A new efficient iron catalyst for olefin epoxidation with hydrogen peroxide

Article abstract of DOI:10.1039/c1cc15935f

A new aminopyridine ligand derived from bipiperidine (the product of full reduction of bipyridine, bipy) coordinates to iron(ii) in a cis-α fashion, yielding a new selective catalyst for olefin epoxidation with H ₂O₂ under limiting substrate conditions.

Full text of DOI:10.1039/c1cc15935f

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 687–689
www.rsc.org/chemcomm
COMMUNICATION
A new efficient iron catalyst for olefin epoxidation with hydrogen
peroxidew
Elena A. Mikhalyova,ab Olga V. Makhlynets,a Taryn D. Palluccio,a Alexander S. Filatovc
and Elena V. Rybak-Akimova*a
Received 23rd September 2011, Accepted 15th November 2011
DOI: 10.1039/c1cc15935f
A new aminopyridine ligand derived from bipiperidine (the product
of full reduction of bipyridine, bipy) coordinates to iron(II) in a cis-a
fashion, yielding a new selective catalyst for olefin epoxidation with
H2O2 under limiting substrate conditions.
Alkene oxides (epoxides) are often used as starting materials
or intermediates in important technological processes in
chemical or pharmaceutical industry. Therefore, discovery of
Scheme 1 Selected aminopyridine ligands that form redox-active iron
catalysts.4–7
efficient, selective and practical epoxidation methods is an
important goal in synthetic chemistry. While a number of
catalysts of epoxidation are known,1 some problems remain,
including the development of catalytic systems with high con-
version and selectivity that utilize non-toxic, environmentally
benign metals and oxidants. Biomimetic systems that model
non-heme iron oxidative enzymes are particularly attractive,
since these enzymes are proficient at oxygen and peroxide
activation using a cheap, readily available, non-toxic metal
(Fe). Examples of non-native, biomolecule-catalyzed epoxida-
tions include epoxidation of small terminal olefins with methane
monooxygenase (a well-known diiron enzyme responsible for
hydroxylating methane into methanol in bacteria),2a,b and olefin
epoxidation with iron bleomycin (a mononuclear metal-dependent
antibiotic that oxidatively cleaves nucleic acids in the presence of
O2 or H2O2).2c,d Active sites of non-heme iron enzymes usually
contain coordinatively unsaturated iron centers bound to nitrogen
donors from aromatic heterocycles (e.g. imidazole of histidine
sidechain) and carboxylate oxygen donors, and generate metal-
based intermediates that carry out selective oxidations.
selective iron catalysts for biomimetic oxidations. Fenton
chemistry often dominates the reactivity of iron salts or
complexes with peroxides, generating non-discriminatory
hydroxyl radicals that hamper selectivity.3 In contrast, tetra-
dentate aminopyridine ligands (Scheme 1) show promise in
H2O2 activation for selective oxidations;4–6 cis geometry of
two labile sites is important for catalytic activity of their
iron(II) complexes,4c,d,6b and increased rigidity of the ligand
framework (BPBP vs. BPMEN, Scheme 1) prevented leakage
of iron and increased the activity of the catalysts, enabling
regioselective aliphatic and aromatic C–H oxidation.7
In this communication we introduce a new rigid ligand,
rac-PYBP (Fig. 1), and describe a new non-heme iron complex,
[Fe(PYBP)(CH3CN)2](ClO4)2 (1), which proved to be a highly
active and selective catalyst for olefin epoxidation with H2O2.
The new ligand, PYBP (Fig. 1), is structurally related to BPMEN
and BPBP (Scheme 1); the flexibility of the former molecule
is reduced by incorporating methyl-amine moiety into rigid
6-membered piperidine rings. Olefin oxidations with hydrogen
peroxide in the presence of known iron(II) complexes with
BPMEN or BPBP usually yield cis-diols and/or epoxides, often
in mixtures.5 The new ligand, PYBP, affords an iron(II) complex
of superior reactivity and selectivity. This ligand is readily synthe-
sized from a commercially available starting material (2,20-bipyri-
dine, bipy), and it allows for a variety of structural modifications,
because a large range of bipy derivatives are available. Fully
reduced bipyridines serve as a highly promising diamine platform
for expanding the toolbox of biomimetic redox catalysts.
Increasing interest in catalytic applications of iron com-
plexes motivated our search for new, simple, efficient and
a Department of Chemistry, Tufts University, 62 Talbot Avenue,
Medford, Massachusetts 02155, USA. E-mail: erybakak@tufts.edu;
Fax: +1-6176273443; Tel: +1-6176273413
b L. V. Pisarzhevskii Institute of Physical Chemistry of the National
Academy of Sciences of the Ukraine, Prospekt Nauki 31, Kiev,
03028, Ukraine
c Department of Chemistry, University at Albany,
1400 Washington Avenue, Albany, New York 12222, USA.
E-mail: afilatov@albany.edu; Fax: +1-5184423462;
Tel: +1-5184424428
w Electronic supplementary information (ESI) available: Full experi-
mental details for the preparation of the complex 1, experimental
procedures for catalytic epoxidation reactions, and crystallographic
details. CCDC 839714. For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c1cc15935f
Attaching two picolyl fragments to the previously reported
diamine BP8 afforded rac-PYBP (Fig. 1); the identity and purity
of the compound was confirmed by elemental analysis, NMR,
and ESI-MS. The iron(II) complex 1 was synthesized by the
reaction of PYBP and Fe(ClO4)2ꢀ6H2O in acetonitrile (Fig. 1),
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 687–689 687

View Article Online
Table 1 Catalytic epoxidation of alkenes with 1/H2O2/HOAc
Catalyst
loading, Conversion,
Selectivityc,
%
Alkene
mol%
% (time, min)
TON
cis-Cyclooctenea
cis-Cyclooctenea
Cyclohexenea
Octene-1a
0.5
0.1
0.1
0.5
0.1
0.5
0.1
100 (5)
72 (10)
50 (10)
88 (5)
19 (10)
97.7 (5)
17 (10)
100 (10)
29 (10)
8.5 (10)
100 (5)
100
100
94
4200d
715
490
99
200
Octene-1a
100
100
100
N/D
83
190
Decene-1a
195d
165
Decene-1a
Cyclohexadiene-1,4b,e 0.5
Cyclohexadiene-1,4b,e 0.1
Cyclooctadiene-1,3b,e 0.1
Cyclooctadiene-1,3b,e 0.5
4200d
240
70
83
69
4200d
a
Reaction conditions: Olefin (200 mmol), H2O2
(
300 mmol and
b
Fig. 1 Synthesis and molecular structure of 1 (drawn with thermal
ellipsoids at the 50% probability level). A perchlorate anion and
hydrogen atoms have been omitted for clarity. Symmetry transforma-
tion used to generate equivalent atoms: ꢁx + 1, y, ꢁz + 1/2. Fe–N(1)
1.947(4) A; Fe–N(2) 2.046(3) A; Fe–N(3) 1.936(4) A.
600 mmol, 30% aqueous solution), HOAc (1 mmol) in 2 mL of
CH3CN, room temperature. Determined from GCMS. Accurate
e
determination of TON is impossible. Only mono-epoxide.
c
d
Conversion = [products]/(olefin + products); selectivity = [epoxide]/
(all products); turnover number (TON) = amount of epoxide (moles)
per 1 mole of 1 over the entire reaction time (5 or 10 minutes).
and crystallized by slow ether diffusion under argon (see ESIw
for details).
consistent with redox potentials of alkenes, which becomes less
positive upon alkyl substitution.9 Nucleophilic reactivity of olefins
suggests an electrophilic character of the oxidant. Similarly
to other epoxidations catalyzed by non-heme iron complexes,
the most reactive are cyclooctene and cyclohexene and less
reactive are terminal olefins. For several olefins screened in
this work, the new complex 1 reported herein shows very high
epoxide yields and turnover numbers at low catalyst loadings
compared with other known complexes.1a,6a,b Another beneficial
feature of 1 is its high selectivity with respect to epoxide
products. In some known catalytic systems, epoxide becomes
predominant in the presence of acid, but diol is the main
product without it.6b In our case, the major (or sole) product
is always epoxide, not diol. Remarkably, cyclic dienes (cyclo-
octadiene-1,3 and cyclohexadiene-1,4) yielded primarily (if not
exclusively) mono-epoxides; this selectivity is potentially useful
for synthetic applications. Some relatively reactive alkenes, such
as cyclohexene, are known to often favor allylic oxidation; this
was not the case in olefin epoxidations catalyzed by 1, where
epoxides were the major products and only traces (1–3%) of
allylic alcohol and ketone formed. Aromatic olefins were not
tested in this work, since aromatic hydroxylation7c,d (ESIw,
Fig. S1) may interfere with epoxidation.
The X-ray structure of the complex 1 (Fig. 1) revealed
distorted octahedral geometry of iron(II). All four nitrogen atoms
of the ligand PYBP are coordinated, in a cis-a-topology, to the
iron(II) center: two pyridine rings are located in opposite vertices.
Two acetonitrile molecules are in cis-positions to each other (and
trans- to tertiary amines); cis orientation of labile sites is believed
to be critically important for catalysis.4c,d Average Fe–N bond
length is 2.0 A, the value that is typical of low-spin aminopyridine
iron(II) complexes. The overall structure of complex 1 is very
similar to the structures of known oxidation catalysts, such as
[Fe(BPMEN)(CH3CN)2]2+1a and [Fe(BPBP)(CH3CN)2]2+ 7a,b
.
The redox properties of the three complexes are also similar:
the FeIII/FeII redox potentials are 0.656 V, 0.660 V, and
0.728
[Fe(BPMEN)(CH3CN)2]2+, respectively.
V , and
vs. Fc+/Fc for 1, [Fe(BPBP)(CH3CN)2]2+
Complex 1 was investigated as a catalyst in alkene epoxidation
with H2O2, and demonstrated excellent activity at low catalyst
loadings (Table 1). At 0.5 mol% catalyst with respect to substrate
in the presence of 500 mol% of acetic acid (HOAc), nearly
quantitative conversion of olefins occurred in 5 to 10 minutes at
room temperature, affording ca. 1000 mol of respective epoxides
per 1 mol of 1 (under similar conditions, 5 mol% of the catalyst
was used in [Fe(BPMEN)(CH3CN)2]2+-catalyzed epoxidations).1a
High activity of the new catalyst, especially in the presence of
HOAc, allowed us to drastically lower catalyst loadings and
observe reasonable to high olefin conversions in 10 minutes with
only 0.1 mol% of 1 (Table 1). Although the conversion of alkenes
under these conditions was not quantitative, increasing the
reaction time to 1 hour did not improve the product yields, but
increasing the amount of catalyst to 0.5 mol% did.
In view of known effects of carboxylic acids on the out-
comes of catalytic olefin oxidations with H2O2,1a,6 we varied
the amounts of HOAc in cis-cyclooctene and 1-decene epoxi-
dations with 1/H2O2 (Fig. 2). While TON with respect to
epoxide uniformly increased in the presence of HOAc, the
shapes of the plots of TON vs. the amount of HOAc depend
on the nature of olefin, and on relative amounts of reactants.
Under synthetically relevant conditions (slight excess of the
oxidant with respect to substrate), the turnover number goes
through the maximum at ca. 5000 equivalents of HOAc for
cis-cyclooctene, while TON for oxidation of 1-decene con-
tinues to increase (albeit slowly) at high concentrations of acid.
For the system with limiting H2O2, TON for 1-decene epoxidation
goes through the maximum at ca. 400 equivalents of acid
(Fig. S2, ESIw). Decrease of TON at high concentrations of
Epoxidation results presented in Table 1 demonstrate vari-
able reactivity of alkene substrates: cyclooctene and cyclohexene
afford higher epoxide yields, while aliphatic terminal alkenes
(1-decene and 1-octene) are less reactive. The susceptibility of
tested olefins with respect to catalytic epoxidation with 1/H2O2
increases as the electron density of the double bond increases
due to substituent inductive effects (CH3 4 CH2 4 H), a trend
c
688 Chem. Commun., 2012, 48, 687–689
This journal is The Royal Society of Chemistry 2012

View Article Online
Scheme 2 Proposed epoxidation pathway.6b,7c
Fe(V)(OH)2 in excess HOAc, and agrees with exclusive formation
of epoxides rather than diols,5b,c,6b–d which was now observed in
catalytic oxidations with 1/H2O2/HOAc.
In conclusion, we have discovered a new iron catalyst of
very high selectivity and efficiency in epoxidation of aliphatic
olefins with hydrogen peroxide; no diol formation was observed.
The epoxide yield strongly depends on the amount of HOAc in
the reaction mixture. Acid-promoted O–O bond cleavage in
the observed Fe(III)–OOH intermediate generates an electro-
philic metal-based oxidant.
Fig. 2 Epoxide TON (Table 1) depends on the amount of HOAc
added (with respect to 1; logarithmic scale). Olefin (200 mmol), 1
(0.2 mmol), H2O2 (300 mmol, 30% aqueous solution), 25 1C, 10 min.
HOAc is explained by the decomposition of complex 1 and the
formation of less active binuclear carboxylato-bridged iron(III)
compounds, and by a possible competing catalase reaction.
Control experiments with iron(II) salts in place of complex 1
did not yield epoxide products.
This work was supported by the US Department of Energy
(DE-FG02-06ER15799); instrumentation was supported by
the NSF (CHE-9723772, 0639138, and MRI CHE 0821508).
Stopped-flow studies of the direct reaction between 1 and H2O2
identified a short-lived intermediate with lmax = 560 nm (Fig. 3).
The optical spectrum of this intermediate is similar to the spectra
of known mononuclear FeIII(OOH) complexes that typically have
an intense band at B500–550 nm. The EPR spectrum of the
frozen sample prepared by manual rapid mixing of 1 and a 10-fold
excess H2O2 (g = 2.185, 1.98, and 1.955) is also very similar to the
spectra of known LS iron(III) hydroperoxides10 and confirms that
FeIII(OOH) is present in the reaction mixture. By analogy
with reaction mechanisms proposed for related systems,6b,7c
HOAc-assisted O–O bond heterolysis would generate a reactive
electrophilic intermediate, Fe(V) = O(OAc), that transfers a single
oxygen atom to olefins, generating epoxide (Scheme 2). Isotope
labeling experiments showed the lack of incorporation of 18O
from water in epoxidations of 1-decene and cyclooctene with
1/H2O2/HOAc, and incorporation of oxygen from H218O2 into
epoxide products (see ESIw for details). Slow exchange of
metal-based oxidant with water implies low accumulation of
Notes and references
1 (a) C. M. White, A. G. Doyle and E. N. Jacobsen, J. Am. Chem.
Soc., 2001, 123, 7194–7195; (b) I. Garcia-Bosch, A. Company,
X. Fontrodona, X. Ribas and M. Costas, Org. Lett., 2008, 10(11),
2095–2098; (c) M. B. Andrus and B. W. Poehlein, Tetrahedron Lett.,
2000, 41, 1013–1014; (d) M. Dakkach, M. I. Lo
M. Rodrıguez, A. Atlamsani, T. Parella, X. Fontrodona and
´
pez, I. Romero,
´
A. Llobet, Inorg. Chem., 2010, 49, 7072–7079; (e) M. Abrantes,
T. R. Amarante, M. M. Antunes, S. Gago, F. A. A. Paz,
I. Margiolaki, A. E. Rodrigues, M. Pillinger, A. A. Valente and
I. S. Gonc¸ alves, Inorg. Chem., 2010, 49, 6865–6873.
2 (a) S. Friedle, E. Reisner and S. J. Lippard, Chem. Soc. Rev., 2010, 39,
2768–2779; (b) L. J. Murray and S. J. Lippard, Acc. Chem. Res., 2007,
40, 466–474; (c) S. M. Hecht, Acc. Chem. Res., 1986, 19, 383–391;
(d) J. Chen and J. Stubbe, Curr. Opin. Chem. Biol., 2004, 8, 175–181.
3 (a) C. Walling, Acc. Chem. Res., 1998, 31, 155; (b) H. J. H. Fenton,
J. Chem. Soc., 1894, 65, 889–910.
4 (a) S. Enthaler, K. Junge and M. Beller, Angew. Chem., Int. Ed.,
2008, 47, 3317–3321; (b) A. Correa, O. G. Mancheno and C. Bolm,
Chem. Soc. Rev., 2008, 37, 1108–1117; (c) L. J. Que and
W. B. Tolman, Nature, 2008, 455, 333–340; (d) P. D. Oldenburg
and L. J. Que, Catal. Today, 2006, 117, 15–21.
5 (a) S. Taktak, W. Ye, A. M. Herrera and E. V. Rybak-Akimova,
Inorg. Chem., 2007, 46(7), 2929–2942; (b) K. Chen and L. Que Jr.,
Chem. Commun., 1999, 1375–1376; (c) C. Duboc-Toia, S. Menage,
C. Lambeaux and M. Fontecave, Tetrahedron Lett., 1997, 38,
3727–3730; (d) T. K. Paine, M. Costas, J. Kaizer and L. J. Que,
JBIC, J. Biol. Inorg. Chem., 2006, 11, 272–276; (e) Y. M. Chiou
and L. J. Que, J. Am. Chem. Soc., 1992, 114(19), 7567–7568.
6 (a) S. Taktak, S. V. Kryatov, T. E. Haas and E. V. Rybak-Akimova,
J. Mol. Catal. A: Chem., 2006, 259, 24–34; (b) R. Mas-Balleste
L. Que Jr., J. Am. Chem. Soc., 2007, 129, 15964–15972;
(c) L. Gomez, X. Fontrodona, X. Ribas and M. Costas,
Chem.–Eur. J., 2008, 14, 5727–5731; (d) A. Company, Y. Feng,
´
and
´
M. Guell, X. Ribas, J. M. Luis, L. Que Jr. and M. Costas,
¨
Chem.–Eur. J., 2009, 15, 3359–3362.
7 (a) M. S. Chen and M. C. White, Science, 2007, 318, 783–787;
(b) K. Suzuki, P. D. Oldenburg and L. Que Jr., Angew. Chem., Int.
Ed., 2008, 47(10), 1887–1889; (c) O. V. Makhlynets and E. V. Rybak-
Akimova, Chem.–Eur. J., 2010, 16, 13995–14006; (d) O. V. Makhlynets,
Fig. 3 Time-resolved UV-vis spectra of the FeIII(OOH) formation at
27 1C in acetonitrile ([1] = 1 mM, [H2O2] = 30 mM). Maximum
accumulation of FeIII(OOH) was seen in 2 s after mixing reagents (red-
spectrum of 1, purple-spectrum of FeIII(OOH)). Inset: EPR spectrum
(120 K) of the FeIIIOOH, generated by mixing 1 and H2O2 in
acetonitrile at 0 1C and freezing in B7 s after reagents were mixed
([1] = 5 mM, [H2O2] = 50 mM after mixing). Spectrum simulated
with SimFonia with g values 2.185, 1.98, and 1.955. Another low-spin
species (g = 2.40, 2.17, 1.90) most likely results from FeII - FeIII
oxidation without OOH coordination.
P. Das, S. Taktak, M. Flook, R. Mas-Balleste, E. V. Rybak-Akimova
´
and L. Que Jr., Chem.–Eur. J., 2009, 15(47), 13171–13180.
8 (a) C. R. Smith, J. Am. Chem. Soc., 1928, 50, 1936–1938;
(b) G. Lunn and E. B. Sansone, J. Org. Chem., 1986, 51, 513–517.
9 J. Grimshaw, Electrochemical reactions and mechanisms in organic
chemistry, Elsevier Science B.V., 2000.
10 (a) J.-J. Girerd, F. Banse and A. J. Simaan, Struct. Bonding, 2000,
97, 145; Y. Jiang, J. Telser and D. P. Goldberg, Chem. Commun.,
2009, 6828–6830.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 687–689 689