A novel azulene synthesis from the Ramirez ylide involving two different modes of its reaction with activated alkynes

Article abstract of DOI:10.1039/b316759c

The Ramirez ylide undergoes electrophilic substitution with acetylenedicarboxylates to form Z and E adducts. The latter can react by cycloaddition with another equivalent of the alkyne to provide a new route to novel tetra-substituted azulenes, which show

Full text of DOI:10.1039/b316759c

A novel azulene synthesis from the Ramirez ylide involving two different
modes of its reaction with activated alkynes†
a
b
a
a
c
Lee J. Higham, P. Gabriel Kelly, David M. Corr, Helge Müller-Bunz, Brian J. Walker and Declan
G. Gilheany*
a
a
Chemistry Department, The Centre for Synthesis and Chemical Biology, Conway Institute of Biomolecular
and Biomedical Sciences, University College Dublin, Dublin 4, Ireland. E-mail: declan.gilheany@ucd.ie;
Fax: +353-1-706-2127; Tel: +353-1-716-2308
Chemistry Department, National University of Ireland, Maynooth, Co. Kildare, Ireland
b
c
School of Chemistry, Queens University Belfast, Stranmillis Road, Belfast, N. Ireland BT9 5AG
Received (in Cambridge, UK) 23rd December 2003, Accepted 22nd January 2004
First published as an Advance Article on the web 11th February 2004
Another aspect to azulene 3 (R /R2 = Me) relates to its bond
lengths. There is currently interest in the solid-state structures of
azulenes9 in relation to the propensity of substituents to cause
1
The Ramirez ylide undergoes electrophilic substitution with
acetylenedicarboxylates to form Z and E adducts. The latter can
react by cycloaddition with another equivalent of the alkyne to
provide a new route to novel tetra-substituted azulenes, which
show interesting bond localisation and crystal packing effects.
,10
1
1
alternation of C–C bonding distances. In azulene itself the
peripheral bond lengths are almost equivalent (1.38–1.40 Å, ~ 0.02
Å difference) but alternation is evident in the 5,6-furanyl adduct of
azulene10 with an average difference in lengths of adjacent bonds of
0.06 Å (1.36 Å for the short bonds, 1.42 Å for the long ones). For
3, the shortest peripheral bond is 1.366(3) (C8–C9) and the longest
1.413(3) (C5–C6), a maximum difference of 0.047 Å showing that
the substitution pattern of 3 causes a moderate alternation effect.
The five-membered ring of 3 demonstrates bond angle distortions
The ylide cyclopentadien-1-ylidenetriphenylphosphorane 1, origi-
1
nally reported by Ramirez and Levy, is notable for being probably
the least reactive phosphorus ylide, failing to undergo the Wittig
1
,2
reaction and unaffected by boiling alkali.
This inertness is
attributed to the aromatic nature of the cyclopentadienyl moiety.
1 3
from 120° with C and C values of 109.59(17)° and 109.25(18)°
respectively, similar to those found for azulene-1,3-dipropionic
acid.12 The crystal packing diagram of 3 (Fig. 2) is also unusual in
that the seven-membered rings are stacked above each other in the
bc plane, as are the five-membered rings with a planar separation
between the azulene molecules of the order 3.5–3.8 Å. This is a
3
Consistent with this, Yoshida and co-workers noted that 1 reacts
with electrophiles to undergo substitution at the 2 position of the
five-membered ring. In one reaction they used an activated alkyne
(diethyl acetylenedicarboxylate) to give a putative vinyl adduct 2
but its structure was not established securely. We recently re-
4
investigated their work and found that a photochemically sensitive
Z/E mixture is produced. More significantly we also found that the
E adducts can react with a further molecule of activated alkyne to
give moderate yields (20–30%) of the novel tetra-substituted
1
2
t
azulenes 3 (R /R = Me, Et, Bu). Their striking blue-green colour
and high R values enabled them to be isolated relatively easily by
f
chromatography from the reaction mixtures, which also contained
a number of other products. The structures of the azulenes 3 could
not be established unambiguously by spectroscopy and we had to
rely on a single crystal X-ray structure determination for the methyl
Fig. 1 Structure of 3 (R /R2 = Me) (50% probability ellipsoids). Selected
bond distances (Å) and angles (°): C(1)–C(2) 1.393(3), C(1)–C(9) 1.414(3),
C(3)–C(2) 1.388(3), C(3)–C(10) 1.404(2), C(4)–C(5) 1.394(3), C(4)–C(10)
1
1
2
ester case (R /R = Me, Figs. 1 and 2).‡
1
.377(17), C(5)–C(6) 1.413(3), C(6)–C(7) 1.378(3), C(7)–C(8) 1.405(3),
C(8)–C(9) 1.366(3), C(9)–C(10) 1.495(3), C(1)–C(2)–C(3) 109.28(18),
C(3)–C(10)–C(4) 125.01(18), C(4)–C(5)–C(6) 126.62(18).
Azulenes are important compounds used in applications as
5
a,b
5c
5d
varied as pharmaceuticals,
cosmetics, infrared absorbers,
liquid crystals5 and optical films.5f However their preparation is
e
not facile in many cases: low yields are not uncommon and many
6
substitution motifs remain difficult to synthesise. The very short
7
and unusual new synthesis reported here is significant, providing
8
a convenient route to this rare 1,3,5,6-substitution pattern in what
is a respectable yield for an azulene synthesis.
†
Electronic supplementary information (ESI) available: full experimental
procedures and characterisation for 1–3. See http://www.rsc.org/suppdata/
cc/b3/b316759c/
Fig. 2 Packing of 3 viewed down the crystallographic a axis.
6
84
C h e m . C o m m u n . , 2 0 0 4 , 6 8 4 – 6 8 5
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

surprising observation because the strongly dipolar azulene moiety
usually stacks five-membered rings above seven-membered.13 The
opposite arrangement here implies that the dipolar contribution in 3
is small. Another notable feature of Fig. 2 is that it shows the
molecules of 3 within the bc plane to be > 5 Å apart.
Curie Development Host Fellowship HPMD-CT-2000-00053
(L.J.H.).
Notes and references
‡
Crystal data: 3: C17
16 7 1
H O , M = 332.3, orthorhombic, space group Pna2 ,
A surprisingly large number of possible mechanisms can be
advanced for the formation of an azulene from E-2 and an alkyne.
A number of difficulties attend each of them, the most notable
3
a = 7.3017(7), b = 19.2542(18), c = 11.0940(17) Å, U = 1559.7(3) Å ,
23
T = 293(2) K, Z = 4, D = 1.415 Mg m , crystal dimensions 0.77 3 0.4
c
3 0.5 mm, Mo-Ka radiation, l = 0.71073 Å. Data were collected on a
14
being that phosphorus ylides do not readily react with esters and
Bruker Smart Apex diffractometer and a total of 3019 of the 10842
cyclopentadienyl ylides do not undergo Wittig reaction at all.1 We
,2
2
reflections were unique [R(int) = 0.0195]. Refinement on F , wR2 =
have performed an extensive set of mechanistic investigations15
0.0908 (all data), R1 = 0.0376 for [I > 2s(I)]. CCDC 225162. See
http://www.rsc.org/suppdata/cc/b3/b316759c/ for crystallographic data in
that enabled us to eliminate five of the potential mechanisms to
arrive at our best proposal shown in Scheme 1. The most important
observation was that we could isolate a bis-alkyne adduct 4 from
the reaction mixture in low yield but this compound did not lead to
azulene under any conditions tried, thermal or photochemical. Also
relevant is that adduct E-2 is thermally stable in the absence of
alkyne.
.cif or other electronic format.
1
2
F. Ramirez and S. Levy, J. Am. Chem. Soc., 1957, 79, 6167; F. Ramirez
and S. Levy, J. Am. Chem. Soc., 1957, 79, 67.
W. Johnson, Ylides and Imines of Phosphorus, Wiley & Sons, Inc., New
York, 1990, p. 79.
3 Z. Yoshida, S. Yoneda, Y. Murata and H. Hashimoto, Tetrahedron Lett.,
971, 1523; Z. Yoshida, S. Yoneda, Y. Murata and H. Hashimoto,
1
Tetrahedron Lett., 1971, 1527; Z. Yoshida, S. Yoneda and Y. Murata, J.
Org. Chem., 1973, 38, 3537.
4
5
By single crystal X-ray crystallography: L. J. Higham, P. G. Kelly, H.
Müller-Bunz and D. G. Gilheany, submitted for publication.
(a) T. Tomiyama, I. Tomiyama and M. Yokota (Kotobuki Seiyaku K.
K.); JP2000007611, 2000; (b) W.-G. Friebe, F. Grams, R. Haag, H.
Leinert and J. Dickhaut (Boehringer Mannheim GmbH); US6121322,
2
000; (c) C. S. Slavtcheff, S. R. Barrow, V. D. Kanga, M. C. Cheney and
We suggest therefore that the azulene forming reaction involves
a different mode of reaction of the second alkyne molecule and we
propose: first, its cycloaddition across the C4–C5 bond of the
cyclopentadiene ring, then ring expansion to a seven-membered
ring and, last, Wittig reaction to form azulene.
An alternative scenario is that the Wittig step precedes the ring
expansion. In both cases we reason that the Wittig reaction is now
more likely as a result of both the loss of aromaticity in the five-
membered ring at that stage in the reaction and additionally the
resonance energy gained from formation of the azulene. One
drawback to the proposed mechanism is that the [2+2] addition is
disallowed thermally, although it may be tentatively suggested that
the reaction proceeds in a polar stepwise manner as distinct from a
concerted process.
A. Znaiden (Unilever); WO9503779, 1995; (d) Y. Nakamura, S. Yaoi,
T. Tanaka, Y. Katagiri, S. Ishimaru, S. Takezawa and S. Yoshida
(Fujitsu Limited); US2003129516, 2003; (e) S. E. Estdale, R. Brettle, D.
A. Dunmur and C. M. Marson, J. Mater. Chem., 1997, 7, 391; (f) M.
Findlater, M. Heeney, M. Shkunov, L. Farrand, M. Giles, I. McCulloch,
D. Sparrowe, M. Thompson and S. Tierney (Merck Patent GmbH);
EP1318185, 2003.
D. Lloyd, The Chemistry of Conjugated Cyclic Compounds, John Wiley
&
Houk, J. Am. Chem. Soc., 1979, 101, 251 and references therein; R. L.
Danheiser and D. A. Becker, J. Am. Chem. Soc., 1989, 111, 389; K.
Hafner and K.-P. Meinhardt, Organic Syntheses, Wiley & Sons, Inc.,
New York, Collect. Vol. VII, p. 15, 1990; H. Wakabayashi, P.-W.
Yang, C.-P. Wu, K. Shindo, S. Ishikawa and T. Nozoe, Heterocycles,
1992, 34, 429 and references therein; J. L. Kane, Jr, K. M. Shea, A. L.
Crombie and R. L. Danheiser, Org. Lett., 2001, 3, 1081.
7 An arsenic methylcyclohepta-2,4,6-trienone ylide forms azulenes with
dimethyl acetylenedicarboxylate, but the phosphorus analogues showed
no such reactivity in that reaction: M. Nitta and Y. Mitsumoto, J. Chem.
Soc., Perkin Trans. 1, 2002, 2268.
6
Sons: Chichester, 1989, Ch. 13; D. Mukherjee, L. C. Dunn and K. N.
An investigation into the scope of this reaction and a detailed
mechanistic study15 will be the subject of a future report.
The authors thank the National University of Ireland, Maynooth,
University College Dublin and Enterprise Ireland for Scholarships
(P.G.K. and D.M.C.) and the European Commission for a Marie
8
Two 6-ethynyl-5-alkyl-1,3-dicarboxylates have been reported: T.
Morita, T. Fujita and K. Takase, Bull. Chem. Soc. Jpn., 1980, 53,
1
647.
9
K. Yamamura, N. Kusuhara, Y. Houda, M. Sasabe, H. Takagi and M.
Hashimoto, Tetrahedron Lett., 1999, 40, 6609.
1
0 Y. Lu, D. M. Lemal and J. P. Jasinski, J. Am. Chem. Soc., 2000, 122,
440 and references cited therein.
2
1
1
1 A. W. Hansen, Acta Crystallogr., 1965, 19, 19.
2 H. L. Ammon and M. Sundaralingam, J. Am. Chem. Soc., 1966, 88,
4
794.
1
3 A recent example is [4-(dimethylamino)phenyl]-3-guaiazulenyl-methy-
lium tetrafluoroborate which displays a p-stacking separation of 4.2 Å
that is arranged so as to mutually negate dipole moments: M. Sasaki, M.
Nakamura, T. Uriu, H. Takekuma, T. Minematsu, M. Yoshihara and S.
Takekuma, Tetrahedron, 2003, 59, 505.
1
4 K. B. Becker, Tetrahedron, 1980, 36, 1717; H. J. Bestman, G. Schade,
H. Lutke and T. Monius, Chem. Ber., 1985, 118, 2640; M. Hatanaka, Y.
Himeda and I. Ueda, J. Chem. Soc., Chem. Commun., 1990, 526; W.
Ding, P. Zhang and P. Cao, Tetrahedron Lett., 1987, 28, 81.
5 L. J. Higham, P. G. Kelly, D. M. Corr and D. G. Gilheany, manuscript
in preparation.
1
Scheme 1 Proposed mechanism for the formation of 3.
C h e m . C o m m u n . , 2 0 0 4 , 6 8 4 – 6 8 5
685