Reaction of furans with trithiazyl trichloride: New synthesis of isothiazoles

Article abstract of DOI:10.1039/a608345e

Trithiazyl trichloride 1 converts 2,5-disubstituted furans into isothiazoles (e.g. 3, 6, 7) regiospecifically and in good yield, providing a new, one-step synthesis of isothiazoles, for which a novel mechanism involving formation and ring-opening of a β-t

Full text of DOI:10.1039/a608345e

View Article Online / Journal Homepage / Table of Contents for this issue
Reaction of furans with trithiazyl trichloride: new synthesis of isothiazoles
Xiao-Lan Duan, Charles W. Rees and Tai-Yuen Yue
Department of Chemistry, Imperial College of Science, Technology and Medicine, London, UK SW7 2AY
Trithiazyl trichloride 1 converts 2,5-disubstituted furans
into isothiazoles (e.g. 3, 6, 7) regiospecifically and in good
yield, providing a new, one-step synthesis of isothiazoles, for
which a novel mechanism involving formation and ring-
opening of a b-thiazylfuran 4 is proposed.
electrophilic substitution of the furan ring at an unsubstituted
b-position to give a b-thiazyl derivative 4, which could then
rearrange to the isothiazole by a ring-opening, ring-closing
mechanism (Scheme 3). We hoped to distinguish between these
mechanisms by reacting the trimer 1 with polarised, un-
symmetrical 2,5-disubstituted and 2,3,5-trisubstituted furans.
2-(4-Methoxyphenyl)-5-(4-nitrophenyl)furan 5a in boiling
THF gave 5-(4-methoxybenzoyl)-3-(4-nitrophenyl)isothiazole
6 (85%) very cleanly and regiospecifically (Scheme 4). This
agreed well with the substitution mechanism (Scheme 3) since
the aryl groups at C-2 and C-5 would strongly direct
electrophilic attack to C-3, which leads to the observed isomer
6. However, this result is harder to reconcile with the Diels–
Alder mechanism since the expected orientation, based on the
polarities of the furan 5a and thiazyl chloride (N NS –Cl)
would most reasonably lead to the undetected isothiazole
isomer. The more nucleophilic furan ring position (C-3) of 5a
was then ‘blocked’ by bromination to give 5b, which was
treated with the trimer 1 under the same conditions. The reaction
of 5b was considerably slower than that of 5a, but it too gave
only one product, 4-bromo-2-(4-methoxyphenyl)-5-(4-nitro-
benzoyl)isothiazole 7 (65%) with the opposite regiochemistry.
This follows from the substitution mechanism, where the
thiazyl substituent is introduced at the only available (deacti-
Trithiazyl trichloride 11 is a stable but moisture-sensitive,
yellow crystalline solid, prepared by heating ammonium
chloride and sulfur with disulfur dichloride, and chlorination of
2
the resulting salt, S
3 2
N Cl
2
. The cyclic trimer 1 is in thermal
equilibrium with the highly reactive monomer, thiazyl chloride
(Scheme 1). It is reactive towards many simple organic
2
substrates and has considerable potential in the synthesis of
sulfur-nitrogen compounds.3
+
2
We thought that the monomer 2 might be selectively
4
intercepted by furans in a Diels–Alder reaction. Trithiazyl
trichloride 1 reacted very cleanly with 2,5-diphenylfuran in
boiling CCl , converting 3 equiv. into 5-benzoyl-3-phenyliso-
4
thiazole 3a in high yield (90%) (Scheme 2). 2,3,5-Triphenyl-
furan and 3-bromo-2,5-diphenylfuran reacted similarly to give
the analogous isothiazoles 3b (75%) and 3c (85%). No isomers
of the products were detected and so cleavage of the furan ring
and incorporation of the S–N unit is apparently regiospecific.
This represents a new opening of the furan ring and a new, mild,
one-step synthesis of isothiazoles. Closely related, but fully
substituted furans, such as 3,4-dibromo-2,5-diphenyl- and
N
Cl
N
N
S
S
3
-bromo-2,4,5-triphenyl-furan did not react with the reagent 1;
S
R
R
clearly a ring hydrogen is required for the formation of HCl,
which is evolved during the reactions. 2,5-Bis-(4-methyl-
phenyl)-, 2,5-bis(4-methoxyphenyl)- and 2,5-di-tert-butyl-
furan all gave the analogous isothiazoles in boiling CCl
toluene, although in somewhat lower yields (50–60%).
••
R
O
R
R
O
R
O
4
4
or
R
R
•
•
All of these products could have been formed by Diels–Alder
••
N•
R
N
R
•
reactions with thiazyl chloride 2, followed by rearrangement of
S
S
3
the adduct with loss of HCl, but the above results led us to
O
O
consider an alternative mechanism. This is based upon
Scheme 3
Cl
S
R
N
S
N
S
O
Cl
N
Cl
1
MeO
NO2
5a R = H
b R = Br
R = Br
OMe
R = H
_
N
+
••
slow
fast
•
S
R
Cl
3
N
S
Cl
R
S
Cl
2
NO2
Scheme 1
Br
Ph
N
N
O
O
1
S
S
3
Ph
N
Ph
O
Ph
S
O
3
a R = H, 90%
b R = Ph, 75%
c R = Br, 85%
NO2
OMe
7
6
Scheme 2
Scheme 4
Chem. Commun., 1997
367

View Article Online
1
vated) site, which leads to the observed isomer (Scheme 3). The
Diels–Alder mechanism would require the complete reversal of
orientation of cycloaddition, on passing from furan 5a to 5b,
which seems unlikely.
Each of these highly polarised furans gave only one
isothiazole; it seemed likely that if the degree of polarisation
was reduced both possible isomeric isothiazoles might be
formed, and this proved to be so. Thus 2-(4-methylphenyl)-
(Scheme 5). Detailed analysis of the H NMR spectra showed
that each gave mixtures of two isomeric isothiazoles which
were inseparable by chromatography; these were 9 and 10
(3.3:1) (56%) and 11 and 12 (3.9:1) (65%) respectively. Thus
in each case the more electron-releasing aryl group in the furan
becomes part of the 5-aroyl group of the major isomer, in
agreement with the reaction of the more highly polarised furan
5a, where only the ‘major’ isomer was observed. On the
mechanism of Scheme 3, substitution of SN should predominate
at the ring position adjacent to the more electron-releasing
group, which then transforms into the 5-aroyl group, as
observed. The expected orientation in the Diels–Alder reaction
would have led to the minor rather than the major isomer.
From all this evidence we favour the mechanism of Scheme
5
-phenylfuran 8a and 2-(4-methoxyphenyl)-5-(4-methyl-
phenyl) furan 8b were treated with the trimer in boiling THF
Ph
Tol
+
+
O
O
N
O
N
S
S
3
. The initially puzzling step, ready opening of the b-thiazyl
Ar = Ph
Tol
Ph
furan ring, is comprehensible when the nitrenoid character 13 of
this intermediate is considered. The highly reactive thiazyl
substituent could induce cleavage of the furan ring (arrows in 4
· 13) to give the delocalised intermediate 14 which then
collapses to the isothiazole (Scheme 6). This unusual ring-
opening is somewhat analogous to the well established cleavage
reactions of nitrenes (and carbenes) directly attached to furan
9
10
Ar
O
Tol
Tol
Anis
8
a Ar = Ph
b Ar = Anis
Ar = Anis
N
O
N
S
S
Anis = 4-MeOC6H4
Tol = 4-MeC6H4
Anis
Tol
12
11
5
and other five-membered heterocyclic rings (cf. 15 and 16).
Scheme 5
The structures of all the new compounds reported here are
based on a detailed analysis and comparison of their spectra,
particularly NMR and mass spectra, with model compounds.
These data, and syntheses of the compounds, will be presented
in a full paper.
•
•
•
N
•
S
We thank Zeneca Specialities, the Commonwealth Scholar-
ship Commission and MDL Information Systems (UK) Ltd for
financial support, and the Wolfson Foundation for establishing
the Wolfson Centre for Organic Chemistry in Medical Science
at Imperial College.
R
O
R
R
13
–
S
••
•N
N
N
S
•
S
etc.
R
••
R
+
••
R
O
R
O
R
O
14
References
1
Gmelin Handbook of Inorganic Chemistry, 8th edn., Sulfur-Nitrogen
Compounds, Part 2, Springer-Verlag, Berlin, 1985, p. 92.
•
•
•
O
O
O
N•
2 W. L. Jolly and K. D. Maguire, Inorg. Synth., 1967, 9, 107.
N
O
1
5
3
X.-G. Duan, X.-L. Duan, C. W. Rees and T.-Y. Yue, J. Heterocycl.
Chem., 1996, 33, 1419.
•
•
N
N•
4 cf. J. Passmore and M. J. Schriver, Inorg. Chem., 1988, 27, 2749; A. Haas
and T. Mischo, Can. J. Chem., 1989, 67, 1902.
•
+
5
T. L. Gilchrist, Adv. Heterocycl. Chem., 1987, 41, 41.
1
6
Scheme 6
Received, 12th December 1996; Com. 6/08345E
368
Chem. Commun., 1997