On the diverse outcomes of base-induced cyclisations of 2-alkynylphenylhydroxamic acids

Article abstract of DOI:10.1016/S0040-4039(02)02254-2

Base-induced cyclisations of 2-alkynylphenylhydroxamic acids occur by attack of nitrogen onto the alkyne group in either exo or endo fashions to give the corresponding isoindol-1(2H)-ones or 1(2H)-isoquinolinones depending upon the alkyne substituent.

Full text of DOI:10.1016/S0040-4039(02)02254-2

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 9187–9189
On the diverse outcomes of base-induced cyclisations of
2-alkynylphenylhydroxamic acids
David W. Knight,a,* Paul B. M. Lewis,a K. M. Abdul Malik,a Elena V. Mshvidobadzeb and
Sergei F. Vasilevskyb
aChemistry Department, Cardiff University, PO Box 912, Cardiff, CF10 3TB, UK
bInstitute of Chemical Kinetics and Combustion, Siberian Branch of the Russian Academy of Sciences, 630090,
Novosibirsk, Russia
Received 8 August 2002; revised 26 September 2002; accepted 4 October 2002
Abstract—Base-induced cyclisations of 2-alkynylphenylhydroxamic acids occur by attack of nitrogen onto the alkyne group in
either exo or endo fashions to give the corresponding isoindol-1(2H)-ones or 1(2H)-isoquinolinones depending upon the alkyne
substituent. © 2002 Elsevier Science Ltd. All rights reserved.
Since the original observations of Stephens and Castro
on the facile cyclisation of 2-alkynylbenzoic acids 1 to
the ylidenephthalides 2,1 this has been expanded into a
useful heterocyclic synthesis, especially by modification
of the carboxylic acid into the corresponding amides2
or hydrazides.3 There is however a potential drawback
associated with this methodology: in many versions, the
isomeric isochromenones 3 are formed, sometimes as
mixtures along with the phthalides 2 (Fig. 1).1
Our combined interests both in this area and also in the
reverse Cope elimination of unsaturated hydroxyl-
amines4 led us to speculate that it might be possible to
carry out such cyclisations but using 2-alkynylphenyl
hydroxamic acids 4. At the outset, the viability of such
cyclisations was far from clear as was the nature of the
products, as these could in principle proceed via three
different nucleophilic centres, the carbonyl or hydroxyl-
amine oxygens or the nitrogen in either an endo or an
exo fashion. Herein, we report on a surprising diversity
of outcomes to what turned out to be extremely viable
and selective.
The necessary starting materials were all prepared from
methyl 2-bromobenzoate 5 by Sonogashira coupling5
with the appropriate 1-alkyne (Fig. 2). Under relatively
standard conditions [(Ph3P)4Pd, CuI (cat.) in degassed
1:3 THF–Et2NH, reflux 3 h], couplings of ester 5 with
1-alkynes provided 82-95% isolated yields of the
expected 2-alkynylbenzoates 6. Attempts to generate
the key hydroxamic acids 4 by similar couplings but
using 2-bromophenylhydroxamic acid were, perhaps
not surprisingly, unsuccessful.
Figure 1.
A straightforward method for the conversion of such
benzoate esters 6 into the corresponding hydroxamic
acids involves heating the ester with hydroxylamine.
Starting with the hexynyl derivative 7, we heated this
with excess hydroxylamine in methanol until transfor-
mation to what appeared to be a single product was
complete (Scheme 1).6 Typically this took around three
hours. This first experiment gave a single product, m.p.
Figure 2.
* Corresponding author.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02254-2

9188
D. W. Knight et al. / Tetrahedron Letters 43 (2002) 9187–9189
Scheme 1.
Scheme 3.
108–110°C, in 86% isolated yield. This was tentatively
identified as the isoquinolinone 8 from its 1H NMR
data which featured a singlet at lH 6.45 and a triplet
(J=7Hz) centred at lH 2.88. An alternative structure,
formation of which could reasonably be expected in the
light of some of the foregoing results,1–3 the ylidene-
isoindolone 9 which would arise from exo-cyclisation
was ruled out in the light of these data. The latter
would be expected to display an olefinic proton as a
triplet together with an allylic methylene with a double
triplet or apparent quartet splitting pattern. This con-
clusion was confirmed by comparison with literature
data (lit.7 m.p. 108–109°C) and also by an X-ray crys-
tallographic determination.8
Similar exo-cyclisations occurred when the propargylic
alcohol derivatives 13 were exposed to excess hydroxyl-
amine in hot methanol:6 only the (Z)-isoindolones 14a
(m.p. 94°C) and 14b (m.p. 96°C) were isolated in 65%
and 60% yields respectively (Scheme 3). In both cases,
1H NMR data unambiguously distinguished these
structures from those of the isomeric isoquinolinones
15.9 Both 1H and especially 13C data showed both
products to be single geometric isomers, again identified
as having (Z)-stereochemistry from NOE enhancements
of 4-5% between the alkene protons and aromatic
doublets. These data clearly exclude the alternative
formulation as isoquinolinones 15.
Thinking that we had discovered a new isoquinolinone
synthesis, we treated the phenylalkyne derivative 10
with hydroxylamine under the same conditions6 and
isolated a single crystalline product; m.p. 172°C, in 81%
yield (Scheme 2). However, this was proven to be the
isoindolone 11 instead, rather than the isomeric iso-
quinolinone 12. While spectroscopic data did not
provide completely unambiguous proof of structure 11,
the lower melting point (137–138°C)7 of isoquinolinone
12 indicated this was not the structure of the product.
Moreover, an NOE enhancement between the new
olefinic proton singlet at lH 6.90 and an aromatic
doublet at lH 7.80 indicated the (Z)-stereochemistry
shown. Any lingering uncertainty was removed when
an X-ray crystallographic determination proved the
correctness of structure 11.8
However, such a structure 17 was obtained as a single
isomer in 72% isolated yield when the thiophenyl
derivative 16 was heated with hydroxylamine (Scheme
4).
Scheme 4.
A final example produced some more complex chem-
istry: when the silylated alkyne derivative 18 was
treated similarly,6 the only product isolated was the
benzoazinone 19, m.p. 163–164°C, in 64% yield
(Scheme 5). This was identified by comparison with
authentic material derived from 2-acetylbenzoic acid.10
A possible mechanism for formation of this unexpected
product involves the assumed ester to hydroxamic acid
conversion followed by exo-cyclisation to give the
methylideneisoindolone 20, during which desilylation
also takes place. Tautomerism would then lead to the
more stable nitrone 21 which could rearrange to the
observed product 19 via the oxaziridine alkoxide 22.
In summary, we have defined new and straightforward
syntheses of some isoquinolinones and isoindolones
Scheme 2.

D. W. Knight et al. / Tetrahedron Letters 43 (2002) 9187–9189
9189
References
1. (a) Castro, C. E.; Gaughan, E. J.; Owsley, D. C. J. Org.
Chem. 1966, 31, 4071; (b) Castro, C. E.; Stephens, R. D.
J. Org. Chem. 1963, 28, 2163; (c) Liao, H.-Y.; Cheng,
C.-H. J. Org. Chem. 1995, 60, 3711.
2. Vasilevsky, S. F.; Schvartsberg, M. S. Izv. Akad. Nauk
SSSR, Ser. Khim. 1990, 2089 and references cited therein.
3. (a) Vasilevsky, S. F.; Shvartsberg, M. S. Izv. Akad. Nauk
SSSR, Ser. Khim. 1983, 1686; (b) Vasilevsky, S. F.;
Pozdyakov, A. V. Izv. Sib. Otd. Akad. Nauk SSSR, Ser.
Khim, 1985, 1367. See also (c) Vasilevski, S. F.;
Tretyakov, E. V. Annalen, 1995, 775; (d) Tretyakov, E.
V.; Knight, D. W.; Vasilevsky, S. F. J. Chem. Soc.,
Perkin Trans. 1, 1999, 3721.
4. Hanrahan, J.; Knight, D. W.; Salter, R. Synlett 2001,
1587.
5. Sonogashira, K.; Tohda, Y.; Hagihara, N. A. Tetra-
hedron Lett. 1975, 4467.
6. General method: Warm solutions of hydroxylamine
hydrochloride (6 eq.) in methanol (0.5 ml mmol−1) and
potassium hydroxide (9 eq.) in methanol (0.25 ml
mmol−1) were mixed and the resulting solution cooled to
below 40°C when potassium chloride precipitated. This
was filtered off and a methyl 2-alkynylbenzoate (1 equiv.)
was added. The resulting mixture was then refluxed until
the reaction appeared to be complete according to TLC
(ca. 3 h). The cooled mixture was then evaporated. Excess
ice-cold 2 M hydrochloric acid was added to the residue
and the product extracted into dichloromethane. The
combined organic extracts were washed with water and
brine then dried, filtered and evaporated. Crystallization
of the resulting residue, usually from ethyl acetate or
ethyl acetate-petrol mixtures then secured the pure prod-
ucts in the yields quoted.
Scheme 5.
which, while clearly limited to a particular substituent,
are remarkable for their selectivity in favour of either
endo or exo-cyclisation. In each case, examination of
crude material did not reveal the presence of the alter-
native product. At present we can only speculate on the
origins of this very high selectivity which may be associ-
ated with the fact that all isoindolones are formed when
the alkyne is substituted with an electron-withdrawing
group, albeit not a powerful one, while the isoquinoli-
nones 8 and 17 have electron-donating substituents.
However, these relatively weak influences of such sub-
stituents does not seem entirely sufficient to account for
such a high level of regioselectivity. Hopefully further
studies along with theoretical calculations will provide a
better insight.11
7. Moriconi, E. J.; Creegan, F. J. J. Org. Chem. 1966, 31,
2090.
8. These data have been deposited on the Cambridge Crys-
tallographic Database: Compound
8
[CCDC No.
193931], compound 11 [CCDC No. 193932].
9. Isoindolinone 14a showed lH (CDC13) 4.70 (1H, d, J 6.3)
and 5.91 (t, J 6.3) while the related protons in analogue
14b resonated at 4.58 and 5.83.
Acknowledgements
10. (a) Gabriel, S. Ber. 1883, 16, 1995; (b) Kampschmidt, L.
W; Wibaut, J. P. Rec. Trav. Chem. 1954, 73, 431; (c)
Sudoh, Y; Jin, Z.-T; Matsamura, H. J. Heterocycl. Chem.
1982, 19, 525.
11. Full analytical and spectroscopic data consistent with the
proposed structures have been obtained for all com-
pounds reported herein.
We are grateful to EPSRC for support of the X-ray
crystallographic facilities in Cardiff and also thank the
EPSRC Mass Spectrometry Service, University College,
Swansea for the provision of high resolution mass
spectra and INTAS and Cardiff University for financial
support.