Deprotection of N-tosylated indoles and related structures using cesium carbonate

Article abstract of DOI:10.1016/j.tetlet.2006.06.132

A very mild, efficient, and convenient method for deprotection of N-tosylated indoles and related structures by cesium carbonate in THF-MeOH is described.

Full text of DOI:10.1016/j.tetlet.2006.06.132

Tetrahedron Letters 47 (2006) 6425–6427
Deprotection of N-tosylated indoles and related structures
using cesium carbonate
*
Joginder S. Bajwa, Guang-Pei Chen, Kapa Prasad, Oljan Repi cˇ and Thomas J. Blacklock
Process R&D, Chemical and Analytical Development, Novartis Pharmaceuticals Corporation, One Health Plaza,
East Hanover, NJ 07936, USA
Received 12 June 2006; revised 22 June 2006; accepted 23 June 2006
Abstract—A very mild, efficient, and convenient method for deprotection of N-tosylated indoles and related structures by cesium
carbonate in THF–MeOH is described.
Ó 2006 Elsevier Ltd. All rights reserved.
Many compounds in medicinal chemistry contain
heteroaromatics as part of the structure and the NH-
functionality, that is, present in some of these molecules
needs to be protected during the synthesis by a suitable
group, that is, easily removable. Tosyl group is one such
blocking group, and its deprotection is usually accom-
plished by one of following methods: dissolving metal
reductions (Li or Na) in ammonia, alcohol or HMPA;
single electron transfer reagents such as sodium nap-
Cs CO3
2
+
p
3
-MePhSO CH3
N
THF/CH OH
N
3
Ts
H
MeOH
1
2
p
-MePhSO H
+
CH OCH3
3
3
Scheme 1.
thalenide, Na–Hg, n-Bu SnH; reducing agents L-Select-
ride, Red-Al; photolysis. Other reagents used for this
experiments was performed with N-tosyl-5-bromoindole
9, and the results are shown in Table 1. The reactions
were carried out in a mixed solvent THF–MeOH (2:1).
Ideally, the solvent of choice is methanol but simple
N-tosyl indoles (such as 9) are highly liphophilic and
are not soluble in methanol. When the reactions were
carried out in a single solvent methanol or ethanol, the
reactions times were longer. The reaction did not work
when carried out in iso-propanol.
3
1
,2
deprotection are highly nucleophilic Gilman’s reagent
1
PhMe SiLi, highly basic NaOH (or KOH) in alcohol
2
3
,4
5
solvents at high temperature, KF on alumina under
microwaves, n-Bu NF in refluxing THF, Mg–MeOH,
polymer-supported potassium thiophenolate, and
HSCH COOH/LiOH.
6
7
4
8
9
2
Most of these methods, while effective in a number
of cases, have serious incompatibilities with other
functional groups. While making an indole derived drug
substance, we needed to effect N-detosylation. Several
literature methods were tried which invariably gave
low yields due to the formation of many by-products.
This led to the development of a new and a very mild
method for N-detosylation of indoles and related struc-
tures using cesium carbonate (Scheme 1). To determine
the optimum stoichiometry of cesium carbonate, a set of
It is apparent from the results in Table 1 that 3 equiv of
cesium carbonate are required to achieve a reasonable
reaction rate for N-detosylation of indole 9. It is also
a
Table 1. Reaction of N-tosyl-5-bromoindole 9 with varying amounts
of cesium carbonate in THF–MeOH
b
Entry
Amount of
Cs CO (equiv)
Reaction
time (h)
Conversion
10/9
2
3
1
2
3
1.0
2.0
3.0
15
15
15
69/31
95/5
99/1
Keywords: N-detosylation; Azaindoles; Indoles; Imidazoles; Cesium
carbonate.
a
2 3
A mixture of 9 (0.25 mmol) and Cs CO (1–3 equiv) in THF–MeOH
*
Corresponding author. Tel.: +1 862 778 3474; fax: +1 973 781
384; e-mail: joginder.bajwa@novartis.com
(3 mL, 2:1) was stirred at 22 °C and followed by HPLC.
Determined by HPLC.
b
4
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.06.132

6426
J. S. Bajwa et al. / Tetrahedron Letters 47 (2006) 6425–6427
a
a
Table 2. Reaction of N-tosyl-5-bromoindole 9 with various alkali
Table 3. Deprotection of N-tosyl indoles and related heterocycles
metal carbonates in THF–MeOH in the presence/absence of water
12
Substrate
Temperature (°C)/
Product
b
Entry
Alkali metal
Additive
(equiv)
Conversion
10/9
reaction time (h);
b
carbonate (3.0 equiv)
conversion
1
2
3
4
5
Li
Na
CO
CO
CO
2
CO
CO
3
—
—
—
—
h0.1/i 99.9
h0.1/i 99.9
24/76
99.8/0.2
h0.1/i 99.9
h0.1/i 99.9
h0.1/i 99.9
h0.1/i 99.9
23/77
2
3
N
H
N
22/90; 95/5 (2/1)
K
2
3
Ts
6
4/0.5; >99/1 (2/1)
Cs
2
3
1
2
Li
2
3
H
2
H
2
H
2
H
2
H
2
H
2
H
2
H
2
O (1.0)
O (40.0)
O (1.0)
O (40.0)
O (1.0)
O (40.0)
O (1.0)
O (40.0)
Me
Me
6
7
8
Na
2
CO
3
N
N
2
6
2/70; 2/98 (4/3)
4/48; 97/3 (4/3)
Ts
H
K
2
CO
3
3
4
0.4/99.6
99.7/0.3
7/93
Me
Me
Cs
2
CO
3
a
22/70; 12/88 (6/5)
64/8; 98/2 (6/5)
A mixture of 9 (0.25 mmol), alkali metal carbonate (3 equiv) and
water (0, 1 or 40 equiv) in THF–MeOH (3 mL, 2:1) was stirred at
N
N
H
Ts
2
2 °C for 15 h.
5
6
b
Determined by HPLC.
MeO
MeO
N
Ts
22/18; 7/93 (8/7)
N
interesting to note that the initial by-product observed
in this reaction is methyl p-toluenesulfonate (Scheme
H
6
4/2.5; >99/1 (8/7)
7
8
1
). If the reaction rate is slow enough as it is in the pres-
ent case, the p-MePhSO Me formed reacts further with
Br
Br
3
methanol to give p-MePhSO H and MeOMe. On the
3
N
22/15; >99/1 (10/9)
N
other hand, p-toluenesulfonic acid and dimethyl ether
are the only by-products observed when the reaction
was carried out at reflux temperature.10
Ts
H
9
10
CO Et
COOEt
2
Next we examined N-detosylation of N-tosyl-5-bromo-
indole 9 with different alkali metal carbonates. The effect
of water on this reaction was also studied. The results
are shown in Table 2.
2
2/1; >99/1 (14/13)
N
N
Ts
H
11
12
O N
2
O N
2
It is clear from the results in Table 2 (entries 1, 2, 5, and
6
) that Li CO and Na CO with or without water are
2 3 2 3
N
0–5/5; >99/1 (12/11)
22/2; >99/1 (16/15)
22/0.5; >99/1 (18/17)
N
H
Ts
totally ineffective in deprotection of tosyl group in 9.
Potassium carbonate does effect detosylation of 9 (entry
3
While the use of one equivalent of water in the reaction
mixture has practically no effect, excess of water
c
13
14
) but it is much less effective than Cs CO (entry 4).
2 3
N
N
N
N
(
7
40 equiv) virtually shuts down the reaction (entry 3 vs
and entry 4 vs 8).
Ts
H
1
5
16
NHMe
NHMe
Based upon the optimized conditions with respect to the
solvent, the stochiometry, and the alkali metal carbon-
ate described above (Tables 1 and 2), a variety of substi-
tuted indoles and azaindoles were subjected to
N-detosylation with Cs CO . The results are shown in
N
N
N
N
N
N
Ts
H
2
3
17
18
Ph
1
1
Table 3.
N
N
Ph
N
N
2
2/0.5; >99/1 (20/19)
The reaction of unactivated N-tosyl indole 1 with cesium
carbonate in THF–MeOH at room temperature was
rather slow and gave 95% conversion after 90 h. How-
ever, when the reaction mixture was heated at reflux
Ts
H
1
9
20
a
2 3
A mixture of N-tosylindole (0.5 mmol) and Cs CO (1.5 mmol) in
THF–MeOH (3 mL, 2:1) was stirred at a specified temperature and
progress of the reaction was followed by HPLC.
Determined by HPLC.
(
64 °C), complete deprotection of the tosyl group was
achieved in just 0.5 h. The reaction is slower in MeOH
alone as compound 1 has limited solubility in MeOH.
The reaction rate as expected is very sensitive to both
electronic and steric effects of the substituents. For
example, with indole 3 there is no reaction at ambient
temperature for up to 70 h. Even at reflux, the reaction
b
c
Unidentified by-product (9.6%) was also formed.
was slow, and it took 48 h to effect 97% conversion of 3
to 4. In contrast, the deprotection of N-tosyl-3-methyl-

J. S. Bajwa et al. / Tetrahedron Letters 47 (2006) 6425–6427
6427
indole 5 was essentially complete in 8 h at reflux. Elec-
tron-donating substituents such as methoxy group in 7
also slow down the reaction. Nevertheless, complete
deprotection of the tosyl group was achieved at reflux
in 2.5 h. Electron withdrawing substituents such as
bromo, vinyl ester, and nitro groups greatly facilitate
the nucleophilic attack. For example, N-tosyl-5-bromo-
indole 9 and the unsaturated ester derivative 13 were con-
verted into the corresponding free indoles 10 and 12 in
quantitative yields in 15 h and 1 h, respectively. Depro-
tection of 11 was carried out in THF–EtOH rather than
in THF–MeOH to avoid any trans-esterification by-
products. Also, it is interesting to note that no Michael
addition by-product(s) was observed.5,9 The deprotec-
tion of N-tosyl-5-nitroindole 13 was complete in 0.5 h
at 22 °C but it also gave an unidentified by-product.
To minimize the formation of the by-product the reac-
tion was carried out at 0–5 °C and the product 14 was
formed in 90.4% yield.
2. (a) Green, T. W.; Wuts, P. G. M. Protective Groups in
Organic Chemistry, 3rd ed.; Wiley Interscience: New York,
2
005; (b) Kocie n´ ski, P. J. Protective Groups, 3rd ed.;
Thieme: New York, 2003.
3
4
5
. Gilbert, E. J.; Chisholm, J. D.; Van Vranken, D. L. J. Org.
Chem. 1999, 64, 5670–5676.
. Garg, N. K.; Sarpong, R.; Stoltz, B. M. J. Am. Chem. Soc.
2
002, 124, 13179–13184.
. Sabitha, G.; Abraham, S.; Subba Reddy, B. V.; Yadav, J.
S. Synlett 1999, 1, 1745–1746.
6. Witulski, B.; Alayrac, C. Angew. Chem., Int. Ed. 2002, 41,
3281–3284.
7
. (a) Nyasse, B.; Grehn, L.; Ragnarsson, U. Chem. Com-
mun. 1997, 1017–1018; (b) Muratake, H.; Natsume, M.
Hetrocycles 1989, 29, 783–794.
8
9
. MacCoss, R. N.; Henry, D. J.; Brain, C. T.; Ley, S. V.
Synlett 2004, 675–678.
. Haskins, M. C.; Knight, D. W. Tetrahedron Lett. 2004, 45,
5
99–601.
1
1
0. The formation of p-MePhSO Me and p-MePhSO H in the
3 3
reaction mixture was confirmed by comparison of HPLC
analysis of the reaction mixture with authentic samples.
The formation of dimethyl ether was assumed on the basis
of the corresponding reaction carried out in n-butanol
where the formation of the dibutyl ether by-product was
confirmed by GC–MS.
We have successfully extended this methodology to N-
detosylation of azaindoles. Azaindoles have lower pKa
1
3
values compared to indoles and thus are expected to
be better leaving groups. Indeed, detosylation of aza-
indoles 15 and 17 were complete at ambient temperature
in only 2 h and 0.5 h, respectively.
1. A representative procedure is as follows: N-Tosyl 5-
bromoindole 9 (2.1 g, 6.0 mmol) was dissolved in a
mixture of THF (50 mL) and MeOH (25 mL) at ambient
temperature. Cesium carbonate (5.85 g, 18.0 mmol) was
added to the clear solution. The resulting mixture was
stirred at ambient temperature and the progress of the
reaction was monitored by HPLC. When the reaction was
complete (18 h), the mixture was evaporated under
vacuum. To the residue was added water (25 mL) and
the mixture was stirred at ambient temperature for 10 min.
The solids were filtered, washed with water (15 mL) and
dried at 45 °C/1.5 mbar/18 h to give crude product 10
Imidazole also acts as a good leaving group and as a
result, detosylation of 2-phenylimidazole derivative 19
proceeded very rapidly at room temperature to give
product 20 in quantitative yield. It is interesting to note
that the reaction time (0.5 h) is much shorter than 2.5 h
9
as reported by the thioglycolate method.
In summary, we have developed a very mild and efficient
method for detosylation of a wide range of indoles,
azaindoles, and imidazoles. Cesium carbonate is readily
available, inexpensive, and easy to handle. This method
should prove useful for deprotection of tosyl groups in
indoles, azaindoles, and in situations where the other
methods are not selective.
(
1.156 g, 98.3%). An analytically pure sample was
obtained in 88.2% yield and 100% purity (HPLC) by
recrystallization of crude product (1.156 g) from n-heptane
(
15 mL).
12. All N-tosyl derivatives were prepared by following the
procedure described in Poissonnet, G.; Th e´ ret-Bettiol,
M.-H.; Dodd, R. H. J. Org. Chem. 1996, 61, 2273–
2
282.
1
3. The pKa of indole is reported to be 20.95 according to:
Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456–463; The
pKa’s of azaindoles are reported to be 4.50–8.26 according
to: LeHyaric, M.; Vieira de Almeida, M.; Nora de Souza,
M. V. Quim. Nova 2002, 25, 1165–1171.
References and notes
1
. Fleming, I.; Frackenpohl, J.; Ila, H. J. Chem. Soc., Perkin
Trans. 1 1998, 1229–1235, and Ref. 8 cited there in.