Unexpected formation of 2,1-benzisothiazol-3-ones from oxathiolano ketenimines: A rare tandem process

Article abstract of DOI:10.1021/ol9001416

A rare one-pot reaction, a tandem [1,5]-H shift/1,5 electrocyclization/[3 + 2] cycloreversion process, leading from N-[2-(1,3-oxathiolan-2-yl)]phenyl ketenimines to 1-(β-styryl)-2,1-benzisothiazol-3-ones and ethylene, is disclosed and mechanistically unra

Full text of DOI:10.1021/ol9001416

ORGANIC
LETTERS
2009
Vol. 11, No. 6
1365-1368
Unexpected Formation of
2,1-Benzisothiazol-3-ones from
Oxathiolano Ketenimines: A Rare
Tandem Process
Mateo Alajarin,*,† Baltasar Bonillo,† Pilar Sanchez-Andrada,† Angel Vidal,† and
Delia Bautista‡
Departamento de Quimica Organica, Facultad de Quimica, and SerVicio UniVersitario
de Instrumentacion Cientifica, UniVersidad de Murcia, Campus de Espinardo, 30100
Murcia, Spain
alajarin@um.es
Received January 26, 2009
ABSTRACT
A rare one-pot reaction, a tandem [1,5]-H shift/1,5 electrocyclization/[3 + 2] cycloreversion process, leading from N-[2-(1,3-oxathiolan-2-
yl)]phenyl ketenimines to 1-(ꢀ-styryl)-2,1-benzisothiazol-3-ones and ethylene, is disclosed and mechanistically unraveled by means of a
computational DFT study. The two latter stages of the tandem process are calculated to occur in a single mechanistic step via a transition
structure of pseudopericyclic characteristics.
In the course of our studies on the chemistry of ketenimines,1
we have reported a number of sigmatropic rearrangements of
atoms or groups of atoms to the central carbon atom of the
heterocumulenic function.2 Particularly relevant for the contents
of this report are the recently unveiled hydricity-promoted [1,5]-
H shifts in acetalic ketenimines and carbodiimides (Scheme 1).2e
The cyclic (dithio)acetal function of ketenimines 1,2 and
carbodiimides 3 promotes the [1,5] shift of the acetalic proton
to the central carbon atom of the heterocumulene fragment,
under thermal conditions, for yielding carbonyl-protected 4-qui-
nolones 5 (Y ) CR2) or 4-quinazolinones 5 (Y ) NAr), via
further 6π-electrocyclic ring closure of the initially formed
transient o-quinomethanimines 4.
† Facultad de Quimica.
‡ Servicio Universitario de Instrumentacion Cientifica.
(1) For reviews in the chemistry of ketenimines, see: (a) Krow, G. R.
Angew. Chem., Int. Ed. Engl. 1971, 10, 435–449. (b) Gambaryan, N. P.
Usp. Khim. 1976, 45, 1251–1268. (c) Dondoni, A. Heterocycles 1980, 14,
1547–1566. (d) Barker, M. W.; McHenry, W. E. In The Chemistry of
Ketenes, Allenes and Related Compounds; Patai, S., Ed.; Wiley-Interscience:
Chichester, UK, 1980; Part 2, pp 701-720. (e) Alajarin, M.; Vidal, A.;
Tovar, F. Targets Heterocycl. Syst. 2000, 4, 293–326.
Scheme 1. Reaction Path Converting 1-3 into 5
(2) (a) Alajarin, M.; Vidal, A.; Ortin, M.-M. Tetrahedron 2005, 61,
7613–7621. (b) Alajarin, M.; Ortin, M.-M.; Sanchez-Andrada, P.; Vidal,
A.; Bautista, D. Org. Lett. 2005, 7, 5281–5284. (c) Alajarin, M.; Ortin,
M.-M.; Sanchez-Andrada, P.; Vidal, A. J. Org. Chem. 2006, 71, 8126–
8139. (d) Alajarin, M.; Vidal, A.; Ortin, M.-M. Synthesis 2007, 590–596.
(e) Alajarin, M.; Bonillo, B.; Ortin, M.-M.; Sanchez-Andrada, P.; Vidal,
A. Org. Lett. 2006, 8, 5645–5648.
10.1021/ol9001416 CCC: $40.75
Published on Web 02/26/2009
2009 American Chemical Society

While exploring the ability of other acetalic functions for
activating similar processes, we tested ketenimines 9 bearing
1,3-oxathiolane rings, which were easily prepared from the
respective 2-azidobenzaldehydes 6. Surprisingly, after heating
toluene solutions of ketenimines 9 under reflux for a few hours,
these compounds did not convert into the expected quinolines
11, but instead 1-(ꢀ-styryl)-2,1-benzisothiazol-3-ones 10 were
obtained in fair to good yields (Scheme 2, Table 1).
converting 9 into the transient 12, would be followed by a
rare 1,5 cyclization of the latter involving S-N bond
formation, thus providing the sulfur ylide 13, which finally
should undergo a [3 + 2] cycloreversion step to yield 10
plus ethylene (Scheme 3, 9a is used for simplicity).
Scheme 3. Mechanistic Proposal for the Conversion 9 f 10
Scheme 2. Preparation of 2,1-Benzisothiazol-3-ones 10
Whereas the first step of this mechanistic proposal is prece-
dented in the introduction above, the atom connectivity of
the final products apparently leaves no room for mechanistic
speculations other than the two final steps outlined in Scheme
3.
Taking into account that the thioanalogues of 12, structures
4 (X ) S), did not experience similar transformations but
instead electrocyclize to quinolines 5, it seems clear that the
generation of a strong CdO bond in 10 should be a crucial
fact for explaining its formation.
In order to gain insight into the mechanistic path of the
transformation of ketenimines 9 into ethylene plus ben-
zisothiazolones 10, and also to answer why analogous
dithiolanes 2 do not experience a similar transformation, we
have carried out a computational DFT study at the B3LYP/
6-31+G** level of theory using the Gaussian03 package
(see Supporting Information).
Table 1. 1-(ꢀ-Styryl)-2,1-benzisothiazol-3-ones 10
compd
R1
R2
R3
yield (%)
10a
10b
10c
10d
10e
H
CH3
H
H
H
H
H
Cl
CH3
H
Ph
Ph
Ph
Ph
Me
73
92
74
92
81a
First we approached the transformation of the oxathiolane
ketenimine 14a into the initially expected spiroquinoline 15a,
and also into the actual product benzisothiazolone 16a (Scheme
4). The first step, common to both reaction channels, consists
of a [1,5]-H shift, as anticipated, leading to the intermediate
o-quinomethanimine Z-17a which further can follow the two
alternative pathways (a and b) depicted in Scheme 4.
The most relevant outcome of these calculations is that
formation of benzisothiazolone 16a plus ethylene does not
involve a transient sulfur ylide similar to 13; instead,
intermediate E-17a undergoes simultaneously the two fol-
lowing events, 1,5-electrocyclization and [3 + 2] cyclor-
eversion, via the transition state TS4a. The optimized
geometries of the four transition states involved in these
transformations are depicted in Figure 1. The computations
establish the same mechanistic paths for the conversion of
the dithiolane ketenimine 14b into the spiroquinoline 15b
or the benzisothiazole-3-thione 16b. However, the energy
barriers are quite different as shown in Table 2.
a Total yield for the E + Z isomers; ratio E/Z ) 4:1.
The structural determination of the 1-(ꢀ-styryl)-2,1-ben-
zisothiazol-3-ones 10 was achieved following their analytical
and spectral data, and unequivocally established by the X-ray
structure determination of a monocrystal of 10a (R1 ) R2 )
H; R3 ) Ph) (see Supporting Information).
We obviously envisaged that the transformation 9 f 10
should occur with the concomitant formation of ethylene,
although we made no efforts to detect it in the reaction
exhausts. However, the results of the computational mecha-
nistic study of such transformation (see below) support this
assumption. The presence of the original acetalic proton of
9 at the styryl ꢀ-carbon of benzisothiazolones 10, which was
the central keteniminic carbon in 9, suggest the involvement
of a [1,5]-H shift as the initial step of the sequence leading
from 9 to 10. In our first mechanistic proposal, this step,
1366
Org. Lett., Vol. 11, No. 6, 2009

Scheme 4
.
Alternative Mechanistic Paths for the Conversions of
Table 2. Free Energy Barriersa (kcal·mol-1) Calculated for the
Conversions 14a-cf15a-c/16a-c at the B3LYP/6-31+G**
Level
Ketenimines 14a-c into 15a-c and into 16a-c Plus Ethylene
14f15/16 ∆Gq ∆Gq ∆Gq ∆Gq ∆Grxn path a ∆Grxn path b
1
2
3
4
a
b
c
31.7 21.1
33.3 21.1 10.2 27.5
26.9 26.4 9.5 20.5
9.5 19.2
-21.2
-17.4
-1.4
-12.5
+2.0
-10.7
a See Scheme 4 for the notation of the energy barriers ∆Gq
(these
1-4
energies are relative to those of the preceding stationary points).
kinetically controlled product when starting from 14b, whereas
the benzisothiazolone 16a is the kinetically controlled product
when starting from 14a. These conclusions are in accordance
with the experimental results obtained in the thermal treatment
of dithiolane ketenimines 2, which results only into the
corresponding spiroquinolines,2e whereas the calculations predict
a small kinetic preference for the formation of benzisothiazolone
16a versus spiroquinoline 15a.
It is conceivable that replacing H/H by Ph/Ph or Ph/Me
at the sp2 carbon atom of the ketenimine 14a, as in the
experimental substrates 9, should considerably increase the
barrier of the 6π-ERC step due to steric constrains. This is
why we also studied the transformations 14cf15c/16c
involving a ketenimine fragment bearing two phenyl groups.
The chief outcomes of this study are the increase of the
energy barrier for the 6π-ERC step when compared with the
unsubstituted case, which is now 5.3 kcal·mol-1 higher,
whereas that of the last step in path b only increases by 1.3
kcal·mol-1. In addition, spiroquinoline 15c is higher in energy
than benzisothiazolone 16c by 9.3 kcal·mol-1. Therefore, 16c
is clearly predicted to be the thermal and kinetically
controlled product, accordingly with the experiments.
It is worth commenting on the special features of the
mechanistic step converting E-17 into 16. It is a complex,
apparently pericyclic process having no precedent in the
literature. It looks like a 1,5-electrocyclization and a [3 +
2] cycloreversion taking place simultaneously through the
transition state TS4. However, we have found in the literature
neither 1,5-electrocyclizations of pentadienyl skeletons simi-
lar to the 1-thia-5-aza-2,4-pentadiene fragment of E-17 nor
[3 + 2] cycloreversions involving 1,3-oxathiolane rings, and
these two transformations do not seem, independently, easily
amenable. By contrary, the combination of both processes
in a unique mechanistic step results in a process of a discrete
activation energy, close to 20 kcal·mol-1. Probably the two
simultaneous transformations cooperate in a favorable way.
For instance, now the 1,5-electrocyclization might be viable
thanks to the formation of the strong CdO bond, occurring
by virtue of the simultaneous ethylene extrusion, which in
turn is favored by the initial S-N bond formation. Moreover,
the aromaticity recovery at the benzene ring along the way
to 16 and the inherent entropic assistance of this mechanistic
step must be also decisive factors for determining its success.
We analyzed in detail some geometric and electronic
parameters of the transition state TS4a. Its geometry does not
resemble that expected for a [3 + 2] cycloreversion or an 1,5-
Several main conclusions can be inferred by comparing the
first two entries of this table: (1) for both reaction paths in each
entry (14a,bf15a,b/16a,b) the rate limiting step is the [1,5]-
H shift; (2) the energy barrier for the 6π-electrocyclization of
Z-17 (path a) is identical in both entries (21.1 kcal·mol-1); (3)
concerning path b, the ZfE C)N isomerization involves
always an easily surmountable barrier; (4) the barrier for the
conversion of E-17a leading to the benzisothiazolone 16a is
considerably lower than that calculated for the thioanalogous
transformation E-17bf16b; (5) the energy barrier of this latter
conversion is higher than that of the alternative 6π-ERC leading
to 15b, whereas just the opposite occurs for the transformations
of intermediate Z-17a, that is, the energy barrier for the 6π-
ERC (Z-17af15a) is a little higher than that leading to 16a;
(6) the spiroquinoline 15b is predicted to be the thermal and
Figure 1. Optimized B3LYP/6-31+G** geometries of the transi-
tion structures involved in the transformations TS1-4a.
Org. Lett., Vol. 11, No. 6, 2009
1367

ERC. The three rings are in the same plane, in fact TS4a is
nearly planar and only the exocyclic CdC double bond attached
to the nitrogen atom projects out of the molecular plane. In a
classical [3 + 2] cycloaddition the interacting π-systems in the
transition state are parallel,3 whereas in TS4a the developing
π system of ethylene and that of the remaining molecular frame
are perfectly perpendicular, indicating that, on going backward
from 16a plus ethylene to TS4a, the nonbonding lone pairs at
the heteroatoms interact with the p orbitals of ethylene. In
accordance, the second order perturbation analysis4 along the
IRC coordinate shows as the more relevant donor-acceptor
interactions Lp1Ofπ*C8-C11 and πC8-C11fσ*N-S. The
geometry of the forming thiazole ring in TS4a does not
correspond to that expected for a disrotatory 6π-electron
5-center electrocyclizations as all the atoms are in the molecular
plane.5 Going forward from E-17a to TS4a the most relevant
donor-acceptor interactions along the IRC are LpNfσ*S-C11,
Lp1Ofπ*C3-C4 and π*C3-C4fπ*N-C2. Therefore, the
electronic movements in the direct and inverse reaction paths
passing through TS4a could be represented as shown in Figure
2. Accordingly, the planar geometry and the orbital topology
bonding orbitals interchange roles), both confer pseudopericy-
clic6 characteristics to this transition state.
As the electronic reorganization only takes place at the
periphery of TS4a, and the single C4-S bond not interven-
ing, the transformation of E-17a into 16a could also be
viewed as a particular case of vinylogous retro-thia-ene
reaction (Figure 2) in which the enophile component is
ethylene and the vinylogous ene partner is triheterosubsti-
tuted, the migrating atom being sulfur.
In conclusion, we have experimentally unveiled and com-
putationally scrutinized an unprecedented tandem reaction
composed of a [1,5]-H shift and a further mechanistic step of
pseudopericyclic characteristics which is structurally related with
vinylogous ene reactions. As these latter processes are scarcely
known,7 the reactions reported here would require more indepth
studies, a part of which is currently underway in our laboratory.
Acknowledgment. This work was supported by the
Ministerio de Educacion y Ciencia of Spain and FEDER
(Projects CTQ2005-02323/BQU and CTQ2008-05827/BQU),
and Fundacion Seneca-CARM (Project 08661/PI/08). B.B.
also thanks Fundacion Seneca-CARM for a fellowship.
Supporting Information Available: Experimental details
for the synthesis of compounds 7, 8 and 10. Spectral data
(NMR, IR, MS, elemental analyses) for compounds 7, 8,
and 10. CIF file of 10a. 1H and 13C NMR spectra of
compounds 10. Details of computational procedures, geom-
etries, Cartesian coordinates, IRC calculations on TS4a, and
energies for all the stationary points. This material is available
free of charge via the Internet at http://pubs.acs.org.
OL9001416
(6) (a) Ross, J. A.; Seiders, R. P.; Lemal, D. M. J. Am. Chem. Soc.
1976, 98, 4325–4327. (b) Birney, D. M. J. Org. Chem. 1996, 61, 243–251.
(c) de Lera, A. R.; Alvarez, R.; Lecea, B.; Torrado, A.; Cossio, F. P. Angew.
Chem., Int. Ed. 2001, 40, 557–561. (d) Birney, D. M. Org. Lett. 2004, 6,
851–854. (e) Jones, G. O.; Xuechen, L.; Hayden, A. E.; Houk, K. N.;
Danishefsky, S. J. Org. Lett. 2008, 10, 4093–4096. (f) Alajarin, M.; Ortin,
M.-M.; Sanchez-Andrada, P.; Vidal, A.; Bautista, D. Org. Lett. 2005, 7,
5281–5284.
Figure 2. Electronic reorganization leading to TS4a.
(7) We could only locate two documents reporting on vinylogous ene
reactions: (a) Gollnick, K.; Griesbeck, A. Tetrahedron 1984, 40, 3235–
3250. (b) Maas, H.; Roeper, M. Preparation of alkoxydodecatrienes and
analogous by a Vinylogous ene reaction; BASF AG: Germany, DE 1995069,
A1 20001005; CAN 2000, 133, 281536. It is not surprising that all-carbon
vinylogous ene reactions are so rarely reported. Whereas they would involve,
if concerted, cyclic transition states involving eight electrons, they could
hardly compete with the alternative Diels-Alder reaction and other classical
six-electron ene processes. Only the special structural and electronic features
of the transformation E-17 into 16 (heteroatoms, lone-pair interactions,
pseudopericyclic character, and aromaticity recovery) seem to explain why
this process is easily amenable.
of TS4a, containing orbital disconnections in the cyclic array
of overlapping orbitals (where orthogonal bonding and non-
(3) Cossio, F. P.; Morao, I.; Jiao, H.; Schleyer, P. v. R. J. Am. Chem.
Soc. 1999, 121, 6737–6746, and references therein. .
(4) (a) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys.
1985, 83, 735–746. (b) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem.
ReV. 1988, 88, 899–926. (c) Reed, A. E.; Schleyer, P. v. R. J. Am. Chem.
Soc. 1990, 112, 1434–1445.
(5) Houk, K. N.; Li, Y.; Evanseck, J. D. Angew. Chem., Int. Ed. Engl.
1992, 31, 682–708.
1368
Org. Lett., Vol. 11, No. 6, 2009