Catalytic hydroxylation of alkanes by immobilized mononuclear iron carboxylate

Article abstract of DOI:10.1039/a706178a

Mononuclear iron carboxylate complex immobilized and isolated on a modified silica surface (1) catalyzes oxidation of hexane to a mixture of hexan-1-ol, -2-ol and -3-ol without ketone formation in the presence of mercaptan, acetic acid, triphenylphosphine and O₂ at ambient conditions.

Full text of DOI:10.1039/a706178a

View Article Online / Journal Homepage / Table of Contents for this issue
Catalytic hydroxylation of alkanes by immobilized mononuclear iron
carboxylate
Keiji Miki* and Takeshi Furuya
National Institute for Resources and Environment, 16-3 Onogawa, Tsukuba, Ibaraki 305, Japan
Mononuclear iron carboxylate complex immobilized and
isolated on a modified silica surface (1) catalyzes oxidation
of hexane to a mixture of hexan-1-ol, -2-ol and -3-ol without
ketone formation in the presence of mercaptan, acetic acid,
triphenylphosphine and O2 at ambient conditions.
molecules per iron, while the elemental analysis did not show
the presence of nitrogen derived from NO3. The EPR spectrum
revealed a broad signal at g = 4.2 which was assigned to a high
spin mononuclear iron(iii). We thus propose the structure
shown in Scheme 1 for synthesized complex 1.
Oxidation of alkanes was carried out in an acetonitrile
solution as described in Table 1. No oxidation took place if
either O2 or PDT was omitted. In the presence of PDT alone or
AcOH/PDT without PPh3, hexane was not oxidized, but
cyclohexane yielded a trace of cyclohexanol. By contrast, assay
containing PPh3 remarkably improved the reaction. Without
AcOH, conversion was low, but in the complete system, hexane
was selectively oxidized to hexanols with a total turnover
number of 59 in the ratio of 1-ol (14%), 2-ol (44%) and 3-ol
(42%). Similarly cyclohexane was converted to cyclohexanol
alone with a turnover of 109. Catalyst recovered by acetonitrile
washing after the assay showed a slight decrease in iron
content,‡ but there was no change in the hydroxylation activity.
The catalytic turnover is evidently due to the immobilized
mononuclear iron carboxylate. Recovered catalyst, however,
showed an iron:methanol molar ratio of 1:1 which differed
from the original 1:3. Since this phenomenon was not observed
in the catalyst recovered from an anaerobic assay, the change in
the iron environment can be accounted for through the
Biomimetic oxidation studies on non-heme iron enzymes have
been extensively reported for the direct hydroxylation of
saturated hydrocarbons at ambient conditions.1,2 Most of the
efforts have been focused on the modeling of dinuclear iron
complexes, because of the structural correspondence to the
active site of methane monooxygenase (MMO). In this report
we describe a novel catalytic oxidation yielding only alcohols
from alkanes by mononuclear iron carboxylate complex
immobilized and isolated on a modified silica surface (1) in the
presence of mercaptan (propane-1,3-dithiol, PDT), acetic acid
(AcOH), triphenylphosphine (PPh3) and O2.
For the effective catalytic turnover, it was necessary to
protect the complex against decomposition. Immobilization and
isolation on an amorphous silica powder (Aerosil 200, specific
surface area = 200 m2 g21) was employed for this purpose. A
dilute solution of masked ligand, isopropylidene ketal and tert-
butyl ester of 3-[N-(2,3-dihydroxypropyl)-N-carboxymethyl]-
aminopropylmethylsilyl group,† was first anchored on the
surface and the remnant silanols were subsequently blocked by
diethoxydodecylmethylsilane. Removal of the protecting
groups resulted in the attachment of highly dispersed ligands at
a concentration of 4.8 3 1022 mmol g21 (3.2% of total silanols)
as quantified by the measurement of nitrogen content. Loaded
dodecylmethylsilyl groups, 3.75 3 1021 mmol g21 based on
carbon content, were estimated to cover 25.3% of total silanols.
The high coverage implies that long alkylsilyl groups are tightly
packed in the cylindroid arrangement on the surface.3
O(H)Me
Me(H)O
(H)O
O(H)Me
O
Fe
O(H)
O
Complexation in 2.5 mm Fe(NO3)3–methanol, followed by
thorough washing with methanol yielded immobilization of 4.3
3 1022 mmol g21 iron (determined colorimetrically) which
occupied 90% of the ligands previously anchored. The
formation of complex was confirmed by the appearance of
asymmetric CO vibration at 1570 cm21 and a decrease in the
signal representing free carboxyl groups at 1715 cm21 with
FTIR. Taking into account the remainder of a trace amount of
ester (1740 cm21), the complexation seemed to be accompli-
shed in almost quantitative yield. In addition treatment of the
white powder in 200 mm HCl–dioxane gave three methanol
N
SiMe
OH
SiMe
SiMe
O
O
O
O
OH
O
O
1
Scheme 1 Proposed structure for immobilized iron carboxylate
Table 1 Hydroxylation of alkanes by immobilized mononuclear iron carboxylate (1)a
Run
Substrate
Additives
Products (yield)b
Total yield
1
Hexane
Hexane
Hexane
Cyclohexane
Cyclohexane
PDT, PPh3
1-ol (1.0), 2-ol (1.7), 3-ol (1.4)
1-ol (8.1), 2-ol (26.0), 3-ol (25.3)
1-ol (8.3), 2-ol (25.5), 3-ol (24.8)
Cyclohexanol (10.0)
4.1
59.4
58.6
10.0
109.2
2
3c
4
AcOH, PDT, PPh3
AcOH, PDT, PPh3
PDT, PPh3
5
AcOH, PDT, PPh3
Cyclohexanol (109.2)
a
Catalyst (0.5 mmol iron), substrate (3.5 mmol), AcOH (0.25 mmol), PDT (0.25 mmol), PPh3 (0.25 mmol) were placed into a 20 ml vial containing
acetonitrile (2.6 ml) and sealed with a Teflon septum. After replacing the gas phase by O2, the vial was shaken at 180 strokes min21 and 25 °C for 4 h using
a Bioshaker. b Moles of product per mole of iron. c Catalyst recovered from the assay scaled up tenfold was used.
Chem. Commun., 1998
97

View Article Online
FeII
FeIII
porphyrin iron complex was reduced with an equivalent of PDT
and then exposed to O2. Addition of H2O2 gave an identical
EPR signal. The same intermediate has been reported in the
model studies of bleomycin with mononuclear iron coordinat-
ing nitrogen ligands.8 By contrast this phenomenon was not
found in the investigation using 1, and there was no effect of the
intermediate on the hydroxylation of alkanes even with the
addition of PPh3 or AcOH–PPh3. The fact may reflect a
characteristic nature of the iron complex with oxygen-rich
ligands involving carboxylate different from those bearing a
porphyrin or nitrogen ligands.
RSH
O2
R′CH2OH
R′CH3
RSSR
PPh3
+
[Ph3P-SR] –SR
RS–
RSH
FeIII-OO
FeV=O
RS–
It now appears that mononuclear iron carboxylate can
catalyze highly selective hydroxylation of alkanes by reductive
dioxygen activation. The key reaction is assumed to be the
efficient heterolytic scission of the O–O bond by the PPh3
participated deoxygenation. However, regarding the generation
of putatively unstable FeVNO species in a mononuclear iron
core, there is an argument against delocalization of the
oxidizing equivalents.2 This point contrasts to a FeIVNO
dinuclear cluster which has been reported recently as the critical
intermediate of MMO.9 The results obtained in this work could
be attributed to unique properties of the complex with
predominantly oxygen ligation in a hydrophobic micro-
environment analogous to the active site of an enzyme buried in
a protein matrix.
PPh3O
FeIII-OO–
Scheme 2 Hypothetical catalytic cycle of immobilized mononuclear iron
carboxylate (1) for alkane hydroxylation
participation of oxygen during the catalytic cycle. After release
of oxygen, acetonitrile may serve as alternative ligands.
When oxidation was carried out without substrate, PPh3 was
converted to PPh3O with a turnover number of 136. Also we can
not rule out unavoidable oxidation of the alkane chains adjacent
to the active center in 1, considering the strong oxidative
capability toward primary C–H. Such competitive oxidation
explains the rapid consumption of PPh3 causing unexpected
termination of the catalytic turnover as well as the requirement
of high concentration of substrate for effective hydroxylation in
our system.
Footnotes and References
* E-mail: miki@nire.go.jp
A hypothetical mechanism for this notable catalytic hydrox-
ylation is illustrated in Scheme 2. The reduction of FeIII to FeII
by mercaptan initiates the reaction. Mercaptan consumed at this
stage is converted to alkyl disulfide which is subsequently
attacked by the PPh3 nucleophile to form thioalkoxyphosphon-
ium cation intermediate.4 Since the intermediate is known to be
readily trapped by a carboxylic acid in refluxing acetonitrile to
yield mercaptan, thioester and PPh3O,5 a similar reaction could
be possible in the presence of AcOH. On the other hand, the
assay using 18O2 demonstrated almost quantitative 18O in-
corporation into both PPh3O and hexanols regardless of the
AcOH addition. Accordingly the intermediate is directly
attacked by a dioxygen– metal adduct, presumably a peroxoiron
species, to release PPh3O. The resulting O–O bond scission is
thought to allow the generation of an iron–oxo species from
which oxygen is transferred to the substrate. The role of AcOH
is therefore assumed to protonate the thiolate anion and promote
its dissociation from the intermediate ion pair.
† Ligand was prepared from 3-aminopropyldiethoxymethylsilane via 2
steps: (a) condensation with 2,2-dimethyl-4-chloromethyl-1,3-dioxolan at
220 °C; (b) treatment with chloroacetic acid tert-butyl ester in acetonitrile
at 50 °C in the presence of triethylamine, followed by extraction with
hexane and purification using an activated carbon column. Purity was
checked by GC, GCMS and NMR spectroscopy.
‡ Iron content of the recovered catalyst after hexane oxidation was 3.7 3
1022 mmol g21 (86% of the initial catalyst).
1 J. B. Vincent, J. C. Huffman, G. Christou, Q. Li, M. A. Nancy,
D. N. Hendrickson, R. H. Fong and R. H. Fish, J. Am. Chem. Soc., 1988,
110, 6898; R. M. Buchanan, S. Chen, J. F. Richardson, M. Bressan,
L. Forti, A. Morvillo and R. H. Fish, Inorg. Chem., 1994, 33, 3208;
A. L. Feig and S. J. Lippard, Chem. Rev., 1994, 94, 759; A. L.
Nivorozhkin and J.-J. Girerd, Angew. Chem., Int. Ed. Engl., 1996, 35,
609; B. J. Wallar and J. D. Lipscomb, Chem. Rev., 1996, 96, 2625.
2 L. Que, Jr. and R. Y. N. Ho, Chem. Rev., 1996, 96, 2607.
3 D. W. Sindorf and G. E. Maciel, J. Phys. Chem., 1982, 86, 5208; K. Miki
and Y. Sato, Bull. Chem. Soc. Jpn., 1993, 66, 2385.
4 L. E. Overman and E. M. O’Connor, J. Am. Chem. Soc., 1976, 98,
771.
5 T. Mukaiyama, Angew. Chem., 1976, 88, 111.
6 J. T. Groves, Y. Watanabe and T. J. McMurry, J. Am. Chem. Soc., 1983,
105, 4489; A. M. Khenkin and A. A. Shteinman, J. Chem. Soc., Chem.
Commun., 1984, 1219.
7 K. Tajima, M. Shigematsu, J. Jinno, K. Ishizu and H. Ohya-Nishiguchi,
J. Chem. Soc., Chem. Commun., 1990, 144.
The participation of PPh3 in the O–O bond cleavage is similar
to that of acylating reagents in cytochrome P450 model studies.6
In fact we could detect the same hexanols in the hydroxylation
of hexane using 1, KO2–18-crown-6 and acetic anhydride in
benzene although the yield was low (total turnover
number
=
0.12). Furthermore [tetrakis(pentafluoro-
phenyl)porphyrinate] iron(iii) hydroxide instead of 1 oxidized
hexane to only hexanols in comparable product yield and
distribution in our system. The results again support the
formation of a high-valent iron–oxo species being responsible
for the hydroxylation of alkanes by 1.
8 R. J. Guajardo and P. K. Mascharak, Inorg. Chem., 1995, 34, 802;
M. Lubben, A. Meetsma, E. C. Wilkinson, B. Feringa and L. Que, Jr.,
Angew. Chem., Int. Ed. Engl., 1995, 34, 1512.
9 L. Shu, J. C. Nesheim, K. Kauffmann, E. Mu¨nck, J. D. Lipscomb and
L. Que, Jr., Science, 1997, 275, 515.
Interestingly a hydroperoxoiron(iii) species was identified by
EPR spectroscopy at 77 K (g = 2.27, 2.20, 1.97)7 when the
Received in Cambridge, UK, 26th August, 1997; 7/06178A
98
Chem. Commun., 1998