Amino acid/copper-catalyzed enantioselective allylic benzoyloxylation of olefins in water promoted by diethylene glycol

Article abstract of DOI:10.1016/S0957-4166(03)00318-5

Amino acids, mainly L-proline, in association with diethylene glycol and copper salts have been used in water for a recyclable enantioselective version of the Kharasch-Sosnovsky reaction with e.e.s up to 42%.

Full text of DOI:10.1016/S0957-4166(03)00318-5

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 14 (2003) 1911–1915
Amino acid/copper-catalyzed enantioselective allylic
benzoyloxylation of olefins in water promoted by diethylene
glycol
Jean Le Bras* and Jacques Muzart
Unit e´ Mixte de Recherche ‘R e´ actions S e´ lectives et Applications’, CNRS-Universit e´ de Reims Champagne-Ardenne, BP 1039,
51687 Reims cedex 2, France
Received 24 March 2003; accepted 8 April 2003
Abstract—Amino acids, mainly
L
-proline, in association with diethylene glycol and copper salts have been used in water for a recyclable
enantioselective version of the Kharasch–Sosnovsky reaction with e.e.s up to 42%. © 2003 Elsevier Science Ltd. All rights reserved.
1
. Introduction
observed. The main strategy to obtain suitable water-
soluble ligands was the attachment of ionic or polar
substituents to standard ligands.
9
The development of methodologies for the asymmetric
functionalization of CꢀH bonds remains a challenge.
The Kharasch–Sosnovsky reaction, in which the allylic
carbon of an alkene is acyloxylated, offers access to
chiral allylic alcohols and has been the subject of
In the above context, we have quite recently developed
the aqueous non-enantioselective Kharasch–Sosnovsky
reaction by the mean of the hydrophilic ligand
[(HOCH CH NHCOCH ) NCH ] : mixing this ligand
1
,2
several studies. The initial asymmetric version of the
Kharasch–Sosnovsky reaction was a diastereoselective
oxidation involving the use of tert-butyl hydrogen per-
oxide as oxidizing reagent and a-ethyl camphorate as
the chiral auxiliary; oxidation products were obtained
2
2
2 2
2 2
and Cu(MeCN) BF in water led to an efficient and
4
4
reusable catalytic system for the allylic oxidation of
10
olefins by tert-butyl perbenzoate.
3
with low d.e., 57%. Later, tert-butyl peroxybenzoate
In the course of investigations to develop asymmetric
aqueous oxidation reactions, we thought to exploit the
high solubility of amino acids in water to obtain enan-
tioselective water-soluble catalysts. Indeed, the use of
simple commercial water-soluble molecules has been
rather neglected. Our studies lead us to report the first
enantioselective Kharasch–Sosnovsky reaction in water
(Scheme 1).
and the catalytic L-proline/copper system were found to
be superior; optimum selectivity being obtained using a
/1 ligand/copper ratio. Optimized oxidation of cyclo-
hexene using bis-prolinato-copper catalyst with
PhCO t-Bu and PhCO H in refluxing benzene yielded
9% cyclohexenyl benzoate with 45% e.e. Subse-
quently, bicyclic amino acid analogs of proline were
used. Such ligands were superior in both reactivity and
4
2
3
2
5
5
6
selectivity when a mixture of copper(II)/copper bronze
was used as catalyst.
In recent years, organometallic catalysis in aqueous
media has developed into an important field of
7
,8
research. These procedures make them valuable alter-
natives and have high economic and environmental
impact. Indeed, they do not call for dry solvents, they
may allow recycling of catalysts and furthermore
unique reactivity and selectivity have been sometimes
*
Corresponding author. Tel.: 00 33 (0)3 26 91 32 46; fax: 00 33 (0)3
6 91 31 66; e-mail: jean.lebras@univ-reims.fr
2
Scheme 1.
0
957-4166/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0957-4166(03)00318-5

1
912
J. Le Bras, J. Muzart / Tetrahedron: Asymmetry 14 (2003) 1911–1915
2
. Results and discussion
equiv.) and copper bronze (0.5 equiv.) as the source of
Cu , the reaction was complete in 18 h at room temper-
I
2.1. Preliminary investigation
ature yielding 95% allylic ester but with very low e.e.
0
(
entry 11). When lower amounts of Cu (from 0.5 to 0.1
Initial experiments in water were performed at 80°C
equiv.) were used in association with Cu(OAc) , the
reaction rates decreased regularly and enantioselectivi-
ties were not improved.
2
with the water-soluble complex Cu(MeCN) BF4 and
4
L
-proline as ligand (Table 1, entry 1). After 40 h,
cyclohexenyl benzoate was obtained in 13% yield; a
white solid isolated during work-up was identified as
benzoic acid. Since allylic oxidations in organic medium
with copper–amino acid complexes may be improved in
In conclusion, in these preliminary experiments, the
oxidation of cyclohexene with copper/proline associa-
tions in water affords the corresponding allylic ester in
fair yield but with low e.e. However, some experiments
have shown that after catalysis the water phase keeps a
blue color and was still catalytic active, this result urges
us to pursue the study to improve the enantioselectivity.
After a great deal of experimentation, we have found
out that the addition of diethylene glycol to the reac-
tion mixture was able to fulfil in part our objectives.
5
,11
the presence of an organic carboxylic acid,
the effect
of the addition of benzoic acid to the biphasic system
was then studied. With 0.5 equiv. of this acid, the
reaction was faster and afforded a better yield of 55%,
however with a very low e.e. (entry 2). Increasing the
amount of PhCO H to 1.5 equiv. improved slightly the
2
yield to 62% and the e.e. to 12% (entry 3). Under these
conditions, the water phase was catalytically active for
recycling experiments.
2.2. Effect of diethylene glycol
II
The use of a Cu complex, Cu(
L
-prolinato) ·2H O
,
Oxidation of cyclohexene could be performed at room
temperature in 18 h using Cu(OTf)2, L-proline, NaOH,
2
2
I
increased the e.e. to 18% (entry 4). As with Cu ,
decreasing the amount of benzoic acid afforded lower
e.e. (entry 5). A catalyst in situ prepared from
and PhCO H in a mixture of water (30 equiv.)/
2
diethylene glycol (9.4 equiv.), the allylic ester being
isolated in low yield (31%) but with 38% e.e. (Table 2,
entry 1). When the amount of diethylene glycol was
reduced to 5.2 equiv., reaction required 8 days to go to
the completion and e.e. decreased to 34% but yield was
improved to 62% (entry 2). Monitoring the reaction by
TLC has shown that only traces of allylic ester were
produced during the first 5 days; after this induction
period, the product was generated regularly until com-
plete disappearance of tert-butyl peroxybenzoate. This
observation led us to look for a procedure to reduce or
suppress this induction period.
Cu(OTf)2,
entry 6). Lower reaction rates were obtained in the
absence of NaOH (entry 7). In contrast to reactions
L-proline and NaOH led also to 18% e.e.
(
12
performed in an organic medium,
anthraquinone-2-sulfonic acid has no effect under these
aqueous conditions (entry 8). Decrease of the reaction
temperature with Cu or Cu catalysts (entries 9, 10)
afforded only traces of product even if usual co-sol-
vents used in biphasic process such as acetone or etha-
addition of
I
II
1
3
nol, were added. In contrast, when the catalytic
system was a syn-proportionation of Cu(OAc)2 (0.1
Table 1. Allylic oxidation of cyclohexene in water
Entry
Cu (0.1 equiv.)
L
-Proline
NaOH
PhCO H
T (°C)
Time (h)
Yield (%)a
E.e. (%)b
2
(
equiv.)
(equiv.)
(equiv.)
1
2
3
4
5
6
7
Cu(MeCN) BF
0.25
0.25
0.25
0
0
0
0
0
0
0.20
0
0.20
0.20
0.20
0
0
80
80
80
80
80
80
80
80
50
50
rt
40
17
17
17
17
17
40
17
72
72
18
13
55
62
50
44
70
71
66
nd
5
4
4
4
4
Cu(MeCN) BF
0.5
1.5
1.5
0.5
1.5
1.5
1.5
1.5
1.5
1.5
4
Cu(MeCN) BF
12
18
13
18
18
17
nd
nd
5
4
Cu(L
-prolinato) ·2H O
2
2
Cu(L
-prolinato) ·2H O
0
2
2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
0.45
0.45
0.45
0.45
0.45
0.45
c
8
9
Traces
Traces
95
1
1
0
1
Cu(CH CN) BF
3 4 4
0
d
II
Cu /Cu
a
b
c
Isolated yield calculated on the amount of PhCO t-Bu introduced.
S-Configuration of the main enantiomer.
3
Reaction carried out in the presence of anthraquinone-2-sulfonic acid (4.0 equiv.).
d
0
Cu(OAc) ·H O (0.1 equiv.) and Cu (0.5 equiv.) were used.
2
2

J. Le Bras, J. Muzart / Tetrahedron: Asymmetry 14 (2003) 1911–1915
1913
Table 2. Allylic oxidation of cyclohexene in water in the presence of diethylene glycol
Entrya
Cu (0.1 equiv.)
NaOH (equiv.)
Diethylene glycol (equiv.)
Time (h)
Yield (%)b
E.e. (%)c
1
2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
0.20
0.20
0.20
0
0.20
0.20
9.4
5.2
5.2
5.2
5.2
5.2
18
192
72
96
96
31
62
70
69
72
60
38
34
34
29
33
33
d
3
d
4
e
5
Cu(OTf)2
6
Cu(CH CN) BF
72
3
4
4
a
Conditions: L-proline (0.45 equiv.), PhCO H (1.5 equiv.), room temperature.
2
b
Isolated yield calculated on the amount of PhCO t-Bu introduced.
3
c
d
e
S-Configuration of the main enantiomer.
The system was activated with 0.1 equiv. of tert-butyl peroxybenzoate at 80°C for 30 min.
The system was activated with 0.05 equiv. of tert-butyl peroxybenzoate at 80°C for 30 min.
Scheme 2.
The system was activated by heating at 80°C for 30 min
a solution of the catalyst in water/diethylene glycol in the
presence of a catalytic amount of tert-butyl peroxyben-
zoate (0.1 equiv. based on the total amount used later),
the turn of the initial blue color to blue–green or yellow
These results suggest that under these conditions, the real
active species is a copper(I); this is generally admitted for
Kharasch–Sosnovsky reactions and proposed mecha-
1
4
nisms start most of the time with a copper(I) (Scheme 2).
Under the present aqueous conditions, diethylene glycol
plays an important role, since in its absence no reaction
I
indicated that Cu species were probably produced. After
1
5
cooling to room temperature, cyclohexene and tert-butyl
peroxybenzoate were added. Following the reaction by
TLC has shown no induction period and the allylic ester
was then produced over 3 days at room temperature in
occurs at room temperature. In a previous work, we
have suggested that supramolecular associations of active
species could lead to a stable and non-catalytic active
copper complex, thisphenomenon being suppressed in the
presence of an excess of substrate. In the reactions
summarized in Tables 1 and 2, substrate and catalyst are
in two separate phases and auto-association of copper
species in the water phase could be promoted. However,
this phenomenon seems to be prevented in elevating the
temperature (Table 1) or in adding diethylene glycol
(Table 2). Hence, we suspect that diethylene glycol either
coordinates to copper species or forms a third catalyst-
7
0% yield and 34% e.e. (entry 3). Slight reduction of
activation and e.e. were obtained in the absence of base
entry 4). When 0.05 equiv. of oxidizing agent was used
for initiation, reaction time was increased to 4 days (entry
), while no activation was observed in the absence of
(
5
I
tert-butyl peroxybenzoate. The use of Cu was also
examined in the presence of diethylene glycol, no induc-
tionperiodwasobservedandsimilarresultswereobtained
1
6
(
entry 6).
rich phase as observed with PEG (Scheme 2).

1
914
J. Le Bras, J. Muzart / Tetrahedron: Asymmetry 14 (2003) 1911–1915
Surprisingly, concentration of ligand had an effect on
5,11
the enantioselectivity (Fig. 1). In organic medium
and in the experiments performed in water in the
absence of diethylene glycol, best ratio for -proline/
L
copper was around 2. In the presence of diethylene
glycol, increasing the amount of the chiral ligand from
2
5 to 100 mol%, improved the e.e. from 20 to 42% but
yield dropped from 80% to 51% (Fig. 1).
Figure 2. Recycling experiments under argon.
3
. Conclusion
In conclusion, enantioselective allylic oxidation could
be performed in a biphasic-water-organic medium with
amino acids and copper complexes. The presence of
diethylene glycol is required to obtain active catalysis at
room temperature, which led to the formation of the
allylic ester with fair yield and enantiomeric excesses up
to 42%. Performing the reaction under an argon atmo-
sphere allows the recycling of catalyst; yields were not
affected while enantioselectivity decreased slightly from
run to run. Work is in progress to improve the e.e. with
other water-soluble molecules and to understand the
influence of the different parameters on the activity and
enantioselectivity.
Figure 1. Evolution of e.e. and yield with L-proline/Cu ratio.
Non-cyclic amino acids were tested and afforded lower
enantioselectivities as already observed in organic
4
media. For example, when
L-proline was substituted
by
L-valine and oxidation performed as reported in
entry 3, Table 2, the reaction was completed in 24 h
yielding 62% allylic ester with 13% e.e.
4
. Experimental
Two other substrates were subjected to the procedure.
The oxidation of cyclopentene under conditions of
entry 3, Table 2, was faster than with cyclohexene (44 h
instead of 72 h) affording (S)-cyclopentenyl benzoate in
4
.1. General
All reactions were monitored by TLC (TLC plates, 60
F254, Merck); detection was effected by UV light and
phosphomolybdic acid indicator. Water and diethylene
glycol were freeze–pump–thaw degassed three times.
Petroleum ether was either used as received or was
freeze–pump–thaw degassed three times when used in
recycling experiments. Cyclohexene was distillated over
CaH2 under Ar; all other commercially available
reagents were used as received. Cu(MeCN) BF was
7
0% yield and 32% e.e. The oxidation of cyclooctadiene
was not efficient and no traces of expected compound
were observed even after 9 days.
Recycling experiments were performed under atmo-
sphere of argon. First cycle was run according to entry
3
, Table 2, the aqueous phase recovered from work-up
remains blue colored and was catalytically active as
shown in Figure 2. More interestingly, activity of cata-
lyst was improved after the first cycle and reaction time
passed from 72 to 48 h. Yields of allylic ester were quite
constant but e.e.’s decreased regularly from 34 to 22%.
When experiments were performed in air, activity of the
catalyst decreased and reaction required 4 days for the
second cycle and 8 days for run 3 with, respectively, 70,
4
4
prepared from Cu O, MeCN and HBF just prior to
2
4
17
use. Cu(L-prolinato) ·2H O was prepared by treat-
2 2
ment of copper(II) carbonate with a hot water solution
of (S)-proline followed by filtration and slow evapora-
18
tion of the resulting solution. Enantiomeric excesses
were determined by chiral HPLC: Chiracel OD, hex-
ane-isopropanol (99.9:0.1), flow rate 1 mL/min, UV
detection at 220 nm, cyclohexenyl benzoate: t=12 (S),
t=13 (R) min, cyclopentenyl benzoate: t=14 (S), t=18
(R). Perkin–Elmer-241 automatic polarimeter has been
used for optical rotation measurements. Absolute
configuration was determined by comparison of specific
7
1, 57% yield and 33, 30, 26% e.e.
The very low coloration of organic phases and their low
catalytic activities when they were used for recycling
experiments have shown that most of the copper com-
plex remains thus in the aqueous phase.
5
rotations with literature data.

J. Le Bras, J. Muzart / Tetrahedron: Asymmetry 14 (2003) 1911–1915
1915
II
I
4
.2. Catalysis with Cu and Cu at high temperature
Community (COST D12/0028/99).
A Schlenk flask was charged with L-proline (103 mg, 0.9
mmol), NaOH (16 mg, 0.4 mmol) and water (1 mL). After
stirring for 5 min, catalyst (0.2 mmol) was added against
References
an Ar counterflow. Five minutes later, PhCO H (366 mg,
1. For reviews, see: (a) Brunel, J.-M.; Legrand, O.;
Buono, G. C.R. Acad. Sci. Paris, S e´ rie II C 1999, 2,
19–23; (b) Eames, J.; Watkinson, M. Angew. Chem.,
Int. Ed. 2001, 40, 3567–3571; (c) Andrus, M. B.; Lash-
ley, J. C. Tetrahedron 2002, 58, 845–866.
2. For recent references, see: (a) Andrus, M. B.; Zhou, Z.
J. Am. Chem. Soc. 2002, 124, 8806–8807; (b) Fache, F.;
Piva, O. Tetrahedron: Asymmetry 2003, 14, 139–143.
3. Denney, D. B.; Napier, R.; Cammarata, A. J. Org.
Chem. 1965, 30, 3151–3153.
2
3
.0 mmol), cyclohexene (1.0 mL, 10.0 mmol) and t-
BuOOC(O)Ph (0.38 mL, 2.0 mmol) were successively
added. The mixture was heated (oil bath 80°C) until the
disappearance of the perbenzoate (TLC analysis). After
cooling to rt, the mixture was extracted with petroleum
ether (3×20 mL), dried over MgSO , filtered. Partial
4
evaporation of the organic phase afforded a white solid
(
PhCO H). After filtration, the mother liquor was concen-
2
trated and subjected to flash-chromatography (SiO2,
petroleum ether/ethyl acetate 95/5, R =0.55).
4. (a) Araki, M.; Nagase, T. Ger. Offen. 26 25 030, 1976;
Chem. Abstr. 1977, 86, 120886r; (b) Muzart, J. J. Mol.
Catal. 1991, 64, 381–384.
F
II
4
.3. Catalysis with Cu , diethylene glycol and activa-
tion process
5. Levina, A.; Muzart, J. Tetrahedron: Asymmetry 1995, 6,
147–156.
A Schlenk flask was charged with
mmol), NaOH (16 mg, 0.4 mmol) and water (1 mL). After
L
-proline (103 mg, 0.9
6. S o¨ dergren, M. J.; Andersson, P. G. Tetrahedron Lett.
1996, 37, 7577–7580.
stirring for 5 min, Cu(OTf) (72 mg, 0.2 mmol) was added
against an Ar counterflow. Five minutes later, diethylene
7. (a) Li, C.-J.; Chan, T.-H. Organic Reactions in Aqueous
Media; John Wiley & Sons: New York, 1997; (b)
Organic Synthesis in Water; Grieco, P. A., Ed.; Blacky
Academic and Professional: London, UK, 1998.
8. For recent reviews on asymmetric catalysis in aqueous
media, see: (a) Sinou, D. Adv. Synth. Catal. 2002, 344,
221–237; (b) Lindstr o¨ m, U. M. Chem. Rev. 2002, 102,
2751–2772; (c) Manabe, K.; Kobayashi, S. Chem. Eur.
J. 2002, 8, 4095–4101.
2
glycol (1 mL), PhCO H (366 mg, 3.0 mmol) and t-
2
BuOOC(O)Ph (0.04 mL, 0.2 mmol) were successively
added. The mixture was placed in an oil bath at 80°C for
30 min. After cooling to rt, cyclohexene (1.0 mL, 10.0
mmol) and t-BuOOC(O)Ph (0.38 mL, 2.0 mmol) were
added and the mixture was stirred until the disappearance
of the perbenzoate. Work-up was performed as described
above.
´
9. Jo o´ , F.; Kath o´ , A. J. Mol. Catal. 1997, 116, 3–26.
0. Le Bras, J.; Muzart, J. Tetrahedron Lett. 2002, 43, 431–
1
I
4
.4. Catalysis with Cu and diethylene glycol
433.
11. Levina, A.; Muzart, J. Synth. Commun. 1995, 25, 1789–
A Schlenk flask was charged with
L
-proline (103 mg, 0.9
1794.
mmol), NaOH (16 mg, 0.4 mmol) and water (1 mL). After
stirring for 5 min, the solution was transferred via cannula
to a Schlenk containing Cu(MeCN) BF (62.9 mg, 0.2
12. Zondervan, C.; Feringa, B. L. Tetrahedron: Asymmetry
1996, 7, 1895–1898.
13. Cornils, B.; Herrmann, W. A. Aqueous-Phase
Organometallic Catalysis-Concepts and Applications;
Wiley-VCH: Weinheim, Germany, 1998; pp. 181–192.
14. Beckwith, A. M. J.; Zavitsas, A. A. J. Am. Chem. Soc.
4
4
mmol). Five minutes later, diethylene glycol (1 mL),
PhCO H (366 mg, 3.0 mmol), cyclohexene (1.0 mL, 10.0
2
mmol) and t-BuOOC(O)Ph (0.38 mL, 2.0 mmol) were
added successively and the mixture was stirred at room
temperature until the disappearance of the perbenzoate.
Work-up was performed as described above.
II
1986, 108, 8230–8234. In contrast, Cu –tris(oxazolines)
complexes have been reported to be more effective than
I
Cu –tris(oxazolines) complex: Kawasaki, K.; Tsumura,
S.; Katsuki, T. Synlett 1995, 1245–1246. Studies of sta-
II
4.5. Recycling experiments
bility of Cu –prolinato complex by UV–vis spec-
troscopy have indicated that the complex is rather
stable throughout the oxidation process: Levina, A.;
H e´ nin, F.; Muzart, J. J. Organomet. Chem. 1995, 494,
165–168.
The reaction medium is extracted with degassed
petroleum ether (3×20 mL). Work up of the organic phase
was performed as described above. PhCO H (366 mg, 3.0
2
mmol), cyclohexene (1.0 mL, 10.0 mmol) and t-
BuOOC(O)Ph (0.38 mL, 2.0 mmol) were successively
added to the aqueous phase under an Ar flow and the
mixture was stirred until the disappearance of the
perbenzoate.
15. Le Bras, J.; Muzart, J. J. Mol. Catal. A: Chem. 2002,
185, 113–117.
16. Starks, C. M.; Liotta, C. L.; Halpern, M. Phase-Trans-
fer Catalysis, Fundamentals Applications and Industrial
Perspectives; Chapman & Hall: New York; London,
1994; pp. 123–206.
17. Kubas, G. J. Inorg. Synth. 1979, 19, 90–92.
Acknowledgements
18. Sartoris, R. P.; Ortigoza, L.; Casado, N. M. C.; Calvo,
R.; Castellano, E. E.; Piro, O. E. Inorg. Chem. 1999,
38, 3598–3604.
This work was supported by CNRS and the European