Photochemistry of 5-nitro-1,2-benzisothiazole derivatives: effects of substituents, solvents and excitation wavelength

Article abstract of DOI:10.1016/j.tetlet.2008.03.104

3-R-5-Nitro-1,2-benzisothiazole derivatives (1, R = substituents) in solution, undergo photochemical isomerization to produce 2-R-5-nitro-1,2-benzothiazole derivatives. Here, generalizations and limitations by the substituent, solvent, and excitation wave

Full text of DOI:10.1016/j.tetlet.2008.03.104

Available online at www.sciencedirect.com
Tetrahedron Letters 49 (2008) 3444–3448
Photochemistry of 5-nitro-1,2-benzisothiazole derivatives:
effects of substituents, solvents and excitation wavelength
Hiroharu Tanikawa, Kazuhiro Ishii, Shun Kubota, Shiki Yagai,
Akihide Kitamura, Takashi Karatsu *
Department of Applied Chemistry and Biotechnology, Graduate School of Engineering, Chiba University,
1
-33 Yayoi-cho, Inage-ku, Chiba-shi 263-8522, Japan
Received 29 January 2008; revised 18 March 2008; accepted 21 March 2008
Available online 28 March 2008
Abstract
-R-5-Nitro-1,2-benzisothiazole derivatives (1, R = substituents) in solution, undergo photochemical isomerization to produce 2-R-5-
3
nitro-1,2-benzothiazole derivatives. Here, generalizations and limitations by the substituent, solvent, and excitation wavelength for the
photoisomerization reaction of 5-nitro-1,2-benzisothiazoles are reported.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: Photoisomerization; Benzisothiazole; Benzothiazole; Substituent effect; Photochemistry
8
Benzisothiazoles (BIT) and benzothiazoles (BT) are
often found in pharmaceutical and agricultural chemicals,
and possess key functions. For example, ziprasidone which
testing of ziprasidone, therefore a better understanding
of this reaction may be important in the design of new
drugs. This Letter deals with nitrobenzo-isothiazoles, as
reactions involving the nitro group are easy to detect,
and allow conversion to various functional groups, such
as amines, imines, and amides and halides.
1
has a BIT skeleton, is approved by US-FDA and is now
being used as an antipsychotic drug, while the BT riluzole
2
is an authorized medicine for ALS. The BT IKF-230 is
3
9
also used as an antifungal agricultural chemical. If BIT-
When a solution of 5-nitro-1,2-benzisothiazole (1a,
À2
BT photoisomerization occurs, the reaction could become
a valuable tool in the synthesis of BIT and BT derivatives
having similar skeletons, as these are currently synthesized
by completely separate routes.
R = H) in acetonitrile (10 mL, 1.5 Â 10 M) was exposed
to a 400 W Xe-arc lamp through a quartz tube using UV30
glass filter (>300 nm), no photochemical reactions were
observed; however, using no filter (>200 nm) for 8 h,
UV–vis absorption analyses showed clear isosbestic points
at 228, 268, and 286 nm, and almost all of the reactants
Five-membered aromatic compounds consisting of two
4
–7
heteroatoms, such as isothiazoles–thiazoles
or isooxaz-
4
–7
oles-oxazoles have been studied in depth and are known
to undergo the reversible photoisomerization. However,
the photochemical behavior of their condensed ring ana-
logues has not been widely studied. The present work will
focus on the BIT-BT isomerization reaction, which is
becoming increasingly important. For example, partial
photoisomerization was observed during photostability
were consumed, giving a product showing IR absorption
assigned to CN triple bond at 2240 cm
À1 10
.
Previously,
1
1
Ohashi et al. showed that 1,2-benzoisothiazole under
the same conditions yields bis(2-cyanophenyl)disulfide
generated by the dimerization reaction of sulfinil radicals,
which are generated by homolytic cleavage of the S–N
bond in the isothiazole ring. Similar phenomena occurred
1
2
with 1a to give bis(2-cyano-4-nitrophenyl)disulfide (3a).
*
Substituents effects at the 3-position of 1a were
Corresponding author. Tel.: +81 43 290 3366; fax: +81 43 290 3401.
E-mail address: karatsu@faculty.chiba-u.jp (T. Karatsu).
then investigated (Fig. 1). First, we examined typical
0
040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.03.104

H. Tanikawa et al. / Tetrahedron Letters 49 (2008) 3444–3448
3445
Fig. 1. Scheme of the substitution effect on photochemistry of 5-nitro-1,2-benzisothiazole derivatives.
electron-withdrawing and -donating substituents (cyano
and methoxy groups, respectively). Under similar reaction
conditions as used for compound 1a, 3-cyano-5-nitro-1,2-
9
benzisothiazole (1b, R = CN) was photolyzed (either at
>200 nm or >300 nm) for 8 h, the compound was very
stable and did not yield any photochemical products. On
the other hand, although the irradiation of 3-methoxy- 5-
9
nitro-1,2-benzisothiazole (1c, R = MeO) with >300 nm
light gave no reaction products, irradiation with >200 nm
light induced gradual degradation of the reactant, and after
1
4 h, a complex mixture of the products was obtained. The
complexity of this mixture inhibited isolation and complete
identification of the products.
The effect of electron-donating amino groups at the 3-
position of 5-nitro-1,2-benzisothiazole was then examined,
and marked variations in the photochemistry of 5-nitro-
Fig. 2. Spectral change by photoirradiation of 1d at >200 nm in
acetonitrile.
1
,2-benzisothiazole derivatives were observed. Under simi-
with those under argon. Therefore, this reaction was not
quenched by oxygen and may occur via a singlet excited
state having short lifetime. In addition, the isomerization
reaction did not proceed in non-polar solvents such as n-
hexane, but was accelerated by twofold in methanol.
The effects of amino groups were then examined. Benz-
isothiazole derivatives with tertiary amino groups, such as
piperidino and piperazino groups, have been isomerized to
give benzothiazoles. Therefore, the effect of amino groups
was investigated, using 3-n-butylamino- (secondary, 1e)
and amino (primary, 1f) groups as substituents.
lar reaction conditions as used for compound 1a, 5-nitro-3-
1
3
piperidino-1,2-benzisothiazole (1d, R = piperidino) was
photolyzed for 160 min. The UV–vis spectral change dur-
ing the irradiation is shown in Figure 2. This spectral
change showed clear isosbestic points at 248 and 322 nm,
thus indicating this reaction is very clean. The final spec-
trum measured at 160 min almost completely matched
the absorption spectrum of separately synthesized 5-
nitro-2-piperidinobenzothiazole (2d). The H NMR and
GC–MS spectra also support this identification. The aro-
matic proton signals shift to a higher magnetic field indi-
cating lower electron density in the benzene ring of BT
when compared to that of BIT. The isolated yield was
1
4
1
The reaction of 3-n-butylamino-5-nitro-1,2-benziso-
1
3
thiazole (1e) showed remarkable dependence on excita-
tion wavelength. Excitation at >400 nm using L-40 glass
filter resulted in the dealkylation of the amino group with
quantitative yield (91%). However, this dealkylation was
not simple, as the product was not 3-amino-5-nitro-1,2-
benzisothiazole (1f), but rather was 3-amino-6-nitro-1,2-
8
8%. Irradiation of 2d did not produce 1d; therefore, this
isomerization is not reversible.
This reaction was also examined under oxygen saturated
acetonitrile, however, the spectral change was identical

3446
H. Tanikawa et al. / Tetrahedron Letters 49 (2008) 3444–3448
benzisothiazole (3e). This positional isomerization of nitro
group may be induced by the isomerization of isothiazole
ring; however, the detailed mechanism is unclear at present.
In addition, the formation of 3e did not occur by a stepwise
reaction, as the intermediate dealkylation product 1f did
not give 3e during photolysis. We have no experimental
data at this moment to indicate the intermediacy of a sulfur
atom migrated intermediates for 3e, however, DFT calcu-
lation result for 2d showed an energetic possibility via a sul-
fur atom migration from 1e to 3e, as similar route as shown
dashed arrows for the mechanism of 2d in Figure 5. More
calculation for intermediates of 3e is underway.
Irradiation of 1e at >200 nm in acetonitrile gave 2-n-
1
4
butylamino-5-nitrobenzothiazole (2e) as an isomerization
product. Figure 3 shows the UV–vis spectral changes of 1e
by irradiation in acetonitrile.
Irradiation (>400 nm) of a primary amine, 3-amino-5-
nitro-1,2-benzisothiazole (1f) did not lead to an isomeri-
zation reaction. Figure 4 shows the spectral changes of 1f
using >400 nm light. The spectral change showed clear
isosbestic points at 270, 308, and 392 nm. This change
was very clean, and thought to be due to isomerization,
1
À1
but H NMR and IR (2260 cm ) measurements revealed
thiazole ring opening and the formation of a cyano group.
Similarities between this product and a standard sample
separately prepared indicate that the product was 2-
1
5
amino-5-nitrobenzonitrile (3f). The mechanism for this
reaction was proposed by Ohashi et al., however, addi-
tional experiments are underway.
Fig. 3. Spectral change by photoirradiation of 1e at >200 nm.
11
The reactions of 1d and 1f were examined in the pres-
ence of trifluoroacetic acid (TFA) as a protonation reagent
for amino group. For 1d, TFA had no effect on the photo-
chemical isomerization reaction; however, the formation of
3f was completely quenched by the presence of TFA. Qua-
ternarization of the amino group acts as an important step
in the mechanism to inhibit transamination. Further
detailed investigations are underway.
Figure 5 indicates a plausible reaction mechanism of 1d–
2d, and this scheme is based on the experimental results
and the relative energies of intermediates, as shown in
Figure 6. In this case, the reaction starts from singlet
excited state (energy was estimated from UV absorption)
generated by excitation, followed by the formation of a
Dewar-type isomer or spiroazacyclopropene (azirine).
Other isomeric benzothiazole intermediates may not be
formed, as these have higher energies than those of the
Fig. 4. Spectral change by photoirradiation of 1f at >400 nm in
acetonitrile.
Fig. 5. Plausible mechanism of the isomerization of 1d.

H. Tanikawa et al. / Tetrahedron Letters 49 (2008) 3444–3448
3447
Fig. 6. Relative energies calculated by DFT method for the reactant, product, and intermediates considered in the photoisomerization of 1d.
Dewar isomer and azirine. The ground state energy of 2d is
calculated to be 12.6 kcal mol lower than 1d.
2. Howard, H. R.; Lowe, J. A., III; Seeger, T. F.; Seymour, P. A.; Zorn,
À1
16
S. H.; Maloney, P. R.; Ewing, F. E.; Newman, M. E.; Schmidt, A. W.;
Furman, J. S.; Robinson, G. L.; Jackson, E.; Johnson, C.; Morrone,
J. J. Med. Chem. 1996, 39, 143–148.
Figure 1 summarizes the substituent effects on the photo-
chemistry of the benzisothiazole derivatives examined in
this study. The reaction between nitrobenzisothiazole to
nitrobenzothiazole was not reversible, however, the isomer-
ization mechanism of the 3-piperidino-derivative (1d,
R = piperidino) was elucidated, as shown in Figure 5. This
study thus revealed the usefulness and limitations of nitro-
benzisothiazole to nitrobenzothiazole photochemistry.
3
. Miyake, Y.; Sakai, J.; Shibata, M.; Yonekura, N.; Miura, I.;
Kumakura, K.; Nagayama, K. J. Pestic. Sci. 2005, 30, 390–396.
4. Pavlik, J. W.; Kurzweil, E. M. J. Org. Chem. 1991, 56, 6316.
5
6
. Pavlik, J. W.; Tongcharoensirikul, P. J. Org. Chem. 2000, 65, 3626.
. Pavlik, J. W. CRC Handbook of Organic Chemistry and Photochem-
istry, 2nd ed.; CRC Press LLC, 2004.
7
. D’Auria, M. Tetrahedron 2002, 58, 8037–8042.
8. Sharp, T. R.; Leeman, K. R.; Bryantb, D. E.; Horana, G. J.
Tetrahedron Lett. 2003, 44, 1559–1561.
9
. 5-Nitro-1,2-benzisothiazole 1a. 3-cyano-5-nitro-1,2-benzisothiazole
b, and 3-methoxy-5-nitro-1,2-benzisothiazole 1c, were synthesized
Acknowledgments
1
by the methods reported in Ref. 17. 3-Amino-5-nitro-1,2-benziso-
The authors thank Analytical Center of Chiba Univer-
sity for their assistance of characterization of the com-
pounds. This research project was financially supported
by Grants-in-Aid for Scientific Research (No. 17350090),
Scientific Research on Priority Areas (417, No. 17029013)
from the Ministry of Education, Culture, Sports, Science,
and Technology (MEXT) to T.K.
thiazole 1f was prepared by the methods reported in Ref. 18.
10. Irradiation was performed by a 400-W Xe-lamp using appropriate
1
glass filters. H NMR, FAB-MS, and Elemental Analyses were
performed by instruments in Analytical Center of Chiba University.
1
1. Ohashi, M.; Ezaki, A.; Yonezawa, T. J. Chem. Soc., Chem. Commun.
974, 617.
1
12. 3a was identified by comparison of spectral data to those of standard
sample prepared by the method reported; Hartmans, H.M.A., Rec.
1
Trav. Chim., 1946, 65, 468. H NMR 400 MHz (CDCl
3
/DMSO-d
6
was used because of low solubility) d = 8.52 (1H, d, J = 2.2 Hz), 8.41
References and notes
(
2
1H, dd, J = 8.0, 2.2 Hz,), 8.32 (1H, d, J = 7.6 Hz), IR (KBr)
240 cm (CN). Mp 138–139 °C. Unfortunately, M signal could
À1
+
1
. Seeger, T. F.; Seymour, P. A.; Schmidt, A. W.; Zorn, S. H.; Shulz, D.
W.; Lebel, L. A.; McLean, S.; Guanowsky, V.; Howard, H. R.; Lowe,
J. A., III; Heym, J. J. Pharmacol. Exp. Ther. 1995, 275, 101–
not be detected by FAB-MS using NBA (m-nitrobenzyl alcohol)
matrix.
13. Preparation of 3-piperidiono-5-nitro-1,2-benzisothiazole, 1d.
A
113.
mixture of piperidine (6 ml, 101 mmol) and 3-chloro-5-nitrobenziso-

3448
H. Tanikawa et al. / Tetrahedron Letters 49 (2008) 3444–3448
thiazole (310 mg, 1.4 mmol) prepared by the method reported in Ref.
6 was refluxed overnight, and removed excess piperidine under
vacuum. Purification by crystallization (dichloromethane:hexane)
(CDCl
2.2 Hz, H6), 7.65 (1H d, J = 8.6 Hz, H7), 3.65 (4H, m, NCH
13 3 2
(6H, m, –CH –). GC–MS m/z: 263. Anal. Calcd for C12H N O
3
) d = 8.31 (1H, d, J = 2.2 Hz, H4), 7.91 (1H, dd, J = 8.6,
–), 1.73
S: C,
1
2
2
after the column chromatography gave orange crystals, 121 mg,
54.74; H, 4.98; N, 15.96. Found: C, 54.66; H, 4.88; N, 15.75.
1
yield 32.8%.
H
NMR 400 MHz (CDCl
3
), d = 8.79 (1H, d,
Compound 2e was prepared by a similar method with 2d, 310 mg
1
J = 2.4 Hz, H4), 8.30 (1H, dd, J = 9.0, 2.0 Hz, H6), 7.89 (1H, d,
J = 9.0 Hz, H7), 3.53 (4H, t, J = 5.4 Hz, –NCH –), 1.86–1.84 (4H, m,
CH –), 1.74–1.71 (2H, m, –CH –). GC–MS m/z: 263. Anal. Calcd
for C12 S: C, 54.74; H, 4.98; N, 15.96. Found: C, 54.56; H,
.83; N, 16.03. Compound 1e was prepared by a similar method with
reactant gave 207 mg 2e, 58.8%. H NMR 400 MHz (CDCl
3
) d = 8.32
2
(1H, d, J = 2.2 Hz, H4), 7.96 (1H, dd, J = 9.0, 2.2 Hz, H6), 7.67 (1H,
d, J = 9.0 Hz, H7), 5.91 (1H, br s, NH), 3.48–3.44 (2H, t, J = 7.3 Hz,
–
2
2
H
13
N
3
O
2
NCH
J = 7.3, 7.3 Hz –CH
251. Anal. Calcd for C11
Found: C, 52.54; H, 5.15; N, 16.53.
2
–), 1.75–1.67 (2H, quin, J = 7.3 Hz, –CH
–), 0.98 (3H, t, J = 7.3 Hz, –CH
S: C, 52.57; H, 5.21; N, 16.72.
2
–), 1.51–1.42 (2H, qt,
4
2
3
). GC–MS, m/z:
1
1
d, 310 mg reactant gave, 311 mg, 84.4%. H NMR 400 MHz
), d = 8.60 (1H, d, J = 1.9 Hz, H4), 8.31 (1H, dd, J = 9.0,
.9 Hz, H6), 7.87 (1H, d, J = 9.0 Hz, H7), 5.14 (1H, br s, NH), 3.63–
.58 (2H, m, –NCH –), 1.78–1.70 (2H, quin, J = 7.3 Hz, –CH –),
.53–1.44 (2H, sex, J = 7.3 Hz, –CH –), 1.00 (3H, t, J = 7.3 Hz, –
). GC–MS m/z: 251. Anal. Calcd for C11 S: C, 52.57; H,
.21; N, 16.72. Found: C, 52.60; H, 5.23; N, 16.56.
13 3 2
H N O
(CDCl
3
1
1
3
1
CH
5
6
15. 3f Yield 81%; H NMR 400 MHz (DMSO-d ) d = 6.87 (1H, d,
2
2
J = 9.4 Hz), 7.50 (2H, br s), 8.13 (1H, dd, J = 2.7 Hz, 9.4 Hz), 8.39
(1H, d, J = 2.7 Hz), IR (KBr) 2260 (CN). Manh, G. T.; Bakkali, H.;
Maingot, L.; Pipelier, M.; Joshi, U.; Pradere, J. P.; Sabelle, S.;
Tuloup, R.; Dubreuil, D. Tetrahedron Lett. 2004, 45(30), 5913–5916.
16. Quantum calculation was performed by DFT B3LYP, 6-31G* level
using Win. Spartan’06.
17. Kameswaran, V. U.S. Patent, US 6355796, 2002.
18. Harusako, S.; Satoh, T.; Nakamura, M.; Mizuno, M.; Mori, S.
Patent, WO03878072, 2003.
2
3
13 3 2
H N O
1
4. Preparation of 3-piperidiono-5-nitro-1,2-benzothiazole, 2d. A mixture
of piperidine (4 ml, 68 mmol) and 2-chloro-5-nitro-1,3-benzothiazole
19
(
3
420 mg, 1.9 mmol) prepared by the method reported was refluxed
h, and removed excess piperidine under vacuum. Purification by
crystallization (CH Cl /hexane) after the column chromatography
gave yellow crystals, 408 mg, yield 81.5%. H NMR 400 MHz
2
2
1
19. Zhu, L.; Zhang, M.; Dai, M. J. Heterocycl. Chem. 2005, 42, 727.