Selective epoxidation in dense phase carbon dioxide

Article abstract of DOI:10.1039/a801655k

Selective epoxidations with transition metal catalysts (V, Ti, Mo) and Bu^tOOH proceed with high conversions and high selectivity in dense phase carbon dioxide.

Full text of DOI:10.1039/a801655k

View Article Online / Journal Homepage / Table of Contents for this issue
Selective epoxidation in dense phase carbon dioxide
a
b
c
b
David R. Pesiri, David K. Morita, William Glaze and William Tumas* †
a
Department of Environmental Sciences and Engineering, University of North Carolina, Chapel Hill, NC 27599, USA
Chemical Science and Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545, USA
Carolina Environmental Program, University of North Carolina, Chapel Hill, NC 27599, USA
b
c
Selective epoxidations with transition metal catalysts (V, Ti,
anhydrous conditions of our experiments the reactions were
t
Mo) and Bu OOH proceed with high conversions and high
visibly homogeneous and the product epoxides were found to be
stable and could be isolated in high yields as their acetate
derivatives (up to 91%).6
selectivity in dense phase carbon dioxide.
t
The selective catalytic oxidation of olefins constitutes one of
the most fundamentally important classes of synthetically
useful reactions and is typically carried out in aromatic or
chlorinated hydrocarbon solvents.1 We report here the first
demonstration that dense phase carbon dioxide‡ is an effective
solvent for the selective oxidation of activated alkenes using
We have found that the reactivity of olefins with Bu OOH
i
and VO(OPr )
3
as a catalyst parallels that reported in conven-
tional solvents; activated olefins react faster than simple
,2
7
8
olefins and allylic alcohols react faster than homoallylics. At
25 °C most homoallylic and allylic alcohols reacted completely
i
within 24 h with VO(OPr )
3
, while no detectable reaction was
t
Bu OOH and a number of transition metal catalysts. High valent
observed for oct-1-ene and cyclohexene under the same reaction
conditions. Allylic alcohols were found to react about four times
i
t
3
oxovanadium(v) tri(isopropoxide) [VO(OPr ) ] with Bu OOH
catalyzes the epoxidation of a wide range of allylic and
homoallylic alcohols with high selectivities and catalytic
faster than homoallylics.
We studied the kinetics of the VO(OPr )
tion of olefins to benchmark the reactivity in CO
i
3
-catalyzed epoxida-
relative to that
turnovers in CO
catalysts such as VO(acac)
dense phase CO . In the presence of chiral ancillary tartrate
2
. We have also investigated other known
2
2
and Mo(CO) for oxidations in
6
in organic solvents since no rate data in any solvent have been
reported for epoxidation reactions with this catalyst to date. The
rate law was found to be first order in olefin, oxidant and
2
ligands, titanium isopropoxide catalysts lead to epoxides with
high enantioselectivity in this medium in what is believed to be
t
catalyst with an inverse, non-integral dependence on Bu OH.
the first example of stereoselective oxidation in a supercritical
Table 2 summarizes our kinetic data and shows the overall
solvent effect on the epoxidation reaction is not great, as
expected for such a non-polar (and non-ionic) reaction. At low
constant catalyst concentrations, and during initial conditions
i
fluid. We have found that the VO(OPr )
3
-catalyzed epoxidation
of allylic and homoallylic alcohols proceeds three times faster
in CO than in hexane, a solvent to which CO solubility and
reactivity have been compared.
Several recent publications have demonstrated the potential
of supercritical CO as an alternative reaction medium for
2
2
3
t
(low [Bu OH]), we can consider the observed rate constant to be
pseudo-second order. The rate constant, kA, for (Z)-non-3-en-
2
1 21
2
1-ol in CO
2
was determined to be 9 m
s
at 25 °C. Rates for
4
i
synthetic transformations. Although there are several descrip-
tions of mostly non-selective, low conversion, heterogeneous
epoxidations using VO(OPr )
CH Cl by a third and are roughly equal to those observed in
toluene and acetonitrile. The rate constant is about three times
larger in CO than in hexane, suggesting that aromatic solvents
3
are slower than those observed in
2
2
5
oxidations, there have been no reports on selective homo-
geneous metal-catalyzed oxidations in dense phase carbon
2
dioxide despite the fact that CO
therefore should be an ideal solvent for this chemistry. Table 1
contains our results from a wide range of allylic and homoallylic
2
is oxidatively stable and
may be better models than alkanes for solubility in dense phase
carbon dioxide.§ The rate of the epoxidation reaction using
i
VO(OPr )
3
was found to be at least 20-fold slower when run
t
t
alcohols which were oxidized with high conversions and
with 90% Bu OOH in water rather than anhydrous Bu OOH–
i
i
selectivities to the corresponding epoxides using VO(OPr )
3
and
decane.¶ VO(OPr )
subject to hydrolysis of the V–OR bonds to form stable, and
apparently CO -insoluble, complexes in the presence of H
thus limiting its activity.
VO(acac) is an effective epoxidation catalyst in organic
solvents but we have been unable to obtain significant
3
is a moisture sensitive catalyst and is
t
Bu OOH in liquid CO
2
at 25 °C. We could find no inherent
since
advantage of operating above the critical point of CO
2
2
2
O
these reactions are often carried out at or below room
temperature in conventional solvents. Therefore, the majority of
our studies focused on liquid carbon dioxide. Under the
2
Table 1 Epoxidation of allylic and homoallylic alcohols with VO(OPri)
in liquid CO a
3
2
R1
R2
R1
H
R1
R2
R2
O
H
n
Conversion %
Selectivity (%)
(CH2)nOH
(CH2)nOH
Epoxide
Aldehyde
Me
Me(CH
Me
Me
Me
H
H
H
Me
Me
Me
Et
1
> 99
> 99
> 96
> 99
> 99
> 99
> 99
> 99
85
> 99
86
> 95
> 99
89
15
—
—
—
—
—
—
—
2
)
2
1
1
1
1
2
2
2
2
CNCH(CH
2
)
2
2
2 2
CNCH(CH )
Et
H
H
89
99
Me(CH
2
)
4
a
2 3
4 h reaction in liquid CO (3.5 mol%), 0.42 mm substrate, 100 mm ROOH, magnetically stirred and analyzed by GC.
(10.3 bar), 1.47 mm VO(OPri)
2
Chem. Commun., 1998
1015

View Article Online
i
Table 2 Rate of epoxidation of (Z)-non-3-en-1-ol with VO(OPr )
3
in
Directed Research and Development (LDRD) grant. We would
like to acknowledge many helpful discussions with Dr Tom
Baker and Dr Steve Buelow.
a
serveral solvents
Solvent
kA/m2 s21
1
Relative rate
CH
MeCN
PhMe
2
Cl
2
30
18
17
9
5
3
3.3
2.0
1.9
1.0
0.6
0.3
Notes and References
† E-mail: tumas@lanl.gov
‡ Dense phase fluids refer to systems that would be considered gases at their
temperature of use, but are compressed to the point that they have liquid-like
CO
CCl
n-C
2
4
2
3
6
H
14
densities (r = 0.3 - 1.0 g cm ) including systems above their critical point
e.g. supercritical fluids).
(
a
[
VO(OPri)
t
3
] = 1.47 mm (3.5 mol%), [Bu OOH] = 100 mm, [olefin] = 42
§ The solubility and reactivity of dense phase CO
2
has been compared to
mm, T = 24 °C, CO
2
reactions run at 29 bar. kA = kinitial,pseudo second order
hydrocarbons, aromatics and fluorocarbons.
i
[VO(OPr )
3
]. Rates reported in this work are accurate to within 10%.
¶ It is important to note that the use of decane to introduce the alkyl
hydroperoxide to sc CO reactions constituted less than 1% of the solvent
2
t
volume (600 ml of 5.6 m Bu OOH in decane, or 265 ml decane in 33 ml of
conversions in dense phase CO
presumably due to the limited solubility of VO(acac)
We are in the process of examining other, more ‘CO
acetylacetonate-based ligands including highly fluorinated
systems. We have also examined the Mo(CO)–Bu OOH
oxidation system and found that unactivated alkenes such as
cyclohexene can be oxidized to their corresponding diols in CO
at elevated temperatures [95 °C, eqn. (1)]. We were surprised
2
under similar conditions,
in CO
-philic’,
solvent).
2
2
.
∑ Preliminary results on the Mo(CO) -catalyzed oxidation of cyclic olefins
6
t
25
2
with Bu OOH (aqueous and anhydrous in decane) yielded kA = 5 3 10
1
2
23 21
s
for anhydrous ROOH, k A = 1310
s
for aqueous ROOH.
t
** The overall experimental setup and the general experimental procedures
including sampling and product recovery have been described in detail
elsewhere.4 All high pressure reactions were carried out in custom-built
stainless steel high pressure cells (33 ml volume) with sapphire view
windows which enabled the direct visual observation of reactions. Proper
safety precautions were taken for these high pressure reactions. An
homogeneous solution was observable and complete conversion to the
corresponding epoxide was detected, usually within 2 h. Yields were
determined by gas chromatography of letdown solutions using authentic
,11
2
9
O
OH
OH
OH
Mo(CO)6
+
+
(1)
t
Bu OOH(aq.)
1
standards when available. GC–MS and H NMR analysis were also used to
sc CO2
(73%)
(10%)
(10%)
confirm product identification.
t
to find that oxidations utilizing aqueous Bu OOH (e.g. 90%
t
Bu OOH–H
those run with dry Bu OOH (e.g. Bu OOH–decane) for the
Mo(CO) system suggesting an interesting effect of water.∑
2
O) gave significantly higher conversions than
1 Organic Syntheses by Oxidation with Metal compounds, ed. W. J. Mijs,
and C. R. H. I. de Jonge, Plenum Press, New York, 1986; R. A. Sheldon
and J. K. Kochi, Metal Catalyzed Oxidation of Organic Compounds,
Academic Press, New York, 1981.
t
t
6
High concentrations of Mo catalysts (200 mg, 12.5 mol%) were
necessary for this reaction. The oxidation of cyclohexene
resulted in a 73% selectivity to cyclohexene-1,2-diol at 74%
conversion (presumably from the hydrolysis of cyclohexene
oxide) during a 12 h reaction using aqueous Bu OOH with
additional products derived from allylic oxidation [eqn. (1)].
Using anhydrous Bu OOH resulted in only 15% conversion of
cyclohexene to oxidized products.
We have found that Ti(OPr )
tartrate ligands, results in enantioselective epoxidation catalysis
2
K. B. Sharpless, Chemtech., 1985, 692; E. D. Mihelich, K. Daniels and
D. J. Eickhoff, J. Am. Chem. Soc., 1981, 103, 7690; D. J. Berrisford, C.
Bolm and K. B. Sharpless, Angew. Chem., Int. Ed. Engl.,1995, 34, 1059;
K. A. Jorgensen, Chem Rev., 1989, 89, 431.
t
3
4
J. A. Hyatt, J. Org. Chem., 1984, 49, 5097.
A number of reactions have been carried out in dense phase CO , for
2
t
example see: P. G. Jessop, T. Ikariya and R. Noyori, Science, 1995, 269,
1065; D. A. Morgenstern, R. M. LeLacheur, D. K. Morita, S. L.
Borkowski, S. Feng, G. H. Brown, L. Luan, M. F. Burk and W. Tumas,
in Green Chemistry, Designing Chemistry for the Environment, ed. P. T.
Anastas and T. C. Williamson, American Chemical Society Sympo-
sium, 1996, No. 626, p. 132; P. G. Jessop, T. Ikariya, and R. Noyori,
Chem. Rev., 1995, 95, 259; J. W. Rathke, R. J. Klingler and T. B.
Krause, Organometallics, 1991, 10, 1350; J. B. McClain, D. E. Betts,
D. A. Canelas, E. T. Samulski, J.M. DeSimone, J. D. London, H. D.
Cochran, G. D. Wignall, D. Chillura-Martino and R. Triolo, Science,
1996, 274, 2049.
i
4
, in the presence of chiral
in CO
2
[eqn. (2)]. Although the limited solubility of diethyl
O
i
Ti(OPr )4
OH
OH (2)
*
*
t
Bu OOH
diisopropyl L-tartrate
5
L. Zhou and A. Akgerman, Ind. Eng. Chem. Res., 1995, 34, 1588; L.
Zhou, C. Erkey and A. Akgerman, Ind. Eng. Chem. Res., 1995, 41,
2
tartrate in liquid CO limits the activity of this system, the use
2
1
122; K. M. Dooley and F. C. Knopf, Ind. Eng. Chem. Res., 1987, 26,
910; P. Srinivas and M. Mukhopadhyay, Ind. Eng. Chem. Res., 1997,
of diisopropyl tartrate in its place resulted in high conversions to
epoxide. Only 16% enantiomeric excess (ee) was obtained for
36, 2066; R. N. Occhiogrosso and M. A. McHugh, Chem. Eng. Sci.,
1987, 42, 2481; G. Suppes, R. N. Occhiogrosso and M. A. McHugh, Ind.
Eng. Chem. Res., 1997, 36, 2066.
the epoxidation of(E)-hex-2-en-1-ol with the titanium–diisopro-
i
pyl tartrate (0.322 mmol Ti(OPr )
4
, 0.389 mmol diisopropyl
6
7
8
9
T. Itoh, K. Jitsukawa, K. Kaneda and S. Teranishi, J. Am. Chem. Soc.,
tartrate) catalyst at 25 °C (93% conversion). At 0 °C the
enantioselectivity increased to 87% with the same substrate
during a 72 h reaction ( > 99% conversion). This unoptimized
result approaches the 94% ee reported by Katsuki and
Sharpless.10
1
979, 101, 159.
K. Takai, K. Oshima and H. Nozaki, Bull. Chem. Soc. Jpn., 1983, 56,
791.
R. B. Dehnel and G. H. Whitman, J. Chem. Soc., Perkin Trans. 1, 1979,
, 953.
3
4
2
It is clear that dense phase CO is an effective solvent for
2
J. Kolis, et al., Mo catalyzed oxidations in dense phase CO (personal
these olefin epoxidation reactions, with rates and selectivities
similar to those observed in organic solvents. Given its
oxidative stability, CO shows promise as an environmentally
2
benign solvent alternative for chemical oxidation. We are
currently exploring a number of other catalytic reactions,
including heterogeneous catalytic oxidation, in this important
medium.**
communication).
10 T. Katsuki and K. B. Sharpless, J. Am. Chem. Soc., 1980, 102, 5976.
11 S. Buelow, P. Dell’Orco, D. K. Morita, D. R. Pesiri, E. Birnbaum, S.
Borkowski, G. Brown, S. Feng, L. Luan, D. A. Morgenstern and W.
Tumas, in Frontiers in Benign Chemical Synthesis and Processing, ed.
P. T. Anastas and T. C. Williamson, Oxford University Press, in the
press.
This work was supported as part of the Los Alamos Catalysis
Initiative by The Department of Energy through a Laboratory
Received in Cambridge, UK, 27th February 1998; 8/01655K
1016
Chem. Commun., 1998