Full text of DOI:10.1016/S0040-4039(01)02053-6

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 299–301
Intramolecular Wittig reactions with esters utilising
triphenylphosphine and dimethyl acetylenedicarboxylate
Lyndsay A. Evans, Kimberley E. Griffiths, Holger Guthmann and Patrick J. Murphy*
Department of Chemistry, University of Wales, Bangor, Gwynedd LL57 2UW, UK
Received 9 August 2001; revised 17 October 2001; accepted 26 October 2001
Abstract—The intramolecular Wittig olefination of a-hydroxy- and a-amino esters has been effected in high yield using a
combination of triphenylphosphine and dimethyl acetylenedicarboxylate. © 2002 Elsevier Science Ltd. All rights reserved.
1
4
We previously reported that reaction of a-hydroxybu-
protons. In the case of 6/7a, these values were found to
tyrolactones, for example 1, with triphenylphosphine
TPP) and dimethyl acetylenedicarboxylate (DMAD) in
be J=4.6 and 1.8 Hz, respectively (Scheme 2).
(
dioxane at reflux led to the formation of two bicyclic
products 3 and 4 in a 2:1 ratio via the ylide intermedi-
ate 2 (Scheme 1).
When the reaction was repeated using ethyl lactate 5b
(Table 1, entry 2) a similar result was observed with the
cyclisation products 6/7b being isolated in 90% yield
and in a 5:1 ratio, together with a smaller amount of 8b
In order to test the generality of this reaction, we
investigated the corresponding reaction of esters and2
thus treated ethyl glycolate 5a under similar conditions
and were pleased to observe the formation of the
expected Wittig product 6a in 57% yield together with
(8%). These again were identified on the basis of the
5
magnitude of their J
values. The relative ratio of
H–H
these two compounds remained constant (ca. 4.5–5:1)
over several reactions indicating that the selectivity for
the cis-product probably has its origins in the step
leading to the formation of the intermediate ylide. This
might also suggest that the products are stable to
epimerisation under the reaction conditions. Finally,
reaction of ethyl mandelate 5c (Table 1, entry 3) under
identical conditions gave 6/7c in 78% yield and in a 25:1
ratio, together with the addition product 8c in 15%
yield (Scheme 2).
8
a, the product from the conjugate addition of ethyl
glycolate to DMAD, in 26% yield (Table 1, entry 1).
This was a particularly encouraging result as we had
envisaged that more forcing conditions might have been
required to effect the cyclisation, as is often the case
3
with esters. We attempted to minimise the formation
of the by-product 8a by increasing the relative amounts
of TPP and DMAD and allowing the reaction to stir at
room temperature for longer periods of time, however,
this had little effect on the overall outcome of the
process. Structural determination of the product 6a was
We next turned our attention to the formation of the
corresponding dihydropyrroles and initially looked at
the series of Boc-protected amino acid esters 9a–c
5
5
aided by the characteristic JH–H coupling found in
(Table 2, entries 1–3 and Scheme 3). We were encour-
2
,5-dihydrofurans, which is of the order of 4–5 Hz for
aged to see that reaction of N-Boc glycine ethyl ester 9a
under the standard conditions gave a respectable 46%
the cis-protons and 1–2 Hz for the corresponding trans-
Scheme 1. (a) PPh /DMAD/dioxane/10°C, 1 h, then reflux 3 h 84% (2:1).
3
Keywords: intramolecular Wittig reaction; 2,5-dihydrofurans; 2,5-dihydropyrroles.
*
Corresponding author. E-mail: paddy@bangor.ac.uk
0
040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)02053-6

3
00
L. A. E6ans et al. / Tetrahedron Letters 43 (2002) 299–301
Table 1.
Entry
5
R
6/7 (%)
6:7
8 (%)
5JH–H cis (Hz)
5JH–H trans (Hz)
1
2
3
a
b
c
H
Me
Ph
57
90
78
–
26
8
15
J=4.6
J=3.9
J=4.1
J=1.8
J=1.5
J=2.6
a
b
5:1
b
25:1
a
Ethyl (S)-(−)-lactate was used.
Inseparable by column chromatography.
b
Scheme 2. (a) PPh /DMAD/dioxane/10°C, 1 h, then reflux 16 h.
3
yield of the expected 2,5-dihydropyrrole 10a, together
with a considerable amount of recovered 9a which raised
dihydropyrroles 10/11d–f in 75–96% overall yield (Table
2, entries 4–6), with only 9e (entry 5) being recovered in
any appreciable amount (Scheme 3). As observed with
the dihydrofuran examples, in all these reactions, there
appeared to be a bias towards the formation of the
cis-products 10, however, the effect was not as pro-
nounced.
the yield, based on conversion, to 90%. Again, structural
5
determination of 9a was aided by the characteristic J
H–H
3
coupling found in 2,5-dihydropyrroles. Following on
from this, we attempted a similar reaction with N-Boc
alanine ethyl ester 9b and were disappointed to achieve
only a 13% yield of the desired products 10/11b in ca. 1:1
ratio. Again, based on recovered starting material this
yield rose to 93%. Finally, we reacted N-Boc-2-phenyl-
glycine methyl ester 9c under identical conditions to give
We next investigated the reaction of ethyl thioglycolate
12 under these conditions and not surprisingly obtained
only the addition product 13 in high yield. We attempted
to prevent the formation of this product by firstly
10/11c in 46% yield in a 1.4:1 ratio. Again the yield was
6
found to be very high when recovered starting material
was taken into account.
generating the vinyl phosphonium salt 14 in situ from
DMAD and triphenylphosphonium hydrobromide. This
was then reacted with 12 in the presence of pyridine or
DBU to effect the formation of the intermediate ylide.
Despite considerable effort, we were unable to isolate any
of the required 2,5-dihydrothiophene, the only major
product obtained being triphenylphosphine sulphide,
suggesting that desulfurisation of the starting material
and/or product is a competetive process (Scheme 4).
We felt that the low conversion of substrates 9a–c was
probably due to a steric effect related to the Boc
protecting group. We thus performed a series of reactions
on the corresponding series of Z-protected amino acid
esters 9d–f. We were pleased to find that these substrates
were far more suitable for the reaction leading to the
a
Table 2.
Entry
9
R
R1
P
10/11 (%)
10:11
5JH–H-cis (Hz)
J=4.4
5JH–H-trans (Hz)
J=1.7
1
2
3
4
5
6
a
b
c
d
e
f
H
Et
Et
Me
Et
Et
Boc
Boc
Boc
Z
Z
Z
46 (90)
13 (93)b
46 (99)
91
–
c
c
Me
Ph
H
Me
Ph
ca. 1:1
1.4:1
–
ca. 1:1
1.2:1
–
–
J=4.6
J=4.2
J=ꢀ0
J=1.3
75 (96)b
96
–
c
–
c
Me
J=4.4
J=ꢀ0
a
Yields in parentheses are based on recovered starting material.
Inseparable by column chromatography.
Signals obscured.
b
c
Scheme 3. (a) PPh /DMAD/dioxane/10°C, 1 h, then reflux 16 h.
3

L. A. E6ans et al. / Tetrahedron Letters 43 (2002) 299–301
301
Scheme 4. (a) PPh , DMAD, dioxane, 10°C, 1 h; then reflux, 16 h.
3
In conclusion, we have demonstrated that the
intramolecular Wittig reaction of the ylide generated
from the reaction of a-hydroxy or Z-protected a-
aminoesters leads to the formation of 2,5-dihydrofurans
or 2,5-dihydropyrroles, respectively, in high yields. In
addition, the preferential formation of the cis-heterocy-
cles in these reactions and their apparent stability to
epimerisation would suggest that this reaction could be
utilised in an asymmetric manner. We intend to investi-
gate this eventuality, as well as other transformations,
for example preliminary results include the aromatisa-
tion of 10a to the pyrrole 15 and the deprotection of
which was followed by the addition of silica gel (8–10 g)
and evaporation of the solvent. The resulting free flowing
powder was applied to a silica gel column and the prod-
ucts eluted with a diethyl ether/petroleum ether mix. All
compounds displayed satisfactory analytical data.
4,5-Dicarbomethoxy-3-ethoxy-2,5-dihydrofuran 6a.
1
H
NMR: l 5.25 (dd, 1H, J=4.6, 1.8 Hz, CH), 4.88 (dd, 1H,
J=13.7, 4.6 Hz, CHaH ), 4.70 (dd, 1H, J=13.7, 1.8 Hz,
6
b
CH H6 ), 4.15 (q, 2H, J=7.0 Hz, CH ), 3.90 (s, 3H, CH ),
a
b
2
3
13
3.60 (s, 3H, CH ), 1.37 (t, 3H, J=7.0 Hz, CH ).
C
3
3
NMR: l=171.52 (C), 166.06 (C), 162.31 (C), 99.51 (C),
83.33 (CH), 71.96 (CH ), 68.64 (CH ), 52.31 (CH ), 51.28
2
2
3
1
0/11b to give 16a/b, both of which proceeded in high
(CH ), 15.37 (CH ). IR: wmax=2984, 2953, 2911, 2860
3 3
yield (Scheme 5).
(C-H), 1745, 1724 (CꢀO), 1650 (CꢀC). MS (CI): m/z=171
+ +
(
10%, [M−CO Me] ), 231 (100%, [M+H] ), 248 (80%, [M+
2
+
+
NH ] ). HRMS (CI): C H O ([M+H] ) requires
4
10 15
6
231.0868; found: 231.0868.
N-Boc-4,5-dicarbomethoxy-3-ethoxy-2,5-dihydropyrrole
0a. (The Boc and Z-protected derivatives were found to
1
exist as atropisomers, figures in parentheses represent
chemical shift values for the alternate isomer where given.)
1
H NMR: l 5.05 (5.09) (dd, 1H, J=4.4, 1.7 Hz, CH), 4.52
(
1
(
4.49) (dd, 1H, J=16.0, 4.4 Hz, CH6 aH ), 4.40 (4.32) (dd,
b
H, J=16.0, 1.7 Hz, CH H6 ), 4.18 (m, 2H, CH ), 3.74
a b 2
3.76) (s, 3H, CH ), 3.72 (3.73) (s, 3H, CH ), 1.44 (m, 12H,
3
3
13
t-Bu+CH3). C NMR: l=171.92 (C), 164.59 (C), 162.52
C), 152.98 (C), 100.44 (C), 81.17 (C), 67.95 (CH ), 63.72
(
2
Scheme 5. (a) DDQ, toluene, reflux 16 h; (b) CF COOH,
DCM, 2 h.
(CH), 52.37 (CH ), 51.24 (CH ), 50.33 (CH ), 28.28 (3×
CH ), 15.27 (CH ). IR: wmax=2979, 2951, 2868 (C-H),
3 3
3 3 2
3
1
748, 1708 (CꢀO), 1639 (CꢀC). MS (CI): m/z=230 (40%,
+
+
Acknowledgements
[M+H-Boc] ), 291 (100%, [M+NH
4
−C
M+H] ), 347 (35%, [M+NH ] ). HRMS (CI): C H NO
7
4 8
H ] ), 330 (30%,
+
+
[
4
15 24
+
(
[M+H] ) requires 330.1553; found: 330.1556.
Thanks are given to the ERASMUS/SOCRATES
scheme (H.G.) and to the EPSRC Mass spectrometry
centre at Swansea.
3. (a) Murphy, P. J.; Brennan, J. Chem. Soc. Rev. 1988, 17, 1;
(b) Lee, S. E.; Murphy, P. J. J. Chem. Soc., Perkin Trans.
1 1999, 3049.
4
. (a) Anet, E. F. L. Aust. J. Chem. 1965, 18, 837; (b) Anet,
E. F. L. Carbohydr. Res. 1962, 2, 448.
References
5. The protected amino acid esters where prepared using a
combination of the methods found in references (a)–(e): (a)
Roglans, A.; Marquet, J.; Moreno-Manas, M. Synth.
Commun. 1992, 22, 1249; (b) Minder, B.; Schuerch, M.;
Mallat, T.; Baiker, A.; Heinz, T.; Pfaltz, A. J. Catal. 1996,
160, 261; (c) Dondoni, A.; Perrone, D.; Semola, T. Synthe-
sis 1995, 181; (d) Barkdoll, A. E.; Ross, W. F. J. Am.
Chem. Soc. 1944, 66, 951; (e) Kashima, C.; Tsuruoka, S.;
Mizuhara, S. Tetrahedron 1998, 54, 14679.
1
. Heys, L.; Murphy, P. J.; Coles, S. J.; Gelbrich, T.; Hurst-
house, M. B. Tetrahedron Lett. 1999, 40, 7151.
. General procedure: The required a-hydroxy ester or pro-
tected amino acid ester (1 equiv., typically 10 mmol) was
dissolved in dry dioxane (20 ml) together with
triphenylphosphine (1.1–3.0 equiv.) and cooled in ice water
2
(
to ca. 10°C). DMAD (1.0–1.5 equiv.) was then added
dropwise over 5 min and the reaction warmed to rt over 1
6. Shaw, M. A.; Tebby, J. C.; Ward, R. S.; Williams, D. H.
h. The reaction was refluxed overnight (16 h), cooled to rt
J. Chem. Soc. (C) 1968, 2796.