2-Naphthalenesulfonyl as a tosyl substitute for protection of amino functions. Cyclic voltammetry studies on model sulfonamides and their preparative cleavage by reduction

Article abstract of DOI:10.1021/jo990695q

With the aim to develop a practically useful, reductively more labile alternative to tosyl for protection of amino functions, initially a number of N-arenesulfonyl-protected heterocycles (pyrroles, imidazoles, indole, and carbazole) have been prepared and

Full text of DOI:10.1021/jo990695q

J . Org. Chem. 1999, 64, 7135-7139
7135
2
-Na p h th a len esu lfon yl a s a Tosyl Su bstitu te for P r otection of
Am in o F u n ction s. Cyclic Volta m m etr y Stu d ies on Mod el
Su lfon a m id es a n d Th eir P r ep a r a tive Clea va ge by Red u ction †
Barth e´ l e´ my Nyasse,‡,§ Leif Grehn, Hernani L. S. Maia, Luis S. Monteiro, and
§
|
|
Ulf Ragnarsson*,
§
Department of Organic Chemistry, Faculty of Science, University of Yaound e´ I,
Box 812, Yaound e´ , Cameroon, Department of Biochemistry, University of Uppsala, Biomedical Center,
PO Box 576, SE-751 23 Uppsala, Sweden, and Departamento de Quimica, Universidade do Minho,
Gualtar, P-4700-320 Braga, Portugal
Received April 26, 1999
With the aim to develop a practically useful, reductively more labile alternative to tosyl for protection
of amino functions, initially a number of N-arenesulfonyl-protected heterocycles (pyrroles, imida-
zoles, indole, and carbazole) have been prepared and studied by cyclic voltammetry (CV). The
recorded activation potentials vary from -1.32 to -1.99 V (vs SCE). In N-sulfonylazolides such as
tosylindole the cathodic potentials are shifted by over 0.5 V compared to simple sulfonamides. An
additional effect of the sulfonic acid component is also indicated. Among the compounds studied,
1
- and 2-naphthalenesulfonylindole give CV peaks at about 0.4 and 0.2 V, respectively, less negative
potential than tosylindole. To further investigate naphthalenesulfonyl for this purpose, we have
also prepared a variety of simple 1- and 2-naphthalenesulfonyl derivatives and studied them
similarly. They have activation potentials above -2.14 V and are all smoothly cleaved by Mg/
MeOH. The latter reagent is capable of cleaving N-arenesulfonyl derivatives that give CV peaks
above -2.30 V, whereas Al(Hg) requires potentials above about -1.7 V. Selective cleavage of
2
-naphthalenesulfonyl in the presence of tosyl by Mg/MeOH is demonstrated. Several examples of
reductive cleavage of arenesulfonyl derivatives with Mg/MeOH, Al(Hg), and electrolysis on a
preparative scale are given.
8
In tr od u ction
silane. Nevertheless, in our opinion there is still a need
for improvements in this area, and we have therefore
been engaged in the development of an alternative to
tosyl that can be cleaved under milder conditions.
In connection with attempts to cleave tosyl reductively
The tosyl moiety is a classical and in many cases useful
protecting group for various nitrogen functions, including
amines.1 Nevertheless it is often not an ideal choice,
because it normally requires drastic conditions for its
cleavage so that many other sensitive functional groups,
simultaneously present, are affected. Various approaches
at eliminating this shortcoming, such as making it more
labile to acid by introduction of suitable substituents in
the benzene ring, were attempted and found to be
useful.1 Several new alternative procedures for tosyl
cleavage with the common feature that they shed light
on relevant structural effects were also proposed recently,
4
a
from various tosylcarbamates by Mg/MeOH, we also
studied 1-tosylindole (1a ) and found that, contrary to
tosylamides in general, it was rapidly and completely
cleaved, as briefly reported for a substituted 1-tosylindole
9
derivative by Yokoyama et al. As we had previously
investigated many arenesulfonamides and, especially, the
corresponding tert-butyl sulfonylcarbamates by cyclic
voltammetry (CV),10 we therefore determined the activa-
tion potential of compound 1a by this technique. We
obtained a value of -1.96 V (vs SCE) as compared to
a
2
a-c
using a wide variety of reagents such as SmI
2
,
Bu
3
-
3
4a-c
SnH, magnesium powder in methanol,
lithium pow-
-2.52 V for tosylamide, which amply demonstrated the
5
der in the presence of naphthalene, tetrabutylammoni-
powerful influence of the indole moiety. Moreover, be-
cause benzenesulfonyl has been cleaved from a substi-
tuted pyrrole derivative with Mg/MeOH,11 we decided to
prepare a few additional N-tosylated heterocycles to
measure their activation potentials similarly.
6
7
um fluoride, 2,2-dimethoxypropane, and iodotrimethyl-
*
Corresponding author: fax: +46 18 55 21 39; e-mail: urbki@
bmc.uu.se.
†
This paper is dedicated in friendship and with affection to Gotfryd
Kupryszewski on the occasion of his 70th birthday.
‡
University of Yaound e´ I.
University of Uppsala.
Universidade do Minho.
§
(7) Chandrasekhar, S.; Mohapatra, S. Tetrahedron Lett. 1998, 39,
695.
|
(
1) For an overview, see: (a) Greene, T. W.; Wuts, P. G. M. Protective
Groups in Organic Chemistry, 3rd ed.; Wiley-Interscience: New York,
999. (b) Kocienski, P. J . Protecting Groups; Thieme: New York, 1994.
2) (a) Vedejs, E.; Lin, S. J . Org. Chem. 1994, 59, 1602. (b) Goulaouic-
Dubois, C.; Guggisberg, A.; Hesse, M. J . Org. Chem. 1995, 60, 5969.
(8) Sabitha, G.; Subba Reddy, B. V.; Abraham, S.; Yadav, J . S.
Tetrahedron Lett. 1999, 40, 1569 and references therein.
(9) Yokoyama, Y.; Matsumoto, T.; Murakami, Y. J . Org. Chem. 1995,
60, 1486.
(10) (a) Grehn, L.; Gunnarsson, K.; Maia, H. L. S.; Montenegro, M.
I.; Pedro, L.; Ragnarsson, U. J . Chem. Res. 1988, (S) 399; (M) 3081.
(b) Maia, H. L. S.; Monteiro, L. S.; Degerbeck, F.; Grehn, L.; Ragnars-
son, U. J . Chem. Soc., Perkin Trans. 2 1993, 495. (c) Nyasse, B.; Grehn,
L.; Ragnarsson, U.; Maia, H. L. S.; Monteiro, L. S.; Leito, I.; Koppel,
I.; Koppel, J . J . Chem. Soc., Perkin Trans. 1 1995, 2025 and references
therein.
1
(
(
c) Knowles, H.; Parsons, A. F.; Pettifer, R. M. Synlett 1997, 271.
(
(
3) Parsons, A. F.; Pettifer, R. M. Tetrahedron Lett. 1996, 37, 1667.
4) (a) Nyasse, B.; Grehn, L.; Ragnarsson, U. Chem. Commun. 1997,
1
(
017. (b) Grehn, L.; Nyasse, B.; Ragnarsson, U. Synthesis 1997, 1429.
c) Alonso, D. A.; Andersson, P. G. J . Org. Chem. 1998, 63, 9455.
(
(
5) Alonso, E.; Ramon, D. J .; Yus, M. Tetrahedron 1997, 53, 14355.
6) Yasuhara, A.; Sakamoto, T. Tetrahedron Lett. 1998, 39, 595.
(11) Okabe, K.; Natsume, M. Tetrahedron 1991, 47, 7615.
1
0.1021/jo990695q CCC: $18.00 © 1999 American Chemical Society
Published on Web 08/31/1999

7
136 J . Org. Chem., Vol. 64, No. 19, 1999
Nyasse et al.
Ta ble 1. CV P ea k P oten tia ls for Som e N-Ar en esu lfon yl-
a n d Tw o N-Ar oyl-P r otected Heter ocyclesa
These were found to give peaks at higher potentials than
those of 1b-e, 0.77-0.94 V higher than that of tosyla-
mide. Finally, 1-(4-cyanobenzenesulfonyl)-indole (1k ) and
9-(4-cyanobenzenesulfonyl)-carbazole (1l) were also made
and studied. As could be expected from our previous
work,1 their CV peaks were further shifted in compari-
son with the previously discussed compounds (to -1.36
and -1.32 V, respectively). These syntheses were after-
ward followed up with 9-(2-naphthalenesulfonyl)-carba-
zole (1i), which was used in the preparative cleavage
experiments below.
b
-
EP /V
compd
arenesulfonyl derivative
1-tosylindole
(vs SCE)
1
1
1
1
1
1
1
1
1
1
1
1
1
1
a
b
c
d
e
f
g
h
i
j
k
l
m
n
1.96
1.98
1.86
1.84
1.78
1.99
1.58
1.75
1.59
1.59
1.36
1.32
1.91
1.74
2.52
0c
1-tosylpyrrole
1-tosylpyrrole-2-COOEt
1-tosylimidazole
c
1-tosylbenzimidazole
1-(4-fluorobenzenesulfonyl)-indole
1-(1-naphthalenesulfonyl)-indole
1-(2-naphthalenesulfonyl)-indole
c
c
c
c
9-(2-naphthalenesulfonyl)-carbazole
Results such as those described in Table 1, together
with previously available data, allow some conclusions
to be drawn. Because, in addition to the electronic effects
of substituents in the arenesulfonyl group observed
earlier,1 the aromatic systems and the amine moieties
obviously have a significant influence on the activation
potential of sulfonamides as measured by CV, a wide
variation in stability of sulfonamides toward reduction
should be expected. This must be kept in mind in the
fine-tuning of such protecting groups. Series of com-
pounds with known activation potentials are obviously
useful in determining the scope of various reducing
agents. Because seven of the N-arenesulfonylcarbamates
cleaved by magnesium powder in methanol4 had earlier
been studied by CV and found to give peaks in the range
from -1.44 to -2.28 V,1 at this stage it appears that
Mg/MeOH is able to efficiently cleave compounds with
activation potentials down to about -2.3 V (vs SCE), i.e.,
close to the normal potential of magnesium, whereas for
mercury-activated aluminum, as judged from the com-
pounds studied in this paper, the corresponding value
seems to be -1.7 to -1.8 V.
c
1-dansylindole
c
1-(4-cyanobenzenesulfonyl)-indole
9-(4-cyanobenzenesulfonyl)-carbazolec
1-benzoylindole
1-(1-naphthoyl)-indole
tosylamide
0c
a
All compounds prepared gave interpretable H and 13C NMR
1
b
spectra. Cathode, vitreous carbon; solvent, DMF; supporting
electrolyte, Bu4NBF4 0.1 mol dm ; substrate concentration,
0.005 mol dm
elemental analysis.
-
3
-
3 c
∼
. New compound which gave a satisfactory
A reexamination of some electrochemical data in the
light of our present experiences indicated that in the
a
1
0c
series of tert-butyl sulfonylcarbamates recently studied
the peak potentials were shifted by 0.19-0.30 V in
comparison with those of the related sulfonamides. On
the other hand, within the same two series of compounds,
the variations due to the arenesulfonyl moieties spanned
0c
0
.97 and 0.99 V, respectively. Considering the effect of
tert-butoxycarbonylation of arenesulfonamides with re-
spect to their reductive cleavage as illustrated by Mg/
MeOH,4 we set out to find an inexpensive alternative to
tosyl that would not require tert-butoxycarbonylation of
the sulfonamide but would allow reductive cleavage of
the sulfonamide itself. In this paper we present two
moieties that both seem to chemically fulfill these
requirements, i.e., 1- and 2-naphthalenesulfonyl, of which
the latter is the least expensive in use. As will be seen
below, these groups first came out from the heterocycle
work.
a
As shown in Table 1, 1- and 2-naphthalenesulfonyl
derivatives of indole have activation potentials 0.2-0.4
V higher than that of tosylindole, corresponding ap-
proximately to the effect of the carbamate contribution
in arenesulfonylcarbamates.1 Therefore, a number of
model sulfonamides derived from them have been pre-
pared, studied by CV, and subsequently used in depro-
tection experiments (Table 2).
0c
Inspection of Table 2 shows that, for the three pairs of
1
- and 2-naphthalenesulfonyl derivatives made (a , b, and
Resu lts a n d Discu ssion
e), the activation potentials are higher within the former
series but that the differences are at the most 0.11 V.
Larger differences in this respect exist between the
isopropylamides and anilides (0.19 and 0.23 V, respec-
tively) and, especially, between the isopropylamides and
indole derivatives (1g and h , Table 1; 0.45 and 0.39 V,
respectively). The activation potentials for the hydrazines
are close to those of the anilides. In the remaining
compounds studied, the 2-naphthalenesulfonyl groups
are attached to aliphatic nitrogens, as a consequence of
which their activation potentials approach those of 3a
and 3f.
To explore the effect of the heterocyclic and the
arenesulfonyl parts of the molecules on the activation
potentials as determined by CV, a few tosylated deriva-
tives (1b-e), other sulfonylated indoles (1f-h , j, k ), and
two carbazoles (1i, l) were prepared.12 These substances
are listed in Table 1 together with their activation
P
potentials (E ).
As shown in Table 1, compounds 1b-e all displayed
CV peaks at about the same or slightly higher (less
negative) potential as that of 1a , i.e., the constituent
N-heterocycles lowered the cathodic potentials by 0.54-
0
.74 V in comparison with unsubstituted tosylamide.
All experimental potentials measured for 2 and 3 are
well above -2.30 V. Consequently, with Mg/MeOH these
compounds should cleave, and we have confirmed chro-
matographically that they do so completely within 60
min. As a rule, these reactions were carried out under
ultrasonic conditions, and spontaneous evolution of gas
was initiated within a few min. Nonvolatile amines could
generally be isolated in over 90% yields (Table 2). Similar
results were obtained under simple stirring conditions,
but in this case activation of the metal by addition of at
Because nevertheless some minor influence of electron
delocalization was indicated in these experiments, we
decided to extend our work and make three simple,
readily accessible naphthalenesulfonyl indoles (1g, h , j).
(
12) All new compounds in Table 1 were made under phase-transfer
catalysis conditions with tetrabutylammonium hydrogen sulfate as
catalyst in benzene with solid NaOH/K CO (1c, i, and l) or 50%
2
3
KOH as base (1e-h , j, and k ). They were all obtained as crystalline
solids. For a typical procedure used, see: Hodson, H. F.; Madge, D. J .;
Slawin, A. N. Z.; Widdowson, D. A.; Williams, D. J . Tetrahedron 1994,
1
1
5
0, 1899.
4 4
least catalytic amounts of NH Cl was required. NH -

Amino Protection by 2-Naphthalenesulfonyl
J . Org. Chem., Vol. 64, No. 19, 1999 7137
Ta ble 2. CV P ea k P oten tia ls a n d Yield s on Mg/MeOH Red u ction of Som e Na p h th a len esu lfon a m id esa
-EPb (vs SCE)
time (min)
yield (%)
c
compd
naphthalenesulfonamide
2
3
2
3
a
a
b
b
N-isopropyl-1-naphthalenesulfonamided
2.03
2.14
1.84
1.91
2.18
2.05
2.41
1.96
1.87
1.89
2.32
2.02/
55
60
40
45
e
e
96
94
d
N-isopropyl-2-naphthalenesulfonamide
d
1-naphthalenesulfonanilide
2-naphthalenesulfonanilide
d
Ts-anilide
3
c
N-benzyl-2-naphthalenesulfonamided
N-benzyl-Ts-amide
60
e
f
3
2
3
d
e
e
N-benzyl-N-methyl-2-naphthalenesulfonamide
60
40
50
e
90
92
f
1-naphthalenesulfonyl-NHNH-Cbz
2-naphthalenesulfonyl-NHNH-Cbzf
Ts-NHNH-Cbz
3
f
2-naphthalenesulfonyl-NH(CH2)2NH-Tsf
40
93
∼2.48
1.97
2.02
3
3
g
h
2-naphthalenesulfonyl-L-phenylalaninef
60
45
e
86
2-naphthalenesulfonyl-L-phenylalanine Me esterf
a
1
13
b
All compounds prepared gave interpretable H and C NMR spectra. Cathode, vitreous carbon; solvent, DMF; supporting electrolyte,
-
3
-3 c
Bu4NBF4 0.1 mol dm ; substrate concentration, ∼0.005 mol dm
. Reductions performed under sonication with 10 equiv of Mg powder
d
in MeOH, generally on a 1 mmol scale. TLC indicated that all starting material reacted. Compound previously reported. Mp recorded:
1
24-125 °C (2a ); 114-115 °C (3a ); 157-159 °C (2b); 130-131 °C (3b); 120.5-122 °C (3c); from EtOAc/heptane (2/3a and b) or EtOAc/
e
light petroleum (3c). TLC indicated that all starting material had reacted; no yields were determined owing to the volatility of the
products (or zwitterionic character of the product in the experiment with 3g). New compound which gave a satisfactory elemental analysis.
f
Mp recorded: 94-95 °C (3d ); 149-149.5 °C (2e); 129.5-130.5 °C; (3e); 173-175 °C (3f); 148-149 °C (3g); 159-160 °C (3h ); from EtOAc/
light petroleum (3d ), CH2Cl2/(C2H5)2O (2/3e), EtOAc (3f and h ) or EtOAc/heptane (3g).
Cl, however, consumes significant amounts of magne-
sium, and therefore after 1 h, according to TLC, some-
times as much as 50% of the starting material remained,
requiring the addition of more magnesium. We have
found that typically a total reaction time of 4 h and
several magnesium additions (up to a total of 20 equiv)
were required to drag reductions to completion under
these conditions. This would explain the large amount
of magnesium used to cleave the pyrrole derivative in
the pioneering work by Okabe and Natsume. However,
in the case of our hydrazine derivatives (2/3e), cleaner
reaction mixtures were obtained when reactions were
carried out under ultrasonic conditions in the presence
of this additive but the yields obtained were the same.
Recently, in connection with work on multisubstituted
hydrazines,13 we made extensive use of the previously
1
0c
studied 4-cyanobenzenesulfonyl protecting group, which
could be cleaved under very mild reductive conditions
with mercury-activated aluminum [Al(Hg)] in 92-99%
yield. Therefore preparative cleavage experiments with
1k and 1l using this procedure have now also been
undertaken. Both compounds were cleaved, and from the
latter carbazole was isolated in quantitative yield, whereas
that of indole was a bit lower. It should be noted in this
context that the three naphthalenesulfonylazolides 1g-i
and in fact 1c could also be deprotected with Al(Hg), but
a larger excess of reducing agent was required for
complete cleavage in these cases. However, despite a very
large excess of reagent (60 equiv), this procedure failed
to cleave 1a , as only a minor amount of indole was
detected in the crude reaction mixture. We therefore
believe that Al(Hg) is going to be a useful, milder
alternative to Mg/MeOH for cleavage of certain com-
pounds of this type. For comparison, a few successful
preparative electrolysis experiments were also carried out
at potentials just below that of the CV peaks, thus
testifying to the significance of the data in Table 1. Also,
with this method 1l was cleaved and subsequently
carbazole was isolated in quantitative yield. Because of
its simplicity and mildness, cathodic reduction should,
of course, always be kept in mind in this context. Only
further work will tell us the relative merits of the
discussed cleavage procedures.
4
We anticipate that NH Cl can occasionally have a favor-
able effect by lowering the basicity of the reaction
medium. The increased lability of 2-naphthalenesulfonyl
in relation to tosyl is highlighted in the data for 3f and
its clean selective cleavage with Mg/MeOH as described
at the end of the Experimental Section. No trace of
starting material was visible chromatographically in the
cleavage product. More importantly, no significant byprod-
uct(s) could be detected in this key experiment.
Several further cleavage experiments on compounds
1
-3 have been performed. Because the activation poten-
tials of benzamides and benzoyl carbamates as measured
by CV are similar to those of tosylamides and tosylcar-
bamates,1 we also prepared 1-benzoylindole (1m ) and
found that it gave a CV peak at -1.91 V, i.e., at about
the same potential as that of 1a . It could also be cleaved
by magnesium powder in methanol, but because acylin-
doles are known to cleave under basic conditions, in this
case a control experiment was also carried out which
0b
Con clu sion s
In this paper we have explored CV as a tool for
investigating the cathodic stability of various sulfonyl
moieties attached to nitrogen. This allowed us to correlate
semiquantitatively the effect of structure with the per-
formance of Mg/MeOH as a reducing agent for sulfona-
mides and to develop naphthalenesulfonyl for protection
of amino groups. Naphthalenesulfonamides are signifi-
cantly more easily reduced than tosylamides and can be
-
1
showed that 1m with amide absorption at 1677 cm is
cleaved by Mg(OMe) alone. In compound 1n the CV peak
2
is shifted to -1.74 V, i.e., a slightly more negative value
than that for 1g. Its amide absorption band appears at
-
1
1
697 cm , i.e., closer to the ester region, indicating an
additional effect of 1-naphthyl in comparison with
phenyl.
(
13) Grehn, L.; L o¨ nn, H.; Ragnarsson, U. Chem. Commun. 1997,
1381.

7
138 J . Org. Chem., Vol. 64, No. 19, 1999
Nyasse et al.
cleaved efficiently by Mg/MeOH within 1 h under ultra-
sonic conditions or by simple stirring, especially when
the metal has been activated by addition of NH Cl.
4
) 7.7 Hz, J
≈ 0.7 Hz, 2 H), 8.46 (pert. d, J ≈ 1 Hz, 1 H); C NMR (100.4
MHz, CDCl ) δ 115.10, 120.02, 121.20, 123.95, 126.35, 127.42,
27.56, 127.73, 128.22, 129.11, 129.40, 129.43, 131.66, 134.85,
35.18, 138.38. Anal. Calcd for C22 S: C, 73.93; H, 4.23;
2
) 1.3 Hz, J
3
≈ 0.7 Hz, 2 H), 8.40 (dt, J
1
) 8.4 Hz,
1
3
J
2
3
1
1
Ordinary tosylamides are stable under these conditions
and require prior conversion to tosylcarbamates to be
cleaved. Of the two required reagents, 2-naphthalene-
sulfonyl chloride is particularly inexpensive. As we have
not noticed any significant difference in the reactivities
between 1- and 2-naphthalenesulfonyl groups so far, at
this stage we prefer to use the latter instead of tosyl as
an amino protecting group.
H
15NO
2
N, 3.92. Found: C, 73.8; H, 4.2; N, 3.9.
-Na p h th a len esu lfon a n ilid e (3b). (Typ ica l P r oced u r e).
Aniline (465 mg, 5.0 mmol) in CH Cl (100 mL) was chilled in
2
2
2
ice with the exclusion of moisture, and then triethylamine (550
mg, 5.45 mmol) was added. The resulting solution was treated
dropwise under stirring with 2-naphthalensulfonyl chloride
(
0
1.13 g, 5 mmol) also dissolved in CH
°C, and the mixture was left overnight at ambient temper-
ature. After concentration to ca. 30 mL the solution was
partitioned between EtOAc (150 mL) and 1 M KHSO (50 mL),
and the organic phase was washed successively with 1 M
KHSO , 1 M NaHCO , and brine (three times each). The
extract was dried (Na SO ) and then evaporated to give 3b as
2 2
Cl (30 mL) over 1 h at
Exp er im en ta l Section
4
Gen er a l. All melting points were determined on a Gallen-
kamp melting point apparatus and are uncorrected. TLC
analyses were carried out on 0.25-mm-thick precoated silica
plates (Merck Fertigplatten Kieselgel 60F254) using systems
4
3
2
4
a white solid (1.35 g, 95%). The product was recrystallized first
from EtOAc/heptane and then from EtOH to give small
(
A) toluene/MeCN 2:1, (B) light petroleum/Et
CH Cl /Me CO/HOAc 40:10:1. Spots were visualized under UV
light or, for hydrazines preferentially, by alcoholic H
P(Mo ] spray and subsequent heating (blue spots). Pre-
parative chromatography was carried out on Merck Kieselgel
0 (70-230 mesh). Magnesium slurries were subjected to
ultrasound treatment at 35 kHz/120-240 W (Bandelin, Berlin,
2
O 2:1, and (C)
14
1
needles: mp 132.5-133.5 °C (lit. mp 132 °C); H NMR (400
MHz, CDCl ) δ 7.04-7.08, 7.11-7.13 and 7.17-7.21 (compl.
sign., 1 H + 2 H + 2 H), 7.31 (br s, 1 H), 7.54 (pert. dt, J
.4 Hz, J ≈ 7.5 Hz, J ≈ 1.4
≈ 1.4 Hz, 1 H), 7.60 (pert. dt, J
Hz, 1 H), 7.79 (dd, J ) 8.7 Hz, J ) 1.9 Hz, 1 H), 7.84 and
2
2
2
3
7
-
≈
1
[
2 7 6
O )
7
2
1
2
1
2
6
7
.86 (2 overlapping d, J ≈ 8.7 Hz, 1 H+2 H), 8.39 (pert. d, J
13
)
1
1
3
1.5 Hz, 1 H); C NMR (100.4 MHz, CDCl ) δ 121.60, 122.22,
1
13
type RK106) at room temperature. H and C NMR spectra
were recorded at 400 and 100.4 MHz in ∼5% CDCl solu-
tion at 25 °C, unless otherwise stated. All shifts are given in
δ ppm using δ (TMS) ) 0 and δ (CDCl ) ) 77.02, respectively,
25.38, 127.49, 127.84, 128.90, 129.29, 129.31, 129.41, 131.99,
34.89, 135.89, 136.39.
3
1
-(2-Na p h th a len esu lfon yl)-NH(CH
compound was made from Ts-NH(CH NH
and 2-naphthalensulfonyl chloride (1.13 g, 5 mmol) in CH
2
)
2
NH-Ts (3f). This
(1.07 g, 5 mmol)
Cl
H
C
3
2
)
2
2
as reference. Assignments were made by comparison of chemi-
cal shifts and peak multiplicities. Elemental analyses of
crystalline derivatives were carried out by Mikro Kemi AB
2
2
in the presence of triethylamine (550 mg, 5.45 mmol). After
the reaction was allowed to proceed overnight at ambient
temperature as for 3b, a white precipitate was formed. The
mixture was then filtered, and the resulting white solid
obtained was washed five times with EtOAc before it was
(Uppsala).
1
-(2-Na p h th a len esu lfon yl)-in d ole (1h ). (Typ ica l P r o-
ced u r e). Compound 1h was made from indole (1.17 g, 10
mmol) and 2-naphthalenesulfonyl chloride (2.38 g, 10.5
mmol) in benzene in the presence of TBAHS (250 mg) and 50%
KOH (w/v; 15 mL) by stirring for 3 h. Water containing NaCl
was added, and the benzene separated, whereupon it was
washed with water (three times) and dried (MgSO ). After
4
evaporation, the crude solid was recrystallized from EtOH to
give 1h (2.27 g, 74%): mp 101-102.5 °C; pure by TLC (UV;
recrystallized from EtOAc/light petroleum to afford 3f (1.90
1
g, 94%): mp 173-175 °C; H NMR (400 MHz, DMSO-d
6
) δ
2
.30 (s, 3 H), 2.73 and 2.78 (2m, 2 H + 2 H), 7.26 and 7.54
2d, J ) 8.2 Hz, 2 H + 2 H), 7.60 (br t, J ≈ 5.7 Hz, 1 H), 7.68
and 7.72 (2dt, J ) 6.9 Hz, J
≈ 1 Hz, 1 H + 1 H), 7.76 (dd, J
8.8 Hz, J
(
1
2
1
)
2
) 1.8 Hz, 1 H), 7.81 (br t, J ≈ 5.7 Hz, 1 H), 8.05
and 8.16 (2 pert. d, J ≈ 7.7 Hz, 1 H + 1 H), 8.12 (d, J ) 8.6
1
A); H NMR (400 MHz, CDCl
0
3
) δ 6.66 (dd, J
≈ 7.5 Hz, J
≈ 1.3 Hz, 1 H), 7.50 (pert.
≈ 1 Hz, 1 H), 7.56 and 7.59 (2 overlapping
1
) 3.7 Hz, J
2
2
)
13
Hz, 1 H), 8.40 (d, J ≈ 1.5 Hz, 1 H); C NMR (100.4 MHz,
.7 Hz, 1 H), 7.20 (perturbed dt, J
H), 7.32 (pert. dt, J
dt, J
dt, J
1
≈ 1.1 Hz, 1
DMSO-d ) δ 20.86, 42.12, 42.14, 122.08, 126.33, 127.32, 127.58,
6
1
≈ 7.8 Hz, J
2
1
1
6
27.81, 128.72, 129.17, 129.41, 129.52, 131.66, 134.13, 137.22,
42.62. Anal. Calcd for C19 : C, 56.42; H, 4.98; N,
.93. Found: C, 56.3; H, 5.0; N, 6.9.
Al(Hg)-Med ia ted Clea va ge of 1i. Recrystallized 1i (715
mg, 2 mmol) was dissolved in Et O (160 mL), and H O (1 mL)
was added followed by freshly prepared Al(Hg) (0.54 g, 20
mmol) in small portions under CO (g) with rapid stirring.
1
1
) 7.7 Hz, J
≈ 7 Hz, J
2
20 2 4 2
H N O S
2
≈ 1.6 Hz, 1 H + 1 H), 7.64 (d, J ) 3.7 Hz, 1
H), 7.75 (dd, J
Hz, 2 H), 7.93 (pert. dd, J
dd, J ) 8.3 Hz, J
1
) 8.8 Hz, J
2
1
) 2.0 Hz, 1 H), 7.80 (pert. t, J ≈
) 7.4 Hz, J
≈ 1.6 Hz, 1 H), 8.06
9
(
2
2
2
1
2
≈ 0.8 Hz, 1 H), 8.52 (pert. d, J ≈ 2 Hz, 1
H); C NMR (100.4 MHz, CDCl ) δ 109.21, 113.51, 121.41,
21.44, 123.34, 124.64, 126.39, 127.76, 127.88, 128.55, 129.37,
29.44, 129.68, 130.74, 131.86, 134.84, 135.11, 135.19. Anal.
S: C, 70.34; H, 4.26; N, 4.56. Found: C,
15
1
3
3
2
1
1
After 4 h, when most of the Al had dissolved and TLC indicated
only minor amounts of remaining 1i, more Al(Hg) (0.27 g, 10
mmol) was added as above and allowed to react overnight, at
which time 1i could no longer be detected. The grayish solid
material was filtered off with suction and rinsed thoroughly.
The clear filtrate was concentrated to 25 mL, diluted with light
petroleum (25 mL), and further concentrated to precipitation.
After brief cooling, the white solid was collected by filtration,
rinsed with light petroleum, and dried to give pure carbazole
Calcd for C18
0.6; H, 4.3; N, 4.6.
-(2-Na p h th a len esu lfon yl)-ca r ba zole (1i). (Alter n a tive
H13NO
2
7
9
P r oced u r e). Carbazole (1.67 g, 10 mmol) in benzene (40 mL)
was treated with freshly ground NaOH (0.60 g, 15 mmol), dry
K
2
CO
one portion with vigorous stirring. After a few minutes,
-naphthalenesulfonyl chloride (3.40 g, 15 mmol) in benzene
25 mL) was added dropwise, and the mixture was left with
3
(4.14 g, 30 mmol), and TBAHS (0.68 g, 2.0 mmol) in
2
(
(313 mg, 94%) with mp 248-249 °C.
4
Con tr olled P oten tia l Electr olysis of 1l. A solution of Et -
NCl (0.1 M; supporting electrolyte) and Et NHCl (0.05 M;
3
stirring overnight, at which time TLC (A) indicated complete
reaction. Most of the solvent was stripped off at reduced
pressure, and the solid residue was partitioned between EtOAc
proton donor) in MeCN was added to a three-electrode cell of
10a
batch type. To its cathodic compartment 1l (332 mg, 1.00
and 1 M KHSO
KHSO , 1 M NaHCO
of the solvent left a white solid (3.50 g, 98%): pure by TLC
A); heavy, lustrous crystals; mp 169-170 °C [from EtOAc/
4
. The organic extract was washed with 1 M
mmol) was added, and a cyclic voltammogram was recorded.
The potential was adjusted to a value 50 mV more negative
4
3
, and brine and dried (Na SO ). Removal
2
4
(
1
(14) Curtius, T.; Bottler, H.; Raudenbusch, W. J . Prakt. Chem. 1930,
light petroleum 1:1 (80 mL/g)]; H NMR (400 MHz, CDCl
3
) δ
≈ 1 Hz, 2 H), 7.45-7.52 (compl. sign.,
H), 7.64 (pert. d, J ≈ 0.5 Hz, 2 H), 7.66-7.68 (compl. sign.,
H), 7.82 (pert. dd, J ≈ 2 Hz, 1 H), 7.85 (ddd, J
≈ 7 Hz, J
1
25, 380.
15) Smith, M. In Reduction. Techniques and Applications in
7
4
1
.32 (dt, J
1
) 7.5 Hz, J
2
(
Organic Synthesis; Augustine, R. L., Ed.; Marcel Dekker, New York,
1968; pp 136-137.
1
2
1

Amino Protection by 2-Naphthalenesulfonyl
J . Org. Chem., Vol. 64, No. 19, 1999 7139
than that for the CV peak, and the electrolysis was started.
Monitoring was by HPLC. When all starting material had
disappeared, the content of the cathodic compartment was
evaporated at reduced pressure, and the residue was chro-
Selective Clea va ge of 3f w ith Mg/MeOH. Compound 3f
(404 mg, 1 mmol) was added to a suspension of Mg powder
(250 mg, 10 mmol) and NH
(8 mL). The resulting mixture was sonicated for 40 min at
room temperature, at which time TLC on silica (A and CH
Cl /MeOH 3:1) indicated that all the starting material had
4
Cl (54 mg, 1 mmol) in dry methanol
matographed on silica using CH
2
Cl
2
as eluent to give 164 mg
2
-
(98%) of carbazole with mp 246.5-248 °C. Similar experiments
2
with 1a , 1k , and 1m on a 0.1 mmol scale without isolation of
indole indicated yields of 92%, 80%, and 84%, respectively, as
determined by HPLC.
been consumed. The reaction mixture was then extracted five
times with EtOAc (40 mL each), and the organic phase was
dried (Na
2 4
SO ) before being evaporated to dryness. The white
Clea va ge of 2b/3b w ith Mg/MeOH. A. (Typ ica l Exp er i-
powder obtained (200 mg, 93%) was suspended in dry Et O
2
m en t w ith ou t NH
mmol) in anhydrous methanol (8 mL) was added Mg powder
250 mg, 10 mmol). The resulting mixture was sonicated for
0 min, at which time TLC (A) indicated that all the starting
material had disappeared. The reaction was quenched by
addition of a saturated solution of NH Cl (10 mL) and
extracted three times with EtOAc (30 mL). The organic phase
was dried (Na SO ) and evaporated to dryness. The resulting
4
Cl). To a solution of 2b/3b (284 mg, 1
and sonicated for 5 min to afford upon standing a very fine
white solid with mp 120-121 °C corresponding to Ts-NH-
2 2 2
(CH ) NH .
(
4
Ack n ow led gm en t. This work was supported by the
Swedish Natural Science Research Council (NFR), the
Swedish Research Council for Engineering Sciences
(TFR), Carl Tryggers Stiftelse, Astra Draco AB, and the
Funda c¸ ao para a Ci eˆ ncia e a Tecnologia (Portugal). B.N.
gratefully acknowledges the RSC for a journals grant
for international authors and the ISP for a fellowship,
as well as the University of Yaound e´ I for a leave of
absence.
4
2
4
brown residue was purified by flash chromatography (EtOAc/
cyclohexane 1:3) on a silica gel pad to afford 90 mg (96%) of a
liquid corresponding to aniline.
B. (Typ ica l Exp er im en t w ith NH
ment was carried out in the presence of NH
4
Cl). The same experi-
Cl (54 mg, 1 mmol)
4
and stirring at room temperature. When gas evolution ceased,
TLC indicated remaining 2b/3b, and an extra 5 equiv of Mg
powder was therefore added. This was repeated every time
gas evolution had ceased until the complete disappearance of
the starting material. Workup as above afforded the same
amount of aniline.
Su p p or tin g In for m a tion Ava ila ble: Full experimental
details for all new compounds 1-3 (9 pages). This material is
available free of charge via the Internet at http://pubs.acs.org.
J O990695Q