Table 1. Oxidation of adamantane by H2O2 catalyzed by FeIII–dpaqR complexes.[a]
tammograms of [FeIII(dpaqR)
ACHTUNGTRNENUNG ACHTUNRGTEGUN(NN MeCN)]ACHTUNGTREN(NNGU ClO4)2 deriv-
atives measured in MeCN/H2O (9:1) clearly reflect
the electronic nature of the ligand substituents: the
redox potentials corresponding to the Fe3+/Fe2+
and Fe4+/Fe3+ couples fall in the ranges ꢀ0.10 to
0.09 V versus Fc+/Fc and 0.78 to 1.33 V versus
Fc+/Fc, respectively, and both follow linear rela-
tionships with the Hammett substituent constants,
FeIII–dpaqR
Product yield [%][b]
TON[c]
38/28[d]
1-ol
1,3-diol
2-ol
2-one
R=OMe
45.1ꢂ0.8
42.3ꢂ1.1
42.6ꢂ3.0
47.3ꢂ2.4
10.1ꢂ0.2
12.5ꢂ1.5
17.2ꢂ1.6
25.2ꢂ1.6
5.4ꢂ0.3
4.0ꢂ0.1
4.3ꢂ0.4
4.4ꢂ0.4
2.2ꢂ0.1
2.1ꢂ0.1
2.2ꢂ0.2
3.0ꢂ0.3
75.1ꢂ1.7
75.7ꢂ4.4
86.0ꢂ7.0
108.1ꢂ6.6
21.8ꢂ0.7
26.7ꢂ0.8
27.3ꢂ1.0
29.2ꢂ0.7
with 1 values of 176 and 480 for the Fe3+/Fe2+ and R=H
Fe4+/Fe3+ couples, respectively (Figure 2 and S1,
Supporting Information). As the substituent group
becomes more electron-donating, the redox poten-
tials become more negative, probably because of
the stronger coordination of the amido group to the
iron center.
R=Cl
R=NO2
[a] Catalyst/H2O2/substrate=1:120:100. H2O2 was introduced by a syringe pump over
30 min. All reactions were performed in triplicate. [b] Based on substrate. [c] TON:
([1-adamantanol]+2ꢁ[1,3-adamantandiol]+[2-adamantanol]+2ꢁ[2-adamantanone])/
[FeIII–dpaqR]. [d] 38/28: 3ꢁ([1-adamantanol]+[1,3-adamantandiol])/([2-adamanta-
nol]+[2-adamantanone]).
HꢁCl<NO2; Table 1). The 38/28 value for FeIII–dpaqNO
2
ranks in the top group among nonheme iron catalysts
(Table S1, Supporting Information).[4n,7] In addition, the
FeIII–dpaqNO complex showed the best performance in
2
terms of TON. The electron-deficient ligand might suppress
the decomposition of the catalyst, similarly to the situation
for electron-deficient metalloporphyrins exhibiting high
TONs for catalytic oxidation reactions.[8] Thus, FeIII–dpaqNO
2
complex showed the best selectivity in the adamantane hy-
droxylation among the series of FeIII–dpaqR derivatives. If
the principle of reactivity versus selectivity is operative in
Figure 2. Hammett plot for the redox potentials of FeIII–dpaqR complexes
(R: OMe, H, Cl, and NO2). Filled circles and squares indicate Fe3+/Fe2+
and Fe4+/Fe3+ couples, respectively. The values are averages of oxidation
and reduction peak potentials determined by differential pulse voltam-
metry (see Figure S1, Supporting Information).
the current series, these results indicate that the dpaqNO
2
ligand produces the mildest oxidant toward alkane hydroxy-
lation. However, these results are not understood simply
from the viewpoint of redox potentials, because the
electron-deficient dpaqNO ligand should destabilize the
2
high-valent iron species, and hence, produce the strongest
oxidant.
The Fe4+/Fe3+ redox potentials were obtained by means
of spectropotentiometry for oxoiron(IV) complexes support-
ed by a series of neutral pentadentate ligands, and they
range from 0.83 to 1.23 V versus Fc+/Fc in MeCN with 0.1m
Selectivity of epoxidation: To better understand the proper-
ties of the oxidants generated from FeIII–dpaqR complexes
with H2O2, we performed intramolecular competitive epoxi-
dation reactions using (2E)-3,7-dimethyl-2,6-octadien-1-yl 4-
methoxybenzoate as a substrate containing two C=C bonds
with different electron densities at the 2- and 6-positions.
H2O.[5] These values correspond to oxoiron(IV)/oxoiron
ACTHNUTRGNE(NUG III)
couples, whereas the Fe4+/Fe3+ redox potentials obtained in
this study probably correspond to aquairon(IV)/aquairon-
ACHTUNGTRENNUNG(III) couples. This difference is a possible reason why similar
1
Fe4+/Fe3+ redox potentials were obtained despite the exis-
tence of anionic amido ligation in Fe–dpaqR complexes.
Product analysis by H NMR spectroscopy revealed that ep-
oxidation occurred more preferentially at the C=C bond lo-
cated at the 6-position than at the one located at the 2-posi-
tion (Table 2). The highest selectivity was obtained with the
FeIII–dpaqOMe derivative, and as the substituent group
became more electron-withdrawing, the selectivity de-
ꢀ
Selectivity of C H hydroxylation: Adamantane was chosen
as the substrate to evaluate the selectivity of hydroxylation
catalyzed by FeIII–dpaqR complexes, because it contains
2
ꢀ
twelve secondary and four tertiary C H bonds having
creased (R: OMe>Cl>H>NO2). Thus, FeIII–dpaqNO com-
ꢀ
different bond dissociation energies (BDEs: 28 C H,
plex showed the worst selectivity in the intramolecular com-
petitive epoxidation among the series of FeIII–dpaqR deriva-
tives, although it showed the best selectivity in alkane hy-
droxylation (Figure 3).
ꢀ1 [6]
100.2 kcalmolꢀ1; 38 C H, 96.2 kcalmol ). The product dis-
ꢀ
tribution was examined under conditions whereby the ratio
of catalyst/H2O2/substrate was 1:120:100. Product analysis
by gas chromatography revealed that the preference for the
ꢀ
ꢀ
hydroxylation of a tertiary C H bond over a secondary C
H bond (38/28) increased up to 29.2 as the substituent group
R became more electron-withdrawing in nature (R: OMe<
Proposed mechanism: Similar opposing effects of electron
donation from axial ligands on an oxo-transfer reaction
versus a H-atom abstraction reaction have been reported
14698
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 14697 – 14701