Efficient addition of acid enediolates to epoxides

Article abstract of DOI:10.1002/ejoc.200300795

We report new conditions to facilitate the addition of dianions of carboxylic acids to epoxides as an alternative method to the use of aluminum enolates. These conditions require the use of a sub-stoichiometric (10%) amount of amine for dianion generation and the previous activation of the epoxide with LiCl. Other Lewis acids have been shown to be less effective. Yields are good but only low diastereoselectivity is attained, which has not been controlled despite attempts at optimization. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejoc.200300795

FULL PAPER
Efficient Addition of Acid Enediolates to Epoxides
[a]
[a]
[a]
[a]
´
Salvador Gil, Mercedes Torres, Natalia Ortuzar, Richard Wincewicz, and
Margarita Parra*[a]
Keywords: Lactones / Lithium chloride / Nucleophilic addition / Regioselectivity / Diastereoselectivity
We report new conditions to facilitate the addition of dianions
of carboxylic acids to epoxides as an alternative method to
the use of aluminum enolates. These conditions require the
use of a sub-stoichiometric (10%) amount of amine for dian-
ion generation and the previous activation of the epoxide
with LiCl. Other Lewis acids have been shown to be less ef-
fective. Yields are good but only low diastereoselectivity is
attained, which has not been controlled despite attempts at
optimization.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2004)
Introduction
instance, addition of phenylacetic acid 1 to two equivalents
of LDE (lithium diethyl amide) affords the corresponding
lithium enediolate (as an equilibrium system LDE/car-
boxylate enediolate/amine), which on treatment (Scheme 1):
The opening of epoxide rings by nucleophiles is a fre-
quently required transformation for the synthesis of organic
compounds,[1] usually in the formation of γ-lactones.[2]
However, typical lithium enolates from esters do not satis-
factorily react with epoxides [3] and high yields are generally
only achieved by using the corresponding aluminum ester
enolates.[1,4] It is likely that the presence of a Lewis acid
such as the aluminum cation leads to an electrophilic as-
sisted opening of the epoxide.[5]
Addition of the enediolates of carboxylic acids to epox-
ides was reported a few years ago with irregular results.[6]
The lack of recently published results leads us to think that
unsatisfactory results from other enolates and epoxides
have discouraged a systematic study of the selectivity of
these reactions.
Lithium dialkylamides are the bases usually used to gen-
erate lithium enediolates[7] owing to their strength as bases
and their low nucleophilicity, especially when they are de-
rived from sterically hindered amines, and to their solubility
in non-polar solvents.[8,9] It is well known that, in these sol-
vents, lithium enolates exist as complex ion-pair aggregate
structures. The metal centre may be coordinated to solvent
molecules or other chelating ligands, such as the amines
Scheme 1
a) with a highly reactive epoxide, such as styrene oxide,
resulting from the deprotonation of the acid by the lithium yields the corresponding amino alcohol 2 in 77% yield, even
amide. The available data confirms the complexity present though the amide is the only effective nucleophile present
in these aggregated reactive species. Many different factors in the reaction medium.
can affect those aggregates and consequently their reactiv-
ity.
b) with a less reactive epoxide, such as 1,2-epoxidecane,
yields a mixture of amino alcohol 3 and γ-lactone from ad-
These highly basic conditions are critical to the results of dition of the enediolate.
the addition of carboxylic acid dianions to epoxides. For
c) with a secondary epoxide, such as cyclododecene ox-
ide, gives no addition products until it is heated for one
hour under reflux. This treatment led to the isolation of
allylic alcohol 4, which results from the amide acting as a
base, as has been already described.[10]
[a]
Department of Organic Chemistry, University of Valencia,
Dr. Moliner, 50, 46100 Burjassot Valencia, Spain
E-Mail: salvador.gil@uv.es
2160
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200300795
Eur. J. Org. Chem. 2004, 2160Ϫ2165

Efficient Addition of Acid Enediolates to Epoxides
FULL PAPER
Previous studies by our group on reactions of enediolates used are given in Table 1. In every case only one re-
with several electrophiles[11] led us to develop new con- gioisomer, formed from addition to the most accessible
ditions for the generation of dianions of carboxylic acids, epoxide site, was observed. Workup allows isolation of the
which, in some cases, improved the yield and selectivity of γ-lactones 5 or of the corresponding hydroxy acid 6, de-
the reaction.[12] We have optimized a complete generation pending on the intermediate. In each case the hydroxy acids
of dianions of carboxylic acids by using an equimolecular lead to the corresponding γ-lactones quantitatively by re-
amount of nBuLi combined with a less than stoichiometric fluxing in toluene for 2 h.
amount of amine. As can be seen in Scheme 2, a catalytic
cycle is possible because the carboxylate and the dianion
can be held together without self-condensation. This is an
advantage of diendiolate chemistry over the corresponding
enolates. The amount and nature of the amine can be
changed, but 0.6 equivalents of amine has evolved as a
standard in our laboratory because it has proved to be an
efficient method of generating dianions while avoiding ad-
dition of nBuLi to the acid or the appearance of any self-
condensation products.
Following the trend in our group, LDE was used as a
base to generate the dianion except for reaction with epox-
ide d (Entry 10). LDA is slightly less basic than LDE but
its greater bulk makes it more effective (see Entries 9 and
10) in the reaction of the dianion with epoxide d. Despite
this, the corresponding amino alcohol is produced even
when a sub-stoichiometric amount of LDE is used to gener-
ate the dianion. Using LDA may lower the amount of the
unwanted side reaction, such as the attack of the amide on
the epoxide to give the amino alcohol.
In the reaction of phenylacetic acid with styrene oxide (c)
30% of an additional product, 2,4-diphenyl-3-butenoic acid,
was observed in the acidic fraction along with the expected
γ-lactones 5c. This product is formed on dehydration of
hydroxy acid 6c. Attempts to purify products 5c and 6c by
column chromatography led to their total decomposition.
As shown in Table 1, results were good for the addition of
phenylacetic dianion to primary epoxides, with a noticeable
increase in yield compared to those described previously.[1,2]
Creger et al.[2] indicated that a stoichiometric amount of
LDA was successfully used in some cases (steroidal trans-
formations). More rigorous conditions (i.e. reflux for 18 h)
were needed in order to achieve a 75% conversion, but this
is not of general use.
Scheme 2
The relative configuration for products 5a, 5b, 5d and 6g
were determined by NOE NMR experiments. The hydroxy
Results and Discussion
We report here the results obtained when a sub-stoichio- acid 6g was converted into the corresponding γ-lactone by
metric amount of amine is tested with the addition of the refluxing it in toluene for 2 h in order to confirm its con-
phenylacetic acid (1) dianion to several epoxides (aϪg). figuration. The diastereomeric ratio is very poor under
(Scheme 3). Factors that might have affected the yield were these conditions (see Table 1). Taking into account that al-
optimized, namely: the amount and nature of the lithium koxyamines can lead to an increase of the (R*,S*)-dia-
[9c]
amide used as a base to generate the dianion, the tempera- stereoisomer in alkylations of dianions
ture and the reaction time.
we tested the ef-
fect on the stereoselectivity of their addition to epoxides. In
The optimized standard conditions were found to be 3 h order to get good induction, larger amounts of amine are
at room temperature using 10 mol % of the corresponding required but, on optimization, we reached the conclusion
amine. The corresponding yields along with any additives that not more than a 30% amount of amine may be used.
Scheme 3
Eur. J. Org. Chem. 2004, 2160Ϫ2165
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2161

´
S. Gil, M. Torres, N. Ortuzar, R. Wincewicz, M. Parra
FULL PAPER
Table 1. Addition of phenylacetic dianion to primary and secondary epoxides (Scheme 3)
Entry
Epoxide
Product
Yield
RR:RS
Amine (equiv.)
Observations
1
2
3
4
5
6
7
8
9
10
11
12
13
14
a
a
a
a
a
b
b
c
d
d
d
e
f
5a
5a
5a
5a
5a
5b
5b
5c
5d
5d
5d
6e
Ϫ
76
74
0
55:45
40:60
Ϫ
51:49
57:43
55:45
LDE (0.2)
LDE (0.2)
LDE (0.2)
LDE (0.2)
NBHPA (0.5)
LDE (0.2)
NBHPA (0.5)
LDE (0.2)
LDE (0.2)
LDA (0.2)
NBHPA (0.5)
LDE (0.2)
LDE (0.2)
LDE (0.2)
DMI (2 equiv.)
DMI (10 equiv.)
DDOMG
61
74
76
63
77
71
73
73
58
Ϫ
51:49
[a]
30% 2,4-diphenyl-3-butenoic acid
20% 2d
Ϫ
48:52
45:55
74% starting material
Reflux for 3 h
f
4
60
[a]
Undetermined decomposes on purification.
Accordingly, a sub-stoichiometric amount of lithium N- din-2-one (DMI) and 1,4,3,6-tetrahydro-di-O-methyl--glu-
benzyl-2-hydroxypropanamide (NBHPA) was used as a citol (DDOMG) but they did not affect this reaction (En-
base (see Entries 5, 7 and 11).[11] Similar yields to those tries 2, 3, 4). The amount of DMI used allows it to act as
obtained with LDE were observed, but with no increase a lithium chelating agent (Entry 2) but when it was used as
in diastereoselectivity.
a co-solvent (Entry 3) no product was isolated.
On the other hand, α/γ-regioselectivity in the alkylation
The less reactive secondary epoxides do not react under
of dienediolates, from unsaturated carboxylic acids, may be these conditions and the elimination process described
controlled by the addition of specific lithium chelating com- above prevents the use of higher temperatures.[13] In order
pounds. This probably occurs by modification of the aggre- to circumvent this problem we turned to the well-known
gation states.[12b] Here we have used 1,3-dimethylimidazoli- use of Lewis acids for activating carbonϪcarbon bond for-
Table 2. Lewis acid-catalysed addition of phenylacetic dianion to cyclohexene oxide (e)
Entry
Lewis acid
Lewis acid equivalents
Crude yield
Proportion
6e
starting material
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Ϫ
1 equiv.
1 equiv.
1 equiv.
1 equiv.
1 equiv.
1 equiv.
1 equiv.
1 equiv.
1 equiv.
2 equiv.
1 equiv.
2 equiv.
1 equiv.
2 equiv.
1 equiv.[a]
1 equiv.[a]
2 equiv.[a]
2 equiv.
58
75
82
61
39
34
25
37
53
38
61
48
76
39
95
64
32
51
26
24
35
8
2
0
74
76
65
92
98
100
100
Ϫ
60
93
99
99
47
17
13
49
23
31
CeCl3·THF
CeCl3 (powder)
CeCl3 (granules)
GaCl3
TiCl3·THF
AlCl3·THF
BF3·OEt2
MgCl2
MgCl2
LiBr
LiBr
LiCl
LiCl
LiCl
LiCl
LiClO4
LiClO4
0
Ϫ
40
7
1
1
53
83
87
51
77
69
[a]
Inverse addition
2162
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 2160Ϫ2165

Efficient Addition of Acid Enediolates to Epoxides
FULL PAPER
mation through the ring opening of epoxides.[14] We then (see Entries 13 to 16) and the higher yield obtained when
tested the compatibility of this method of epoxide acti- the dianion is added to a mixture of epoxide and LiCl sug-
vation with dienediolate chemistry.
gests that activation of the epoxide by the LiCl prior to the
In order to study this, secondary epoxides were reacted addition of the dianion is crucial.
with dianions in the presence of a variety of Lewis acids.
At this point, we want to show the results of the addition
Only those Lewis acids compatible with the basic con- of the dienediolate of phenylacetic acid 1 to several epox-
ditions of the reaction were used which has not been pre- ides under the combined optimized conditions: sub-stoi-
viously reported.
chiometric amount of amine for base generation and pre-
As a model reaction, we present here the result of the vious addition of LiCl to the epoxide. The results shown
addition of the enediolate of phenylacetic acid to cyclohex- in Table 3 indicate that this combination may represent an
ene oxide (e) in the presence of several Lewis acids (Table 2) efficient procedure for the direct addition of dianions of
under the optimized conditions described above.
carboxylic acids to epoxides, in a way not described before.
Very irregular results were obtained. As expected, some
Under these conditions, the use of lithium N-benzyl-2-
of the Lewis acids produced a complete reprotonation of hydroxypropanamide had no effect on the diastereoselec-
the dianion and only starting material was recovered. tivity of the addition, except for the addition to 1,2-epoxy-
Others, usually lithium salts, led to an increase in the yield 2-methyl-3-butene (g), the most hindered epoxide used in
of the addition product. The best results were obtained with this study.
LiCl under inverse addition. Inverse addition involved the
addition of the dianion solution to a mixture of the epoxide
with LiCl in THF (Entry 15, Table 2). Surprisingly, LiBr
had no effect under similar conditions (Entries 11 and 12).
Conclusion
It is well known that LiCl is a disaggregating agent of
enolates[15] and that lithium amides form mixed dimers with
LiCl, this being the species responsible for the highly en-
antioselective deprotonation of ketones by chiral lithium
amides.[10a] Similar behaviour is observed with LiBr. Thus
in the intramolecular alkylation of cyclohexanones the use
of LiBr as the additive in place of LiCl leads to slightly
better results.[16] Taking this into account, the big difference
that LiCl and LiBr exhibit in the activation of the addition
of dienolates to epoxides may be related to the LiCl playing
a double role. It could be acting firstly as a disaggregating
agent of the enolate and secondly as an activating agent in
the opening of the epoxide through its coordination to the
oxygen. Thus, the different behaviours of the two salts
could be related to the second role, but we have no expla-
nation for such a large difference in reactivity with such a
small change in the nature of the Lewis acid.
We present new conditions for the addition of dianions of
carboxylic acid to epoxides as an alternative to aluminum
enolate chemistry. The changes involved are the use of a
sub-stoichiometric (10%) amount of amine for dianion gen-
eration and activation of the epoxide with LiCl. These
modifications show that this reaction has a much wider
range than was previously reported.[1]
Experimental Section
General: Melting points were determined with a Cambridge Instru-
ments Hot Plate Microscope and are uncorrected. IR spectroscopic
data were obtained for liquid film or KBr discs; the measurements
were carried out by the SCSIE (Servei Central de Suport a la In-
´
vestigacio Experimental de la Universitat de Valencia) with a Mat-
In addition, the dependence of the yield on the relative
concentrations of the enediolate, the epoxide and the salt
teson Satellite FTIR 3000 model spectrophotometer. NMR spectra
were recorded for CDCl3 solutions, with Varian Unity 300 or
Bruker Unity AC-300, AC-400 or AC-500 spectrometers. High-res-
olution mass spectra were determined with a Fison VG Autospec
spectrometer. Flash Column Silica Gel of 230Ϫ400 mesh (manufac-
turer: Scharlau) was used for flash column chromatography, with
hexane/ethyl acetate mixtures for elution.
Table 3. LiCl catalysed addition of phenylacetic dianion to epox-
ides under inverse addition conditions
All reactions were carried out under argon, using standard con-
ditions for exclusion of moisture, in oven-dried glassware, in THF
freshly distilled from blue benzophenone ketyl and with diethyl-
amine and diisopropylamine distilled from CaH2. 1-Benzylamino-
2-propanol[9c] was placed under vacuum and left for 24 h before
use. The DMI was distilled (106 °C at 17 Torr) and collected over
Epoxide
Product
Yield
R*R*:R*S*
Amine
˚
a
a
d
d
g
g
e
5a
5a
5d
5d
6g
6g
6e
6e
71
71
82
65
63
78
71
52
47:53
57:43
38:62
38:62
57:43
28:72
Et2NH
NBHPA
Et2NH
NBHPA
Et2NH
NBHPA
Et2NH
NBHPA
3/4-A molecular sieves. The distilled DMI was stored over molecu-
lar sieves and kept under Ar.
The BuLi used was 1.6  in hexane. Exact determination of the
solution’s concentration was periodically checked before use. Usu-
ally, the molarity quoted is not the true concentration of the solu-
tion (e.g. 1.18  for fresh new Aldrich bottles and 1.19  for
Merck ones).
e
Eur. J. Org. Chem. 2004, 2160Ϫ2165
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2163

´
S. Gil, M. Torres, N. Ortuzar, R. Wincewicz, M. Parra
FULL PAPER
The reaction temperature (Ϫ78 °C) was achieved by cooling with
a CO2/acetone bath and 0 °C with an ice/water bath. Organic ex-
tracts were dried with anhydrous MgSO4, and solutions were eva-
porated under reduced pressure with a rotary evaporator and a
bath at 40 °C.
CH2CH3), 0.94 (t, J ϭ 6.6 Hz, 3 H, 4Ј-CH3) ppm. 13C NMR
(100 MHz, CDCl3): (R*S*): δ ϭ 177.1 (CϭO), 136.9 (CAr), 129.1
(2 CHAr) 128.3 (2 CHAr), 127.8 (CHAr), 79.2 (CHϪO), 45.9
(CHCϭO), 36.7 (CH2CHCϭO), 35.4 (CHCH2CH2), 27.7
(CHCH2CH2), 22.7 (CH2CH3) 14.2 (CH3) ppm; (R*R*): δ ϭ 177.5
(CϭO), 136.9 (CAr), 129.1 (2 ϫ CHAr) 128.3 (2 ϫ CHAr), 127.8
(CHAr), 78.9 (CHϪO), 47.5 (CHCϭO), 38.4 (CH2CHCϭO), 35.3
(CHCH2CH2), 27.6 (CHCH2CH2), 22.7 (CH2CH3), 14.2 (CH3)
ppm. MS (EI): m/z (%) ϭ 218 (1.4) [Mϩ], 174 (28), 104 (100)
[PhCHCH2ϩ], 77 (12) [C6H5ϩ]. HRMS: m/z calcd. for C14H18O2 ϭ
218.1307, found Mϩ ϭ 218.1306.
General Procedure for Addition Reactions: Carboxylic acid
(2.25 mmol) in THF (2 mL) was slowly added to stirred lithium
amide (4.8 mmol for stoichiometric amount, 0.5 mmol for sub-stoi-
chiometric LDE or LDA or 1.3 mmol for sub-stoichiometric 1-
benzylamino-2-propanol) in THF (2 mL) at Ϫ78 °C, according to
the method already described.[9c][12a] The solution was stirred for
30 min at 0 °C and cooled again to Ϫ78 °C. Epoxide (2.25 mmol)
in THF (2 mL) was added dropwise (5 min), and the solution
stirred for 1 h at room temperature. The reaction was quenched
with water (20 mL) and the mixture extracted with diethyl ether (3
ϫ 15 mL). The aqueous layer was acidified under ice-bath cooling
by careful addition of conc. hydrochloric acid, and then extracted
with ethyl acetate (3 ϫ 15 mL). The organic layer was washed with
brine and dried (MgSO4). Evaporation of solvent gave the crude
acid reaction mixture. For analytical purposes the products were
isolated by column chromatography.
5-Octyl-3-phenyltetrahydrofuran-2-one (5b): M.p. 44Ϫ45 °C. IR:
ν˜max. ϭ 3075 (ArϪH), 2922 (CϪH), 1760 (CϭO), 1610, 1522, 1450
(ArϪH), 1388, 1185, 1012, 752, 697 cmϪ1 1H NMR (400 MHz,
.
CDCl3): (R*S*): δ ϭ 7.20 (m, 5 H, Ar-H), 4.63 (m, 1 H, CHϪO),
3.89 (m, 1 H, CHCϭO), 2.49 (ddd, J ϭ 6.0, 9.6, 12.8 Hz, 1 H,
CH2CHCϭO), 2.39 (ddd, J ϭ 6.4, 6.8, 13.2 Hz, 1 H, CH2CHCϭ
O), 1.79 (m, 1 H, CHCH2CH2), 1.64 (m, 1 H, CHCH2CH2),
2.85Ϫ1.15 (m, 12 H, 6CH2), 0.89 (t, J ϭ 6.8 Hz, 3 H, CH3) ppm;
(R*R*): δ ϭ 7.20 (m, 5 H, Ar-H), 4.48 (m, 1 H, CHϪO), 3.88 (m,
1 H, CHCϭO), 2.77 (ddd, J ϭ 5.2, 8.8, 12.4 Hz, 1 H, CH2CHCϭ
O), 2.01 (ddd, 1 H, J ϭ 10.4, 12.0, 12.4 Hz, CH2CHCϭO), 1.82
(m, 1 H, CHCH2CH2), 1.68 (m, 1 H, CHCH2CH2), 2.85Ϫ1.15 (m,
12 H, 6CH2), 0.89 (t, J ϭ 6.8 Hz, 3 H, CH3) ppm. 13C NMR
(100 MHz, CDCl3): (R*S*): δ ϭ 177.2 (CϭO), 137.5 (CAr), 129.2
(2 ϫ CHAr), 128.3 (2 ϫ CHAr), 127.8 (CHAr), 79.3 (CHϪO), 45.9
(CHCϭO), 36.6 (CH2CHCϭO), 35.6 (CHCH2CH2), 32.1 (CH2),
29.6 (CH2), 29.4 (CH2), 25.6 (CH2), 25.5 (CH2), 22.9 (CH2CH3),
14.3 (CH3)ppm; (R*R*): δ ϭ 177.5 (CϭO), 136.9 (CAr), 129.0 (2
ϫ CHAr), 127.8 (2 ϫ CHAr), 127.7 (CHAr), 78.9 (CH-O), 45.5
(CHCϭO), 38.4(CH2CHCϭO), 35.7 (CHCH2CH2), 32.1 (CH2),
29.6 (CH2), 29.4 (CH2), 25.6 (CH2), 25.5 (CH2), 22.9 (CH2CH3),
14.3 (CH3) ppm. MS (EI): m/z (%) ϭ 274 (1.1) [Mϩ], 230 (27), 117
(45), 104 (100) [PhCHCH2ϩ], 91 (16) [C7H7ϩ]. HRMS: m/z calcd.
for C18H26O2 ϭ 274.1933, found Mϩ ϭ 274.1942.
Standard Addition Procedure Modifications
Using the Lewis Acid BF3·Et2O: BF3·Et2O (0.3 mL, 2.25 mmol) was
added to the solution of epoxide (2.25 mmol) in THF (1 mL) af-
fording a blood-red complex. This solution was then introduced
dropwise into the reaction vessel at Ϫ78 °C and the general pro-
cedure continued as described.
Using a Solid Lewis Acid: The solution of epoxide (2.25 mmol) in
THF (1 mL) was added dropwise to the dianion solution at Ϫ78
°C. The reaction flask was opened and the solid Lewis acid
(2.25 mmol) was added via a solids funnel (as one sample). The
reaction flask was then resealed and the general procedure con-
tinued as described.
(3RS,5SR)-5-Phenoxymethyl-3-phenyltetrahydrofuran-2-one (5d):
M.p. 119Ϫ120 °C. IR: ν˜max. ϭ 3059 (ArϪH), 2916 (CϪH), 1756
Using a Chelating Agent: Before addition of the epoxide to the
dianion, a solution of the ligand (number of equivalents stated in
Table 1) in THF (2 mL) was added to the dianion solution at 0 °C.
The solution was maintained at 0 °C, stirring for 15 min. The solu-
tion was cooled to Ϫ78 °C and the addition of the epoxide con-
tinued as described.
(CϭO), 1600, 1585, 1497 (ArϪH), 1244, 1061, 939, 752, 698 cmϪ1
.
1H NMR (400 MHz, CDCl3): δ ϭ 7.33 (m, 7 H, Ar-H), 7.00 (t,
J ϭ 7.6 Hz, 1 H, Ar-H), 6.92 (d, J ϭ 8.0 Hz, 2 H, 2 ArϪH), 4.87
(m, 1 H, CHϪO), 4.26 (dd, J ϭ 3.6, 10.4 Hz, 1 H, CH2ϪO), 4.20
(dd, J ϭ 4.4, 10.4 Hz, 1 H, CH2ϪO), 3.97 (dd, J ϭ 9.6, 12.3 Hz, 1
H, CHCϭO), 2.84 (ddd, J ϭ 6.0, 9.2, 12.8 Hz, 1 H, CH2CHCϭ
O), 2.48 (ddd, J ϭ 10, 12.4, 12.4 Hz, 1 H, CH2CHCϭO) ppm. 13C
NMR (100 MHz, CDCl3): δ ϭ 176.5 (CϭO), 158.4 (CArЈ), 136.7
(CAr), 129.8 (2 CHArЈ), 129.2 (2 ϫ CHAr), 128.3 (2 ϫ CHAr),
128.0 (CHAr), 121.8 (CHArЈ), 114.9 (2 ϫ CHArЈ), 76.1 (CHϪO),
68.8 (CH2ϪO), 46.8 (CHCϭO), 33.7 (CH2CHCϭO) ppm. MS
(EI): m/z (%) ϭ 268 (83) [Mϩ], 174 (100), 131 (78), 103 (84), 77
(59) [C6H5ϩ]. HRMS: m/z calcd. for C17H16O3 ϭ 268.1100, found
Mϩ ϭ 268.1091.
Using a Chelating LiCl and Inverse Addition: Instead of adding the
LiCl (94.4 mg, 2.25 mmol) to the reaction flask at the beginning of
the reaction it was added to a clean flask. The dianion mixture was
diluted with THF (4 mL), as the concentrated dianion mixture was
too viscous to pass easily through the needle, and then transferred
at Ϫ78 °C on top of the epoxide mixture under Ar(g). Each reaction
was extracted in the same way after being quenched.
5-Butyl-3-phenyltetrahydrofuran-2-one (5a): M.p. 26Ϫ31 °C. IR:
ν˜max. ϭ 3064 (Ar-H), 2958 (CϪH), 1769 (CϭO), 1603, 1498, 1467
(ArϪH), 1355, 1180, 1002, 935, 754, 698 cmϪ1
.
1H NMR
(3RS,5RS) 5-Phenoxymethyl-3-phenyltetrahydrofuran-2-one (5d):
(400 MHz, CDCl3): (R*S*): δ ϭ 7.31 (m, 5 H, Ar-H), 4.63 (m, 1 M.p. 90Ϫ91 °C. IR: ν˜max. ϭ 3036 (ArϪH), 2923 (CϪH), 1749 (Cϭ
H, CH-O), 3.92 (m, 1 H, CHCϭO), 2.50 (ddd, J ϭ 6, 9.2, 13.2 Hz,
O), 1589, 1547, 1494 (ArϪH), 1235, 1156, 1078, 742, 684 cmϪ1
;
1 H, CH2CHCϭO), 2.38 (ddd, J ϭ 6.8, 7.2, 13.6 Hz, 1 H, δ ϭ 7.33 (m, 7 H, Ar-H), 7.01 (t, J ϭ 7.6 Hz, 1 H, Ar-H), 6.94 (d,
CH2CHCϭO), 1.79 (m, 1 H, CHCH2CH2), 1.64 (m, 1 H, J ϭ 8.0 Hz, 2 H, 2 ArϪH), 4.95 (m, 1 H, CHϪO), 4.25 (dd, J ϭ
CHCH2CH2), 1.50 (m, 2 H, CHCH2CH2), 1.40 (m, 2 H, CH2CH3), 3.3, 10.2 Hz, 1 H, CH2ϪO), 4.17 (m, 1 H, CH2ϪO), 4.16 (m, 1 H,
0.94 (t, J ϭ 6.6 Hz, 3 H, CH3) ppm; (R*R*): δ ϭ 7.31 (m, 5 H, CHCϭO), 2.81 (ddd, J ϭ 3.6, 9.6, 13.2 Hz, 1 H, CH2CHCϭO),
Ar-H), 4.48 (m, 1 H, CHϪO), 3.88 (m, 1 H, CHCϭO), 2.75 (ddd,
2.60 (ddd, J ϭ 8.4, 8.4, 13.2 Hz, 1 H, CH2CHCϭO) ppm. 13C
J ϭ 5.2, 8.8, 13.6 Hz, 1 H, CH2CHCϭO), 2.05 (ddd, J ϭ 6, 10.4, NMR (100 MHz, CDCl3): δ ϭ 177.4 (CϭO), 158.0 (CArЈ), 137.6
13.6 Hz, 1 H, CH2CHCϭO), 1.83 (m, 1 H, CHCH2CH2), 1.68 (m, (CAr), 129.9 (2 ϫ CHArЈ), 129.3 (2 ϫ CHAr), 128.1 (2 ϫ CHAr),
1 H, CHCH2CH2), 1.50 (m, 2 H, CHCH2CH2), 1.40 (m, 2 H,
127.9 (CHAr), 121.9 (CHArЈ), 114.9 (2 CHArЈ), 76.1 (CHϪO),
2164
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 2160Ϫ2165

Efficient Addition of Acid Enediolates to Epoxides
FULL PAPER
69.6 (CH2ϪO), 45.8 (CHCϭO), 33.7 (CH2CHCϭO) ppm. MS
(EI): m/z (%) ϭ 268 (100) [Mϩ], 131 (40), 120 (35), 105 (46), 91
(31) [C7H7ϩ], 77 (29) [C6H5ϩ]. HRMS: m/z calcd. for C17H16O3 ϭ
268.1100, found Mϩ ϭ 268.1092.
[1]
[1a] S. K. Taylor, Tetrahedron 2000, 56, 1149Ϫ1163. [1b] P. Arga,
H. Qin, Tetrahedron 2000, 56, 917Ϫ947.
[2] [2a]
[2b]
P. L. Creger, J. Org. Chem. 1972, 37, 1907Ϫ1918.
S.
Danishefsky, M-Y. Tsai, T. Kitahara, J. Org. Chem. 1977, 42,
(2-Hydroxycyclohexyl)phenylacetic Acid (6e): M.p. 164Ϫ166 °C. IR:
394Ϫ396.
ν˜max. ϭ 3274 (OH), 2921 (CϪH), 2855 (CϪH), 1672 (CϭO) cmϪ1
.
[3]
[4]
[5]
A. G. Myers, L. McKinstry, J. Org. Chem. 1996, 61,
2428Ϫ2440.
T-J. Sturm, A. E. Marolewski, J. Org. Chem. 1989, 54,
2039Ϫ2040.
1H NMR (500 MHz, CD3OD): δ ϭ 7.40Ϫ7.20 (m, 5 H, Ar-H),
4.93 (s, 2 H, OH), 3.85 (d, J ϭ 7.2 Hz, 1 H, CHCϭO), 3.20 (dt,
J ϭ 4.2, 9.9 Hz, 1 H, CH-OH), 2.11 (m, 1 H, CHCHCϭO), 1.92
(m, 1 H, HCHeqCHOH), 1.73 (m, 1 H, HCHeqCHCHOH), 1.64
(m, 2 H, CH2), 1.4 (m, 1 H, HCHaxCHOH), 1.2 (m, 2 H, CH2),
0.7 (m, 1 H, CH2) ppm. 13C NMR (125 MHz, CD3OD): δ ϭ 167.4
(COOH), 129.3 (CAr), 128.8 (2 ϫ CHAr), 127.9 (2 ϫ CHAr), 126.6
(CHAr), 72.5 (CϪOH), 53.4 (Ar-CHϪCO2H), 46.7 (CHCHOH),
35.1, 27.2, 24.8 and 24.3 (4 ϫ CH2) ppm. MS (EI): m/z (%) ϭ 234
(0.2) [Mϩ], 216 (11) [Mϩ Ϫ H2O], 173 (15), 172 (100) [Mϩ Ϫ H2O
Ϫ CO2], 143 (12), 136 (33) [PhCCO2H2]ϩ, 104 (41) [PhCHCH2ϩ],
91 (24) [C7H7ϩ].
S. K. Taylor, J. A. Fried, Y. N. Grassl, A. E. Marolewski, E.
A. Pelton, T-J. Poel, A. S. Rezanka, M. R. Whittaker, J. Org.
Chem. 1993, 58, 7304Ϫ7307.
[6]
[7]
[6a] F. F. Blicke, P. E. Wright, J. Org. Chem. 1960, 25, 693Ϫ698.
[6b]
T. Fujita, S. Watanabe, K. Suga, Australian J. Chem. 1974,
27, 2205Ϫ2218.
C. M. Thomson, Dianion Chemistry in Organic Synthesis, CRC
Press: Boca Raton (Florida), 1994; pp. 88Ϫ129.
[8] [8a]
H. B. Mekelburger, C. S. Wilcos, Formation of enolates; in
Comprehensive Organic Synthesis (Eds.: B. M. Trost, I. Flem-
ing), Pergamon Press: Oxford, 1991, pp. 99Ϫ131. [8b]E. Juaristi,
A. K. Beck, J. Hansen, T. Matt, T. Mukhopedhyay, M. Simon,
D. Seebach, Synthesis 1993, 1271Ϫ1290.
4-Hydroxy-4-methyl-2-phenylhex-5-enoic Acid (6g): Pale yellow oil.
IR: ν˜max. ϭ 3381 (OH), 2946 (CϪH), 2834 (CϪH), 1762 (CϭO)
cmϪ1. 1H NMR (400 MHz, CDCl3): (R*S*): δ ϭ 7.30 (m, 5 H, Ar-
H), 6.04 (dd, J ϭ 11.2, 16.8 Hz, 1 H, CH2ϭCH), 5.35 (d, J ϭ
16.8 Hz, 1 H, CH2ϭCH), 5.19 (d, J ϭ 11.2 Hz, 1 H, CH2ϭCH),
4.04 (dd, J ϭ 9.2, 11.2 Hz, 1 H, PhCH), 2.60 (dd, J ϭ 9.2, 11.2 Hz,
1 H, CH2COH), 2.36 (t, J ϭ 11.2 Hz, 1 H, CH2COH), 1.56 (s, 3
H, CH3) ppm; (R*R*): δ ϭ 7.30 (m, 5 H, Ar-H), 5.93 (dd, J ϭ
10.8, 17.2 Hz, 1 H, CH2ϭCH), 5.38 (d, J ϭ 17.2 Hz, 1 H, CH2ϭ
CH), 5.22 (d, J ϭ 10.8 Hz, 1 H, CH2ϭCH), 3.89 (dd, J ϭ 8.8,
12.8 Hz, 1 H, PhCH), 2.67 (dd, J ϭ 8.8, 12.8 Hz, 1 H, CHCH2),
2.26 (t, J ϭ 12.8 Hz, 1 H, CHCH2), 1.58 (s, 3 H, CH3) ppm. 13C
NMR (100 MHz, CDCl3): (R*S*): δ ϭ 176.8 (CO2H), 141.0 (CHϭ
CH2), 136.6 (CAr), 130.4 (2 CHAr), 126.2 (2 CHAr), 114.4
(CHAr), 112.4 (CHϭCH2), 83.5 (COH), 46.7 (PhCH), 43.1
(CHCH2), 25.5 (CH3) ppm; (R*R*): δ ϭ 177.2 (CO2H), 139.7
(CHϭCH2), 136.6 (CAr), 130.4 (2 CHAr), 126.2 (2 CHAr), 114.4
(CHAr), 114.3 (CHϭCH2), 83.4 (COH), 46.5 (PhCH), 43.3
(CHCH2), 27.3 (CH3) ppm. MS (EI): m/z (%) ϭ 220 (0.5) [Mϩ],
202 (17) [Mϩ Ϫ H2O], 158 (55) [Mϩ Ϫ H2O Ϫ CO2], 157 (29), 143
(20) [Mϩ Ϫ C6H5], 143 (100) [Mϩ Ϫ CO2 Ϫ H2O Ϫ CH3], 129 (38),
128 (38), 118 (20) [PhCH2COϩ], 91 (26) [C7H7ϩ], 77 (19) [C6H5ϩ].
[9] [9a]
A. Streitwieser, Z-R. Weng, J. Am. Chem. Soc. 1999, 121,
[9b]
6213Ϫ6219.
Y. Balamraju, C. D. Sharp, W. Gammill, N.
[9c]
Manuel, L. M. Pratt, Tetrahedron 1998, 54, 7357Ϫ7366.
M. Brun, S. Gil, M. Parra, Tetrahedron: Asymmetry 2001, 12,
E.
915Ϫ921.
[10] [10a]
P. O ЈBrien, J. Chem. Soc., Perkin Trans.
1 1998,
[10b]
1439Ϫ1457.
1625Ϫ1642.
D. M. Hodgson, E. Gros, Synthesis 2002,
[11] [11a]
[11b]
S. Gil, M. Parra, Curr. Org. Chem. 2002, 6, 283Ϫ302.
S. Gil, M. Parra, Recent Res. Devel. Organic Chem. 2002, 6,
449Ϫ481.
[12] [12a]
E. M. Brun, I. Casades, S. Gil, R. Mestres, M. Parra,
[12b]
Tetrahedron Lett. 1998, 39, 5443Ϫ5446.
Gil, M. Parra, Synlett 2001, 156Ϫ159.
E. M. Brun, S.
[13] [13a] P. C. Brookes, D. J. Milne, P. J. Murphy, B. Spolaore, Tetra-
[13b]
hedron 2002, 58, 4675Ϫ4680.
S. K. Bertilsson, P. G. And-
ersson, Tetrahedron 2002, 58, 4665Ϫ4668.
[14]
See for example: [14a] P. R. Likhar, M. P. Kumar, A. K. Bondy-
[14b]
opadhyay, Tetrahedron Lett. 2002, 43, 3333Ϫ3335.
Reddy, M. A. Reddy, N. Bhanumathi, K. R. Rao, Synthesis
G. Prestat, C. Baylon, M-P. Heck, C.
Mioskowski, Tetrahedron Lett. 2000, 41, 3829Ϫ3831.
G. Constantino, V. Lacerda Jr., V. Arepao, Molecules 2001, 6,
L. R.
[14c]
2001, 831Ϫ832.
[14d]
M.
HRMS: m/z calcd. for C13H16O3 ϭ 220.1676, found Mϩ
ϭ
220.1672.
770Ϫ776.
[15] [15a]
M. Goto, K. Akimoto, K. Aoki, M. Shindo, K. Koga,
[15b]
Chem. Pharm. Bull. 2000, 48, 1529Ϫ1531.
M. Parra, E.
Acknowledgments
The present research has been financed by DCICYT
(PPQ2002Ϫ00986).
Sotoca, S. Gil, Eur. J. Org. Chem. 2003, 1386Ϫ1388.
J. E. Kropf, S. M. Weinreb, Chem. Commun. 1998, 2357Ϫ2358.
Received December 19, 2003
[16]
Eur. J. Org. Chem. 2004, 2160Ϫ2165
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2165