börzel, P. Comba, R. Ferrari, A. Heydt, M.
Kerscher, S. Kuwata, G. Laurenczy, G. A. Lawrance, A. Lienke,
B. Martin, M. Merz, B. Nuber, H. Pritzkow, Inorg. Chem. 2005,
44, 8145.
[6] M. G. Weller, U. Weser, Inorg. Chim. Acta 1985, 107, 243.
[7] G. H. Jang, D. D. Cox, L. Que Jr, J. Am. Chem. Soc. 1991, 113,
9200.
[8] W. O. Koch, H.-J. Krüger, Angew. Chem. 1995, 107, 2928; An-
gew. Chem. Int. Ed. Engl. 1995, 34, 2671.
[9] M. Duda, M. Pascaly, B. Krebs, Chem. Commun. 1997, 835.
[10] T. Funabiki, T. Yamazaki, A. Fukui, T. Tanaka, S. Yoshida,
Angew. Chem. 1998, 110, 527; Angew. Chem. Int. Ed. 1998, 37,
513.
[45] K. Born, P. Comba, R. Ferrari, S. Kuwata, G. A. Lawrance,
H. Wadepohl, Inorg. Chem. 2007, 46, 458.
[46] For comparison, a geometry optimization and vibrational
analysis were also done with the TZV* basis set; these results
are given in the Supporting Information (Table S1); geometry
optimization with Jaguar 6.5, B3LYP, TZV*; frequency calcu-
lations with Gaussian 09. The important observations are that
the agreement between computed and observed structural pa-
rameters, and the relative bond lengths and spin distributions
(vide infra) do not strongly depend on the basis set.
[47] Note that, under these conditions, there is also some autooxi-
dation of dbcH2 to dbq.
[48] D. T. Sawyer, Oxygen Chemistry, Oxford University Press, New
York, 1991.
[49] T. A. Walsh, D. P. Ballou, R. Mayer, L. Que Jr, J. Biol. Chem.
1983, 258, 1422.
[50] M. G. Weller, U. Weser, J. Am. Chem. Soc. 1982, 104, 3752.
[51] D. D. Cox, L. Que Jr, J. Am. Chem. Soc. 1988, 110, 8085.
[52] M. Y. M. Pau, M. I. Davis, A. M. Orville, J. D. Lipscomb, E. D.
Solomon, J. Am. Chem. Soc. 2007, 129, 1944.
[53] V. Georgiev, H. Noack, T. Borowski, M. R. A. Blomberg,
P. E. M. Siegbahn, J. Phys. Chem. B 2010, 114, 5878.
[54] P. Comba, A. F. Sickmüller, Inorg. Chem. 1997, 36, 4500.
[55] Note that further intermediates along the reaction coordinate
and the corresponding transition states have not been opti-
mized.
[11] P. Mialane, L. Tchertanov, F. Banse, J. Sainton, J.-J. Girerd,
Inorg. Chem. 2000, 39, 2440.
[12] M. Xin, T. D. H. Bugg, J. Am. Chem. Soc. 2008, 130, 10422.
[13] M. K. Panda, A. John, M. M. Shaikh, P. Ghosh, Inorg. Chem.
2008, 47, 11847.
[14] M. Wagner, C. Limberg, T. Tietz, Chem. Eur. J. 2009, 15, 5567.
[15] T. Dhanalakshmi, E. Suresh, M. Palaniandavar, Dalton Trans.
2009, 8317.
[16] R. Mayilmurugan, K. Visvaganesan, E. Suresh, M. Palanian-
davar, Inorg. Chem. 2009, 48, 8771.
[17] P. Brombach, H. H. Richnow, M. Kästner, A. Fischer, Appl.
Microbiol. Biotechnol. 2010, 86, 839.
[18] R. Matta, K. Hanna, T. Kone, S. Chiron, Chem. Eng. J. 2008,
144, 453.
[19] N. Raffard, V. Balland, J. Simaan, S. Létard, M. Nierlich, K.
Miki, F. Banse, E. Anxolabéhère-Mallart, J.-J. Girerd, Comptes
Rendu Chim. 2002, 5, 99.
[20] J. N. Chheda, G. W. Huber, J. A. Dumesic, Angew. Chem. 2007,
119, 7298; Angew. Chem. Int. Ed. 2007, 46, 7164.
[21] J. B. Binder, R. T. Raines, J. Am. Chem. Soc. 2009, 131, 1979.
[22] R. v. Glasow, P. J. Crutzen, in: Treatise on Geochemistry, vol.
4, Elsevier Pergamon, Amsterdam, 2003, p. 21.
[23] S. A. Huber, K. Kotte, H. F. Schöler, J. Williams, Environ. Sci.
Technol. 2009, 43, 4934.
[24] P. Comba, S. Wunderlich, Chem. Eur. J. 2010, 16, 7293.
[25] C. Mannich, P. Mohs, Ber. Dtsch. Chem. Ges. 1930, B63, 608.
[26] P. Comba, M. Kerscher, W. Schiek, Prog. Inorg. Chem. 2007,
55, 613.
[27] P. Comba, B. Nuber, A. Ramlow, J. Chem. Soc., Dalton Trans.
1997, 347.
[28] P. Comba, Coord. Chem. Rev. 2000, 200–202, 217.
[29] P. Comba, A. Lienke, Inorg. Chem. 2001, 40, 5206.
[30] P. Comba, M. Kerscher, M. Merz, V. Müller, H. Pritzkow, R.
Remenyi, W. Schiek, Y. Xiong, Chem. Eur. J. 2002, 8, 5750.
[31] P. Comba, W. Schiek, Coord. Chem. Rev. 2003, 238–239, 21.
[32] P. Comba, M. Merz, H. Pritzkow, Eur. J. Inorg. Chem. 2003,
1711.
[33] P. Comba, Structure and Function, Springer, Heidelberg, 2010.
[34] P. Comba, S. Fukuzumi, H. Kotani, S. Wunderlich, Angew.
Chem. Int. Ed. 2010, 49, 2622.
[56] Note that along the reaction coordinate, transition states could
not be located. Due to the relatively high ΔG values of the [(L)
FeIII(O2)(dbc)]+ intermediates the energy barriers are expected
to be comparably low.
[57] SAINT, Bruker AXS, Madison, Wisconsin, 1997–2008.
[58]
[59]
R. H. Blessing, Acta Crystallogr. Sect. A 1995, 51, 33.
G. M. Sheldrick, SADABS 2004–2008 and SADABS 2008/1,
Bruker AXS, Madison, Wisconsin.
[60]
G. M. Sheldrick, SHELXL-97, University of Göttingen, Ger-
many, 1997.
[61]
[62]
G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112.
Schrödinger, Jaguar v. 5.5, Jaguar v. 6.5, Schrödinger LLC,
New York, NY, 2005.
[63]
[64]
[65]
[66]
A. D. Becke, J. Chem. Phys. B 1993, 98, 5648.
C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
J. P. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 99.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery Jr, T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tom-
[35] D. Wang, K. Ray, E. R. Farquhar, M. J. Frisch, L. Gomez,
T. A. Jackson, M. Kerscher, S. Pandian, A. Waleska, P. Comba,
5248
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 5242–5249

Oxidation versus Dioxygenation of Catechol
asi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratch-
ian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gom-
perts, R. E. Stratmann, O. Yazyev, A. Austin, R. Cammi, C.
Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A.
Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dap-
prich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q.
Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov,
G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayak-
kara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian 03, revision
B.03, Gaussian Inc., Wallingford CT, 2003.
Received: July 31, 2011
Published Online: October 31, 2011
Eur. J. Inorg. Chem. 2011, 5242–5249
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5249