Unexpected reactivity of annulated 3H-benzothiazol-2-ylideneamines as an acyl transfer catalyst

Article abstract of DOI:10.1016/j.tetlet.2006.04.109

Catalytic ability of annulated 3H-benzothiazol-2-ylideneamines and 1H-pyridin-2-ylideneamines for acylation of alcohols was investigated and 3,4-dihydro-2H-9-thia-1,4a-diazafluorene (2b) was found to be an extremely effective catalyst, the reaction with which was faster than that with DMAP.

Full text of DOI:10.1016/j.tetlet.2006.04.109

Tetrahedron Letters 47 (2006) 4347–4350
Unexpected reactivity of annulated
3H-benzothiazol-2-ylideneamines as an acyl transfer catalyst
Megumi Kobayashi and Sentaro Okamoto*
Department of Applied Chemistry, Kanagawa University, 3-27-1 Rokkakubashi, Kanagawa-ku, Yokohama 221-8686, Japan
Received 23 March 2006; revised 20 April 2006; accepted 21 April 2006
Abstract—Catalytic ability of annulated 3H-benzothiazol-2-ylideneamines and 1H-pyridin-2-ylideneamines for acylation of alcohols
was investigated and 3,4-dihydro-2H-9-thia-1,4a-diazafluorene (2b) was found to be an extremely effective catalyst, the reaction with
which was faster than that with DMAP.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Acylation or esterification of alcohols is an important,
n
Ph
n
basic transformation in organic and bioorganic chemis-
try. Non-peptide small organic molecules, such as pyri-
dine and imidazole derivatives, are well known as an
effective catalyst for the transformation. Recently, an
enantioselective variant catalyzed by non-peptide com-
pounds has been of great interest as a versatile means
for acylative kinetic resolution and asymmetric desym-
metrization.1 Many of the effective catalysts developed
so far have a 4-aminopyridine nucleus as a reactive
core, achiral derivatives of which, such as 4-N,N-
dimethylaminopyridine (DMAP), have been used in a
wide range of organic reactions.1 A new aspect has been
shown by Birman et al. who introduced a new class of
acyl transfer catalysts, 2,3-dihydroimidazo[1,2-a]pyri-
dines, 1a and its derivatives, and elegantly applied their
optically active derivatives such as 1c as a catalyst to
kinetic resolution of racemic aryl alcohols (Fig. 1).2
The results prompted us to investigate the possibility
of annulated benzothiazol-2-ylideneamines 2 as a cata-
lyst, expecting that they might have a more nucleophilic
nitrogen atom than compounds of the type 1 and could
exhibit good catalytic activity. In the course of the study
we found that 3,4-dihydro-2H-9-thia-1,4a-diazafluorene
(2b) was an unexpectedly effective catalyst for acyl trans-
fer reaction, the efficiency of which was higher than that
of the compounds illustrated in Figure 1, including
4-N,N-dimethylaminopyridine (DMAP, 4). Herein we
report the results of investigation on catalytic reactivity
and the mechanism of annulated benzothiazol-2-yliden-
N
N
N
N
N
N
S
S
2a: n = 0
2b: n = 1
1a: n = 0
1b: n = 1
3
NMe2
Ph
Ph
N
N
N
F3C
N
S
N
1c
2c
DMAP (4)
Figure 1. Structures of compounds 1, 2, 3, and 4.
amines 2 and pyridin-2-ylideneamines 1 as a reactive
core of the catalyst. During these investigations, quite
recently Birman et al. have reported that tetramisole
(3) and its benzo derivative 2c were efficient enantioslec-
tive acyl transfer catalysts, the latter of which exhibited
a high selectivity factor in the 100–350 range in the
kinetic resolution of secondary benzylic alcohols.3
Annulated benzothiazol-2-ylidenamines 2a and 2b were
synthesized according to the procedure illustrated in
Scheme 1.4,5 Thus, 2-chlorobenzthiazoles were substi-
tuted with 1,2- or 1,3-aminoalcohols followed by chlori-
nation of the resulting alcohols and cyclization under
basic conditions. 2,3-Dihydroimidazo[1,2-a]pyridine 1a
and 1b were prepared by the procedure reported with
minor modification.2
First, we carried out the reaction of 1-phenylethylalco-
hol with acetic anhydride (Ac2O) in the presence of 1a,b,
2a,b or DMAP under the identical reaction conditions
*
Corresponding author. Tel.: +81 45 481 5661; fax: +81 45 413
9770; e-mail: okamos10@kanagawa-u.ac.jp
0040-4039/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.04.109

4348
M. Kobayashi, S. Okamoto / Tetrahedron Letters 47 (2006) 4347–4350
N
S
N
i
OH
Cl
N
n
H
S
n
ii (n = 0)
or
iii (n = 1)
N
iv
2a: 47%
2b: 67%
Cl
N
H
S
n = 0: 67%
n = 1: 86%
Scheme 1. Synthesis of 2: (i) HOCH2(CH2)nCH2NH2, neat, 130 ꢁC; (ii)
SOCl2, DMF, reflux; (iii) SOCl2, CHCl3, reflux; (iv) K2CO3, CHCl3.
(Eq. 1). Thus, to a solution of 1-phenylethylalcohol
(1.0 mmol), Et3N (1.5 mmol) and the catalyst
(5 mol %) in CH2Cl2 (5 mL) was added Ac2O
(1.5 mmol) at 22–24 ꢁC. The TLC analysis of the reac-
tion showed the approximate reaction time for con-
sumption of the substrate alcohol to be >40, >40, 15,
<0.25, and <0.25 h, respectively, with 1a,b, 2a,b, and
DMAP (4).
OAc
OH
catalyst
ð1Þ
+
Ac2O
Et3N, CH2Cl2, rt
Ph
Ph
To confirm the relative efficiency of catalysts, the reac-
tion with 1a,b or 2a was performed with 1-phenylethyl-
alcohol (1.0 mmol), Ac2O (1.0 mmol), Et3N (1.5 mmol),
and a catalyst (5 mol %) in 1 mL of CH2Cl2, while that
with 2a or DMAP was carried out with 1-phenylethyl-
alcohol (1.0 mmol), Ac2O (1.0 mmol), Et3N (1.5 mmol),
and a catalyst (1 mol %) in 10 mL of CH2Cl2. The time
course of the reaction was traced by TLC analysis: thus,
images of stained TLC were allowed to be imported as
digital data to computer by image-scanning and ana-
lyzed by the image processing and analysis software
(NIH image)6 and the amounts of the alcohol and the
acetate were quantified to determine the conversion.
The results are illustrated in Figure 2. As revealed in
Figure 2a, pyridine derivatives 1a,b and dihydroimidazo
derivative of benzothiazole 2a have similar catalytic
activity and the reactions with them were much slower
than that with DMAP and, surprisingly, a dihydropy-
rimido derivative of benzothiazole 2b was found to be
an extremely effective catalyst and the activity was high-
er than that of DMAP. These reactions could be treated
as a pseudo-second order kinetics (Fig. 2b).7 From the
time versus 1/(1-x) plots (x = conversion, 0 6 x 6 1)
shown in Figure 2b, TOF50 [turnover factor at 50%
conversion, hꢀ1 (mol of product/(mol of catalyst · h))]8,9
was calculated for each catalyst: 3.04 for 1a, 3.61 for 1b,
3.57 for 2a, 531 for 2b, and 278 for DMAP. Although a
similar trend that the dihydropyrimido derivative was
more active than the dihydroimidazo derivative was ob-
served for 1 and 2, it was noteworthy that the difference
between 2a and 2b was much larger than that between
1a and 1b.
Figure 2. Kinetic study: The reaction was carried out with 5 mol % of
1a,b or 2a in a 1.0 M solution, or with 1 mol % of 2b or DMAP in a
0.1 M solution (see text). (a) Time course of conversion, (b) time versus
1/(1 ꢀ conversion) plots.
tions shown in Eqs. 2–4, where the reactions of Eqs. 5
and 6 among catalyst compound, AcOH and Et3N
may be involved. Regarding Eqs. 5 and 6, in all cases,
NMR spectra of the 1:1:1 mixture of the catalyst com-
pound, AcOH and Et3N in CDCl3 showed peaks of a
neutral catalyst compound and of an acetic acid salt of
triethylamine quantitatively.
Cat: þ Ac2Oꢀ½Cat:–AcꢁþꢀOAc
ð2Þ
½Cat:–AcꢁþꢀOAc þ ROH ! ROAc þ Cat: þ HOAc ð3Þ
HOAc þ Et3N ! ½Et3NHꢁþꢀOAc
Cat: þ HOAcꢀ½Cat:–HꢁþꢀOAc
ð4Þ
ð5Þ
ð6Þ
ꢀ
½Cat:–Hꢁþ OAc þ Et3NꢀCat: þ ½Et3NHꢁþꢀOAc
Next, NMR experiments of the reaction of catalyst com-
pounds and Ac2O were carried out. The addition of
1 equiv of Ac2O to DMAP in CDCl3 at room tempera-
ture gave no new peak other than DMAP and Ac2O
(data not shown), the results of which indicate that an
acylated intermediate [DMAP–Ac]+-OAc becomes ther-
modynamically much less stable than DMAP, due to
loss of aromaticity of the pyridine ring. Meanwhile, as
shown in Figure 3, the mixture of benzothiazole deriva-
tives 2 and 1 equiv of Ac2O generated a considerable
amount of the corresponding acylated compounds
[Cat.–Ac]+-OAc. It can be explained by assuming that
the acylated compounds of 2a and 2b may be stabilized
by aromaticity of the newly formed thiazole structure
and resonance effect of the nitrogen cation by the benz-
With these results in hand, we next carried out the stoi-
chiometric reaction of catalyst compounds in CDCl3
1
which were analyzed by 500 MHz H NMR. The reac-
tion shown in Eq. 1 might consist of the following reac-

M. Kobayashi, S. Okamoto / Tetrahedron Letters 47 (2006) 4347–4350
4349
pounds, and [DMAP–Ac]+ may be expected to be rela-
tively less stable. These assumptions are in good
agreement with the results of NMR experiments for
Eq. 2 mentioned above. Moreover, it may be considered
that pyrimido derivatives 1b and 2b could be acylated
easier than the corresponding imidazo derivatives 1a
and 2a, respectively. Acylated 2b is much more stable
than acylated 2a, presumably due to a larger ring strain
of acylated 2a than that of acylated 2b. Although the
confirmation of details of the mechanism must await
further study, it may totally be assumed that easy gener-
ation of the acylated intermediate from 2b due to a large
energy gain by delocalizing stabilization of the generated
positive charge might mainly affect the catalytic activity
of 2b.
In summary, we have investigated the reactivity of annu-
lated benzothiazol-2-ylidenamines 2 as an acyl transfer
catalyst and found that the dihydropyrimido derivative
of benzothiazole 2b exhibited extremely high activity,
which might be a good candidate as a reactive core for
development of an asymmetric catalyst.
Acknowledgements
We thank Professor Vladimir B. Birman, Washington
University, St. Louis, for valuable comments and discus-
sion through private communication.
Figure 3. 500 MHz 1H NMR spectra for a 1:1 mixture of 2a and Ac2O
(the upper spectra) or for a 1:1 mixture of 2b and Ac2O (the lower
spectra) in CDCl3 at room temperature.
ene ring. Interestingly, although the spectra of a mixture
of 2b and Ac2O involved one set of peaks corresponding
to an acylated intermediate, those of a mixture of 2a and
Ac2O were much more complicated. The different
behaviors thus observed for 2a and 2b cannot be ex-
plained at this time but may reflect the difference of their
relative catalytic activity.
References and notes
1. Recent reviews: Vedejs, E.; Jure, M. Angew. Chem., Int.
Ed. 2005, 44, 3974; Dalko, P. I.; Moisan, L. Angew.
Chem., Int. Ed. 2004, 43, 5138; France, S.; Guerin, D. J.;
Miller, S. J.; Lectka, T. Chem. Rev. 2003, 103, 2985.
2. Birman, V. B.; Uffman, E. W.; Jiang, H.; Li, X.; Kilbane,
C. J. J. Am. Chem. Soc. 2004, 126, 12226; Birman, V. B.;
Jiang, H. Org. Lett. 2005, 7, 3445.
To consider the large difference of catalytic reactivity
between 2a and 2b, we tried to estimate energy differ-
ences among catalyst compounds and their acylated
intermediates by calculation using molecular mechanics
(MM2).10 From the results shown in Figure 4, acylation
of 2b seems to be much easier than that of other com-
3. Birman, V. B.; Li, X. Org. Lett. 2006, 8, 1351.
4. Young, R. C.; Mitchell, R. C.; Brown, T. H.; Ganellin, C.
R.; Griffiths, R.; Jones, M.; Rana, K. K.; Saunders, D.;
Smith, T. R.; Sore, N. E.; Wilks, T. J. J. Med. Chem. 1988,
31, 656; Abstract for Anisimova, V. A.; Levchenko, M. V.
Khimiya Geterotsiklicheskikh Soedinenii 1987, 59.
5. Compound 2a: 1H NMR (600 MHz, CDCl3) d 7.26
(d, J = 7.6 Hz, 1H), 7.19 (t, J = 7.5 Hz, 1H), 6.95
(t, J = 7.6 Hz, 1H), 6.67 (d, J = 7.6 Hz, 1H), 4.35
(t, J = 8.9 Hz, 2H), 3.85 (t, J = 8.9 Hz, 2H); 13C NMR
(150 MHz, CDCl3) d 162.7, 138.2, 127.3, 124.7, 122.8,
122.5, 110.7, 42.6, 40.9. 2b: 1H NMR (600 MHz, CDCl3) d
7.27 (d, J = 7.6 Hz, 1H), 7.20 (dd, J = 7.6, 8.2 Hz, 1H),
7.00 (t, J = 7.6 Hz, 1H), 6.80 (d, J = 8.2 Hz, 1H), 3.79 (t,
J = 6.2 Hz, 2H), 3.55 (t, J = 5.7 Hz, 2H), 2.01 (m, 2H);
13C NMR (150 MHz, CDCl3) d 159.2, 140.0, 126.2, 122.4
(2C), 121.9, 108.1, 44.1, 42.6, 19.3. Compound 1b: 1H
NMR (500 MHz, CDCl3) d 6.85 (d, J = 6.9 Hz, 1H), 6.67
(t, J = 6.3 Hz, 1H), 6.04 (d, J = 9.8 Hz, 1H), 5.57 (t,
J = 6.9 Hz, 1H), 3.74 (t, J = 6.3 Hz, 2H), 3.28 (t,
J = 6.3 Hz, 2H), 1.77 (m, 2H); 13C NMR (125 MHz,
CDCl3) d 151.0, 136.8, 132.9, 122.7, 103.2, 49.5, 43.3, 20.0.
6. Version 1.63 for Macintosh was used. This is available
via the internet at http://rsb.info.nih.gov/nih-image/
download.html.
O
N+
N
n
N+
N
n
O
N+
N
1a (n=0)
1b (n=1)
-15.4
-16.8
-10.6
-22.1
N+
S
N+
S
n
n
N
N
O
O
O
-9.85
DMAP
-12.0
-23.7
2a (n=0)
2b (n=1)
-14.8
-25.5
Figure 4. The preliminary results of calculation of energy differences
between catalyst compound and the acylated derivative: Cat. to [Cat.–
Ac]+ (kcal/mol).

4350
M. Kobayashi, S. Okamoto / Tetrahedron Letters 47 (2006) 4347–4350
9. Line-fitting was performed using a graph software Cricket
7. d[ROH]/dt = k[ROH][Ac2O]. When [ROH]0 = [Ac2O]0 =
a and x = conversion (0 6 x 6 1), 1/[a(1 ꢀ x)] = kt + 1/a.
Graph 1.3.2 for Macintosh.
If k0 = ak (constant), 1/(1 ꢀ x) = k0t + 1.
10. The MM2 calculations resulting in the data in Figure 4
were performed using CAChe software (Quantum 4.9 for
Macintosh, Fujitsu Ltd.).
8. 0.50 mmol/[time (h) for 50% conversion][mmol of
catalyst].