D. Hédou et al. / Tetrahedron Letters 56 (2015) 4088–4092
4089
Previous work
diamines were obtained but were difficult to isolate, probably
due to iminoquinonic forms generated. Bromination of 8 with N-
bromosuccinimide (NBS) in DMF also failed.
O
N
NC
N
O
O
N
S
O
N
NH
2 steps
H2N
Br
PG
X
NH
N
NH
NC
N
Y
Back to the preceding steps the amino group of compound 7
was protected as a dicarbamate. Methyl 2-[di(tert-butoxycar-
bonyl)amino]-5-nitrobenzoate (12) was prepared from 7 according
to the procedure described for 8. The quantity of Boc2O was
slightly increased (2.2 equiv), and one equivalent of DMAP allowed
the synthesis of the N-di-Boc derivative 12 in very good yield
(96%). The latter was successfully reduced as described above for
9 to give 13 quantitatively. Bromination of 13 was tested in the
conditions described above (Br2/AcOH or NBS/DMF). The procedure
involving bromine led to a complex mixture while the method
using NBS in DMF revealed to be most regioselective and gave
methyl 3-amino-2-bromo-6-[di-(tert-butoxycarbonyl)amino]ben-
zoate (14) in very good yield (90%), along with 10% of its 4-bromo-
substituted isomer 15.8 Despite our efforts, attempts to separate
both isomers by column chromatography only yielded modest
amounts of pure derivative 14 (30%), while recrystallization was
tested without success.
Taking in account these results, the crude mixture of 14 and 15
was directly condensed with Appel salt (1.2 equiv) in dichloro-
methane at room temperature. After 3 h of stirring, pyridine was
added and both isomers of the corresponding imino-1,2,3-dithia-
zoles (16 and 17, respectively) were easily separated and purified.
Major compound 16 was obtained in a good overall 56% yield for
both steps, while its isomeric partner 17 was purified in a low 5%
yield.
Analysis of the spectral data obtained for both imines 16 and 17
revealed that the di-Boc protective group was partially hydrolyzed
during the reaction. This expected result suggests that the release
of hydrogen chloride during the process with Appel salt and/or the
final work-up on an acidic silica gel has induced a partial hydroly-
sis of the N2-protecting group leading to a drastic decrease in the
yields. The residual N-Boc group of imine 16 was cleaved by triflu-
oroacetic acid (TFA) in dichloromethane to furnish 18 in quantita-
tive yield.
Cyclization procedure of 18 was performed in the presence of
copper(I)-iodide (CuI, 1 equiv) in refluxing pyridine under micro-
wave heating. The final methyl 6-amino-2-cyanobenzo[d]thia-
zole-7-carboxylate (4) was purified in a good yield (76%). In an
identical strategy, methyl 6-amino-4-bromo-3-([(5E)-4-chloro-
5H-1,2,3-dithiazol-5-ylidene]amino)-benzoate (19) was obtained
quantitatively from its imino-1,2,3-dithiazole precursor 17 and
was converted into methyl 6-aminobenzo[d]thiazole-5-carboxy-
late (2). Although this route gave access to a valuable and highly
functionalized benzothiazole, product 2, it was only isolated as
side-product of this synthetic sequence. Our group reported a more
S
N
NC
III
: X = N, Y = S
IV: X = S, Y = N
I
II
PG = protectinggroup
NC
CO2H
NH2
CO2Me
NH2
R
S
N
S
N
CN
NC
Br
X
X
NH2
NH2
X = N or CH
R = CO2Me or CN
1: X = N
3
2: X = CH
This work
NC
CO2Me
NH2
S
CO2Me
S
N
N
CO2Me NC
NH2
N
NH2
S
NC
4
5
6
Scheme 1. Previous work on the synthesis thiazolo[4,5-g or -h] and [5,4-f or
-g]quinazolin-4-ones isomers (I–IV), polyfunctionalized methyl 5-amino-2-cyan-
othiazolo[5,4-b]pyridine-6-carboxylate (1), methyl 6-amino-2-cyanobenzo[d]thia-
zole-5-carboxylate (2) and 6-amino-2-cyanobenzo[d]thiazole-7-carbonitrile (3). In
the framed areas, general description of the target molecules of this work is
displayed.
of this synthesis: a regiocontrolled bromination at the desired posi-
tion of the anthranilic acid derivatives; (c) coupling of the thiazole
ring to the benzenic part via Appel salt chemistry and copper(I)-as-
sisted cyclization was preferred to generate a versatile cyano group
on the target structures.1,2 This carbonitrile function can be easily
eliminated (hydrolysis + decarboxylation) or transformed into
amidines, amides, imidates, esters, or acids.3
Our study started from methyl 5-nitroanthranilate (7);10 its
selective bromination in position 4 or 6 should generate key inter-
mediate precursors of benzothiazole 2 and its isomer 4 (Scheme 3).
As described above, the first challenge of our approach was to
find a protective group of the starting aromatic amine, which
would be sufficiently stable to withstand the various stages of
the synthesis.8 Since tert-butylcarbamate (Boc) is stable under
reductive conditions and is easily hydrolyzed under acidic condi-
tions, it was chosen as the most versatile protective group for this
multi-step synthesis. Also, N-arylimino-1,2,3-dithiazoles derived
from the condensation of Appel salt with anilines are, in most
cases, stable in strongly acidic conditions.5,8
Methyl 2-(tert-butoxycarbonylamino)-5-nitrobenzoate (8) was
synthesized in very good yield (90%) by stirring of the starting ester
7 with di-tert-butyl carbonate (Boc2O, 2.0 equiv), N,N-dimethy-
laminopyridine (DMAP, 0.1 equiv) and triethylamine (TEA, 1 equiv)
in dry tetrahydrofuran (THF) at room temperature for 4 h.
Reduction of the NO2 group on N-protected anthranilic ester 8
was successfully performed by using ammonium formate
(HCO2NH4, 5.0 equiv) and palladium on charcoal (Pd/C, 10%) for
efficient method to synthesize
2 starting from methyl 4-
bromoanthranilate.8
Syntheses of methyl 5-amino-2-cyanobenzo[d]thiazole-6-car-
boxylate (5) and methyl 7-amino-2-cyanobenzo[d]thiazole-6-car-
boxylate (6) were inspired by the results described above.
Although reaction conditions and steps’ order were similar to the
preceding work, the stability of the Boc group introduced in N3 is
closely related to the position of the nitro group.
catalytic transfer hydrogenation in refluxing methanol (lw) to give
9 in an excellent yield of 98%. According to previous strategies,
compound 9 was treated with bromine (Br2) in acetic acid to give
the ortho-brominated amines (e.g., 10 and 11 in Scheme 2).
Unfortunately, the reaction afforded a complex mixture that was
difficult to separate. This suggested that N-deprotected 1,4-
Cl
H2N
Br
N
CO2Me
NH-PG
N
S
CO2Me
NH2
N
CO2Me
NH-PG
CO2Me
NH2
5
4
S
O2N
NC
S
Br
4 5
,
6
and
B
A
4- or 5-NO2
PG = Protecting Group
Scheme 2. Retrosynthetic pathway and access to novel linear methyl 6-amino-2-cyanobenzo[d]thiazole-7-carboxylate (4), methyl 5-amino-2-cyanobenzo[d]thiazole-6-
carboxylate (5) and methyl 7-amino-2-cyanobenzo[d]thiazole-6-carboxylate (6) from methyl 4- or 5-nitroanthranilate derivatives.