Iron(III) oxamato-catalyzed epoxidation of alkenes by dioxygen and pivalaldehyde

Article abstract of DOI:10.1039/a705787c

A new iron(III)-carbonato monomeric complex of orthophenylenebis(oxamato) (opba) 1 is synthesized, and spectroscopically and structurally characterized; it is a moderately efficient non-heme catalyst for the aerobic epoxidation of alkenes with co-oxidatio

Full text of DOI:10.1039/a705787c

View Article Online / Journal Homepage / Table of Contents for this issue
Iron(iii) oxamato-catalyzed epoxidation of alkenes by dioxygen and
pivalaldehyde
Rafael Ruiz,a Maria Triannidis,a Ally Aukauloo,a Yves Journaux,*a Isabel Ferna´ndez,b Jose´ R. Pedro,*b
Beatriz Cervera,c Isabel Castroc and M. Carmen Mun˜ozd
a Laboratoire de Chimie Inorganique, URA 420, CNRS, Universite´ de Paris-Sud, 91405 Orsay, France
b Departament de Qu´ımica Orga`nica, Facultat de Qu´ımica, Universitat de Vale`ncia, 46100 Burjassot, Vale`ncia, Spain
c Departament de Qu´ımica Inorga`nica, Facultat de Qu´ımica, Universitat de Vale`ncia, 46100 Burjassot, Vale`ncia, Spain
d Departamento de F´ısica Aplicada, Universidad Polite´cnica de Vale`ncia, 46071 Vale`ncia, Spain
A new iron(III)–carbonato monomeric complex of ortho-
phenylenebis(oxamato) (opba) 1 is synthesized, and spectro-
scopically and structurally characterized; it is a moderately
efficient non-heme catalyst for the aerobic epoxidation of
alkenes with co-oxidation of pivalaldehyde.
situation contrasts to previous descriptions of mononuclear
complexes with this ligand or its derivatives in which all four
donor atoms form a plane (trans geometry).7,8 The distortion
from planarity of the opba ligand is accommodated by a
disrotatory rotation of the C(3)–N(2) amide bond defined as the
angle between the C(6)C(3)N(2) and C(3)N(2)C(4) mean
planes [17 cf. 2° for the equivalent C(1)–N(1) amide bond], as
There has been always a great interest in iron coordination
chemistry owing to the relevant role that this transition metal
ion plays in biology, particularly as the active metal center
embedded in a large number of proteins involved in oxygen
activation chemistry.1 Now, this interest is directed towards
mono- and di-nuclear non-heme iron proteins, owing to the
surprisingly diverse functional properties exhibited by them.2
Amido groups from asparagine or glutamine residues in the
polypeptide backbone of some of these mononuclear iron
proteins are frequently shown to coordinate to the iron metal
center through their oxygen donor atom, like in lipid dioxyge-
nase and isopenicillin N-synthase.2a Moreover, amido nitrogen
coordination has been invoked in bleomycin, a natural iron-
containing glycopeptide with antibiotic and antitumor proper-
ties of fundamental importance in medicine.2a That being so, it
is evident that knowledge of the chemistry and reactivity
properties of iron–amido complexes becomes particularly
appropriate in order to understand the key steps of the
enzymatic mechanism proposed for this class of non-heme iron
proteins and antibiotics. Since the review of Sigel and Martin
devoted to the coordinating properties of the amide bond with
copper, nickel, cobalt and zinc metal ions,3 a considerable effort
has been made in the last fifteen years in order to detect iron ion
interactions with the amide group.4–6 Recently, we have
initiated an investigation program to assess the possible use of
bis-N-substituted oxamato ligands like ortho-phenylenebis-
(oxamato) (opba) in the stabilization of unusual high oxidation
state transition metal ions due to the high donor capacity of the
deprotonated-amido group.7 We report here on the synthesis
and physical characterization,† and the crystal and molecular
structure‡ of its corresponding oxamato iron(iii) complex of
formula [NMe4]3[Fe(opba)(CO3)]·5H2O 1, as well as its
activity toward alkene epoxidation with dioxygen.
previously observed in
a related cobalt(iii) diamido-
N-dialkoxido complex.9 This distortion is also reflected in the
metal–ligand bond lengths, with a significant difference
between the two Fe–O(carboxylate) bond distances of ca. 0.05
Å, the longer Fe–O(5) distance being trans to that of the
carbonate ligand Fe–O(8) which is the longest bond of the
octahedron. The Fe–N(amide) bond distances average 2.05 Å, a
value very close to that observed in high-spin iron(iii)
complexes containing amido nitrogen bonds,5 but significantly
longer than those of the analogous intermediate4b or low-spin6
iron(iii)–amide bonds (1.93–1.96 Å).
In order to uncover the potential reactivity of this new
iron(iii) complex, we have examined its ability to catalyse the
epoxidation of some representative disubstituted alkenes using
dioxygen as oxidant in the presence of an aldehyde.10 The
O(3)
C(2)
O(1)
C(1)
O(2)
C(10)
C(11)
N(1)
Fe
C(5)
O(8)
O(9)
O(7)
C(9)
N(2)
O(5)
C(6)
C(3)
C(4)
Complex 1 consists of mononuclear [Fe(opba)(CO3)]32
complex anions (Fig. 1), tetramethylammonium cations and
crystallization water molecules. The iron atom is bound to the
two deprotonated amido nitrogens and two carboxylate oxygens
of the opba ligand and to two oxygen atoms from the carbonate
in a rhombically distorted octahedral coordination geometry.
The carbonate chelating group is symmetrically bound to the
metal (2.034, 2.045 Å for Fe–O) occupying two cis positions of
the octahedron with a typical small bite angle [63.2° for
O(7)–Fe–O(8)]. This imposes a non-planar conformation for the
tetradentate opba ligand which is the only compatible with an
octahedral environment at the metal ion. In fact, the opba ligand
adopts a cis-b geometry with the N(1), N(2) and O(2) donor
atoms occupying three meridional positions around iron. This
C(8)
C(7)
O(6)
O(4)
Fig. 1 Perspective view of the anionic mononuclear unit of 1 with the atom
numbering scheme (thermal ellipsoids are at the 30% probability level and
hydrogen atoms have been omitted for clarity). Selected bond distances (Å)
and angles (°) with standard deviations in parentheses: Fe–N(1) 2.048(5),
Fe–N(2) 2.045(6), Fe–O(2) 1.986(4), Fe–O(5) 2.035(4), Fe–O(7) 2.034(5),
Fe–O(8) 2.045(5); N(1)–Fe–N(2) 75.9(2), N(1)–Fe–O(2) 78.5(2), N(1)–Fe–
O(5) 125.5(2), N(1)–Fe–O(7) 147.4(2), N(1)–Fe–O(8) 86.9(2), N(2)–Fe–
O(2) 142.8(2), N(2)–Fe–O(5) 77.9(2), N(2)–Fe–O(7) 118.9(2), N(2)–Fe–
O(8) 98.9(3), O(2)–Fe–O(5) 96.0(2), O(2)–Fe–O(7) 97.1(2), O(2)–Fe–O(8)
106.2(2), O(5)–Fe–O(7) 86.9(2), O(5)–Fe–O(8) 144.2(2), O(7)–Fe–O(8)
63.2(2).
Chem. Commun., 1997
2283

View Article Online
Table
1 Results for the epoxidation of alkenes by dioxygen and
would also like to express our gratitude to Dr J. J. Girerd and Dr
P. Mialane for fruitful discussions.
pivalaldehyde catalyzed by 1a
Entry
Alkene
t/h
Yieldb,c (%)
Footnotes and References
1
2
3
4
5
trans-Stilbene
cis-Stilbene
Cyclohexene
1,2-Dihydronaphthalene
1,2-Dihydronaphthalenee
3
95
15
45
50
80
* E-mail: jour@icmo.u-psud.fr
3d
4
† Syntheses and selected data for 1: A solution of the diethyl ester derivative
of opba (1.54 g, 5 mmol) in methanol (100 ml) was charged with a 25%
methanol solution of NMe4OH (10 cm3, 25 mmol) and the resulting mixture
was stirred at 60 °C for 15 min in order to facilitate the hydrolysis of the
ethyl ester groups. A methanolic solution (50 ml) of iron(iii) perchlorate
hydrate (1.77 g, 5 mmol) was then added dropwise via a dropping funnel
under stirring. The resulting intensely red coloured solution was filtered to
eliminate the white precipitate of NMe4ClO4, and reduced to a final volume
of 10 ml on a rotatory evaporator. The concentrated solution was treated
successively with diethyl ether and acetone to give a very hygroscopic
product (solid or oil) which was recuperated with acetonitrile (300 ml). The
suspension obtained was gently heated for 5 h with vigorous stirring and
then filtered to eliminate the remainder of the solid particles. Upon standing
at room temperature, a first crop of a side product appeared which was also
filtered and separated from the mother-liquor. Slow evaporation of the
filtered solution in air afforded, after 2 weeks, red needles of 1 in small
amounts which were picked by hand, dried on filter paper and stored under
vacuum owing to its hygroscopic character. The crystals so obtained were
suitable for X-ray diffraction. Satisfactory elemental analyses were obtained
(C, H, N, Fe). nmax/cm21 (KBr) 1660s (sh), 1635vs (br) and 1571s (CO)
(opba ligand and CO322) and 950s (NC) from NMe4+. UV–VIS (MeCN)
4
4
a Reactions were carried out at room temperature by adding a fluorobenzene
solution (0.2 ml) of the alkene (0.11 mmol) to a stirred mixture of the metal
catalyst (6.5 3 1023 mmol) and pivalaldehyde (0.33 mmol) in fluoro-
benzene (0.2 ml) under dioxygen atmosphere. The consumption of the
alkene and the formation of the corresponding epoxide during the course of
the reaction were monitored by TLC. The obtained epoxide and the
unreacted alkene were separated by flash column chromatography on silica
gel. b Yields refer to isolated and pure compounds (column chromatography
on silica gel). All compounds exhibited spectral data consistent with their
c
structures. In the absence of catalyst some extension of epoxidation was
observed. With a reaction time of 24 h the final conversion was almost
identical (ca. 20%). e In the presence of N-methylimidazole (0.11 mmol).
d
results are summarized in Table 1. Complex 1 catalyses the
epoxidation of trans-stilbene by dioxygen plus pivalaldehyde in
fluorobenzene solution with yields as high as 95% of the
corresponding epoxide, trans-stilbene oxide, after 3 h (entry 1).
By contrast, the percent conversion in the case of cis-stilbene is
of only 15–20% even after a total reaction time of 24 h (entry 2),
indicating that the catalytic epoxidation is greatly stereodepen-
dent. For both cis- and trans-stilbene, however, epoxides were
the only oxidation products as confirmed by 1H NMR
spectroscopy. As a matter of fact, no trace of the corresponding
ketone or alcohol were detected for the epoxidation of
cyclohexene which is typically regarded as a good substrate to
check for competition of alkene epoxidation vs. allylic oxida-
tion (entry 3). This observation suggests that typical free radical
intermediates are not directly involved as potential epoxidizing
agents. On the other hand, it is noteworthy that the amount of
epoxide obtained with 1 increases in the presence of
N-methylimidazole: for instance, almost a twofold yield
enhancement was found for the epoxidation of 1,2-dihydro-
naphthalene, i.e. from 50 to 80% after 4 h (entries 4 and 5,
respectively). Although it is premature to discuss the precise
mechanism at the present stage, N-methylimidazole most likely
acts as a monodentate donor ligand toward the iron(iii) ion by
replacement of the bidentate carbonate ligand, and thereby
affording a vacant coordination site on the iron center for
reaction with the potential oxidant. Furthermore, even if the
exact role of the metal complex in the epoxidation of alkenes by
dioxygen with co-oxidation of aldehydes is still unclear, recent
studies by Valentine and coworkers11 demonstrate that it
coordinates to the acylperoxy radicals generated in the auto-
oxidation of the aldehyde forming a metal-acylperoxo complex.
That being so, iron(iv)–acylperoxo or iron(v)–oxo species
derived by oxygen–oxygen bond cleavage of the acylperoxo
group, seem to us the more probable candidates to play the role
of active epoxidizing agents in our system. In this regard, it is
interesting to note that Collins et al. have isolated stable iron(iv)
complexes with amido-containing ligands which would be a
model for the proposed catalytic entities in our oxidation
system.4a Attempts to characterize these reactive intermediates
species using other transition metal ions with more accessible
high-valent oxidation states such as manganese are in progress
in our laboratory.
l
max/nm [e dm3 mol21 cm21)] 221 (4.0 3 104), 257 (4.2 3 104), 264 (sh),
308 (3.0 3 104), 408 (2.7 3 103). Magnetic moment (room temp.): 5.9 mB.
EPR spectrum (X-band, powder sample, liquid-N2 temperature): a strong
feature at geff = 4.3 which typically arises in a completely rhombic system
(E/D = 1/3) for D > 0.23 cm21
.
‡ X-Ray crystal structure analysis: Enraf-Nonius CAD-4 diffractometer,
Mo-Ka radiation, l 0.71069 Å, graphite monochromator, 293 K.
=
Lorentz and polarization effects but not absorption correction (m = 9.26
cm21). Data collection, solution and refinement: w-q, standard Patterson
methods with subsequent full-matrix least-squares method refinement.
SHELX86, SHELX93.12 C23H50FeN5O14, monoclinic, space group P21/a,
a
= 11.441(2), b = 17.496(2), c = 17.047(2) Å, b = 104.52(2)°,
U = 3303(1) Å3, Z = 4, Dc = 1.36 g cm23, 1 @ q @ 25°, crystal size 0.15
3 0.10 3 0.05 mm. 4456 unique reflections with 3362 assumed as observed
with I ! 2s(I). The hydrogen atoms were located from a difference
synthesis and refined with an overall isotropic thermal parameter.
Refinement on F2 of 390 variables with anisotropic thermal parameters for
all non-hydrogen atoms gave R = 0.079 and Rw = 0.108 with S = 1.8
(observed data). CCDC 182/628.
1 Active Oxygen in Biochemistry, ed. J. S. Valentine, C. S. Foote, A.
Greenberg and J. F. Liebman, Chapman and Hall, London, 1995.
2 (a) A. L. Nivorozhkin and J. J. Girerd, Angew. Chem., Int. Ed. Engl.,
1996, 35, 609 and references therein; (b) L. Que, Jr. and Y. Dong, Acc.
Chem. Res., 1996, 29, 190.
3 H. Sigel and R. B. Martin, Chem. Rev., 1982, 82, 385.
4 (a) T. J. Collins, K. L. Kostka, E. Mu¨nck and E. Uffelman, J. Am. Chem.
Soc., 1990, 112, 5637; (b) K. L. Kostka, B. G. Fox, M. P. Hendrich,
T. J. Collins, C. E. F. Rickard, L. J. Wright, E. Mu¨nck, J. Am. Chem.
Soc., 1993, 115, 6746.
5 Y. Yang, F. Diederich and J. S. Valentine, J. Am. Chem. Soc., 1991, 113,
7195.
6 X. Tao, D. W. Stephan and P. K. Mascharak, Inorg. Chem., 1987, 26,
754.
7 R. Ruiz, C. Surville-Barland, A. Aukauloo, E. Anxolabehere-Mallart,
Y. Journaux, J. Cano and M. C. Mun˜oz, J. Chem. Soc., Dalton Trans.,
1997, 745.
8 M. Tettouhi, L. Ouahab, A. Boukhari, O. Cador, C. Mathoniere and
O. Kahn, Inorg. Chem., 1996, 35, 4932.
9 T. J. Collins and J. M. Workman, Angew. Chem., Int. Ed. Engl., 1989,
28, 912.
10 J. Estrada, I. Fernandez, J. R. Pedro, X. Ottenwaelder, R. Ruiz and
Y. Journaux, Tetrahedron Lett., 1997, 38, 2377.
11 W. Nam, H. J. Kim, S. H. Kim, R. Y. N. Ho and J. S. Valentine, Inorg.
Chem., 1996, 35, 1045 and references therein.
12 G. M. Sheldrick, SHELX. A Program for Crystal Structure Determina-
tion, University of Go¨ttingen, DFR, 1986; G. M. Sheldrick,
SHELXL93. Program for the Refinement of Crystal Structures,
University of Go¨ttingen, 1993.
This work was supported by the DGICYT, Ministerio de
Educacio´n y Ciencia (Spain) through projects PB94-0985 and
PB94-1002. We acknowledge the financial support of the group
Rhoˆne-Poulenc. R. R. and B. C. thank the Ministerio de
Educacio´n y Ciencia (Spain) and the Conselleria de Educacio´ i
Cie`ncia de la Generalitat Valenciana (Spain) for grants. We
Received in Basel, Switzerland, 7th August 1997; 7/05787C
2284
Chem. Commun., 1997