New and efficient conditions for the Z-selective synthesis of unsaturated esters by the Horner-Wadsworth-Emmons olefination

Article abstract of DOI:10.1016/j.tetlet.2004.05.026

The Horner-Wadsworth-Emmons reaction of both 1 (Still's reagent) and 2 (Ando's reagent) with aldehydes was studied. Tris[2-(2-methoxyethoxy)ethyl]amine (TDA-1) was shown to be an efficient additive when the reaction was performed in THF at -78°C. Selectivities close to 95% were obtained in the condensation of 1 with aromatic aldehydes. The use of a 5-fold excess of 18C6 in this reaction had already been reported. However, we have defined conditions in which only catalytic amounts are required.

Full text of DOI:10.1016/j.tetlet.2004.05.026

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 5519–5523
New and efficient conditions for the Z-selective synthesis of
unsaturated esters by the Horner–Wadsworth–Emmons olefination
Francois P. Touchard*
ß
Rhodia Recherches, Centre de Recherches de Lyon, 85, rue des frꢀeres Perret, BP62, 69192 Saint Fons cedex, France
Received 16 February 2004; accepted 5 May 2004
Abstract—The Horner–Wadsworth–Emmons reaction of both 1 (Still’s reagent) and 2 (Ando’s reagent) with aldehydes was studied.
Tris[2-(2-methoxyethoxy)ethyl]amine (TDA-1) was shown to be an efficient additive when the reaction was performed in THF at
)78 °C. Selectivities close to 95% were obtained in the condensation of 1 with aromatic aldehydes. The use of a 5-fold excess of 18C6
in this reaction had already been reported. However, we have defined conditions in which only catalytic amounts are required.
Ó 2004 Elsevier Ltd. All rights reserved.
The Horner–Wadsworth–Emmons (HWE) reaction is a
variant of the Wittig reaction that possesses a number of
significant advantages. Not only is the work-up greatly
facilitated since the only by-product is a water soluble
phosphate salt but the higher reactivity of HWE re-
agents allows them to effect transformations that are
either difficult or impossible using the analogous Wittig
ylides. This reaction, which usually proceeds smoothly
and under mild conditions with phosphonates bearing
stabilizing substituents, typically leads to the more sta-
ble E-disubstituted olefins.1 As it is highly desirable to
be able to prepare the E or Z-isomer at will, Still and
Gennari,2 in 1983, and Ando,3;4 in 1995, developed two
Z-selective reagents: methyl bis(trifluoroethyl)phospho-
noacetate 1 and ethyl diphenylphosphonoacetate 2,
respectively (Fig. 1).5 However, the conditions required
to obtain high Z/E-ratios are fairly stringent and usually
imply the use of low temperatures,6 KHMDS as the base
and a large excess of expensive 18C6 (typically up to
5 equiv). With the aim of establishing an efficient and
scalable protocol for the Z-selective olefination of
aldehydes, we decided to reinvestigate and modulate the
Still/Ando procedures. In this article, we wish to report
some of the salient results of our studies.7
At the onset of this work, both phosphonates 1 and 2
were reacted with benzaldehyde, in THF, at )78 °C, in
the presence of different bases and additives. Some
selected results are collected in Table 1.
Initial experiments confirmed the previously reported
observations that preferential formation of the Z-iso-
mer, using KHMDS alone, occurs with selectivities close
to 90% with both phosphonates (entries 1 and 5).
Adding 5 equiv of 18C6, as described by Still and Ando,
improved the selectivities up to 98% (entries 2 and 6).9
Remarkably, using only 1 equiv of TDA-1, a cheap and
readily available Kþ chelating agent,10 also led to 98%
Z-selectivity with Still’s reagent 1 (entry 3). Replacing
KHMDS by tBuOK did not change the Z/E-ratio (entry
4). To our surprise, TDA-1, even up to 5 equiv, did not
lead to any improvement with Ando’s phosphonate; the
same selectivities were obtained with and without TDA-
1 (entries 5 and 7).
O
P
O
O
O
P
F3C
O
OMe
PhO
OEt
O
OPh
1
2
CF3
Figure 1. Phosphonates developed by Still and Ando.
To evaluate the scope of this new procedure, phospho-
nate 1 has been reacted with other aldehydes in the
presence of TDA-1 (Table 2). The reactions without
Keywords: a,b-Unsaturated esters; HWE reaction; Still’s reagent.
* Tel.: +33-4-72-89-67-59; fax: +33-4-72-89-68-94; e-mail: francois.
touchard@eu.rhodia.com
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.05.026

5520
F. P. Touchard / Tetrahedron Letters 45 (2004) 5519–5523
Table 1. HWE reaction of 1 and 2 with benzaldehyde in THFa
Entry
Phosphonate
Base
Additive
Time (h)
Conversion (%)b
Selectivity (%)c
1
2
3
4
1
1
1
1
KHMDS
KHMDS
KHMDS
tBuOK
None
2
1
2
2
99
98
92 (83)
>98 (98)
98 (90)
98
18C6 (5 equiv)
TDA-1 (1.1 equiv)
TDA-1 (1.1 equiv)
95
100
5
6
7
2
2
2
KHMDS
KHMDS
KHMDS
None
18C6 (5 equiv)
TDA-1 (1.1 equiv)
2
1
3
90
100
95
89 (80)
>98 (90)
89 (80)
a 1 or 2 (0.5 mmol–1 equiv), base (0.5 mmol–1 equiv), benzaldehyde (0.5 mmol–1 equiv), THF (10 mL)––see the typical experimental procedures.8
b Benzaldehyde conversion. Area% calculated by GC.
c Selectivity ¼ Z/(Z+E) calculated by GC. The selectivities obtained at 0 °C are given in brackets.
Table 2. HWE reaction of 1 and aromatic aldehydes in the presence of 1 equiv of TDA-1a
O
P
O
O
R
CO2Me
KHMDS, TDA-1
THF, -78˚C
+
F3C
O
OMe
R
H
O
1
O
O
OMe
OMe
CF3
N
TDA-1 :
O
OMe
Selectivity (%)b
Entry
Aldehyde
Yield (%)c
Without additive
92
With TDA-1
CHO
CHO
1
98
90
CF3
2
95
92
91
91
89
87
99
98
99
98
93
93
95
90
91
95
92
95
F3C
CHO
CHO
3
4
F3C
OMe
CHO
5
CHO
6d
MeO
CHO
7d

F. P. Touchard / Tetrahedron Letters 45 (2004) 5519–5523
5521
Table 2 (continued)
Selectivity (%)b
Entry
Aldehyde
Yield (%)c
Without additive
92
With TDA-1
CHO
8
98
92
93
97
N
S
9d
86
CHO
a 1 (5.5 mmol–1.1 equiv), KHMDS (5.25 mmol–1.05 equiv), aldehyde (5 mmol–1 equiv), THF (90 mL)––see the typical experimental procedures.8
b Selectivity ¼ Z/(Z+E) calculated by 1H NMR integration of the vinyl proton signals and in accordance with GC determination.
c Isolated yields of the Z/E-mixtures of the TDA-1 experiments; the structures have been confirmed by 1H and 13C NMR spectroscopies and by
comparison with literature data.
d The reaction being incomplete after 4 h at )78 °C, the mixture was allowed to warm-up overnight before work-up.
additive have also been performed to compare the
selectivities.
the presence of 1 equiv of 18C6, high Z-selectivities were
obtained at 0 °C (Table 3).
As can be seen, excellent selectivities have been obtained
with all the aromatic aldehydes tested. The Z/E-ratio
does not seem to depend too much upon the substituents
present on the aromatic ring (entries 1–5 and 8). Lower
selectivities have however been obtained with the less
reactive aldehydes (entries 6–7 and 9). In all cases, the
addition of TDA-1 was effective in enhancing the Z/E-
ratios.11
Under these conditions, Still’s reagent 1 proved to be
highly reactive and selective leading to the desired
alkene with up to 98% Z-selectivity and completion
within 3 h in all the solvents tested. Comparatively,
Ando’s reagent was less reactive and also less selective:
the best result was indeed obtained in THF, with a
selectivity close to 90%.
In fact, 18C6 is a powerful Kþ chelating agent, which is
generally recognized to enhance the reactivity of the
phosphonate anion by dissociation of the ion pair, thus
favouring kinetic conditions, and to increase the rate of
phosphate elimination thereby helping to maintain high
Z-stereoselection. In comparison to KHMDS and
tBuOK, K2CO3 is a much weaker base that deproto-
nates the phosphonate slowly and is also less soluble in
organic solvents. We therefore reasoned that if we could
find a suitable solvent in which K2CO3 was unable to
generate the phosphonate anion in the absence of 18C6,
we might be able to employ only catalytic amounts of
this crown ether, under phase transfer conditions.
Among the solvents tested in Table 3, we were delighted
to find that benzaldehyde was not transformed, even
In conclusion, TDA-1 proved to be an efficient additive
for the Z-selective synthesis of a,b-unsaturated esters
from a variety of aromatic aldehydes, using Still’s
reagent. Unfortunately, the reactions have to be carried
out at low temperature to reach selectivities higher than
95%. In contrast, the use of 18C6 allows the olefination
of benzaldehyde to occur with high selectivity, even at
0 °C (Table 1, entry 2). It was therefore decided to
establish suitable reaction conditions under which 18C6
would be used in catalytic amounts. Initial screening
experiments using benzaldehyde as the substrate and
phosphonates 1 and 2, under a variety of conditions,
revealed that K2CO3 was a suitable base that promoted
the HWE condensation in a wide range of solvents. In
Table 3. K2CO3/18C6 mediated HWE reaction of 1 and 2 with benzaldehydea
Entry
Phosphonate
Solvent
Conversion (%)b
Selectivity (%)c
1
2
3
4
2
2
2
2
Toluene
86
85
51
77
87
82
73
89
Dichloromethane
Acetonitrile
Tetrahydrofuran
5
1
1
1
1
1
1
1
Toluene
Dichloromethane
Acetonitrile
98
100
100
100
100
99
97
93
94
98
98
97
97
6
7
8
Tetrahydrofuran
Ethyl acetate
9
10
11
Chlorobenzene
Trifluoromethylbenzene
95
a 1 or 2 (0.5 mmol–1 equiv), K2CO3 (1 mmol–2 equiv), benzaldehyde (0.5 mmol–1 equiv), 18C6 (0.5 mmol–1 equiv), solvent (10 mL)––see the typical
experimental procedures.8
b Benzaldehyde conversion after 3 h at 0 °C. Area% calculated by GC.
c Selectivity ¼ Z/(Z+E) calculated by GC.

5522
F. P. Touchard / Tetrahedron Letters 45 (2004) 5519–5523
after extended periods of time at room temperature, in
chlorobenzene.12 Gratifyingly, addition of 10 mol %
18C6 to this reaction mixture led to the formation of the
desired alkene with 97% Z-selectivity at 0 °C and 94% Z-
selectivity at room temperature.
To evaluate the scope of these novel conditions, a
variety of other aromatic aldehydes were submitted to
the HWE olefination (Table 4). The reactions have also
been performed with a stoichiometric amount of addi-
tive in order to compare the selectivities.
Table 4. K2CO3/18C6 mediated HWE reaction of 1 with aromatic aldehydesa
O
P
O
O
R
CO2Me
K2CO3, 10% 18C6
PhCl, 0˚C
+
F3C
O
OMe
R
H
O
1
CF3
Entry
Aldehyde
Time (h)
Selectivity (%)b
10% 18C6
Yield (%)c
100% 18C6
CHO
CHO
1
18
97
97
98
96
97
98
97
95
97
93
93
CF3
2
3
4
5
6
7
8
9
18
18
18
18
66
42
18
66
99
97
98
98
97
95
97
93
94
100
100
100
98
F3C
CHO
CHO
F3C
OMe
CHO
CHO
MeO
CHO
93
CHO
92
N
S
85
CHO
a 1 (7.62 mmol–1.05 equiv), K2CO3 (14.5 mmol–2 equiv), aldehyde (7.25 mmol–1 equiv), 18C6 (0.72 mmol–10 mol %), chlorobenzene (150 mL)––see
the typical experimental procedures.8
b Selectivity ¼ Z/(Z + E) calculated by 1H NMR integration of the vinyl proton signals and in accordance with GC determination.
c Isolated yields of the Z/E-mixtures of the 10 mol % experiments; the structures have been confirmed by 1H and 13C NMR spectroscopies and by
comparison with literature data.

F. P. Touchard / Tetrahedron Letters 45 (2004) 5519–5523
8. Typical procedures for the HWE reaction:
5523
As can be seen from Table 4, excellent Z-selectivities
were obtained with all the aromatic aldehydes tested.
Once again, the Z/E-ratio does not depend too much
upon the substituents on the aromatic ring. Comparison
of the selectivities between catalytic and stoichiometric
amounts of 18C6 revealed that the crown ether catalysis
was effective in all cases.
With KHMDS (Table 2). In a 250 mL flask were added
under Ar: dry THF (90 mL), 1 (5.5 mmol) and TDA-1
(5.5 mmol). The mixture was then cooled to )78 °C and
treated with KHMDS (10.5 mL of a 0.5 M solution in
toluene-5.25 mmol). The aldehyde (5 mmol) was added
30 min later and the mixture was kept at )78 °C for 4 h.
The reaction mixture was then diluted with 70 mL
MeOtBu and quenched with 50 mL saturated NH4Cl.
The aqueous phase was extracted twice with 20 mL of
MeOtBu and the final organic phase was washed with
water until neutrality. It was then dried over Na2SO4 and
the solvent removed by evaporation under vacuum. The
Z/E-mixture was finally purified by flash chromatography
(mixtures of cyclohexane/AcOEt).
With K2CO3 (Table 4): In a 250 mL flask were added
under Ar: K2CO3 (14.5 mmol), 18C6 (0.72 mmol) and
150 mL of chlorobenzene. The mixture was stirred at room
temperature for 3 h before being cooled to 0 °C. 1
(7.62 mmol) and the aldehyde (7.25 mmol) were then
added and the mixture was maintained at 0 °C until the
work-up. The reaction mixture was quenched with 50 mL
saturated NH4Cl and the organic phase was washed by
saturated NH4Cl and water until neutrality. It was then
dried over Na2SO4 and the solvent removed by evapora-
tion under vacuum. The Z/E-mixture was finally purified
by flash chromatography (mixtures of cyclohexane and
AcOEt). The Z-ester obtained from 2-(trifluoro-
methyl)benzaldehyde has been characterized by 1H, 19F,
13C NMR and HRMS. 1H NMR (CDCl3): d 3.54 (s, CH3),
6.05 (d, J ¼ 12:3 Hz, CHCO2Me), 7.22 (m, J ¼ 12:3 and
2.2 Hz, CHAr), 7.33 (d, J ¼ 7:3 Hz, 1Harom), 7.35 (t,
J ¼ 7:3 Hz, 1Harom) 7.44 (t, J ¼ 7:3 Hz, 1Harom), 7.60
(d, J ¼ 7:3 Hz, 1Harom); 19F NMR (CDCl3): d 15.1 (s,
CF3); 13C NMR (CDCl3): d 51.2 (s, CH3), 122.2 (s,
CHCO2Me), 123.8 (q, JC–F ¼ 273:5 Hz, CF3), 125.5 (q,
JC–F ¼ 5:1 Hz, CHarom), 127.2 (q, JC–F ¼ 29:9 Hz, Car-
om), 127.9 (s, CHarom), 130.0 (s, CHarom), 130.9 (q,
JC–F ¼ 1 Hz, CHarom), 134.6 (q, JC–F ¼ 1:9 Hz, Carom),
140.8 (s, CHAr), 165.6 (s, CO); HRMS calcd for
C11H9O2F3 230.0555, found: 230.0545.
In conclusion, we have shown that TDA-1 was an
efficient additive in the Horner–Wadsworth–Emmons
reaction using Still’s reagent 1 and aromatic aldehydes
in THF at )78 °C. Selectivities up to 92% and even
higher have been obtained in all cases. Moreover, we
have also been able to delineate new reaction conditions
in which the amount of 18C6 could be decreased to 10%
with equal efficiency. K2CO3 appears to be the base
of choice and chlorobenzene the preferred solvent.
The use of these conditions to other phosphonates
bearing various stabilizing groups will be reported in
due course.
References and notes
1. For a general review concerning the Wittig olefination
reaction and modifications see: Maryanoff, B. E.; Reitz,
A. B. Chem. Rev. 1989, 89, 863–927.
2. Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405–
4408.
3. Ando, K. Tetrahedron Lett. 1995, 36, 4105–4108.
4. Since the first publication of Ando,3 many phosphonate
modifications have been reported: (a) Ando, K. J. Org.
Chem. 1997, 62, 1934–1939; (b) Ando, K. J. Org. Chem.
1998, 63, 8411–8416; (c) Ando, K. J. Org. Chem. 1999, 64,
8406–8408; (d) Ando, K.; Oishi, T.; Hirama, M.; Ohno,
H.; Ibuka, T. J. Org. Chem. 2000, 65, 4745–4749; (e)
Kokin, K.; Motoyoshiya, J.; Hayashi, S.; Aoyama, H.
Synthetic Commun. 1997, 27(14), 2387–2392; (f) Kokin,
K.; Iitake, K.; Takaguchi, Y.; Aoyama, H.; Hayashi, S.;
Motoyoshiya, J. Phosphorus, Sulfur Silicon 1998, 133, 21–
40.
5. For recent reviews, see: (a) Ando, K. Yuki Gosei Kagaku
Kyokaishi 2000, 58(9), 869–876; (b) Motoyoshiya, J.
Trends Org. Chem. 1998, 7, 63–73.
6. The phosphonate based on 2,4-difluorophenol described
by Motoyoshiya and co-workers4e;f is the only reported to
date leading to selectivities over 92% at 0 °C with various
aldehydes.
9. The same selectivities were obtained using either 1.1 or
5 equiv of 18C6.
10. Soula, G. J. Org. Chem. 1985, 50, 3717–3721, and
references cited therein. TDA-1 is commercially available
on large scale by RHODIA.
11. We have not observed any effect of the additive with both
octanal and cyclohexane carboxaldehyde. The selectivities
are close to 80% at )78 °C. This is in agreement with the
results of Still reporting modest selectivities with aliphatic
aldehydes when the KHMDS/18C6 system is employed.2
12. Butyl acetate is another solvent in which benzaldehyde
was not transformed, even after extended periods of time
at room temperature.
7. A patent application form was filled: RHODIA FR
0211731.