Highly efficient epoxidation of olefins using aqueous H2O2 and catalytic methyltrioxorhenium/pyridine: Pyridine-mediated ligand acceleration

Full text of DOI:10.1021/ja970623l

J. Am. Chem. Soc. 1997, 119, 6189-6190
Highly Efficient Epoxidation of Olefins Using
6189
Aqueous H2O2 and Catalytic Methyltrioxorhenium/
Pyridine: Pyridine-Mediated Ligand Acceleration
Joachim Rudolph, K. Laxma Reddy, Jay P. Chiang, and
K. Barry Sharpless*
Department of Chemistry, The Scripps Research Institute
10550 N. Torrey Pines Road, La Jolla, California 92037
ReceiVed February 26, 1997
We report here the discovery of a ligand-accelerated1 mode
for methyltrioxorhenium (MTO)-catalyzed olefin epoxidations.2,3
Aqueous H2O2 is the oxidant, and the ligands are pyridine
derivatives. In addition to its beneficial effect on rate, the
pyridine ligand shuts down the acid-catalyzed ring-opening
reactions which are the bane of most epoxidation methods in
use today.2,4 The standard procedure is exemplified in Scheme
1 for 1-phenylcyclohexene whose epoxide is sensitive to acid-
catalyzed destruction and is difficult to prepare by most existing
epoxidation methods.5
Use of aqueous H2O2 as the oxidant in transition metal
catalyzed epoxidations was first described by Venturello who
employed a tungstate catalyst under phase transfer conditions.6
A more effective version was recently published by Noyori,
but this system also presents epoxide-opening problems caused
by the slight acidity of the reaction milieu.7,8
Figure 1. Reaction profile of the MTO-catalyzed epoxidation of
cyclooctene in different solvents with different amounts of pyridine
added (0.5 mol % MTO, 2 equiv of 30% H2O2, csubstrate ) 0.8 mol/L);
analysis via GC after quenching the aliquots with MnO2.
Scheme 1. Epoxidation of 1-Phenylcyclohexene by
MTO/Pyridine (50 mM scale)
Inorganic rhenium compounds such as Re2O7 or ReO3 were
long known to exhibit modest catalytic activity for H2O2-based
oxidations.9 However, real interest in the potential of rhenium
oxidation catalysts began with an extraordinary discovery by
the Herrmann group. They found that organometallic oxorhe-
nium(VII) species (especially MTO)10 are powerful epoxidation
catalysts with H2O2 as oxidant.2a-c Their work focused on the
use of anhydrous H2O2 (particularly in t-BuOH) because water
was detrimental, increasing losses via acid-catalyzed epoxide
destruction pathways.4 This epoxide-instability problem has not
been overcome by Herrmann or others,2 although improvements
have been made.2d,e,i The addition of tertiary nitrogen bases,
including pyridine, was found to suppress epoxide ring-opening
processes but at the expense of a strong detrimental effect on
catalyst actiVity. In any case, amine additives were apparently
not regarded as overall beneficial, since they do not appear in
any of the recommended “general procedures”.2c,d
In accord with observations of Herrmann and Adam, we
found that a range of saturated nonaromatic tertiary amines
strongly inhibit catalyst activity. This effect is independent of
the solvent, the amount of amine, and the presence or absence
of water. However, we were surprised to find that pyridine
and pyridine derivatives exhibited a remarkable acceleration
effect on the epoxidation rate, for pyridine too was found to
have a deleterious effect in the earlier work.2c,11 This accelera-
tion effect of pyridines is most pronounced in aprotic and
noncoordinating solvents (e.g., CH2Cl2 and CH3NO2).
(1) Berrisford, D. J.; Bolm, C.; Sharpless, K. B. Angew. Chem., Int. Ed.
Engl. 1995, 34, 1059.
(2) (a) Hoechst AG; Herrmann, W. A.; Marz, D. W.; Kuchler, J. G.;
Weichselbaumer, G.; Fischer, R. W.; DE 3.902.357, 1989. (b) Herrmann,
W. A.; Fischer, R. W.; Marz, D. W. Angew. Chem., Int. Ed. Engl. 1991,
30, 1638-1641. (c) Herrmann, W. A.; Fischer, R. W.; Rauch, M. U.;
Scherer, W. J. Mol. Catal. 1994, 86, 243. (d) Adam, W.; Mitchell, C. M.
Angew. Chem. 1996, 108, 78-581; Angew. Chem., Int. Ed. 1996, 35, 533.
(e) Boelow, T. R.; Spilling, C. S. Tetrahedron Lett. 1996, 37, 2717. (f)
Al-Ajlouni, A. M.; Espenson, J. H. J. Am. Chem. Soc. 1995, 117, 9243. (g)
Pestovsky, O.; van Eldik, R.; Huston, P.; Espenson, J. H. J. Chem. Soc.,
Dalton Trans. 2 1995, 133. (h) Al-Ajlouni, A. M.; Espenson, J. H. J. Org.
Chem. 1996, 61, 3969. (i) ARCO Chemical Technology; Crocco, G. L.;
Shum, W. P.; Zajacek, J. G.; Kesling, H. S., Jr.; US 5.166.372, 1992.
(3) For an excellent review on rhenium and technetium oxo complexes
in the study of organic oxidation mechanisms, see: Gable, K. P. AdV.
Organomet. Chem. 1997, 41, 127.
(4) Epoxide ring opening is a serious problem in the ligand-free MTO-
catalyzed epoxidation process,2 which is therefore only applicable for olefins
which yield very robust epoxides (e.g., cyclooctene).
(5) Berti, G.; Bottari, F.; Macchia, B.; Macchia, F. Tetrahedron 1965,
21, 3277.
(6) (a) Venturello, C.; Alneri, E.; Ricci, M. J. Org. Chem. 1983, 48,
3831. (b) Venturello, C.; D’Aloisio, R. J. Org. Chem. 1988, 53, 1553. (c)
See also: Prandi, J.; Kagan, H. B.; Mimoun, H. Tetrahedron Lett. 1986,
27, 2617.
(7) Sato, K.; Aoki, M.; Ogawa, M.; Hashimoto, T.; Noyori, R. J. Org.
Chem. 1996, 61, 8310.
While the reaction profile for the epoxidation of cyclooctene
(Figure 1) clearly shows the rate enhancing effect of the pyridine
ligand, it also reveals that catalyst lifetime is critically dependent
on the amount of pyridine present. The initial rates using either
1 mol % or 12 mol % pyridine in CH3NO2 (or CH2Cl2, not
shown) as solvent are nearly the same, but in the former case,
the catalyst decomposes completely within about 5 min,
resulting in poor conversion. In CH3NO2 or CH2Cl2 and about
3 mol % (or more) pyridine present, the catalyst is preserved
and cyclooctene is more than 95% converted to the epoxide
within 15 min.12 In summary, pyridine plays three crucial roles
in enhancing this process: (1) it speeds catalytic turnover, (2)
it prevents decomposition of epoxide products, and (3) in
sufficient concentration, it increases catalyst lifetime. The fact
that pyridine actually accelerates catalyst decomposition until
a threshold concentration is reached, probably explains why
earlier workers did not see its dramatic beneficial effects.
Furthermore, with t-BuOH as the solvent, the enhancing effects
of pyridine are only evident at much higher pyridine concentra-
tions.13 Since both CH3NO2 and CH2Cl2 result in biphasic
(8) (a) Romano, U.; Esposito, A.; Maspero, F.; Neri, C.; Clerici, M. G.
Chim. Ind. (Milan) 1990, 72, 610. (b) Clerici, M. G.; Ingallina, P. J. Catal.
1993, 140, 71.
(9) (a) E. I. DuPont de Nemours & Co.; Parshall, G. W.; US 3,646,130
and 3,657,292, 1972. (b) See references in Applied Homogeneous Catalysis
with Organometallic Compounds Vols. I and II; Cornils, B., Herrmann, W.
A., Eds.; VCH: Weinheim, 1996.
(10) Methyltrioxorhenium (MTO) was first prepared by Beattie and
Jones: Beattie, I. R.; Jones, P. J. Inorg. Chem. 1979, 18, 2318.
(11) It is known that MTO decomposes in basic aqueous solutions, and
the process has been studied in detail by Espenson: Abu-Omar, M. M.;
Hansen, P. J.; Espenson, J. H. J. Am. Chem. Soc. 1996, 118, 4966.
Investigations aimed at resolving this apparent conflict with our system
are currently underway.
(12) Preliminary results show that the amount of water present (delivered
with H2O2 and/or generated in the course of the reaction) has a major
influence on catalyst activity, both turnover rate and lifetime.
S0002-7863(97)00623-9 CCC: $14.00 © 1997 American Chemical Society

6190 J. Am. Chem. Soc., Vol. 119, No. 26, 1997
Table 1. MTO/Pyridine-Catalyzed Epoxidation of Olefinsa
Communications to the Editor
Table 2. MTO/Pyridine-Catalyzed Epoxidation of Dienesa
a Reaction conditions: csubstrate ) 2 mol/L; 0.5 mol % MTO, 12 mol
% pyridine, 2.5 equiv 50% aqueous H2O2, CH2Cl2, rt, water bath, 2
mmol scale, selectivities and rates determined by 1H NMR (400 MHz).
b Reaction performed at 10 °C. c Chemoselectivity with respect to the
epoxidation reaction. d See Supporting Information for details.
at present:19-21 (1) the MTO/pyridine process is comparable in
cost as well as safer than the m-CPBA process; (2) both
selectivity and scope are much greater, because even acid
sensitive epoxides do not suffer ring opening or rearrangement
reactions;21 (3) the new system is more reactive, can be run
with significantly less solvent, and workup/product isolation is
substantially easier; finally, (4) the only byproduct is water.
This ligand-accelerated, rhenium-catalyzed epoxidation pro-
cess may be the first example of rate enhancement by a basic
ligand of an oxidative transformation involving peroxometal
species.22 The mechanistic basis for this pyridine effect deserves
careful study but could take years to work out (cf., OsO4/ligand/
olefin mechanism). HoweVer, the importance of decoupling
epoxidation actiVity from acidity will be immediately apparent
to all synthetic chemists.23 The search for a simple, catalytic
epoxidation process which functions optimally under neutral-
to-basic conditions has been a constant, albeit elusive, goal of
ours for the past 25 years. Having found such a system,24 we
are of course interested in understanding and exploiting the new
reactivity features it offers. For example, experiments aimed
at adding the feature of enantioselectivity are now underway.
a Reaction conditions: csubstrate ) 2 mol/L (except entry 2 where 3
mol/L was used); 0.5 mol % MTO, 12 mol % pyridine, 1.5 equiv 30%
aqueous H2O2, CH2Cl2, rt, water bath. b Determined on a 2 mmol scale
by GLC or 1H NMR (400 MHz) using internal standards. c Determined
on a 50 mmol scale.
Acknowledgment. This paper is dedicated to Anthony Rappe´ and
William Goddard whose pioneering calculations on metal oxo species
inspired this work. We thank the National Institutes of Health (GM
28384), the National Science Foundation (CHE 9531152), the W. M.
Keck Foundation, and the Skaggs Institute for Chemical Biology for
financial support. J.R. is grateful for a NATO postdoctoral fellowship
granted by the Deutscher Akademischer Austauschdienst (DAAD).
conditions, we also determined the reaction profiles shown in
Figure 1 in a homogeneous solvent system (i.e., CH3NO2/t-
BuOH 85:15 v/v) and the results were almost identical.14 In
addition, when the more sterically hindered 2-picoline is used
in place of pyridine, the acceleration feature is lost regardless
of the solvent system employed.14 These control experiments
speak against both phase transfer and/or simple base effects as
the origin of the rate enhancement by pyridine.
The wide scope of this process is revealed in Tables 1 and
2.15 Even epoxides that are very sensitive to ring opening (Table
1, entries 3, 11, and 17) do not undergo hydrolytic ring
opening.16 Epoxidation of selected cyclic dienes resulted in high
diastereoselectivities for the cases studied (Table 2).17,18
We conclude by enumerating some advantages of this new
process over the m-chloroperbenzoic acid (m-CPBA) method,
which is the one in widest use for research-scale epoxidations
Supporting Information Available: Experimental details (6 pages).
See any current masthead page for ordering and Internet access
instructions.
JA970623L
(19) The buffered m-CPBA procedures partly overcome the “acidity
problem” but not as efficiently as this new MTO/pyridine system. See for
example: (a) Camps, F.; Coll, J.; Messeguer, A.; Pujol, F. J. Org. Chem.
1982, 47, 5402. (b) Imuta, M.; Ziffer, H. J. Org. Chem. 1979, 44, 1351. (c)
Anderson, W. K.; Veysoglu, T. J. Org. Chem. 1973, 38, 2267. Only the
Payne epoxidation process20 and the Murray DMDO reagent21 are of
comparable effectiveness.
(20) Payne, G. B. Tetrahedron 1962, 18, 763. The Payne epoxidation is
often the method of choice for industrial batch-type applications, but on a
small scale, the need for continuous pH-control is inconvenient.
(21) Murray, R. W.; Singh, S. Org. Synth. 1996, 74, 91.
(13) J. Rudolph, K. L. Reddy, J. P. Chiang, K. B. Sharpless, unpublished
results, not shown in Figure 1.
(14) See Supporting Information for these results.
(22) An accelerating effect for tertiary amine ligands in catalytic
epoxidations is well documented for systems involving d2, d3, and d4 oxo
metal species. For reviews, see: (a) Meunier, B. Bull. Soc. Chim. Fr. 1986,
578. (b) McMurry, T. J.; Groves, J. T. In Cytochrome P-450: structure,
mechanism, and biochemistry; Ortiz de Montellano, P. R., Ed.; Plenum
Press: New York, 1986; p 1.
(15) (a) Due to practical considerations, we used CH2Cl2 as solvent for
all examples even though CH3NO2 is slightly superior with regard to reaction
rate. (b) See Supporting Information for scale-up experimental procedure.
(16) Such substrates yield only products derived from acid-catalyzed
epoxide opening under the ligand-free conditions of the original MTO-
catalyzed procedure even though anhydrous H2O2 is the oxidant: J. Rudolph,
K. L. Reddy, J. P. Chiang, K. B. Sharpless, unpublished results.
(17) Diastereoselective MTO-catalyzed epoxidations of allylic alcohols
and other chiral olefins using the H2O2-urea complex as oxidant have
already been reported by Adam2d and Boelow.2e
(18) Compare to the peracid-mediated bisepoxidation of 1,4-cyclohexa-
diene and 1,5-cyclooctadiene, respectively: (a) Craig, T. W.; Harvey, G.
R.; Berchtold, G. A. J. Org. Chem. 1967, 32, 3743. (b) Cope, A.; Fisher,
B. S.; Funke, W.; McIntosh, J. M.; McKervey, M. A. J. Org. Chem. 1969,
34, 2231.
(23) With few exceptions (e.g., Payne20 and Murray reagents21), the
existing catalytic and stoichiometric epoxidation systems are most reactive
under acidic conditions, and of course, the latter often cause destruction of
the epoxide product.
(24) The complex path by which we arrived at the present system will
be described elsewhere. Suffice it to say here that the Rappe´-Goddard
“spectator oxo” concept has been a key guiding principle ever since it was
first published in 1980: (a) Rappe´, A. K.; Goddard, W. A., III Nature 1980,
285, 311. (b) Rappe´, A. K.; Goddard, W. A., III J. Am. Chem. Soc. 1982,
104, 448. (c) Rappe´, A. K.; Goddard, W. A., III J. Am. Chem. Soc. 1982,
104, 3287.