Efficient [bis(imino)pyridine-iron]-catalyzed oxidation of alkanes

Article abstract of DOI:10.1039/b712849p

A [bis(imino)pyridine-iron(ii)] complex efficiently catalyses the oxidation of C_sp3-H bonds by dihydrogen peroxide, through the attack of the alkyl substrate by hydroxyl radicals. The Royal Society of Chemistry 2007.

Full text of DOI:10.1039/b712849p

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/dalton | Dalton Transactions
Efficient [bis(imino)pyridine-iron]-catalyzed oxidation of alkanes†
Jinkui Tang, Patrick Gamez and Jan Reedijk*
Received 20th August 2007, Accepted 22nd August 2007
First published as an Advance Article on the web 3rd September 2007
DOI: 10.1039/b712849p
A [bis(imino)pyridine-iron(II)] complex efficiently catalyses
3
the oxidation of Csp –H bonds by dihydrogen peroxide,
through the attack of the alkyl substrate by hydroxyl radicals.
The catalytic oxidation of alkanes is a major challenge in modern
chemistry.1 Indeed, alkanes represent the main component of nat-
ural gas, and are therefore a potential source of valuable products
through their functionalization.2 The activation of primary C–
Fig. 2 Coordination environment expected for the iron catalyst precursor.
H bonds is not straightforward, and very often requires extreme
reaction conditions, such as the use of environmentally unfriendly
oxidants, high temperatures and/or pressures.3
In nature, a number of enzymes are capable of performing the
hydroxylation of alkanes. Thus, non-heme iron systems, like, for
The ligand L can be easily prepared in one step by simple
condensation of 1 equiv. of 2,6-diacetylpyridine with two equiv. of
2-methoxybenzylamine in methanol at room temperature, in the
presence of catalytic amounts of formic acid.§
The new alkane oxidation reported here is based on the simple
Fe(BF4)2L catalyst (obtained in situ from a 1 : 1 Fe(BF4)2–L
mixture). Typically, 2 mmol of the substrate are left to react with
H2O2 30% (H2O2–substrate = 1.5 : 1) in acetonitrile at 50 ◦C. The
oxidation is catalysed by the presence of 1 mol% Fe(BF4)2L
(Fig. 3).¶ It has to be pointed out that the 1.5 equiv. of dihydrogen
peroxide can be added at once, and that no obvious catalase
activity (decomposition of H2O2)14 is observed. Moreover, no
variation of the catalytic activity is observed when H2O2 is added
in four portions. The first substrate tested with this catalytic system
is cyclohexane (Table 1).¶ Thus, the reaction of cyclohexane with
H2O2, catalysed by the in situ generated catalyst (from Fe(BF4)2
and L), leads to a total conversion of the alkane to cyclohexanol
(CyOH) and cyclohexanone (CyO), after a reaction time of 12 h
(Entry 1). The oxidation reaction is highly selective as the sole
formationofCyOHandCyOina1:1ratiois observed. Onlytraces
of other oxidation products are detected. As evidenced in Fig. 4,
the oxidation process is rapid in the early stages, with a conversion
of 33% after a reaction time of only 30 min. After one hour, the
conversion rate slows down (42% conversion is reached after 1 h),
and the oxidation reaction is completed within 12 h (entry 1).
example, methane monooxygenase and bleomycin can efficiently
4,5
3
oxidize Csp –H bonds. Consequently, several biomimetic, non-
heme iron catalysts for the oxidation of alkanes with H2O2 have
been developed during the past fifteen years.6–10
In the study herein presented, a new ligand, namely 6-bis[1-(2-
methylanisolylimino)ethyl]pyridine (Fig. 1; L), has been designed
to generate an iron(II) catalyst for the oxidation of alkyl and alkenyl
substrates in the presence of H2O2. The reaction of an iron(II) salt
with the ligand L is expected to produce a complex where the
planar N3 binding unit is equatorially coordinated to the metal
centre, as depicted in Fig. 2.‡ Indeed, earlier crystal structures
with related bis(imino)pyridine ligands show this type of spatial
arrangement, with the three nitrogen atoms and the iron ion lying
in the same plane.11,12
Fig. 1 Representation of the ligand 2,6-bis[1-(2-methylanisolylimino)-
ethyl]pyridine (L).
Fig. 3 Iron-catalysed oxidation of alkanes with H2O2.
The weak donating methoxy groups are expected to interact
with the iron centre (semi-coordination), playing the role of the
axial ligands found in the active site of iron bleomycin.13
The high oxidation rate noticed at the beginning of the reaction
may be rationalized by the fast production of HO• species through
a Fenton-like process,15 mediated by the iron(II) catalyst. These
hydroxyl radicals then react with cyclohexane to yield CyOH and
CyO. The alcohol : ketone ratios (A : K) determined during the
course of the reaction (Table S1†) remain constant, thus suggesting
that CyOH is not the precursor of CyO. The autooxidation
pathway recently proposed by Hermans et al.16 is unlikely to
occur in the present system (at least not as the major oxidation
route). Indeed, the proposed mechanism for the autooxidation of
Leiden Institute of Chemistry, Leiden University, P. O. Box 9502, 2300 RA
Leiden, The Netherlands. E-mail: reedijk@chem.leidenuniv.nl; Fax: +31
715274671
† Electronic supplementary information (ESI) available: Tables S1 and S2
showing the A : K ratios for two selected oxidation reactions. Fig. S1–S3
showing the ligand field spectra of [iron(II)/Lx] solutions in acetonitrile.
See DOI: 10.1039/b712849p
4644 | Dalton Trans., 2007, 4644–4646
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Table 1 [FeL]-catalyzed oxidation of alkanes and alkenes with H2O2¶
Entry Substrate Catalyst precursors Solvent
Temp/◦C Time/h Conv. (%)a
Major products (%)
A : Kb
1
Cyclohexane Fe(H2O)6(BF4)2 + L Acetonitrile
Cyclohexane Fe(H2O)6(BF4)2 + L Acetonitrile
50
50
12
7
9
100c
100
55f
CyOH (49.4); CyO (49.3)
CyOH (52.5); CyO (47.5)
CyOH (20.4); CyO (15.5)
1.00d
1.11
1.32
2e
3
Cyclohexane Fe(H2O)6(BF4)2 + L Acetonitrile + 50
DMSO
4g
5
6
Cyclohexane Fe(H2O)6(BF4)2 + L Acetonitrile
Cyclohexane Fe(H2O)6(BF4)2 + L Acetone
Cyclohexene Fe(H2O)6(BF4)2 + L Acetonitrile
Cyclohexene Fe(H2O)6(BF4)2 + L Acetonitrile
Cyclohexene Fe(H2O)6(BF4)2 + L Acetonitrile
50
50
50
RT
50
50
50
3
30
20
150
31
22
100h
CyOH (19.9); CyO (56.1)
—
CyhxOH (27); CyhxO (71)
CyhxOH (17.5); CyhxO (7.9)
CyhxOH (23.3); CyhxO (51.4)
Cyclooctene oxide (19.8)
0.35
—
0.38
2.22
0.45
—
No reaction
100i
27j
7
8e
9
86k
30l
Cyclooctene
n-Heptane
Fe(H2O)6(BF4)2 + L Acetonitrile
Fe(H2O)6(BF4)2 + L Acetonitrile
10
54
82m
Heptan-1-ol (5.6); heptanal (0)
—
Heptan-2-ol (7.7); heptan-2-one (17.5) 0.44
Heptan-3-ol (7.1); heptan-3-one (16.2) 0.44
Heptan-4-ol (3.8); heptan-4-one (7.2)
0.53
a Determined by GC. b The alcohol : ketone (A : K) ratio was determined by GC. c 98.7% conversion to cyclohexanol (CyOH) and cyclohexanone (CyO).
Traces of other oxidation products were detected, but not identified. d After 24 h, the A : K ratio is 0.81. e Reaction carried out under argon. f 54.6% of
[CyOH + CyO] were obtained together with 16.5% of cyclohex-3-enol and 2.2% of cyclohexyl hydroperoxide. g Reaction performed with 10 equiv. of
H2O2 instead of 1.5 equiv. h 76% of [CyOH + CyO] were obtained together with 24% of a new product, i.e. cyclohexyl hydroperoxide. i 98% conversion to
cyclohex-2-enol (CyhxOH) and cyclohex-2-enone (CyhxO). Only traces (<0.5%) of 5 other oxidation products were detected by GC. j 25.4% conversion to
[CyhxOH + CyhxO], and 1.6% to other oxidation products. k 74.7% conversion to [CyhxOH + CyhxO], and 11.3% to other oxidation products including
cyclohexane-1,2-diol (6.4%). l Besides cyclooctene oxide, 6 other oxidation products were detected (all below 2%), which have not been identified. m 65.1%
conversion to the mixture of alcohols and ketones. The remaining 16.9% correspond to 5 other oxidation products which have not been identified.
deep purple color of the reaction mixture gradually turns to green
(in the absence of DMSO, the deep purple color is maintained
throughout the oxidation process), with the concomitant increase
of the catalytic activity (for instance, 11% and 27% conversions are
observed after 1 and 2 h, respectively; Fig. 4). It thus appears that
a new active species starts to form in the presence of DMSO after
an incubation time of approximately 30 min. This different species
most likely involves another mechanistic pathway, because two
new oxidation products are obtained, namely cyclohex-3-enol and
cyclohexyl hydroperoxide (entry 3). The mechanism of formation
of these by-products has not been so far investigated. The use of
a large excess of dihydrogen peroxide (10 equiv., entry 4) gives
rise to a notable enhancement of the overall conversion efficiency
(total conversion after 3 h). The alkyl substrate is converted to
20% of CyOH and 56% of CyO. Apparently, the excess of oxidant
favors the formation of the ketone over the alcohol (the K : A
ratio is 2.82). This selectivity towards CyO presumably arises
from the subsequent overoxidation of CyOH (once the alkane
has completely reacted) in the presence of high amounts of H2O2.
In addition, 24% of cyclohexyl hydroperoxide is produced, thus
suggesting a concurrent autooxidation pathway when a surplus
of oxidant is used, which explains the increase in CyO formation.
Remarkably, the use of acetone as a solvent results in a total lack
of reactivity (entry 5). This observation can be rationalized by
the fact that acetone is a known as a good HO• scavenger,18 thus
corroborating the involvement of hydroxyl radicals, in conjunction
with the results observed in the presence of DMSO.
Fig. 4 [Fe(BF4)2L]-catalysed oxidation of cyclohexane at 50 ◦C, under
air (ꢀ), under argon (ꢁ), and in the presence of DMSO (ꢂ).
cyclohexane involves cyclohexylperoxyl radicals (CyOO•) which
are generated from cyclohexyl radicals (Cy•) and dioxygen.16
As a result, the formation of CyOO• is expected to be reduced
under argon. Actually, under argon, the total conversion of
cyclohexane is achieved after a reaction time of only 7 h, instead
of 12 h under aerobic conditions (Table 1, entries 1 and 2;
Fig. 4). Furthermore, a conversion of 44% is reached after 30 min,
while it is only 33% under air. To verify the presence of free
hydroxyl radicals in the reaction mixture, the aerobic oxidation
of cyclohexane has been performed in the presence of DMSO,
which is known to be an efficient hydroxyl radical scavenger.17 As
expected, the oxidation of cyclohexane is significantly retarded
(Fig. 4), and a conversion of only 3% is obtained after 30 min (the
corresponding conversion in the absence of DMSO amounts to
33%). Afterwards, the reaction proceeds smoothly (compared to
the previous reactions carried out without DMSO; Fig. 4) reaching
a plateau after a reaction time of 3 h. A maximum conversion of
55% is achieved after 7 h. It has to be mentioned here that the initial
To further confirm the (HO•)-based mechanism, cyclohexene
has been oxidised with 1.5 equiv. of H2O2 using the typical exper-
imental conditions previously applied for cyclohexane (entry 6).‡
As anticipated, an exclusive formation of products resulting
from the allylic oxidation of cyclohexene, i.e. cyclohex-2-enol
(CyhxOH) and cyclohex-2-enone (CyhxO), is noted (entry 6).
The absence of epoxidation products excludes the participation of
metal-based oxidative species19 during the [Fe(BF4)2L]-catalysed
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 4644–4646 | 4645
©

View Article Online
reaction. Since cyclohexene is susceptible to allylic oxidation,20 the
major production of CyhxOH and CyhxO once again suggests
that the main oxidation pathway brings in hydroxyl radicals.
The high amount of ketone produced (the K : A ratio is 2.21)
is probably due to the favoured formation of the delocalised
a,b-unsaturated carbonyl compound. If the same oxidation is
performed at room temperature (entry 7), a reverse selectivity is
observed with a higher conversion to the alcohol (the A : K ratio
is 2.22). However, the reaction is significantly slower and only 27%
of cyclohexene are converted, after a reaction time of more than 6
d. These observations (entries 6 and 7) indicate that the ketone is
generated from the alcohol. Actually, the A : K ratios determined
during the course of the reaction carried out at 50 ◦C show a
variation of the values, from 1.65 after 30 min to 0.38 after total
conversion (Table S4†), thus supporting the transformation of the
alcohol into the ketone. When the oxidation of cyclohexene is
performed under argon (entry 8), a significant decrease of the
catalytic activity is noticed (see entries 6 and 8). This feature
may indicate the contribution of an auto-oxidation pathway in
the aerobic oxidation of cyclohexane (which is apparently not
the case with cyclohexane; see entries 1 and 2). Furthermore, the
reaction is less selective under argon since new oxidation products
are detected, including cyclohexane-1,2-diol, which is a product
resulting from the oxidation of the double bond.
Next, cyclooctene has been used as substrate to test the catalytic
ability of the [bis(imino)pyridine-Fe(II)] complex to oxidise a
compound known to be easily epoxidized. The production of
cyclooctene oxide is indeed observed (about 20% after 22 h;
entry 9), but the reaction is not selective and is only very
gradual (30% conversion after a reaction time of 22 h; entry 9).
Besides cyclooctene oxide, 6 other oxidation compounds are
produced. This result once more denotes the minor involvement
of metal-based oxidation species with the [bis(imino)pyridine-
Fe(II)] system. Finally, the [bis(imino)pyridine-Fe(II)]-mediated
oxidation of a more inactive alkane (compared to cyclohexane),
namely n-heptane, has been investigated (entry 10). Interestingly,
as much as 82% of the substrate is converted to oxidation products.
About 65% correspond to a mixture of the different primary and
secondary alcohols, and carbonyl compounds. In all cases, about
twice the amount of ketone is obtained (entry 10).
Notes and references
‡ The coordination of FeII to L has been studied by ligand field
spectroscopy (see Fig. S1–S3†). Thus, titration experiments show that the
addition of increasing amounts of Fe(BF4)2 to an acetonitrile suspension
of L first generates an [Fe/L2] complex. An [Fe/L] complex appears to be
produced when 1 equiv. of Fe(BF4)2 is added to 1 equiv. of ligand L (see
Fig. S2 and S3†).
§ Procedure for the preparation of 2,6-bis[1-(2-methylanisolylimino)-
ethyl]pyridine (L): 0.81 g (5 mmol) of 2,6-diacetylpyridine were dissolved
in 10 mL of methanol containing a catalytic amount of formic acid (three
drops). 1.50 g (11 mmol) of 2-methoxybenzylamine were subsequently
added and the resulting reaction mixture was stirred for two hours at room
temperature. The white precipitate obtained was isolated by filtration, and
washed with methanol. 1.80 g (yield = 90%) of pure compound L were
obtained. Elemental analyses (%) calcd for C25H27N3O2: C, 74.79; H, 6.78;
N, 10.47; found: C, 74.28; H, 6.60; N, 10.38. MS (m/z): 401.97. 1H-NMR
(CDCl3): d 8.27–6.89 (m, 11 H, Harom.); 4.79 (s, 4 H, –CH2–); 3.88 (s, 6 H,
CH3O–); 2.54 (s, 6 H, CH3C N) ppm. IR data for L: m/cm 2834 (w),
−1
=
1693 (s), 1635 (s), 1588 (s), 1489 (vs), 1456 (s), 1436 (s), 1358 (s), 1276 (m),
1236 (vs), 1155 (m), 1102 (s), 1046 (s), 1027 (vs), 816 (vs), 754 (vs), 711 (s),
634 (s), 590 (s), 486 (w), 439 (w), 428 (m) cm−1
.
¶ Typical oxidation procedure: The solvents, iron(II) tetrafluoroborate
hexahydrate, the alkyl substrates and the corresponding oxidation products
are commercially available and were used without further purification. The
oxidation of alkanes was carried out under air in a 10 mL round-bottomed
flask equipped with a magnetic stirrer and a reflux condenser. Typically,
8 mg (0.02 mmol) of L were suspended in 2 mL of acetonitrile. Next, 6.8 mg
(0.02 mmol) of Fe(BF4)2·6H2O were added (resulting in the dissolution of
the ligand and the formation of a purple homogeneous solution), followed
by 2.00 mmol of substrate (which corresponds to 168.3 mg in the case
of cyclohexane). The oxidation reaction was initiated by the addition
of 3.00 mmol (1.5 equiv.; 340 mg) of H2O2 30%. The resulting deep
purple reaction mixture was heated to 50 ◦C. Samples of the reaction
mixture were taken out regularly to monitor the reaction by GC. Most
of the products of the reaction were determined by comparison with the
commercially available compounds. The non-commercially accessible ones
were identified by GC-MS analysis.
1 A. A. Fokin and P. R. Schreiner, Chem. Rev., 2002, 102, 1551–1593.
2 N. M. F. Carvalho, A. Horn and O. A. C. Antunes, Appl. Catal. A,
2006, 305, 140–145.
3 R. H. Crabtree, Chem. Rev., 1995, 95, 987–1007.
4 M. Merkx, D. A. Kopp, M. H. Sazinsky, J. L. Blazyk, J. Muller and
S. J. Lippard, Angew. Chem., Int. Ed., 2001, 40, 2782–2807.
5 D. L. Boger, T. M. Ramsey, H. Cai, S. T. Hoehn and J. A. Stubbe, J. Am.
Chem. Soc., 1998, 120, 9139–9148.
6 S. Tanase, C. Foltz, R. de Gelder, R. Hage, E. Bouwman and J. Reedijk,
J. Mol. Catal. A, 2005, 225, 161–167.
7 S. V. Kryatov, E. V. Rybak-Akimova and S. Schindler, Chem. Rev.,
2005, 105, 2175–2226.
8 A. M. I. Payeras, R. Y. N. Ho, M. Fujita and L. Que, Chem.–Eur. J.,
2004, 10, 4944–4953.
9 G. Roelfes, M. Lubben, R. Hage, L. Que and B. L. Feringa, Chem.–
Eur. J., 2000, 6, 2152–2159.
10 G. B. Shul’pin, G. V. Nizova, Y. N. Kozlov, L. G. Cuervo and G.
Suss-Fink, Adv. Synth. Catal., 2004, 346, 317–332.
11 Y. Chen, C. Qian and J. Sun, Organometallics, 2003, 22, 1231–1236.
12 G. J. P. Britovsek, V. C. Gibson, B. S. Kimberley, S. Mastroianni, C.
Redshaw, G. A. Solan, A. J. P. White and D. J. Williams, J. Chem. Soc.,
Dalton Trans., 2001, 1639–1644.
In summary, a new iron-based catalyst has been developed,
which is able to completely convert cyclohexane and cyclohexene
to the corresponding alcohols and ketones. The efficiency of the
catalytic system based on the oxidant (H2O2) is high because
dihydrogen peroxide is totally consumed during the reaction.
It appears that the main oxidation pathway involves hydroxyl
radicals. More detailed studies are currently carried out to better
understand this mechanism. The use of related ligands (with
different substituents on the phenyl rings) as catalyst precursors
is now being examined to evaluate the influence of donating and
withdrawing effects on the catalytic activity and/or selectivity of
the resulting [bis(imino)pyridine-Fe(II)] complexes.
13 S. Taktak, W. H. Ye, A. M. Herrera and E. V. Rybak-Akimova, Inorg.
Chem., 2007, 46, 2929–2942.
14 M. Zamocky and F. Koller, Prog. Biophys. Mol. Biol., 1999, 72, 19–66.
15 C. Walling, Acc. Chem. Res., 1975, 8, 125–131.
16 I. Hermans, P. Jacobs and J. Peeters, Chem.–Eur. J., 2007, 13, 754–
761.
The work described in this paper was supported by the
Graduate Research School Combination “NRSC Catalysis”, a
joint activity of the graduate research schools NIOK, HRSMC
and PTN. Dr Ronald Hage (Unilever) and Dr Stefania Tanase
(Leiden University) are thanked for valuable discussions.
17 C. F. Brayton, Cornell Vet., 1986, 76, 61–90.
18 C. Walling and A. Goosen, J. Am. Chem. Soc., 1973, 95, 2987–2991.
19 X. M. Wang, S. X. Wang, L. J. Li, E. B. Sundberg and G. P. Gacho,
Inorg. Chem., 2003, 42, 7799–7808.
20 A. J. Appleton, S. Evans and J. R. L. Smith, J. Chem. Soc., Perkin Trans.
2, 1996, 281–285.
4646 | Dalton Trans., 2007, 4644–4646
This journal is
The Royal Society of Chemistry 2007
©