5-Nitroisocoumarins from tandem Castro-Stephens coupling-6-endo-dig cyclisation of 2-iodo-3-nitrobenzoic acid and arylethynes and ring-closure of methyl 2-alkynyl-3-nitrobenzoates with electrophiles

Article abstract of DOI:10.1016/j.tet.2006.03.005

Reaction of 2-iodo-3-nitrobenzoic acid with arylalkynyl copper(I) reagent gave 3-aryl-5-nitroisocoumarins. Castro-Stephens coupling was followed by in situ Cu-catalysed ring-closure. ¹H NMR and X-ray crystallography showed the cyclisations to be 6-endo, contrasting with reports of 5-exo cyclisation of analogous 2-iodobenzoate esters with alkynes. Sonogashira couplings of methyl 2-iodo-3-nitrobenzoate with phenylacetylene and with trimethylsilylacetylene gave the corresponding 2-alkynyl-3-nitrobenzoate esters. With HgSO₄, the phenylalkyne underwent 6-endo cyclisation to give 5-nitro-3-phenylisocoumarin. The disubstituted alkyne esters gave 4-phenylselenylisocoumarins with PhSeCl. 5-Nitro-3-phenyl-4-phenylselenylisocoumarin shows significant sterically-driven distortion of the isocoumarin ring. Reaction of methyl 3-nitro-2-phenylethynylbenzoate with ICl gave the 4-iodoisocoumarin. Thus the nitro group tends to direct these electrophile-driven cyclisations towards the 6-endo mode.

Full text of DOI:10.1016/j.tet.2006.03.005

Tetrahedron 62 (2006) 4829–4837
5-Nitroisocoumarins from tandem Castro–Stephens
coupling—6-endo-dig cyclisation of 2-iodo-3-nitrobenzoic
acid and arylethynes and ring-closure of methyl
2-alkynyl-3-nitrobenzoates with electrophiles
Esther C. Y. Woon,a,y Archana Dhami,a Mary F. Mahonb and Michael D. Threadgilla,
*
aDepartment of Pharmacy & Pharmacology, University of Bath, Claverton Down, Bath BA2 7AY, UK
bX-ray Crystallographic Unit, Department of Chemistry, University of Bath, Claverton Down, Bath BA2 7AY, UK
Received 30 January 2006; revised 17 February 2006; accepted 2 March 2006
Available online 30 March 2006
Abstract—Reaction of 2-iodo-3-nitrobenzoic acid with arylalkynyl copper(I) reagent gave 3-aryl-5-nitroisocoumarins. Castro–Stephens
coupling was followed by in situ Cu-catalysed ring-closure. 1H NMR and X-ray crystallography showed the cyclisations to be 6-endo, con-
trasting with reports of 5-exo cyclisation of analogous 2-iodobenzoate esters with alkynes. Sonogashira couplings of methyl 2-iodo-3-nitro-
benzoate with phenylacetylene and with trimethylsilylacetylene gave the corresponding 2-alkynyl-3-nitrobenzoate esters. With HgSO4, the
phenylalkyne underwent 6-endo cyclisation to give 5-nitro-3-phenylisocoumarin. The disubstituted alkyne esters gave 4-phenylselenyliso-
coumarins with PhSeCl. 5-Nitro-3-phenyl-4-phenylselenylisocoumarin shows significant sterically-driven distortion of the isocoumarin
ring. Reaction of methyl 3-nitro-2-phenylethynylbenzoate with ICl gave the 4-iodoisocoumarin. Thus the nitro group tends to direct these
electrophile-driven cyclisations towards the 6-endo mode.
Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction
would be expected to follow the isocoumarin-forming path,
as they lack a reactive a-methylene or a-methyl, they are
not easy to prepare and an alternative route was sought.
As a part of our continuing research on the design and syn-
thesis of heterocyclic compounds as enzyme inhibitors and
potential drugs,1–3 we required a series of 3-substituted-5-
nitroisocoumarins. The synthesis of 5-nitroisocoumarin 3
(Scheme 1) is now well established,4 through conden-
sation of methyl 2-methyl-3-nitrobenzoate 1 with dimethyl-
formamide dimethylacetal, followed by hydrolysis of the
intermediate methyl E-2-(2-dimethylaminoethenyl)-3-nitro-
benzoate 2 and cyclisation catalysed by wet silica. However,
this synthesis cannot be extended to 3-methyl-5-nitroiso-
coumarin 5, as the corresponding initial condensation with
dimethylacetamide dimethylacetal takes an alternate course,
giving a trisubstituted naphthalene 6.5 Compound 5 has been
reported to be synthesised in moderate yield by an alterna-
tive route of Hurtley coupling of 2-bromo-3-nitrobenzoic
acid with pentane-2,4-dione, followed by acyl cleavage
and ring-closure with sodium chloride at very high temper-
ature.6 Although the corresponding condensations of 1 with
dimethylacetals of Ar-substituted N,N-dimethylbenzamides
OMe
CO2Me
Me
NMe2
NO2
i
NO2
1
6
4
iii
ii
CO2Me
Me
NMe2
CO2Me
NMe2
NO2
NO2
2
ii, 53%
from 1
ii, 77%
from 1
O
O
O
O
Me
NO2
NO2
3
5
Scheme 1. Condensation of methyl 2-methyl-3-nitrobenzoate 1 with DMF–
DMA (Ref. 4) and with DMA–DMA (Ref. 5), followed by different modes
of cyclisation. Reagents: (i) HC(OMe)2NMe2, DMF, D; (ii) SiO2, EtOAc
and (iii) MeC(OMe)2NMe2, AcNMe2, D.
Supplementary data associated with this article can be found in the online
version, at doi:10.1016/j.tet.2006.03.005.
* Corresponding author. Tel.: +44 1225 386840; fax: +44 1225 386114;
e-mail: m.d.threadgill@bath.ac.uk
Present address: Department of Pharmaceutical & Biological Chemistry,
School of Pharmacy, University of London, 29-39 Brunswick Square,
London WC1N 1AX, UK.
Several groups have reported the cyclisation of 2-alkynyl-
benzoic acids and 2-alkynylbenzoate esters 7 under electro-
philic conditions, although no examples have a substituent at
y
0040–4020/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2006.03.005

4830
E. C. Y. Woon et al. / Tetrahedron 62 (2006) 4829–4837
the 3-position of the benzoate (corresponding to the
5-position of the target isocoumarins). Moreover, there is
debate about whether the reaction goes 5-exo-dig (giving
ylidenephthalides 8) or 6-endo-dig (giving isocoumarins
9), as shown in Scheme 2 (R3¼H); both are favoured under
Baldwin’s Guidelines. For example, treatment of methyl
2-arylethynylbenzoates with Hg(II) under acidic conditions
is reportedto giveintermediatemercurialsfromwhich3-aryl-
isocoumarins can be isolated by reduction with NaBH4.7,8
Similarly, reaction of the same starting materials with hydro-
gen iodide, electrophilic iodine reagents, bromine, sulfenyl
chlorides and selenyl chlorides gave 3-arylisocoumarins
and their 4-iodo, 4-bromo, 4-arylthio and 4-arylselenyl de-
rivatives, respectively.9–12 Dihydrofuroisocoumarins have
also been synthesised by Ag(I)-promoted 6-endo-dig cyclisa-
tion of the corresponding arylalkynylbenzoate esters.13 In
contrast, methyl 2-ethynylbenzoate undergoes 5-exo-dig
cyclisation with iodine12 and Ag(I)-mediated cyclisation of
2-alkynylbenzoic acids affords the corresponding ylidine-
phthalides.14 Under basic conditions (LiOH), methyl 2-(pent-
1-ynyl)benzoate undergoes exclusive 6-endo-dig cyclisation
whereas methyl 2-(3-hydroxypent-1-ynyl)benzoate gives
a mixture of products from both cyclisation modes.15 Re-
gioisomeric mixtures are also formed during Pd-catalysed
cyclisation of the former alkyne.16 Cherry et al.17 have
shown very recently that treatment of 2-iodobenzoic acid
with allenylstannanes under Pd-catalysed Stille conditions
also gives isocoumarins through 6-endo-dig cyclisation
of the intermediate 2-(3-substituted-allenyl)benzoic acids.
5-exo-Dig cyclisation is also evident at the lower (alcohol)
oxidation level of the intramolecular nucleophilic oxygen
under fluoride-ion catalysis.18
2. Results and discussion
Given the dichotomy of reports for the outcome of the
cyclisations, particularly in the Castro–Stephens tandem
version, we initiated a short study on whether the mode of
cyclisation would be influenced by the presence of the nitro
group ortho to the alkyne. This strongly electron-withdraw-
ing group should influence the electron distribution in the
alkyne and was predicted to favour 6-endo-dig cyclisation
by making the alkyne carbon further from the nitroarene
more electrophilic (Fig. 1). Thus we investigated the tandem
coupling–cyclisation of arylalkynes (as their Cu-acetylides)
with 2-iodo-3-nitrobenzoic acid under Castro–Stephens
conditions.
CO2R1
N
R2
+
O
O –
Figure 1. Proposed influence of the ortho nitro group on the electron-distri-
bution in the alkyne in 2-alkynyl-3-nitrobenzoic acids and 2-alkynyl-3-
nitrobenzoate esters. R1¼H or alkyl, R2¼H, silyl, alkyl and aryl.
The starting material, 2-iodo-3-nitrobenzoic acid 12, was
prepared in two steps from 3-nitrobenzene-1,2-dioic acid
(10, 3-nitrophthalic acid). Decarboxylation/mercuration
with mercury(II) acetate at high temperature in acetic acid
gave the intermediate aryl-mercury 11. Electrophilic re-
placement of the mercury with iodine, using crude 11,
gave 12 conveniently in excellent yield.
To test our hypothesis about the controlling effect of the
nitro group, 12 was exposed to copper(I) phenylacetylide
(prepared from phenylethyne and copper(I) iodide in aq
ethanolic ammonia) under the classical conditions of boiling
pyridine as solvent (Scheme 3). Conventional work-up gave
a high yield of a single cyclisation product, along with
a small amount of a highly non-polar by-product. The latter
was readily identified as 1,4-diphenylbutadiyne 14a,23 de-
rived from an unintended oxidative Glaser homo-coupling
of the alkyne, despite carrying out the reaction under nitro-
R1
O
R3
R2
7
5-exo-dig
6-endo-dig
O
O
O
O
R2
1
R2
R3
R3
gen. Careful examination of the H NMR spectrum of the
8
9
major (cyclisation) product suggested that it was the desired
isocoumarin 13a, rather than the alternative 5-exo-dig prod-
uct 15a (Fig. 2). Firstly, the chemical shift of the signal at
d 7.79 corresponds with the chemical shifts previously
Scheme 2. Possible alternative cyclisation modes of 2-alkynylbenzoic
acids 7 (R1¼OH) and 2-alkynylbenzoate esters (R1¼Oalkyl) via 5-exo-
dig (giving 8) and 6-endo-dig (giving 9) routes. R2¼alkyl, aryl. R3¼H for
previous examples; R3¼NO2 in this paper.
O
While developing their synthesis of arylalkynes from iodo-
arenes and CuI-acetylides, Stephens and Castro19 claimed
that treatment of 2-iodobenzoic acid with Cu–C^C–Ph
gave 3-phenylisocoumarin 9 (R2¼Ph; R3¼H) by Cu-cata-
lysed 6-endo-dig of the initial Castro–Stephens coupling
product 7 (R1¼OH; R2¼Ph; R3¼H). However, this claim
was later withdrawn20 with correction of the characterisation
of the product to be the benzylidine phthalide 8 (R2¼Ph;
R3¼H), resulting from tandem Castro–Stephens coupling
and 5-exo-dig cyclisation. Since then, an isocoumarin has
been synthesised by this method21 and mixtures of these iso-
coumarins and phthalides have been reported from analo-
gous one-pot coupling–cyclisations of 2-iodobenzoic acid
with terminal alkynes under Pd/Zn catalysis.22
CO2H
CO2H
CO2H
I
i, 99%
ii, 65%
O
Hg
NO2
NO2
NO2
iii, 18-74%
O
10
11
12
a: R4' = H
b: R4' = Me
c: R4' = OMe
O
R4'
R4'
+
NO2
R4'
14
13
Scheme 3. Synthesis of 2-iodo-3-nitrobenzoic acid 12 and tandem Castro–
Stephens coupling and 6-endo-dig ring closure to give the 3-aryl-5-nitroiso-
coumarins 13. Reagents: (i) Hg(OAc)2, AcOH, reflux; (ii) I2, KI, AcOH,
0
H2O, reflux and (iii) R4 –C6H4C^CCu, pyridine, reflux.

E. C. Y. Woon et al. / Tetrahedron 62 (2006) 4829–4837
4831
H7
H8
O
O
O
O
O2N
H4
O2N
H1’
15a
13a
Figure 2. Structures of alternative 6-endo-dig (13a) and 5-exo-dig (15a)
cyclisation products, showing the long-range 1H–1H NMR coupling path
in 13a.
Figure 3. X-ray crystal structure of 13b, with crystallographic numbering.
observed by us for the corresponding proton in 5-nitroiso-
coumarin (d 7.39)4 and for 3-phenylisocoumarin (d 6.93)24
better than it does with that observed for the alkene proton
of 3-benzylidenephthalide (d 6.40),25 a close analogue of
15a. Moreover, a small coupling (J 0.8 Hz) was observed
for this proton signal. We have previously observed4 a similar
long-range coupling between H4 and H8 in 5-nitroiso-
coumarin and we propose that it arises from the extended-
numbering scheme are shown in Figure 3. In this structure,
the isocoumarin bicycle is essentially planar. In particular,
the protons and carbons of the extended-W JH–H NMR
5
coupling path are shown to be coplanar, supporting the
assignment of the observed long-range coupling. The 4-
methylphenyl substituent in 13b is only very slightly twisted
out of the isocoumarin plane, with a dihedral angle of 4.8ꢁ;
this closeness to coplanarity presumably reflects the pres-
ence of only one proton on the isocoumarin ortho to the
4-methylphenyl group. However, the nitro group is twisted
some 22.4ꢁ out of the isocoumarin plane, owing to adverse
steric interactions with the peri-hydrogen.
W 5J H4–H8 coupling path shown for 13a in Figure 2, rather
0
than the alternative longer 6J H1 –H7 path in 15a. Moreover,
the IR spectrum of 13a showed an absorption band at
1739 cmꢀ1, which lies in the middle of the range 1730–
1750 cmꢀ1 for six-membered ring lactones but not in the
corresponding range for five-membered ring lactones
(1760–1780 cmꢀ1). These data point strongly towards the
cyclisation having occurred in the 6-endo-dig mode. 2-
Iodo-3-methylbenzoic acid was also treated with the Cu(I)
acetylides of 4-methylphenylethyne and 4-methoxyphenyl-
ethyne under the same conditions. Unexpectedly, the yields
of the cyclised products 13b,c were lower, as were the yields
of the Glaser coupling products 14b,c. However, the NMR
spectra of 13b,c were similar to those of 13a, with H4 reso-
nating at d 7.80 and d 7.66, respectively, indicating the same
mode of cyclisation.
With the mode of Cu-catalysed cyclisation of the 2-(aryl-
alkynyl)-3-nitrobenzoic acids firmly established as 6-endo-
dig, attention was turned to study the mode(s) of cyclisation
of representative methyl 2-(substituted)alkynyl-3-nitrobenz-
oates. To supply the starting material for this part of the
study, 2-iodo-3-nitrobenzoic acid 12 was converted to its
methyl ester 16 in the usual way (Scheme 4). Sonogashira
coupling of this iodoarene with phenylethyne and with tri-
methylsilylethyne gave the disubstituted alkynes 17a,d.
The yield of 17a was excellent but coupling of the more
bulky alkyne with the sterically demanding ortho,ortho0-
disubstituted iodoarene 16 gave a much lower yield (30%).
In the latter case, the difficulty of the coupling was
To confirm unequivocally that the products were indeed the
isocoumarins, the structure of 13b was established by X-ray
crystallography. The structure and X-ray crystallographic
CO2R1
I
CO2Me
CO2Me
ii, 30-80%
NO2
NO2
R2
NO2
18
12: R1 = H
17a: R1 = Ph
14a
i, 95%
16: R1 = Me
17d: R1 = SiMe3
17e: R1 = H
iii, 9% or
iv, 72%
R2 = SiMe3, H, 25-28%
v vi
vii
R2 = Ph, SiMe3, 47-82%
R2 = Ph, 95%
R2 = Ph, 81%
viii, 78%
O
O
O
O
CO2Me
Me
O
O
R2
NO2
O
NO2
13a
NO2
21
I
NO2 SePh
19
22a: R2 = Ph
22d: R2 = SiMe3
O
O
OMe
Me
NO2
20
Scheme 4. Sonogashira couplings of methyl 2-iodo-3-nitrobenzoate 16 and electrophile-driven cyclisations of methyl 2-alkynyl-3-nitrobenzoates 17a,d,e.
Reagents: (i) MeOH H2SO4; (ii) HC^CPh or HC^CSiMe3, (Ph3P)2PdCl2, Pr2iNH THF; (iii) Bu4NF, THF; (iv) AgOTf, acetone, water, CH2Cl2;
(v) HgSO4, H2SO4, Me2CO, D; (vi) ICl, CH2Cl2; (vii) PhSeCl, CH2Cl2 and (viii) HCO2H, Pd(OAc)2, Et3N, DMF.

4832
E. C. Y. Woon et al. / Tetrahedron 62 (2006) 4829–4837
illustrated by the isolation of a 39% yield of the dehalogen-
ated product 18, arising from decomposition of the interme-
diate arylpalladium, which had failed to couple with the
alkyne. Interestingly, despite the presence of copper(I) in
the reaction mixture, no cyclised products were formed,
although some runs with phenylethyne also gave 1,4-di-
phenylbutadiyne 14a, the product of Glaser-like homodi-
merisation of the starting alkyne. Attempted desilylation of
17d under a variety of basic conditions (e.g., potassium car-
bonate/methanol) caused extensive decomposition; indeed
the classical fluoride ion-mediated desilylation gave only
a maximum yield of 9% of 17e. Removal of trimethylsilyl
protection from alkynes with silver(I) nitrate is widely re-
ported to be efficient if cyanide ion is added to break up
the initially formed alkynylsilver(I).26–28 However, Orsini
et al.29 have recently developed a selective desilylation of
1-trimethylsilyl-2-alkylalkynes with silver(I) triflate in a bi-
phasic solvent system at room temperature, which obviates
the use of large molar excesses of silver and of cyanide.
Adapting this method to the current application, prolonged
heating of 17d with silver(I) triflate in a mixture of methanol,
water and dichloromethane gave good yields (>70%) of the
required alkyne 17e. We attribute the need for the higher
temperature and much longer reaction times (days, rather
than hours) to the highly electron-deficient nature of the
starting trimethylsilylalkyne.
philic attack, giving intermediate 24. Hydrolysis would then
afford the major product, the ketone 19. It is not clear whether
the minor side-product 20 is formed from 19 or directly from
intermediate 24.
OMe
CO2Me
O
NO2
R
NO2
23
Hg+
17d: R = SiMe3
17e: R = H
+ OMe
CO2Me
Me
:OH2
O
Hg+
–
NO2
O
NO2
24
19
O
O
OMe
Me
NO2
20
Scheme 5. Proposed routes for the formation of 19 and 20 through 5-exo
attack of the neighbouring ester carbonyl oxygen; the polarisation of the
alkyne is driven by the mercury cation, rather than by the ortho-nitrophenyl
group in this case.
Each of the three methyl 2-alkynyl-3-nitrobenzoates 17a,d,e
(carrying the diverse substituents Ph, SiMe3 and H, re-
spectively) was treated with three different electrophiles,
to investigate whether or not cyclisation would occur and
whether such cyclisation would be 5-exo or 6-endo (Scheme
4). Firstly, treatment of 17a with mercury(II) sulfate under
acidic conditions afforded an almost quantitative yield of
a single product, the isocoumarin 13a, which was identical
to the material prepared by the Castro–Stephens one-pot
method. 6-endo Cyclisation was relatively unsurprising in
this case, in view of the exclusive formation of 3-phenyl-
isocoumarin from methyl 2-phenylethynylbenzoate under
the same conditions, as reported by Nagarajan and
Balasubramanian.7 Similar treatment of the analogous
trimethylsilylalkyne 17d, however, gave only methyl 2-ace-
tyl-3-nitrobenzoate 19, in modest yield after chromato-
graphy. The same ketone 19 was also isolated from the
reaction of 17e with mercury(II) sulfate and acid; the NMR
spectrum also indicated the presence of a trace of the iso-
meric pseudoester 20.30 Formation of 19 from 17d and 17e
clearly involves attack of an oxygen nucleophile on the al-
kyne carbon nearest to the benzene ring, in contrast to the
formation of 13a from 17a, which must arise from 6-endo
attack of the ester carbonyl oxygen on the carbon remote
from the substituted ring. Thus the alkyne must be polarised
in the opposite sense. Scheme 5 shows a mechanistic ration-
alisation for this change in reactivity. By analogy with the
mechanism of the silver(I)-mediated desilylations,29,31 we
proposethattransmetallationof17doccurstoformanalkynyl-
mercury species, such as 23. This intermediate could also be
formed by direct metallation of 17e. In intermediate 23, it
is likely that the polarisation of the alkyne caused by co-
ordination to the mercury would over-ride the opposite polar-
isation induced by the two ortho-electron-withdrawing
substituents (nitro, carbonyl) on the benzene ring. The ester
carbonyl oxygen is then located oppositely for 5-exo nucleo-
The set of three alkynes 17a,d,e was also treated with the
electrophilic iodine reagent iodine monochloride at ambient
temperature in dichloromethane. Only the phenylalkyne 17a
gave an identifiable product, affording 4-iodo-5-nitro-3-
phenylisocoumarin 21 in high yield (Scheme 4). No isocou-
marins or isobenzofuranones could be identified in the NMR
spectra of the crude mixtures of products formed from the
trimethylsilylalkyne 17d or from the monosubstituted
alkyne 17e. The structure of 21 was confirmed as being
the isocoumarin product of 6-endo cyclisation by two
methods. Firstly, the IR spectrum showed an absorption at
1736 cmꢀ1, corresponding to a six-membered ring lactone
of this type. Secondly, palladium-catalysed reductive deiodi-
nation of 21 with formic acid (by the general method of
Rossi et al.12) gave an excellent yield of 5-nitro-3-phenyliso-
coumarin 13a, identical to samples prepared by the one-pot
Castro–Stephens method and the Hg(II)-mediated cyclisation
of 17a.
Finally, the reactions of the electrophile phenylselenyl chlo-
ride with the alkynes 17a,d,e were investigated (Scheme 4).
As with the reaction with iodine monochloride, no iso-
coumarins or isobenzofuranones could be identified as prod-
ucts of the reaction with the monosubstituted alkyne 17e.
However, the disubstituted alkynes 17a,d formed the corre-
sponding 4-phenylselenylisocoumarins 22a,d in moderate-
to-good yields, through 6-endo cyclisation. Again, the IR
spectra indicated 6-membered ring lactones, with bands at
1733 cmꢀ1 for both isocoumarins. In view of the lack of suit-
able methods for reductive removal of the phenylselenyl
group to provide material for comparison with 13a synthe-
sised previously by three independent routes, an X-ray crystal
structure determination was carried out for 5-nitro-3-phenyl-
4-phenylselenylisocoumarin 22a. Large bright orange-red

E. C. Y. Woon et al. / Tetrahedron 62 (2006) 4829–4837
4833
Figure 4. A. X-ray crystal structure of 22a, with crystallographic numbering. B. Axial view of intermolecular and intramolecular p-stacking in the crystal of
22a. C. Side view of intermolecular and intramolecular p-stacking of two molecules in the crystal of 22a. D. View of single molecule of 22a in the plane of the
isocoumarin carbocyclic ring. E. View of single molecule of 22a along the Se–C bond. Grey¼C, white¼H, blue¼N, red¼O and orange¼Se.
crystals were formed from ethyl acetate. The structure and
X-ray crystallographic numbering scheme are shown in
Figure 4A. The most striking observation in this crystal struc-
ture is the intermolecular and intramolecular p-stacking of
all three benzene rings in the molecule. Figure 4B shows
four molecules in two parallel stacks, viewed from the axis
of the stacking. The nature of the stacking is shown in the
side view in Figure 4C, with the C–Ph and SePh rings stacked
intramolecularly; these then stack intermolecularly with the
carbocyclic ring of the isocoumarins. This isocoumarin
also displays interesting conformational features within the
molecule, as shown in Figure 4. The three adjacent substitu-
ents, phenyl, phenylselenyl and nitro, at the 3-, 4- and 5-po-
sitions, respectively, occupy very crowded regions of space.
In particular, there is a severe peri interaction between the ni-
tro and phenylselenyl groups. As for the structure of 13b, the
nitro group in 22a is twisted out of the plane–plane subtended
by atoms C2–C8 and C10 of the isocoumarin, by 36.9ꢁ. How-
ever, the severe stericcrowding has a more profound effect, in
that the pyranone ring of the isocoumarin of 22a is forced out
by interdigitating intra- and intermolecular stacking of the
aromatic rings. The intermolecular centroid–centroid dis-
˚
tance between the aromatic rings is 3.8 A, while the compa-
rable intermolecular aromatic–aromatic distances average
˚
4.1 A. This latter value reflects the fact that the p-stacking
is offset in the intermolecular case.
3. Conclusions
In this paper, we have reported one-pot Castro–Stephens
couplings of arylalkynes with 2-iodo-3-nitrobenzoic acid
12, followed by Cu+-catalysed cyclisation of the intermedi-
ate 2-arylalkynyl-3-nitrobenzoic acids in situ to give exclu-
sively 3-aryl-5-nitroisocoumarins 13a–c. These results
contrast with the results reported by Stephens and Castro
for the analogues lacking the nitro group, which cyclised
in the 5-exo mode to give 5-benzylidene-isocoumarins.20
We interpret this change in regiochemistry of cyclisation
as being due to the nitro group inducing polarisation of the
alkyne, making the remote sp-carbon more electrophilic as
shown in Figure 1. Similarly, Hg(II)-catalysed cyclisation
of pre-formed methyl 2-phenylalkynyl-3-nitrobenzoate 17a
also proceeded in the 6-endo mode to give 2-nitro-3-phenyl-
isocoumarin 13a, again reflecting this polarisation of the
alkyne. Cyclisations of methyl 3-nitro-2-phenylethynylbenz-
oate 17a with iodine monochloride and with phenylselenyl
chloride also followed the 6-endo route to give the isocou-
marins 21 and 22a, respectively, as did cyclisation of methyl
3-nitro-2-trimethylsilylethynylbenzoate 17d with phenyl-
selenyl chloride, affording the isocoumarins 22d. Thus the
regiochemistry of these cyclisations is also likely to be under
˚
of the plane described; above, such that C-1 lies some 0.19 A
above same. This effect is demonstrated even more clearly by
˚
the position of the selenium atom at 0.93 A above this mean
plane, as depicted in Figure 4D (wherein the structure is
viewed from the plane of the benzene ring) and in
Figure 4E (in which the structure is viewed along the Se–C
bond vector). Whereas the 3-aryl substituent was only
twisted out of the isocoumarin plane in 13b by <5ꢁ, the
corresponding 3-phenyl in 22a is twisted out of the isocou-
marin mean plane by 36.5ꢁ. The presence of the adjacent
bulky phenylselenyl is presumably responsible for this
greater lack of coplanarity. The gross structure is dominated

4834
E. C. Y. Woon et al. / Tetrahedron 62 (2006) 4829–4837
the control of the nitro group. In contrast, the formation of
methyl 2-acetyl-3-nitrobenzoate 19 by treatment of 17d,e
with Hg(II) suggests that a 5-exo cyclisation may have been
driven by the change in electron-distribution caused by the
formation of an intermediate alkynylmercury complex.
These studies extend the understanding of electrophile-
driven cyclisations of 2-alkynylbenzoic acids and 2-alkynyl-
benzoate esters to the previously unreported cases where
a powerful electron-withdrawing group is present, influenc-
ing the electron-distribution of the alkyne. This understand-
ing will be useful in predicting modes of cyclisation in the
synthesis of more complex isocoumarins.
tion and chromatography (hexane/EtOAc 12:1) yielded 14a
ꢁ
33
(105 mg, 5%) as a white solid: mp 85–86 C (lit. mp
88 C); IR nmax 2158 cmꢀ1; H NMR d 7.26–7.35 (6H, m,
1
ꢁ
2ꢂ30,40,50-H3), 7.45–7.50 (4H, m, 2ꢂ20,60-H2); MS (EI)
m/z 202.0784 (M) (C16H10 requires 202.0783). Further
elution yielded 13a (2.0 g, 74%) as yellow crystals: mp
1
142–143 C; IR nmax 1341, 1525, 1739 cmꢀ1; H NMR
ꢁ
d 7.50–7.53 (3H, m, 30,40,50-H3), 7.62 (1H, t, J¼7.8 Hz, 7-
H), 7.79 (1H, d, J¼0.8 Hz, 4-H), 7.93–7.97 (2H, m, 20,60-
H2), 8.51 (1H, dd, J¼8.2, 1.2 Hz, 6-H), 8.65 (1H, ddd,
J¼8.2, 1.2, 0.8 Hz, 8-H); MS (EI) m/z 267.0532 (M)
(C15H9NO4 requires 267.0532), 237 (MꢀNO), 184, 150.
4.1.3. 5-Nitro-3-phenylisocoumarin (13a). Method B.
Compound 17a (2.5 g, 8.9 mmol) was boiled under reflux
with HgSO4 (3.0 g, 10 mmol) and concd H2SO4 (4 mL) in
acetone (80 mL) for 48 h. The evaporation residue was
extracted with CHCl3. Evaporation and chromatography
(hexane/Et2O 9:1) gave 13a (2.3 g, 95%) as yellow crystals,
with data as above.
4. Experimental
4.1. General
IR spectra were obtained as KBr discs and NMR spectra
were obtained using solutions in CDCl3, unless otherwise
stated. Mass spectra were obtained using fast atom bombard-
ment in the positive ion mode, unless otherwise stated. Solu-
tions in organic solvents were dried with MgSO4. Solvents
were evaporated under reduced pressure. Mps were obtained
using a Kofler–Galen hot stage microscope and are uncor-
rected.
4.1.4. 5-Nitro-3-phenylisocoumarin (13a). Method C.
HCO2H (30 mg, 0.66 mmol) was added to a degassed mix-
ture of 21 (130 mg, 0.33 mmol), Et3N (100 mg, 1.0 mmol),
Pd(OAc)2 (6.6 mg, 7 mmol) and Ph3P (13.2 mg, 13 ꢁmmol)
in dry DMF (10 mL). The mixture was stirred at 60 C for
4.5 h under Ar before being poured into water (50 mL). Ex-
traction (EtOAc), drying, evaporation and chromatography
(EtOAc/hexane 1:9) gave 13a (69 mg, 78%) as pale yellow
crystals, with data as above.
4.1.1. 2-Hydroxymercuri-3-nitrobenzoic acid (11) and
2-iodo-3-nitrobenzoic acid (12). 3-Nitrobenzene-1,2-dicar-
boxylic acid 10 (21.1 g, 100 mmol) in hot aq NaOH (10%,
80 mL) was added to Hg(OAc)2 (35.0 g, 110 mmol) in hot
AcOH (5 mL) and water (70 mL). The mixture was boiled
under reflux for 70 h, then filtered. The precipitate was
washed (H2O, then EtOH) and dried to give 11 (36.2 g,
99%) as a cream solid. Aq HCl (2 M, 12 mL) was added
slowly to a boiling solution of 11 (36.2 g, 100 mmol) in aq
NaOH (3.5%, 500 mL) and the mixture was allowed to
4.1.5. 3-(4-Methylphenyl)-5-nitroisocoumarin (13b) and
1,4-di(4-methylphenyl)-1,3-butadiyne (14b). Copper(I)
4-methylphenylacetylide was prepared and treated with 12,
as for the synthesis of 13a (Method A), to give 14b (0.4%)
33
ꢁ
as pale yellow crystals: mp 180–181 C (lit.
mp
183 C); IR nmax 2130 cmꢀ1
;
1H NMR d 2.36 (6H, s,
ꢁ
ꢁ
cool to 25 C. AcOH (3 mL) was then added, followed by
2ꢂMe), 7.14 (4H, d, J¼8.1 Hz, 2ꢂ30,50-H2), 7.41 (4H, d,
J¼8.1 Hz, 2ꢂ20,60-H2); MS (EI) m/z 231.1141 (M+H)
(C18H15 requires 231.1174). Further elution gave 13b
KI (19.0 g, 114 mmol) and I2 (29.0 g, 114 mmol) in water
(30 mL). The mixture was boiled under reflux for 24 h,
cooled and neutralised with aq NaOH, before being filtered
and acidified with aq HCl (9 M). The precipitate was col-
lected, dried and recrystallised (EtOH) to give 12 (19.1 g,
ꢁ
(18%) as yellow crystals: mp 161–162 C; IR nmax 1321,
1511, 1618, 1737 cmꢀ1; H NMR d 2.42 (3H, s, Me), 7.29
1
(2H, d, J¼8.2 Hz, 30,50-H2), 7.56 (1H, t, J¼8.2 Hz, 7-H),
7.80 (1H, s, 4-H), 7.81 (2H, d, J¼8.2 Hz, 20,60-H2), 8.46
(1H, dd, J¼8.2, 1.2 Hz, 6-H), 8.60 (1H, dt, J¼8.2, 1.2 Hz,
8-H); MS (EI) m/z 282.0762 (M+H) (C16H11NO4 requires
282.0766).
32
ꢁ
65%) as yellow crystals: mp 203–204 C (lit. mp 204–
1H
205.5 C); IR nmax 1375, 1542, 1712, 2800 cmꢀ1
;
ꢁ
NMR ((CD3)2SO) d 7.66 (t, J¼7.8 Hz, 1H, 5-H), 7.79 (dd,
J¼7.8, 1.7 Hz, 1H, 4-H), 7.92 (dd, J¼7.8, 1.7 Hz, 1H,
6-H); MS m/z 293.9250 (M+H) (C7H5INO4 requires
293.9263).
4.1.6. 3-(4-Methoxyphenyl)-5-nitroisocoumarin (13c)
and 1,4-di(4-methoxyphenyl)-1,3-butadiyne (14c).
Copper(I) 4-methoxyphenylacetylide was prepared and
treated with 12, for the synthesis of 13a (Method A), ꢁto
4.1.2. 5-Nitro-3-phenylisocoumarin (13a) and 1,4-
diphenyl-1,3-butadiyne (14a). Method A. CuI (7.5 g,
39 mmol) in aq NH3 (35%, 100 mL) was slowly added to
phenylethyne (4.0 g, 39 mmol) in EtOH (200 mL). The mix-
ture was stirred for 15 min. The precipitate was collected,
washed withꢁ water (5ꢂ), EtOH (5ꢂ) and Et2O (5ꢂ) and
dried at 50 C (20 torr) for 2 h to afford phenylethynyl-
copper(I) (4.7 g, 73%) as a bright canary-yellow solid. Com-
pound 12 (3.0 g, 10 mmol) and phenylethynylcopper(I)
(1.7 g, 10 mmol) were boiled under reflux in dry pyridine
(100 mL) for 6 h under N2. The mixture was poured into
water (300 mL) and extracted with (Et2O). The extract was
washed (aq HCl (2 M), aq NaHCO3 (5%), water). Evapora-
give 14c (1%) as pale yellow crystals: mp 148–149 C
1
(lit.34 mp 149 C); IR nmax 2145 cmꢀ1; H NMR d 3.82
ꢁ
(6H, s, 2ꢂMe), 6.85 (4H, d, J¼8.9 Hz, 2ꢂ30,50-H2), 7.46
(4H, d, J¼8.9 Hz, 2ꢂ20,60-H2); MS (EI) m/z 262.0995 (M)
(C18H14O2 requires 262.0994), 247 (MꢀMe), 219
(MꢀMeꢀCO). Further elution gave 13c (29%) as yellow
ꢁ
crystals: mp 241–242 C; IR nmax 1346, 1511, 1620,
1738 cmꢀ1
;
1H NMR d 3.88 (3H, s, Me), 6.99 (2H, d,
J¼9.0 Hz, 30,50-H2), 7.54 (1H, t, J¼8.2 Hz, 7-H), 7.76
(1H, s, 4-H), 7.88 (2H, d, J¼9.0 Hz, 20,60-H2), 8.46 (1H,
dd, J¼8.2, 1.2 Hz, 6-H), 8.59 (1H, dd, J¼8.2, 1.2 Hz,

E. C. Y. Woon et al. / Tetrahedron 62 (2006) 4829–4837
4835
ꢁ
8-H); MS (EI) m/z 297.0639 (M) (C16H11NO5 requires
297.0637), 277 (MꢀNO), 135.
solid: mp 78–80 C; IR (film) nmax 1351, 1532, 1738,
1
2219 cmꢀ1; H NMR d 3.67 (1H, s, C^CH), 3.90 (3H, s,
Me), 7.48 (1H, t, J¼7.8 Hz, 5-H), 7.91 (1 H, dd, J¼7.8,
1.2 Hz, 4-H), 8.00 (1H, dd, J¼7.8, 1.2 Hz, 6-H); MS (EI)
m/z 205.0381 (M) (C10H7NO4 requires 205.0375), 191
(MꢀCH2), 175 (MꢀNO), 160 (MꢀNOꢀMe).
4.1.7. Methyl 2-iodo-3-nitrobenzoate (16). Compound 12
(4.0 g, 14 mmol) was boiled under reflux with MeOH
(120 mL) and concd H2SO4 (3 mL) for 48 h, then poured
into ice-water. The precipitate was filtered, dried and recrys-
4.1.11. Methyl 2-ethynyl-3-nitrobenzoate (17e). Method
B. Compound 17d (250 mg, 0.9 mmol) was stirred with
AgOTf (23 mg, 0.09 mmol) in a mixture of acetone
(6.0 mL), water (1.5 mL) and CH2Cl2 (10.5 mL) for 7 d.
Satd aq ammonium chloride (2 mL) was added and the mix-
ture was extracted thrice with CH2Cl2. Evaporation and
chromatography (EtOAc/hexane 1:6) gave 17e (132 mg,
72%) with data as above.
tallised (MeOH) to give 16 (4.0 g, 95%) as yellow crystals:
ꢁ
35
ꢁ
mp 65–66 C (lit. mp 62–64 C); IR nmax 1351, 1533,
1705 cmꢀ1 1H NMR d 3.99 (3H, s, Me), 7.54 (1H, t,
;
J¼7.8 Hz, 5-H), 7.70 (1H, dd, J¼7.8, 1.7 Hz, 4-H), 7.77
(1H, dd, J¼7.8, 1.7 Hz, 6-H); MS (EI) m/z 306.9347 (M)
(C8H6INO4 requires 306.9342), 276 (MꢀOMe).
4.1.8. Methyl 3-nitro-2-(2-phenylethynyl)benzoate (17a)
and methyl 3-nitrobenzoate (18). Compound 16 (3.0 g,
9.8 mmol) was stirred with (Ph3P)2PdCl2 (300 mg,
0.4 mmol) and CuI (400 mg, 2.1 mmol) in dry THF
4.1.12. Methyl 2-acetyl-3-nitrobenzoate (19). Method A.
Compound 17d (110 mg, 0.4 mmol) was boiled under reflux
with HgSO4 (148 mg) and concd H2SO4 (0.1 mL) in acetone
(10 mL) for 48 h. The evaporation residue was extracted
with CHCl3. Evaporation and chromatography (toluene)
i
ꢁ
(120 mL) and dry Pr2 NH (40 mL) at 45 C for 30 min under
Ar. Phenylethyne (1.5 g, 15 mmol) was added during 30 min
and the mixture was stirred for 48 h. Filtration (CeliteÒ),
evaporation and chromatography (hexane/CH2Cl2 1:1)
ꢁ
gave 15 (25 mg, 28%) as a white solid: mp 81–82 C
1
(lit.37 mp 81.5 C); H NMR d 2.71 (3H, s, ArCOMe),
ꢁ
gave 17a (2.2 g, 80%) as reddish-brown crystals: mp 59–
3.93 (3H, s, OMe), 7.65 (1H, t, J¼8.2 Hz, 5-H), 8.33 (1H,
dd, J¼8.2, 1.2 Hz, 4-H or 6-H), 8.37 (1H, dd, J¼8.2,
1.2 Hz, 6-H or 4-H); 13C NMR d 29.66, 53.09, 128.64,
129.47, 129.56, 136.20, 140.19, 146.04, 164.47, 200.03.
1
60 C; IR nmax 1343, 1527, 1737, 2219 cmꢀ1; H NMR
ꢁ
d 4.01 (3H, s, Me), 7.36–7.41 (3H, m, Ph 3,4,5-H3), 7.49
(1H, t, J¼7.8 Hz, 5-H), 7.57–7.62 (2H, m, Ph 2,6-H2),
8.04 (1H, dd, J¼7.8, 1.2 Hz, 4-H), 8.10 (1H, dd, J¼7.8,
1.2 Hz, 6-H); 13C NMR d 52.81, 81.79, 102.61, 117.62,
122.06, 126.81, 127.63, 128.32, 129.44, 131.97, 133.56,
134.94, 151.87, 165.27; MS (EI) m/z 282.0757 (M+H)
(C16H12NO4 requires 282.0766), 266 (MꢀMe), 250
(MꢀOMe), 222 (MꢀCO2Me). Further elution gave 18
4.1.13. Methyl 2-acetyl-3-nitrobenzoate (19). Method B.
Compound 17e was treated with HgSO4, as for Method A,
to give 19 (25%) with properties as above. Also identified
in trace amounts in the NMR spectrum was 3-methoxy-3-
methyl-4-nitroisobenzofuran-1-one 20:30 1H NMR d 2.05
(3H, s, CMe), 3.22 (3H, s, OMe), 7.43 (1H, m, 6-H), 8.22
(1H, dd, J¼8.4, 0.9 Hz, 5-H or 7-H), 8.47 (1H, dd, J¼8.8,
0.9 Hz, 7-H or 5-H).
36
ꢁ
(500 mg, 28%) as yellow crystals: mp 78–79 C (lit. mp
78 C); 1H NMR d 3.99 (3H, s, Me), 7.66 (1H, t,
ꢁ
J¼7.8 Hz, 5-H), 8.37 (1H, ddd, J¼7.8, 2.4, 1.2 Hz, 4-H),
8.42 (1H, ddd, J¼7.8, 2.4, 1.2 Hz, 6-H), 8.87 (1H, t,
J¼2.4 Hz, 2-H). From some runs, 1,4-diphenylbutadiyne
14a was also isolated.
4.1.14. 4-Iodo-5-nitro-3-phenylisocoumarin (21). Com-
pound 17a (114 mg, 0.37 mmol) was stirred with ICl
(90 mg, 0.55 mmol) in CH2Cl2 (1.0 mL) for 2 h in the dark.
The mixture was diluted with Et2O (50 mL), washed with
aq Na2S2O3 and dried. Evaporation and chromatography
(EtOAc/hexane 1:4) gave 21 (130 mg, 81%) as pale yellow
crystals. A sample was recrystallisꢁed (EtOAc/hexane) to give
a yellow powder: mp 154–156 C; IR nmax 1347, 1522,
4.1.9. Methyl 3-nitro-2-(2-trimethylsilylethynyl)benzoate
(17d) and methyl 3-nitrobenzoate (18). Compound 16
(3.0 g, 9.8 mmol) in dry THF (120 mL) was added to
(Ph3P)2PdCl2 (300 mg, 0.4 mmol) and CuI (400 mg,
2.1 mmol) in dry Pr2i NH (40 mL) and the mixture was
ꢁ
1736 cmꢀ1 13H NMR d 7.50 (3H, m, Ph 3,4,5-H3), 7.66
;
stirred at 45 C for 30 min under Ar. Trimethylsilylethyne
(1.1 g, 11 mmol) was added during 30 min. The mixture
(1H, t, J¼7.8 Hz, 7-H), 7.75 (2H, m, Ph 2,6-H2), 8.08 (1H,
dd, J¼7.8, 1.2 Hz, 6-H), 8.54 (1H, dd, J¼7.8, 1.2 Hz, 8-H);
13C NMR d 61.58, 123.47, 128.35, 128.63, 130.46, 130.92,
131.21, 131.95, 133.25, 134.90, 150.06, 158.95, 159.55;
MS (EI) m/z 392 (MꢀH), 265 (MꢀHI). Found: C, 46.1; H,
2.15; N, 3.69; C15H8INO4 requires C, 45.83; H, 2.05; N,
3.56%.
Ò
ꢁ
was stirred for 72 h at 45 C. Filtration (Celite ), evapora-
tion and chromatography (hexane/CH2Cl2 3:2) gave 17d
(810 mg, 30%) as a reddish-brown oil: IR (film) nmax
1351, 1532, 1738, 2219 cmꢀ1 1H NMR d 0.28 (9H, s,
;
SiMe3), 3.96 (3H, s, OMe), 7.49 (1H, t, J¼7.8 Hz, 5-H),
7.96 (1H, dd, J¼7.8, 1.6 Hz, 4-H), 8.01 (1H, dd, J¼7.8,
1.6 Hz, 6-H); MS (EI) m/z 278.2361 (M+H) (C13H16NO4Si
requires 278.2357). Further elution gave 18 (690 mg,
39%), with data as above.
4.1.15. 5-Nitro-3-phenyl-4-phenylselenylisocoumarin
(22a). Compound 17a (100 mg, 0.36 mmol) was stirred
with PhSeCl (100 mg, 0.53 mmol) in CH2Cl2 (5.0 mL)
under N2 for 4 h. The mixture was washed (aq NaHCO3)
and dried. Evaporation and chromatography (EtOAc/hexane
1:7) gave 22a (72 mg, 47%) as yellow crystals. A sample
4.1.10. Methyl 2-ethynyl-3-nitrobenzoate (17e). Method
A. Bu4NF (1.0 M in THF, 5.0 mL, 5.0 mmol) was added to
17d (400 mg,ꢁ 1.4 mmol) in THF (50 mL) and water
ꢁ
(2.5 mL) at 0 C and the mixture was stirred at 20 C for
was recrystallised (EtOAc) to give orange-red crystals: mp
1
ꢁ
16 h. The mixture was diluted with Et2O and washed with
satd aq NH4Cl and water. Evaporation and chromatography
(hexane/EtOAc 7:3) gave 17e (28 mg, 9%) as a pale yellow
188–190 C; IR nmax 1354, 1533, 1733 cmꢀ1; H NMR
d 6.77 (2H, ca. d, J¼8.6 Hz, SePh 2,6-H2), 6.95 (2H, ca. t,
J¼ca. 8 Hz, SePh 3,5-H2), 7.07 (1H, tt, J¼7.4, 1.2 Hz,

4836
E. C. Y. Woon et al. / Tetrahedron 62 (2006) 4829–4837
SePh 4-H), 7.32 (2H, ca. t, J¼ca. 7.4 Hz, CPh 3,5-H2), 7.39
(1H, tt, J¼7.1, 1.5 Hz, CPh 4-H), 7.58 (2H, ca. d, J¼ca.
7.5 Hz, CPh 2,6-H2), 7.64 (1H, t, J¼8.0 Hz, 7-H), 8.14
(1H, dd, J¼8.0, 1.5 Hz, 6-H), 8.55 (1H, dd, J¼8.0, 1.5 Hz,
8-H); 13C NMR d 102.24, 123.12, 127.61, 127.85, 128.04,
128.90, 130.34, 130.60, 131.05, 131.70, 132.18, 133.26,
133.68, 134.38, 137.27, 148.35, 159.90, 160.65; MS (EI)
m/z 422 (M). Found: C, 59.8; H, 3.06; N, 3.32; C21H13NO4Se
requires C, 59.73; H, 3.10; N, 3.32.
crystals of 22a. We also thank the EPSRC Mass Spectrome-
try Centre, University of Wales, Swansea, for some of the
mass spectra.
References and notes
1. Goodyer, C. L. M.; Chinje, E. C.; Jaffar, M.; Stratford, I. J.;
Threadgill, M. D. Bioorg. Med. Chem. 2003, 11, 4189–4206.
2. Chatterjee, P. K.; Chatterjee, B. E.; Pedersen, H.; Sivarajah, A.;
McDonald, M. C.; Mota-Filipe, H.; Brown, P. A. J.; Stewart, K.
N.; Cuzzocrea, S.; Threadgill, M. D.; Thiemermann, C. Kidney
Int. 2004, 65, 499–509.
3. Frixa, C.; Mahon, M. F.; Thompson, A. S.; Threadgill, M. D.
Org. Biomol. Chem. 2003, 1, 306–317.
4. McDonald, M. C.; Mota-Filipe, H.; Wright, J. A.; Abdelrah-
man, M.; Threadgill, M. D.; Thompson, A. S.; Thiemermann,
C. Br. J. Pharmacol. 2000, 130, 843–850.
4.1.16. 5-Nitro-4-phenylselenyl-3-trimethylsilyliso-
coumarin (22d). Compound 17d (110 mg, 0.4 mmol) was
stirred with PhSeCl (114 mg, 0.6 mmol) in CH2Cl2
(3.0 mL) under N2 for 24 h. The mixture was washed (aq
NaHCO3) and dried. Evaporation and chromatography
(EtOAc/hexane 1:4) gave 22d (136 mg, 82%) as a yellow
solid. A sample was recrystallised (toluene) to give a yellow
1
powder: mp 105–107 C; IR nmax 1733 cmꢀ1; H NMR
ꢁ
d 0.34 (9H, s, SiMe3), 7.00 (2H, m, Ph 2,6-H2), 7.14 (3H, m,
Ph 3,4,5-H3), 7.38 (1H, t, J¼7.8 Hz, 7-H), 7.80 (1H, dd,
J¼7.8, 1.4 Hz, 6-H), 8.50 (1H, dd, J¼7.8, 1.4 Hz, 8-H);
13C NMR d 0.34, 110.26, 124.04, 126.67, 127.84, 128.30,
129.10, 129.36, 130.15, 130.33, 132.96, 133.73, 160.47,
175.22; MS (EI) m/z 419.0096 (M) (C18H17NO80Se28Si re-
quires 419.0092), 403 (MꢀCH4). Found: C, 51.8; H, 4.15;
N, 3.33; C18H17NOSeSi requires C, 51.67; H, 4.10; N, 3.35.
5. Wong, S.-M.; Shah, B.; Shah, P.; Butt, I. C.; Woon, E. C. Y.;
Wright, J. A.; Thompson, S.; Upton, C.; Threadgill, M. D.
Tetrahedron Lett. 2002, 43, 2299–2302.
6. Ames, D.; Ribiero, O. J. Chem. Soc., Perkin Trans 1 1976,
1073–1078.
7. Nagarajan, A.; Balasubramanian, T. R. Indian J. Chem., Sect. B
1987, 26, 917–919.
8. Larock, R. C.; Harrison, L. W. J. Am. Chem. Soc. 1984, 106,
4218–4227.
4.1.17. X-ray crystallography. Compound 13b. Crystal
˚
9. Yao, T.; Larock, R. C. J. Org. Chem. 2003, 68, 5936–5942.
10. Oliver, M. A.; Gandour, R. D. J. Org. Chem. 1984, 49, 558–
559.
11. Biagetti, M.; Bellina, F.; Carpita, A.; Stabile, P.; Rossi, R.
Tetrahedron 2002, 58, 5023–5038.
12. Rossi, R.; Carpita, A.; Bellina, F.; Stabile, P.; Mannina, L.
Tetrahedron 2003, 59, 2067–2081.
13. Hesse, S.; Kirsch, G. Tetrahedron Lett. 2003, 44, 97–99.
14. Ogawa, Y.; Maruno, M.; Wakamatsu, T. Heterocycles 1995, 41,
2587–2599.
data, C16H11NO4, M¼281.26, l¼0.71073 A, monoclinic,
space group P21/n, a¼5.8870(2), b¼12.0270(3), c¼
3
ꢁ
˚
˚
18.4880(7) A, b¼96.802(1) , U¼1299.79(7) A , Z¼4, Dc¼
1.437 mg mꢀ3, m¼0.105 mmꢀ1, F(000)¼584, crystal size
0.40ꢂ0.13ꢂ0.08 mm, unique reflections¼2964 [Rint¼
0.0686], observed I>2s(I )¼1523, data/restraints/parame-
ters¼2964/0/192, R1¼0.0548 wR2¼0.1228 (observed
data), R1¼0.1295 wR2¼0.1617 (all data), max peak/hole
0.300 and ꢀ0.279 eAꢀ3, diffractometer¼Nonius kappaCCD,
˚
software used, SHELXS,38 SHELXL39 and ORTEX.40
15. Bellina, F.; Ciucci, D.; Vergamini, P.; Rossi, R. Tetrahedron
2000, 56, 2533–2545.
4.1.18. X-ray crystallography. Compound 22a. Crystal
˚
16. Rossi, R.; Bellina, F.; Biagetti, M.; Catanese, A.; Mannina, L.
Tetrahedron Lett. 2000, 41, 5281–5286.
data, C21H13NO4Se, M¼422.28, l¼0.71073 A, monoclinic,
ˆ
space group P21/n, a¼11ꢁ.8050(1), b¼11.8550(1),
17. Cherry, K.; Parrain, J.-L.; Thibonnet, J.; Duchene, A.; Abarbri,
3
˚
˚
c¼13.3860(1) A, b¼113.969(1) , U¼1711.80(2) A , Z¼4,
Dc¼1.639 mg mꢀ3, m¼2.222 mmꢀ1, F(000)¼848, crystal
size 0.50ꢂ0.30ꢂ0.10 mm, unique reflections¼3928 [Rint¼
0.0549], observed I>2s(I )¼3671, data/restraints/para-
meters¼3928/0/245, R1¼0.0248 wR2¼0.0627 (observed
data), R1¼0.0275 wR2¼0.0643 (all data), max peak/hole
M. J. Org. Chem. 2005, 70, 6669–6675.
18. Hiroya, K.; Jouka, J.; Kameda, M.; Yasuhara, A.; Sakamoto, T.
Tetrahedron 2001, 57, 9697–9710.
19. Stephens, R. D.; Castro, C. E. J. Org. Chem. 1963, 28, 3313–
3315.
20. Castro, C. E.; Gaughan, E. J.; Owsley, D. C. J. Org. Chem.
1966, 31, 4071–4078.
21. Marshall, J. E.; Chobanian, H. R.; Yanik, M. M. Org. Lett.
2001, 3, 4107–4110.
0.380 and ꢀ0.568 eAꢀ3, software used, SHELXS,38
˚
SHELXL39 and ORTEX.40
Crystallographic data (excluding structure factors) for the
structures in this paper have been deposited with the Cam-
bridge Crystallographic Data Centre as supplementary pub-
lication numbers CDCC-263103 and CDCC-288997. Copies
of the data can be obtained, free of charge, on application to
CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [fax: +44
1223 336033 or e-mail: deposit@ccdc.cam.ac.uk].
22. Liao, H.-Y.; Cheng, C.-H. J. Org. Chem. 1995, 60, 3711–3716.
23. Damle, S. V.; Seomoon, D.; Lee, P. H. J. Org. Chem. 2003, 68,
7085–7087.
24. Negishi, E.-I.; Makabe, H.; Shimoyama, I.; Wu, G.; Zhang, Y.
Tetrahedron 1998, 54, 1095–1106.
25. Botero Cid, H. M.; Trankle, C.; Baumann, K.; Pick, R.;
¨
Mies-Klomfass, E.; Kostenis, E.; Mohr, K.; Holzgrabe, U.
J. Med. Chem. 2000, 43, 2155–2164.
Acknowledgements
26. Corey, E. J.; Kirst, H. A. Tetrahedron Lett. 1968, 9, 5041–
5043.
We are grateful to Dr. Victoria A. Phillips (University of
Bath) for helpful discussions and for skillful growing of
27. Johnson, W. S.; Yarnell, T. M.; Myers, R. F.; Morton, D. R.;
Boots, S. G. J. Org. Chem. 1980, 45, 1254–1259.

E. C. Y. Woon et al. / Tetrahedron 62 (2006) 4829–4837
4837
28. Jackson, W. P.; Ley, S. V. J. Chem. Soc., Perkin Trans 1 1981,
1516–1519.
34. Tretyakov, E. V.; Knight, D. W.; Vasilevsky, S. F. J. Chem. Soc.,
Perkin Trans 1 1999, 3713–3720.
´
29. Orsini, A.; Viterisi, A.; Bodlenner, A.; Weibel, J.-M.; Pale, P.
35. Holt, P. F.; Hughes, A. N. J. Chem. Soc. 1960, 3216–3221.
36. Furuya, Y.; Nozawa, H.; Morita, T.; Morita, T.; Kosugi, Y.
Chem. Pharm. Bull. 1982, 30, 1157–1162.
37. Tirouflet, J. Bull. Soc. Sci. Bretagne 1951, S26, 7.
38. Sheldrick, G. M. Acta Crystallogr., Sect. A 1990, 46, 467–473.
39. Sheldrick, G. M. SHELXL-97, A Computer Program for Crys-
Tetrahedron Lett. 2005, 46, 2259–2262.
30. Jones, P. R.; Desio, P. J. J. Org. Chem. 1965, 30, 4293–4298.
31. Carpita, A.; Mannocci, L.; Rossi, R. Eur. J. Org. Chem. 2005,
1859–1864.
32. Whitmore, F. C.; Culhane, P. J. J. Am. Chem. Soc. 1929, 51,
602–605.
tal Structure Refinement; University of Gottingen: Gottingen,
¨
¨
33. Markl, G.; Hauptmann, H.; Merz, A. J. Organomet. Chem.
1997.
40. McArdle, P. J. Appl. Crystallogr. 1995, 28, 65.
¨
1983, 249, 335–363.