Excess sodium ions improve Z selectivity in Horner-Wadsworth-Emmons olefinations with the Ando phosphonate

Article abstract of DOI:10.1016/S0040-4039(03)00935-3

New, improved conditions for Z selective Horner-Wadsworth-Emmons olefinations with Ando's bis(o-methylphenyl)phosphonates are reported. A combination of NaH and NaI affords Z olefins in up to >99:1 selectivity and good yields.

Full text of DOI:10.1016/S0040-4039(03)00935-3

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 4361–4364
Excess sodium ions improve Z selectivity in
Horner–Wadsworth–Emmons olefinations with the Ando
phosphonate
Petri M. Pihko* and Taina M. Salo
Laboratory of Organic Chemistry, Department of Chemical Technology, Helsinki University of Technology, POB 6100,
FIN-02015 HUT, Finland
Received 12 March 2003; revised 1 April 2003; accepted 10 April 2003
Abstract—New, improved conditions for
Z selective Horner–Wadsworth–Emmons olefinations with Ando’s bis(o-
methylphenyl)phosphonates are reported. A combination of NaH and NaI affords Z olefins in up to >99:1 selectivity and good
yields. © 2003 Elsevier Science Ltd. All rights reserved.
Di- and trisubstituted (Z)-olefins bearing an allylic
stereocentre are highly useful intermediates in stereose-
lective synthesis. The Z double bond imposes a consid-
erable conformational bias on the molecule dictated by
allylic 1,3-strain.1 This bias is reflected in the reactions
of these olefins, which are often highly diastereoselec-
tive.2 Synthesis of (Z)-olefins with high selectivities,
yields, and optical purities, is therefore an important
objective. For the Horner–Wadsworth–Emmons
(HWE) olefination3 using (phosphono)acetates and
their derivatives, two Z selective methods have attained
widespread use: the Still–Gennari method4 employing
[bis(trifluoroethyl)phosphono]acetates and the Ando
method5 using (diarylphosphono)acetates such as 1.
While the Ando method has been reported to be more
selective in many cases,6 it has not yet attained the
popularity of the Still–Gennari method.
Still and Gennari. Somewhat surprisingly, with
aliphatic and especially a-branched aldehydes, the use
of NaH was recommended (no selectivity data was
reported for lithium or potassium bases).5b Further-
more, the conditions employed by Ando require as
much as 140 mol% of NaH relative to the phosphonate.
Since using an excess of base might rapidly erode the
enantiopurities of valuable a-chiral aldehydes, we
decided to examine the selectivities using substoichio-
metric amounts of base relative to phosphonate 1
(Table 1). With aliphatic aldehydes (n-octanal 2 and
cyclohexanecarboxaldehyde 3), NaH was clearly the
most Z selective among the bases examined. The potas-
sium and lithium bases (KHMDS and n-BuLi) were
both inferior to NaH. Addition of 18-crown-6 to
KHMDS improved the Z selectivity with cyclohexane-
carboxaldehyde to 85:15, but in comparison, NaH
afforded 93:7 selectivity (entry 6). The same trend held
both at −78°C and at 23°C (entries 5, 7 and 9). Raising
the temperature from −78 to 23°C eroded the Z selec-
tivities in all cases, mostly so with the lithium cation.8
In this paper, we present new, simple conditions that
improve the Z selectivities with the Ando reagent 1 and
afford useful methyl esters in high Z selectivity.7
We began by examining the role of the countercation in
the Ando olefination. With benzaldehyde, Ando5a,b has
conducted a full comparison between lithium, sodium
and potassium bases. The potassium bases (KHMDS
or KHMDS/18-crown-6) were reported to give the best
Z selectivities, in accord with the results obtained by
The next variable to be studied was the base/phospho-
nate stoichiometry. With 3, increasing the amount of
NaH from 90 to 140 mol% improved the Z:E selectivity
from 90:10 to 94:6, equalling the results reported by
Ando5b for the corresponding ethyl ester (Table 1,
footnote d). Increasing the amount of base did not
bring about any further improvement in selectivity.
Keywords: Horner–Wadsworth–Emmons reaction; stereoselectivity;
cations.
At this point, we reasoned that an excess of the coun-
tercation appears to be beneficial for selectivity. As
* Corresponding author. Tel.: +358 9 451 2536; fax: +358 9 451 2538;
e-mail: petri.pihko@hut.fi
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)00935-3

4362
P. M. Pihko, T. M. Salo / Tetrahedron Letters 44 (2003) 4361–4364
Table 1. Effect of countercation on Z selectivity of HWE olefination with 1 and aliphatic aldehydes
Entry
Aldehyde
Basea
Temp. (°C)
Z:E ratiob
Note
1
2
3
4
5
6
7
8
9
n-Octanal (2)
n-Octanal
n-Octanal
CyCHO (3)c
CyCHO
CyCHO
CyCHO
CyCHO
CyCHO
n-BuLi
NaH
KHMDS
n-BuLi
n-BuLi
NaH
NaH
KHMDS
KHMDS
−78
−78
−78
−78
23
−78
23
−78
23
75:25
86:14
78:22
83:17
45:55
93:7d
82:18
76:24e
70:30
32% conversion
40% conversion
a Reactions were conducted in THF (5 mL/mmol aldehyde) using 110 mol% of base and 120 mol% phosphonate 1. Unless otherwise stated, all
reactions proceeded to >90% conversion (determined by 1H NMR and TLC) within 2 h.
b The Z:E ratios were determined from crude reaction mixtures by 1H NMR by integration of the vinyl proton signals.
c Cy=cyclohexyl.
d With 90 mol% NaH, Z:E selectivity drops to 90:10. With 140 mol% NaH, selectivity increases slightly (to 94:6).
e With added 18-crown-6 (500 mol%) the selectivity improved to 85:15.
entries 4–6 and 10, Table 2). These conditions turned
out to be more selective than Still’s KHMDS/18-crown-
6 base system (85:15 Z:E with 3, footnote e in Table 1)
or the more recent DBU/NaI reagent combination
reported by Ando (entries 3 and 9). Addition of 50–100
mol% of NaI appears to be sufficient for good Z
selectivities; with 300 mol% NaI (entry 6) the Z selec-
tivity drops slightly.
discussed above, we wanted to keep the amount of base
to a minimum to allow a-chiral aldehydes to be used as
substrates. As such, we reasoned that we could simply
use one equivalent of base and supply the excess cation
in the form of NaI or LiBr to increase the Z selectivity.
This hypothesis was evaluated using two a-branched
aldehydes, cyclohexanecarboxaldehyde 3 and TBS-pro-
tected lactaldehyde 49 (Table 2). With n-butyllithium as
a base, addition of LiBr or NaI did not appreciably
increase the Z:E ratio. However, with an all-sodium
combination (NaH as a base and NaI as an additive),
very high selectivities were obtained (up to >99:1 Z:E,
A range of aldehydes was screened (Table 3) to probe
the scope of these NaH/NaI conditions. Aliphatic alde-
hydes gave excellent Z:E selectivities ranging from 93:7
Table 2. Effect of added cation source on Z selectivity in HWE olefinations with 1a
Entry
Aldehyde
Base (110 mol%) Additive (mol%) Temperature
Z:E ratiob
Note
(°C)
1
2
3
4
5
6
7
CyCHO (3)
n-BuLi
n-BuLi
DBU
NaH
NaH
LiBr (200)
NaI (200)
NaI (100)
NaI (50)
NaI (100)
NaI (300)
LiBr (200)
−20
−78
−78
−78
−78
−78
−78
60:40
78:22
88:12
95:5c
95:5
60% conversion; very slow at −78°C
95:5 selectivity with the ethyl ester5e
3
3
3
3
3
NaH
n-BuLi
94:6
88:12
8
9
10
4
4
4
n-BuLi
DBU
NaH
NaI (200)
NaI (100)
NaI (100)
−78
−78
−78
92:8
95:5
>99:1
a Reactions were conducted in THF (5 mL/mmol with LiBr, 10 mL/mmol with NaI as an additive), using 110 mol% of base and 120 mol% of
phosphonate. Unless otherwise stated, all reactions proceeded to >90% conversion (as judged by 1H NMR and TLC) within 2 h.
b The Z:E ratios were determined from crude products by 1H NMR.
c A control experiment with 3 using NaH (110 mol%) and NaBr (100 mol%) in THF (15 mL/mmol) afforded 94:6 Z:E selectivity. However, even
at this concentration, a small portion of NaBr remained undissolved. NaI is preferred to NaBr due to its far higher solubility in THF.

P. M. Pihko, T. M. Salo / Tetrahedron Letters 44 (2003) 4361–4364
4363
Table 3. Z selective HWE olefinations with phosphonate 1
using the NaH/NaI combination10
The fact that an excess of sodium ions is required for
highest selectivities strongly suggests that these HWE
reactions take place via cation-chelated intermediates
and transition states.15 In preliminary studies, we have
found that adding an excess of cation gives excellent
results with the Still–Gennari phosphonates, too, when
aliphatic aldehydes are employed. A full account of
these studies is forthcoming.
Acknowledgements
Financial support from HUT and TEKES is gratefully
acknowledged. We thank Professor Ari Koskinen for
material support and valuable discussions. We also
thank Nicolas Desroy, Ainoliisa Pihko and Taru Poh-
jasniemi for their early contributions and Anniina
Erkkila¨ for assistance with the starting materials.
References
1. Hoffmann, R. W. Chem. Rev. 1989, 89, 1841–1860.
2. For examples, see: (a) Koskinen, A. M. P.; Chen, J.
Tetrahedron Lett. 1991, 32, 6977–6980; (b) Sakai, N.;
Ohfune, Y. J. Am. Chem. Soc. 1992, 114, 998–1010; (c)
Kauppinen, P. M.; Koskinen, A. M. P. Tetrahedron Lett.
1997, 38, 3103–3106.
3. For reviews, see: (a) Kelly, S. E. In Comprehensive
Organic Synthesis; Trost, B. M.; Fleming, I., Eds.; Perga-
mon Press: Oxford, 1991; Vol. 1; pp. 730–817; (b)
Maryanoff, B. E.; Reitz, A. B. Chem. Rev. 1989, 89,
863–927.
4. Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24,
4405–4408.
5. (a) Ando, K. Tetrahedron Lett. 1995, 36, 4105–4108; (b)
Ando, K. J. Org. Chem. 1997, 62, 1934–1939; (c) Ando,
K. J. Org. Chem. 1998, 63, 8411–8416; (d) Ando, K. J.
Org. Chem. 1999, 64, 8406–8408; (e) Ando, K.; Oishi, T.;
Hirama, M.; Ohno, H.; Ibuka, T. J. Org. Chem. 2000, 65,
4745–4749; (f) Ando, K. Synlett 2001, 1272–1274.
Motoyoshiya and co-workers have also developed related
bis(2,4-difluorophenyl)phosphonates: (g) Kokin, K.;
Motoyoshiya, M.; Hayashi, S.; Aoyama, H. Synth. Com-
mun. 1997, 27, 2387–2392.
6. Selected examples of the use of Ando phosphonates in
synthetic applications: (a) Enders, D.; Vicario, J. L.; Job,
A.; Wolberg, M.; Mu¨ller, M. Chem. Eur. J. 2002, 8,
4272–4284; (b) Feutrill, J. T.; Lilly, M. J.; Rizzacasa, M.
A. Org. Lett. 2002, 4, 525–527; (c) Masaki, H.;
Maeyama, J.; Kamada, K.; Esumi, T.; Iwabuchi, Y.;
Hatakeyama, S. J. Am. Chem. Soc. 2000, 122, 5216–5217;
(d) Oishi, T.; Iwakuma, T.; Hirama, M.; Itoˆ, S. Synlett
1995, 404–406.
7. Ando reagents based on ethyl esters tend to give slightly
higher Z selectivities compared to methyl esters. See Ref.
5d.
8. Similar temperature effects have been noted with
trimethylphosphonoacetate: Thompson, S. K.; Heath-
cock, C. H. J. Org. Chem. 1990, 55, 3386–3388. The
pronounced Z selectivity with lithium salts at −78°C was
attributed to increased solvation of the lithium salts at
low temperatures.
to 99:1 and very high yields (88–100%). As expected,
a-branched aldehydes (entries 2, 3 and 6) turned out to
give slightly better selectivities than unbranched alde-
hydes (entries 1 and 5). The only exception was alde-
hyde 8 (entry 7), which gave only 84:16 Z selectivity. In
agreement with the results obtained by Ando,5a,b ben-
zaldehyde (entry 4) gave inferior selectivity under these
conditions (70:30). To demonstrate the mildness of the
reaction conditions, the enantiomeric purities of Z-
enoates Z-7 and Z-8 derived from sensitive a-chiral
amino aldehydes 7 and 8 (entries 6 and 7) were deter-
mined by chiral HPLC. In both cases the products were
obtained in excellent enantiomeric purity.
In summary, we present new, highly Z selective condi-
tions for Horner–Wadsworth–Emmons olefinations
employing Ando’s bis(o-cresyl)phosphonoacetate 1.
With aliphatic aldehydes, our new NaH/NaI procedure
gave better selectivities than any of the published proce-
dures (including Ando’s new NaI/DBU conditions5e or
Still’s KHMDS/18-crown-6 conditions4). The condi-
tions are sufficiently mild to preserve the stereochemical
integrity of a-chiral aldehyde substrates and employ
only inexpensive reagents.

4364
P. M. Pihko, T. M. Salo / Tetrahedron Letters 44 (2003) 4361–4364
9. (a) Ito, Y.; Kobayashi, Y.; Kawabata, T.; Takase, M.;
Z-8 colourless oil; Rf (30% EtOAc/hexanes, UV/PMA)=
0.49; [h]D=−56.0 (c 0.72, CHCl3); IR (thin film) 3050,
Terashima, S. Tetrahedron 1989, 45, 5767–5790; (b) The
Still–Gennari phosphonate with KH as a base has been
reported to give 89:11 Z:E selectivity with 4: Annunziata,
R.; Cinquini, M.; Cozzi, F.; Dondio, G.; Raimondi, L.
Tetrahedron 1987, 43, 2369–2380. The enantiomeric
purity of Z-4 was not directly measured but the optical
rotation of Z-4 ([h]D=+46, c 1.2, CHCl3) matches the
literature value ([h]D=+49, c 1.8, CHCl3).9b
1
2930, 1715, 1420, 1220, 1085 cm−1; H NMR ((CDCl2)2,
400 MHz, 360 K) l 7.33 (m, 5H), 6.27 (dd, 1H, J=7.6,
11.7 Hz), 5.84 (d, 1H, J=11.7 Hz), 5.51 (m, 1H), 5.15 (d,
1H, J=12.9 Hz), 5.13 (d, 1H, J=12.9 Hz), 4.30 (dd, 1H,
J=7.0, 8.9 Hz), 3.81 (dd, 1H, J=3.2, 9.1 Hz), 3.69 (s,
3H), 1.66 (s, 3H), 1.57 (s, 3H); 13C NMR ((CDCl2)2, 100
MHz, 360 K) l 165.7, 152.3, 149.8, 136.4, 128.1 (2C),
127.6, 127.4 (2C), 119.6, 94.5, 68.8, 66.6, 55.7, 51.0, 26.5,
24.1; HRMS (CI, NH3) calcd for MH+ (C17H22NO5)
320.1498, found: 320.1487, D=3.4 ppm; HPLC (Chiralcel
OD column, 5:95 i-PrOH/hexanes, 0.9 mL/min) tr=16.8
min; ent-Z-8: tr=12.5 min. Authentic ent-Z-8 was pre-
10. Typical experimental procedure: To a solution of phos-
phonate 1 (435 mg, 1.30 mmol) in THF (12 mL) at 0°C
was added NaI (150 mg, 1.00 mmol). After 5 min, NaH
(60% dispersion, 52 mg, 130 mol%) was added (CAU-
TION: evolution of hydrogen!), and the resulting solution
was cooled to −78°C. The aldehyde (1.00 mmol) was then
added dropwise (either neat or as a solution in 1 mL of
THF). After 2 h at −78°C, half-saturated NH4Cl (10 mL)
was added and the reaction mixture was extracted with
Et2O (2×20 mL). The combined organic extracts were
dried (Na2SO4) and concentrated under reduced pressure
(100 mmHg). The crude product was purified by flash
chromatography.
pared starting from
D-serine.
12. Murphy, J. A.; Rasheed, F.; Roome, S. J.; Scott, K. A.;
Lewis, N. J. Chem. Soc., Perkin Trans. 1 1998, 2331–
2440. See also Ref. 6a.
13. Pihko, P. M.; Koskinen, A. M. P. J. Org. Chem. 1998,
63, 92–98.
14. Beaulieu, P. L.; Schiller, P. W. Tetrahedron Lett. 1988,
29, 2019–2022. In our hands, distillation of 8 led to
appreciable racemization (60–85% ee). However, 8 could
be reliably purified by flash chromatography (silica gel,
20% EtOAc/hexanes) to afford a product of >97% ee.
15. The intimate details of the mechanism of HWE olefina-
tions are still poorly understood. For leading references
to computational studies aimed at elucidating the mecha-
nistic details, see: (a) Ando, K. J. Org. Chem. 1999, 64,
6815–6821 (lithium as the countercation); (b)
Motoyoshiya, J.; Kusaura, K.; Kokin, K.; Yokoya, S.;
Takaguchi, Y.; Narita, S.; Aoyama, H. Tetrahedron 2001,
57, 1715–1721 (potassium cation). For studies of the
HWE reaction in a free anion system, see: (c) Brandt, P.;
Norrby, P.-O.; Martin, I.; Rein, T. J. Org. Chem. 1998,
63, 1280–1289; (d) Norrby, P.-O.; Brandt, P.; Rein, T. J.
Org. Chem. 1999, 64, 5845–5852.
11. Characterization data for new compounds Z-7 and Z-8:
Z-7 (data for pure Z isomer, obtained by recrystallization
from hexanes): white needles; Rf (50% EtOAc/hexanes)=
0.55; [h]D=−124 (c 0.99, MeOH); mp 78–79°C; IR
(CDCl3) 3444, 2981, 1717, 1498, 1438, 1368, 1205, 1163,
1
1047 cm−1; H NMR (CDCl3, 400 MHz) l 6.12 (m, 1H),
5.75 (dd, 1H, J=1.3, 11.5 Hz), 5.17 (m, 1H), 4.75 (m,
1H), 3.71 (s, 3H), 1.42 (s, 9H), 1.26 (d, 3H, J=6.7 Hz);
13C NMR (CDCl3, 100 MHz) l 166.1, 155.2, 152.1,
118.6, 79.5, 51.3, 45.4, 28.3, 20.2. Anal. calcd for
C11H19NO4: C, 57.62; H, 8.35; N, 6.11. Found: C, 57.48;
H, 8.37; N, 6.07. The enantiomeric purity was determined
before recrystallization by HPLC (Chiralcel OD column,
5:95 i-PrOH/hexanes, 0.9 mL/min) tr=19.5 min; E-7:
tr=21.6 min, ent-Z-7: tr=30.2 min. Authentic ent-Z-7
was prepared from Boc-L-alaninal.