A study of vinyl radical cyclization onto the azido group by addition of sulfanyl, stannyl, and silyl radicals to alkynyl azides

Article abstract of DOI:10.1002/(SICI)1099-0690(199806)1998:6<1219::AID-EJOC1219>3.0.CO;2-V

Thermal radical reactions of azidoalkynes 2, 8, 14, and 21a-c with thiols 1a-c afford 2-sulfanylvinyl radicals by selective addition of sulfanyl radicals to the triple bond. 1-Phenylvinyl radicals 23 and 30a, as well as vinyl radical 30b, undergo fast 5-cyclization onto the aromatic azide function to give cyclized indoles. In contrast, both 1-phenyl (15, 17) and 1-alkyl (3a,b, 9) vinyl radicals fail to add to their aliphatic azido substituents and exclusively undergo cyclization onto the aromatic sulfanyl ring and H transfer from the thiol precursor. Azidoalkynes 14 and 21a react with Bu₃SnH and TMSS under radical conditions to give instead the corresponding amines as a result of preferential attack of Bu₃Sn · and (TMS)₃Si · radicals on the azido group rather than on the triple bond. Evidence is provided that alkyl radical cyclizations onto azides are not feasible in the presence of thiol, in contrast with the reported utility of these cyclization reactions in the presence of Bu₃SnH and TMSS.

Full text of DOI:10.1002/(SICI)1099-0690(199806)1998:6<1219::AID-EJOC1219>3.0.CO;2-V

FULL PAPER
A Study of Vinyl Radical Cyclization onto the Azido Group by Addition of
Sulfanyl, Stannyl, and Silyl Radicals to Alkynyl AzidesȠ
Pier Carlo Montevecchi*, Maria Luisa Navacchia, and Piero Spagnolo
`
Dipartimento di Chimica Organica “A. Mangini”, Universita di Bologna,
Viale Risorgimento 4, I-40136 Bologna, Italy
Received January 21, 1998
Keywords: Alkynes / Azides / Cyclizations / Indoles / Radicals
Thermal radical reactions of azidoalkynes 2, 8, 14, and 21a–
Azidoalkynes 14 and 21a react with Bu3SnH and TMSS un-
c with thiols 1a–c afford 2-sulfanylvinyl radicals by selective der radical conditions to give instead the corresponding ami-
addition of sulfanyl radicals to the triple bond. 1-Phenylvinyl
nes as a result of preferential attack of Bu3Sn· and (TMS)3Si·
radicals 23 and 30a, as well as vinyl radical 30b, undergo fast radicals on the azido group rather than on the triple bond.
5-cyclization onto the aromatic azide function to give cycli-
zed indoles. In contrast, both 1-phenyl (15, 17) and 1-alkyl
(3a,b, 9) vinyl radicals fail to add to their aliphatic azido sub-
Evidence is provided that alkyl radical cyclizations onto azi-
des are not feasible in the presence of thiol, in contrast with
the reported utility of these cyclization reactions in the pre-
stituents and exclusively undergo cyclization onto the aro- sence of Bu3SnH and TMSS.
matic sulfanyl ring and H transfer from the thiol precursor.
In recent years, radical cyclizations have become a power- have hitherto been totally unexplored. In the course of our
ful tool for the construction of carbocyclic and heterocyclic investigations, we have studied[11] the fate of 2-sulfanylvinyl
systems; including those occurring in natural products.[1] radicals generated by regioselective addition of sulfanyl rad-
Radical cyclizations most often involve carbonϪcarbon icals to the respective CC triple bonds of alkynyl azides 2,
bond formation;[2] those leading to carbonϪnitrogen bond 8, 14, and 21a؊c.[12] The sulfanyl radicals were in turn pro-
formation have been somewhat less well documented. The duced from the corresponding thiols 1a؊c using the AIBN
reported protocols rely on additions of aminyl[3] and im- method. Sulfanyl radicals are well known to undergo facile
inyl[4] radicals to CϭC and CϭO double bonds, or on ad- addition to alkynes to give 2-sulfanylvinyl radicals.[13] On
ditions of carbon radicals to nitrogen atoms of imines[5] the contrary, thiols have been found to be virtually unreac-
and azides. However, the synthetic potential of azides has tive towards azides under radical conditions.[14]
yet to be fully explored, even though their capacity to act
In the present study, these investigations have been ex-
as radical acceptors has been long known since the pio- tended to the corresponding radical reactions of tributyltin
neering work of Horner and Bauer[6] in 1966 and two ear- hydride (Bu3SnH) and tris(trimethylsilyl)silane [TMSS]. We
lier reports on intramolecular addition af aryl[7a] and thi- were interested in ascertaining whether alkynyl azides might
ocarbonyl[7b] radicals to an aromatic azide yielding cyclized selectively react with Bu3Sn·and/or (TMS)3Si·radicals,
aminyl radicals.
Only very recently, cyclization reactions of the aliphatic
thereby providing further access to azidovinyl radicals.
Like sulfanyl radicals, both stannyl and silyl radicals also
azides with alkyl radicals were investigated and were shown react readily with alkynes to give vinyl radical adducts.[13]
to have considerable utility for the synthesis of N-hetero- The ability of stannyl radicals to add azides to give aminyl
cycles.[8] Alkyl radicals, generated from iodoazides through radicals[15] is well established, whereas support for a similar
iodine abstraction with tributyltin or tris(trimethylsilyl)silyl reaction with silyl radicals has to date only come from EPR
radicals, readily underwent intramolecular addition to the spectral evidence.[16]
azido function to give cyclic aminyl radicals after loss of
nitrogen (eq. 1).
Results and Discussion
Reactions were normally carried out by treating the
appropriate alkynyl azide 2, 8, 14, or 21a؊c with 1 equiv.
of thiol 1a؊c, Bu3SnH or TMSS, and AIBN (0.2 equiv.) in
refluxing fluorobenzene for 3 h. The product mixtures were
Our long interest in the chemical reactivity and synthetic directly analyzed by GC/MS and 1H-NMR spectroscopy
potential of vinyl radicals[9] and azido compounds[10] led us and then subjected to column chromatography on silica gel.
to investigate the behaviour of vinyl radicals containing Yields of the reaction products were calculated on the basis
either an aliphatic or an aromatic azido group in the side of the amount of azide that had reacted as 22Ϫ55% of unre-
chain. In fact, radical reactions of azides with vinyl radicals acted material was recovered.
Eur. J. Org. Chem. 1998, 1219Ϫ1226
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ193X/98/0606Ϫ1219 $ 17.50ϩ.50/0
1219

P. C. Montevecchi, M. L. Navacchia, P. Spagnolo
FULL PAPER
Experiments with sulfanylvinyl radicals 3a,b, generated 9 (n ϭ 1) through 5-membered cyclization onto the benzene
by reacting the pentynyl azide 2 with benzyl- (1a) and 4- ring and the azido group, respectively, were however not
cyanobenzyl mercaptan (1b), did not provide any evidence produced in sufficient quantity to allow their separation
for intramolecular attack on the azido moiety. Both reac- and full characterization (Scheme 2).
tions furnished, in a moderately high overall yield, a mix-
Scheme 2
ture of the corresponding benzyl vinyl sulfides 6a,b and
methyl vinyl sulfides 5a,b (Scheme 1). The former com-
pounds represent the products of H-abstraction from the
radicals 3a,b, whereas methyl sulfides 5a؊b are the expected
rearrangement products following intramolecular 5-(π-en-
do)exo cyclization of these radicals onto the adjacent aro-
matic ring.[9b] The proportion of rearrangement to re-
duction product was found to be significantly enhanced in
the case of the vinyl radical 3b as a result of greater stabili-
zation of the spirocyclohexadienyl intermediate 4 due to the
presence of the 4-cyano substituent.[9h] With benzyl mer-
captan 1a, the alkyne 2 additionally furnished a small
amount of the thiopyran 7a, which was also derived from
radical 3a through competing aromatic ortho cyclization
(Scheme 1).
Similarly, the reaction of the homologue hexynyl azide 8
with thiol 1c yielded the corresponding adduct 10 (n ϭ 2)
Scheme 1
as the only product arising from intermediate vinyl radical
9 (n ϭ 2). The adduct 10 was, however, produced only to a
modest extent owing to concomitant formation of the tria-
zole 11. The latter was derived from azide 8 by a monomo-
lecular 1,3-dipolar cycloaddition reaction. In an indepen-
dent experiment we found that in boiling fluorobenzene the
azide 8 alone gave triazole 11 in nearly quantitative yield
(ca. 50% conversion yield within 3 h) (Scheme 2).
Thus, 1-alkyl-2-sulfanylvinyl radicals 3a,b and 9 proved
to be reluctant to intramolecularly add an alkyl azide in a
5- or 6-(π-exo) mode and preferentially followed alternative
reaction pathways, including intramolecular cyclization
onto the sulfanyl aromatic ring and/or H-abstraction from
the alkane- or arenethiol precursor.
Since the π-exo cyclization of σ-bent 1-alkylvinyl radicals
3a,b, 9 might have been inhibited by unfavourable stereo-
electronic factors, we turned our attention to the fate of π-
linear 1-phenyl analogues 15 and 17, which, in principle,
might cyclize onto the azido group in a π-endo mode. It is
known that cyclizations of π-endo and π-exo mode are sen-
sitive to different stereoelectronic demands.[9h] Moreover,
we have previously shown that sp-hybridized 1-phenylvinyl
radicals can cyclize onto both CC double bonds[9e] and aro-
matic rings[18] more rapidly than their sp2-hybridized 1-
alkylvinyl counterparts. However, radicals 15 and 17, which
were generated from the phenylbutynyl azide 14 upon reac-
The reaction of 2 with benzenethiol 1c, carried out using tion with benzenethiol 1c and benzyl mercaptan 1a, respec-
the syringe pump technique,[17] gave the adduct 10 (n ϭ 1) tively, might have behaved in a fashion not dissimilar from
in good yield (80%). Despite the very low thiol 1c concen- that of the 1-alkyl-substituted congeners 3a,b, 9. In fact, the
tration employed, the intermediate sulfanylvinyl radical 9 alkenyl azide 16 was the only identifiable product of the
(n ϭ 1) was able to undergo hydrogen abstraction as the radical 15, whereas the azido compounds 18, 19, and 20
almost exclusive decomposition pathway. The possible pres- were the only observed products of radical 17 (Scheme 3).
ence of very small amounts of benzothiophene 12 and pyr-
In sharp contrast to the alkynyl azides 2, 8 and 14 de-
role 13 as side-products was additionally suggested by GC- scribed above, both 2-azidodiphenylacetylene (21a) and (2-
1
MS and H-NMR analysis of the crude reaction mixture. azidophenyl)trimethylsilylacetylene (21b) provided un-
These products, which might have been derived from radical equivocal evidence for facile vinyl radical cyclization onto
1220
Eur. J. Org. Chem. 1998, 1219Ϫ1226

Study of Vinyl Radical Cyclization
Scheme 3
FULL PAPER
Scheme 4
the azido function upon reaction with benzyl mercaptan 1a
or benzenethiol 1c. In the presence of thiol 1a, the azidoace-
tylene 21a furnished a mixture of the vinyl sulfides 24 and
25, with the major product (48%) being the indole 28
(Scheme 4). The origin of compounds 24, 25 and 28 can be
readily explained by assuming that non-regioselective (and
non-reversible)[19] addition of benzylsulfanyl radicals to the
alkyne 21a initially resulted in a mixture of comparable
amounts of the isomeric vinyl radicals 22 and 23. Then, by
the expected H abstraction and rearrangement, radical 22
would have given products 24 and 25, while radical 23
would have afforded only the indole 28 as a result of 5-(π-
endo)cyclization onto the azide, subsequent loss of nitrogen
from the thus formed triazenyl radical 26, and eventual re-
duction of the derived aminyl radical 27 (Scheme 4).
Scheme 5
In the presence of thiol 1c, the azidoacetylene 21a fur-
nished the corresponding indole 32a in high yield as the
exclusive product (Scheme 5). Presumably, in this case, a
non-regioselective but reversible[19] addition of benzenesul-
fanyl radical to the CC triple bond led to an equilibrium
mixture of the two sulfanyl radicals 29a and 30a, from
which only the indole 32a could be derived as a result of
fast 5-membered cyclization of vinyl radical 30a onto the
azido group (Scheme 5). Similar results were obtained by
reacting thiol 1c with the azidoacetylene 21b, but for un-
known reasons this reaction furnished the indole product
32b to a significantly lesser extent (Scheme 5). In contrast, zation of the unpaired electron into the indole ring. How-
2-azidophenylhexyne 21c did not undergo an analogous ever, the observed inertness of sulfanylvinyl radicals
cyclization to indole in the presence of benzenethiol 1c. As towards alkyl azides seemed rather curious, especially in
might have been anticipated,[9b][9c][9e][9f][9g][13] the ben- view of the previous findings[8] that alkyl radicals can ef-
zenesulfanyl radical most probably added to this alkylaryl- ficiently cyclize onto these azides, at least in the presence of
acetylene to give exclusively the more stable 1-aryl radical Bu3SnH or TMSS as radical precursors. This prompted us
29c, from which the azidoalkene 31c was derived following to briefly investigate the related radical reactions of tolu-
subsequent H abstraction (Scheme 5).
enethiol 1a with 5-azido-1-pentene (33) and 3-(2-azido-
The vinyl radicals 23, 30a and 30b were thus found to phenyl)propene (37). Our aim was to gain comparative in-
readily undergo 5-membered cyclization onto the aromatic formation concerning the reactivity of azido-containing 2-
azido moiety. It is possible that cyclization reactions of sulfanylalkyl radicals 34 and 38 when produced by sulfanyl
these radicals were favoured by moderate resonance stabi- radical addition to the CC double bond under conditions
lization of the generated triazenyl radicals due to delocali- analogous to those employed with our vinyl radicals.
Eur. J. Org. Chem. 1998, 1219Ϫ1226
1221

P. C. Montevecchi, M. L. Navacchia, P. Spagnolo
FULL PAPER
With the alkene 33, the derived benzylsulfanylalkyl rad-
ical 34 yielded mainly the H-abstraction product 35, which
was curiously accompanied by a small amount of the thio-
lactam 36 (Scheme 6).
Scheme 6
confirmed by subsequent desilylation with tetrabutylam-
monium fluoride to give phenylindole 44.
Scheme 8
With the azidophenylpropene 37, the corresponing sul-
fide 39 was similarly produced as the sole product of alkyl
radical 38. In this case, 2-methylindole 40 was additionally
produced due to thermal cycloaddition of the aromatic
azide to the adjacent double bond[20] (Scheme 7).
Scheme 7
Moreover, the butynyl azide 14 reacted with TMSS to
give the amine 41 and the pyrroline 46 in a 3:1 ratio, as
1
indicated by H-NMR and GC/MS analysis of the crude
mixture. Subsequent column chromatography led to the iso-
lation of, besides the amine 41, a 2:1 mixture of 46 and its
desilylated derivative 47 in 15% overall yield (Scheme 9).
The formation of pyrroline 46 points to successful cycliza-
tion of intermediate vinyl radical 45 onto the azido group.
Presumably, radical 45 was prevented from abstracting a
hydrogen from the bulky silane due to steric factors and
was consequently able to add the relatively inert alkyl
azido group.
Thus, under our reducing conditions, sulfanylalkyl rad-
icals were found to be incapable of attacking an aliphatic
or even an aromatic azide. This fact, while pointing to a
lower reactivity of alkyl compared to vinyl radicals towards
the azido group, suggests that the successful outcome of the
reported alkyl radical cyclizations onto aliphatic azides is
primarily attributable to the use of radical precursors such
as Bu3SnH or TMSS, which are less effective as radical
scavengers than thiols.[21]
Scheme 9
One further objective was to investigate azidovinyl rad-
icals by analogous alkyne reactions with Bu3SnH and
TMSS. However, such investigations were largely frustrated
by the observation that stannyl and silyl radicals can add
to the azido group in preference to the triple bond. In fact,
both 1-phenylbutynyl azide 14 and 2-azidodiphenylace-
tylene 21a were found to be cleanly reduced to the corre-
sponding amines 41 and 42 by Bu3SnH (eqs. 2 and 3).
Similar results were obtained with the acetylene 21a in
In conclusion, 2-sulfanylvinyl radicals 3a,b, 9, 15, and 17,
the presence of TMSS, but under these circumstances trace generated by regioselective sulfanyl radical addition to alky-
amounts of the silylated indole 43 were also produced as a nyl azides 2, 8, and 14, did not undergo 5- or 6-exo cycliza-
result of vinyl radical cyclization onto the aromatic azido tion onto the aliphatic azido group owing to preferred cycli-
group (Scheme 8). The structure of 43 was unequivocally zation onto the aromatic sulfanyl substituent and/or H-ab-
1222
Eur. J. Org. Chem. 1998, 1219Ϫ1226

Study of Vinyl Radical Cyclization
FULL PAPER
nitrite in hydrochloric acid and subsequent treatment of the di-
azonium salt with 1.1 equivalent of sodium azide at 5°C. The reac-
tion mixture was then stirred at room temperature for ca. 40 min.
(unless disappearance of the diazonium salt was indicated by a β-
naphthol test) and then extracted with diethyl ether. The organic
layer was separated and dried, and the solvent was evaporated.
Flash silica gel column chromatography of the residue led to the
isolation of the pure products in ca. 60Ϫ70% yields.
straction from the thiol precursor 1a؊c. In contrast, rad-
icals 23 and 30a,b underwent fast 5-cyclization onto the
aromatic azido group, possibly as a result of resonance sta-
bilization of the thus formed triazenyl radical adducts. Fast
5-exo cyclization onto the aliphatic azido group was ob-
served with the silylvinyl radical 45, which was produced
from azide 14 in the presence of TMSS. Unfortunately, un-
like sulfanyl radicals, (TMS)3Si·radicals were found to add
to the azido group in preference to the alkyne triple bond,
thus making them of limited utility for the production of
vinyl radicals from azidoacetylenes. Analogous regioselec-
tivity was displayed by Bu3Sn·radicals, use of which led to
the exclusive reduction of the azido group. Finally, evidence
was obtained that alkyl radical cyclizations onto azides are
not feasible in the presence of a thiol, in contrast with the
reported usefulness of these cyclization reactions in the
presence of Bu3SnH or TMSS.
1
21a: M.p. 36Ϫ37°C (dec. at 150°C). Ϫ H NMR: δ ϭ 7.1Ϫ7.6
(aromatic protons). Ϫ MS; m/z (%): 219 [Mϩ] (15), 191 (100), 163
(20). Ϫ MS (HR); m/z: 219.0799 (C14H9N3, [Mϩ], calcd. 219.0796).
Ϫ IR: ν
ϭ 2150 cmϪ1
.
˜
max
21b: Oil. Ϫ 1H NMR: δ ϭ 0.28 (s, 9 H), 7.0Ϫ7.5 (m, 4 H). Ϫ
MS; m/z (%): 215 [Mϩ] (30), 187 (70), 186 (65), 172 (100), 145 (50).
Ϫ MS (HR); m/z: 215.0876 (C11H13N3Si, [Mϩ], calcd. 215.0879).
Ϫ IR: ν˜max ϭ 2150 cmϪ1
.
1
21c: Oil. Ϫ H NMR: δ ϭ 0.85 (t, J ϭ 7 Hz, 3 H), 1.5Ϫ1.7 (m,
4 H), 2.48 (t, J ϭ 7 Hz, 2 H), 7.0Ϫ7.4 (m, 4 H). Ϫ MS; m/z (%):
199 [Mϩ] (70), 171 (30), 170 (40), 156 (60), 143 (55), 130 (85), 129
(100), 128 (80), 115 (75), 102 (80). Ϫ MS (HR); m/z: 199.1112
`
This work was supported by the Ministero dellЈUniversita e della
Ricerca Scientifica e Tecnologica (MURST), the Consiglio Nazio-
nale delle Ricerche (CNR, Rome) and the University of Bologna
(1997 Funds for selected research topics).
(C H N , [Mϩ], calcd. 199.1109); IR: ν
˜
max
ϭ 2150 cmϪ1
.
12 13
3
2-Aminodiphenylacetylene (42), (2-Aminophenyl)(trimethylsilyl)-
acetylene and 1-(2-Aminophenyl)hex-1-yne were obtained in ca.
80Ϫ90% yield following a procedure reported in the literature[22]
for the preparation of disubstituted alkynes from halobenzenes and
terminal acetylenes. According to this procedure, a solution of 2-
iodoaniline (20 mmol) and tetrakis(triphenylphosphane)pallad-
ium[23] Pd(PPh3)4 (1 mmol, 1.16 g) in piperidine (30 ml) was stirred
at room temperature under nitrogen atmosphere for 15 min., and
then a solution of the appropriate terminal acetylene (phenylacety-
lene, trimethylsilylacetylene, or hex-1-yne) (20 mmol) in piperidine
(20 ml) was added. The reaction mixture was stirred for 6 h at ca.
30°C under nitrogen, then extracted with diethyl ether, and the
combined extracts were washed several times with saturated am-
monium chloride solution. The organic layer was separated and
dried, the solvent was evaporated, and the crude product was puri-
fied by column chromatography on silica gel.
Experimental Section
Starting Materials: Benzenethiol 1c, toluenethiol 1a, pentynol,
hexynol and pentenol were commercially available. 4-Cyanotol-
uenethiol 1b[9b], 3-(2-azidophenyl)propene 37[20] and 4-phenylbut-
3-yn-1-ol[22] were prepared as described in the literature. TMSS was
a commercially available reagent from Fluka.
The following hitherto unknown compounds were prepared by
treating the corresponding tosylate (20 mmol) (in turn prepared
from the corresponding alcohol and tosyl chloride by standard
methods) with sodium azide (22 mmol) in DMSO (100 ml) at room
temperature for two days. The reaction mixture was extracted with
n-pentane and the combined extracts were washed several times
with brine. The organic layer was dried and the n-pentane was dis-
tilled off in a Claisen apparatus. The residue consisted of pure azide
(ca. 15Ϫ16 mmol), which was utilized without further purification.
1
2-Aminodiphenylacetylene (42): Oil. Ϫ H NMR: δ ϭ 4.2 (br s,
5-Azidopent-1-yne (2): Oil. Ϫ 1H NMR: δ ϭ 1.8 (quintuplet, J ϭ
7 Hz, 2 H), 1.97 (t, J ϭ 2.5 Hz, 1 H), 2.3 (dt, Jt ϭ 7 Hz, Jd ϭ 2.5
Hz, 2 H), 3.42 (t, J ϭ 7 Hz, 2 H). Ϫ MS; m/z (%): 109 [Mϩ] (80),
80 (50), 54 (100). Ϫ MS (HR); m/z: 109.0642 (C5H7N3, [Mϩ], calcd.
2 H), 6.7 (m, 2 H), 7.0Ϫ7.6 (m, 7 H). Ϫ MS; m/z (%): 193 [Mϩ]
(100), 165 (60). Ϫ MS (HR); m/z: 193.0889 (C14H11N, [Mϩ], calcd.
193.0891). Ϫ IR: ν
ϭ 3550, 3400 cmϪ1
.
˜
max
2-Aminophenyl)(trimethylsilyl)acetylene: Oil. Ϫ 1H NMR: δ ϭ
0.3 (s, 9 H), 4.1 (br s, 2 H), 7.5Ϫ7.5 (m, 4 H). Ϫ MS; m/z (%): 189
[Mϩ] (70), 174 (100). Ϫ MS (HR); m/z: 189.9875 (C11H15NSi,
109.0640). Ϫ IR: ν˜max ϭ 2150 cmϪ1
.
6-Azidohex-1-yne (8): Oil. Ϫ 1H NMR: δ ϭ 1.5Ϫ1.8 (m, 4 H),
1.95 (t, J ϭ 2.5 Hz, 1 H), 2.2 (dt, Jt ϭ 7 Hz, Jd ϭ 2.5 Hz, 2 H),
3.3 (t, J ϭ 7 Hz, 2 H). Ϫ MS; m/z (%): 123 [Mϩ] (15), 94 (30), 67
(40), 39 (100). Ϫ MS (HR); m/z: 123.0794 (C6H9N3, [Mϩ], calcd.
[Mϩ], calcd. 189.9878). Ϫ IR: ν˜max ϭ 3550, 3400 cmϪ1
.
1
1-(2-Aminophenyl)hex-1-yne : Oil; H NMR: δ ϭ 0.95 (t, J ϭ 7
Hz, 3 H), 1.4Ϫ1.7 (m, 4 H), 2.48 (t, J ϭ 7 Hz, 2 H), 4.10 (br s, 2
H, NH2), 6.5Ϫ7.7 (m, 4 H). Ϫ MS; m/z (%): 173 [Mϩ] (70), 130
123.0796). Ϫ IR: ν
ϭ 2100 cmϪ1
.
˜
max
4-Azido-1-phenylbut-1-yne (14): Oil. Ϫ 1H NMR: δ ϭ 2.70 (t,
J ϭ 7 Hz, 2 H), 3.48 (t, J ϭ 7 Hz, 2 H), 7.2Ϫ7.4 (m, 5 H). Ϫ
MS; m/z (%): 171 [Mϩ] (10), 115 (100). Ϫ MS (HR); m/z: 171.0798
(100).
173.1204). Ϫ IR: ν
Ϫ
MS (HR); m/z: 173.1207 (C12H15N, [Mϩ], calcd.
ϭ 3550, 3400 cmϪ1
.
˜
max
(C10H9N3, [Mϩ], calcd. 171.0796). Ϫ IR: ν˜max ϭ 2100 cmϪ1
.
Reaction Products: Structural assignments of reaction products
were generally made on the basis of 1H-NMR, IR, MS and HRMS
spectral data. Column chromatography was performed on Merck
silica gel (0.040Ϫ0.063 particle size) by gradual elution with light
petroleum (b.p. 40Ϫ70°C)/diethyl ether. 1H-NMR spectra were
generally recorded at 200 MHz (unless otherwise stated) in CDCl3
5-Azidopent-1-ene (33): Oil. Ϫ 1H NMR: δ ϭ 1.7 (quintuplet,
J ϭ 7 Hz, 2 H), 2.15 (br t, Jt ϭ 7 Hz, 2 H), 3.3 (t, J ϭ 7 Hz, 2 H),
4.95Ϫ5.10 (m, 2 H), 5.55Ϫ5.75 (m, 1 H). Ϫ MS; m/z (%): 111 [Mϩ]
(10), 83 (40), 55 (100). Ϫ MS (HR); m/z: 111.0794 (C5H9N3, [Mϩ],
calcd. 111.0797). Ϫ IR: ν˜max ϭ 2100 cmϪ1
.
2-Azidodiphenylacetylene (21a), (2-Azidophenyl)(trimethylsilyl)- solutions, using Me4Si as internal standard. Mass spectra were de-
acetylene (21b), and 1-(2-Azidophenyl)hex-1-yne (21c) were ob-
tained by diazotization of the corresponding amine with sodium
termined by the electron impact method. Yields of reaction prod-
ucts are based on reacted azide, unless otherwise stated.
Eur. J. Org. Chem. 1998, 1219Ϫ1226
1223

P. C. Montevecchi, M. L. Navacchia, P. Spagnolo
FULL PAPER
Reactions of Alkynyl Azides 2, 14 and 21a and Alkenyl Azides 33
and 37 with Ͱ-Toluenethiols 1a,b Ϫ General Procedure: A solution
4-azido-1-phenyl-2-(Ͱ-toluenesulfanyl)but-1-ene (18) (ca. 70:30 Z/E
ratio), and the benzothiopyran 20 in a 64:32:4 ratio (63% overall
1
of the appropriate alkene or alkyne (2 mmol), the appropriate thiol yield). Ϫ H NMR: 19: δ ϭ 2.1 (s, 3 H), 2.65 (t, J ϭ 7 Hz, 2 H),
1a,b (2 mmol) and AIBN (65 mg, 0.4 mmol) in fluorobenzene (20 3.50 (t, J ϭ 7 Hz, 2 H), 7.1Ϫ7.5 (m, 10 H). Ϫ (Z)-18: δ ϭ 2.6 (br
ml) was refluxed for 3 h. The reaction mixture was analyzed by 1H-
NMR and then chromatographed.
t, J ϭ 7 Hz, 2 H), 3.45 (t, J ϭ 7 Hz, 2 H), 3.8 (s, 2 H), 6.7 (br s,
1 H), 7.1Ϫ7.5 (m, 10 H). Ϫ (E)-18: δ ϭ 2.65 (t, J ϭ 7 Hz, 2 H),
3.45 (t, J ϭ 7 Hz, 2 H), 4.0 (s, 2 H), 6.52 (s, 1 H), 7.1Ϫ7.5 (m, 10
H).; signals at δ ϭ 2.5 (t, J ϭ 7 Hz, 2 H), 3.38 (t, J ϭ 7 Hz, 2 H)
and 3.92 (s, 2 H) were assigned to the benzothiopyran 20. The mass
spectrum showed peaks at m/z ϭ 295 [Mϩ] (25), attributable to the
molecular ions of 18 and 19, and at m/z ϭ 293 (2), attributable to
the molecular ion of 20. Ϫ MS (HR); m/z: 295.1140 (C17H17N3S,
From 2-Azidodiphenylacetylene 21a and Ͱ-Toluenethiol 1a: Chro-
matography gave unreacted azide (42%), unreacted thiol 1a, a 60/
40 Z/E mixture of β-(2-azidophenyl)-α-(toluenesulfanyl)styrene
(24) (20%), 1-(2-azidophenyl)-2-(methylthio)stilbene (25) and 2-
phenyl-3-(α-toluenesulfanyl)indole (28) (48%).
1
[Mϩ], calcd. 295.1143). Ϫ IR: ν˜max ϭ 2150 cmϪ1
.
24: H NMR: δ ϭ 3.6 (s, 1.2 H), 3.8 (s, 0.8 H), 6.78 (s, 0.4 H),
6.85 (s, 0.6 H), 6.9Ϫ7.7 (m, 14 H). Irradiation at δ ϭ 3.8 caused a
27% enhancement of the singlet at δ ϭ 6.78 [(E) isomer]. Ϫ MS;
m/z (%): 343 [Mϩ] (20), 315 (15), 224 (80), 91 (100). Ϫ MS (HR);
From Pentenyl Azide 33 and Ͱ-Toluenethiol 1a: Chromatography
gave unreacted thiol 1a (20%), dibenzyl disulfide (25%), 5-azido-1-
(α-toluenesulfanyl)pentane (35) (0.88 mmol, 80%; yield based on
reacted thiol and determined by H-NMR analysis of the reaction
mixture with anisole as internal standard) and 2-thionotetrahydro-
pyridine (36) (3%).
m/z: 343.1140 (C21H17N3S, [Mϩ], calcd. 343.1143). Ϫ IR: ν˜max
ϭ
1
2150 cmϪ1
.
25: 1H NMR: δ ϭ 1.85 (s, 3 H), 6.9Ϫ7.5 (m, 14 H). Ϫ 13C NMR:
δ ϭ 18 (CH3), 118.4, 124.2, 127.0, 127.2, 127.8, 128.0, 128.2, 129.5,
129.9, 132.2, 134.8 (q), 136.8 (q), 138.0 (q), 138.6 (q), 138.9 (q),
141.6 (q). Ϫ MS; m/z (%): 343 [Mϩ] (5), 315 (5), 300 (30), 149 (60),
91 (100), 84 (50). Ϫ MS (HR); m/z: 343.1140 (C21H17N3S, [Mϩ],
35: 1H NMR: δ ϭ 1.55 (m, 6 H), 2.4 (t, J ϭ 6.7 Hz, 2 H; collaps-
ing to singlet upon irradiation at δ ϭ 1.55), 3.24 (t, J ϭ 6.7 Hz, 2
H; collapsing to singlet upon irradiation at δ ϭ 1.55), 3.7 (s, 2 H),
7.2Ϫ7.4 (m, 5 H). Ϫ MS; m/z (%): 235 [Mϩ] (2), 207 [Mϩ Ϫ 28]
calcd. 343.1143). Ϫ IR: ν˜max ϭ 2150 cmϪ1
.
(3), 91 (100). Ϫ IR: ν˜max ϭ 2120 cmϪ1
.
1
28: H NMR: δ ϭ 3.80 (s, 2 H), 6.9Ϫ7.8 (m, 14 H), 8.3 (br s, 1
H, NH). Ϫ MS; m/z (%): 315 [Mϩ] (40), 224 (100), 149 (20), 91 (20).
Ϫ MS (HR); m/z: 315.1078 (C21H17NS, [Mϩ], calcd. 315.1082). Ϫ
36: 1H NMR: δ ϭ 1.8 (m, 4 H), 2.9 (t, J ϭ 7 Hz, 2 H; collapsing
to singlet upon irradiation at δ ϭ 1.8), 3.38 (br t, J ϭ 7 Hz, 2 H;
collapsing to singlet upon irradiation at δ ϭ 1.8; collapsing to sharp
triplet upon irradiation at δ ϭ 8.24), 8.24 (br s, 1 H, NH). Ϫ MS;
m/z (%): 115 [Mϩ] (100), 114 (60). Ϫ MS (HR); m/z: 115.9754
(C5H9NS, [Mϩ], calcd. 115.9751). Ϫ IR: ν˜max ϭ 3400 and 1350
IR: ν˜max ϭ 3480 cmϪ1
.
From Pentynyl Azide 2 and Ͱ-Toluenethiol 1a: Chromatography
gave unreacted azide (50%), unreacted thiol 1a, and an inseparable
mixture of 5-azido-1-(Ͱ-toluenesulfanyl)pent-1-ene (6a) (1.1 Z/E
ratio), (Z)-5-azido-1-(methylthio)-2-phenylpent-1-ene (5a) and 4-
(3-azidopropyl)-1H-2-benzothiopyran (7a) in a 70:25:5 ratio (84%
cmϪ1
.
From Azidophenylpropene 37 and Ͱ-Toluenethiol 1a: Chromatog-
raphy gave unreacted azide (70 mg, 22%), unreacted thiol 1a, di-
benzyl disulfide (20%), 1-(2-azidophenyl)-3-(Ͱ-toluenesulfanyl)pro-
pane (39) (175 mg, 45%) and 2-methylindole (40) (60 mg, 30%).
1
overall yield) . Ϫ H NMR: 6a: δ ϭ 1.6 (m, 2 H), 2.15 (m, 2 H),
3.20 (m, 2 H), 3.85 (s, 2 H), 5.52 (dt, Jd ϭ 10 Hz, Jt ϭ 7 Hz, 0.5
H), 5.58 (dt, Jd ϭ 14.5 Hz, Jt ϭ 7 Hz, 0.5 H), 5.96 (dt, Jd ϭ 14.5
Hz, Jt ϭ 1.0 Hz, 0.5 H), 5.98 (dt, Jd ϭ 10 Hz, Jt ϭ 1.0 Hz, 0.5 H),
7.3 (m, 5 H). Ϫ 5a: δ ϭ 1.6 (m, 2 H), 2.25 (s, 3 H), 2.55 (dt, Jd ϭ
7.0 Hz, Jt ϭ 1.0 Hz, 2 H), 3.20 (m, 2 H), 6.0 (t, J ϭ 1 Hz, 1 H),
7.3 (m, 5 H); signals at δ ϭ 2.7 (t, J ϭ 7 Hz, 2 H), 3.7 (s, 2 H),
and 6.3 (s, 1 H) were assigned to thiopyran 7a. Ϫ IR: ν˜max ϭ 2150
cmϪ1. The mass spectrum showed a molecular ion at m/z 233 [Mϩ]
(10) attributable to isomeric 5a and 6a. Ϫ MS (HR): 233.0983
(C12H15N3S, [Mϩ], calcd. 233.0987)].
39: 1H NMR: δ ϭ 1.85 (m, 2 H), 2.4 (t, J ϭ 7 Hz, 2 H; collapsing
to singlet upon irradiation at δ ϭ 1.85), 2.6 (t, J ϭ 7 Hz, 2 H;
collaψng to singlet upon irradiation at δ ϭ 1.85), 3.7 (s, 2 H),
7.0Ϫ7.4 (m, 9 H). Ϫ MS; m/z (%): 255 [Mϩ Ϫ 28] (5), 116 (100),
115 (50), 91 (60). Ϫ IR: ν
ϭ 2120 cmϪ1],
˜
max
Reactions of Alkynyl Azides 2, 8, 14, 21a؊c with Benzenethiol
1c Ϫ General Procedure: A solution of benzenethiol (2 mmol) in
fluorobenzene (5 ml) was slowly added with a syringe pump over
a period of 3 h to a boiling solution of the appropriate azide (2
mmol) and AIBN (0.4 mmol) in fluorobenzene (20 ml). The reac-
From Pentynyl Azide 2 and 4-Cyano-α-toluenethiol 1b: Chroma-
tography gave unreacted azide (50%), unreacted thiol 1b, and an
1
tion mixture was refluxed for a further 1 h, then analyzed by H-
inseparable
mixture
of
isomeric
5-azido-1-(4-cyano-α-
NMR and chromatographed.
toluenesulfanyl)pent-1-ene (6b) (15%, ca. 60:40 Z/E ratio) and (Z)-
5-azido-1-(methylthio)-2-(4-cyanophenyl)pent-1-ene (5b) (15%). Ϫ
1H NMR 6b: δ ϭ 1.6 (m, 2 H), 2.15 (m, 2 H), 3.25 (t, J ϭ 7 Hz,
2 H), 3.88 (s, 2 H), 5.58 (dt, Jd ϭ 9.5 Hz, Jt ϭ 7 Hz, 0.6 H), 5.64
(dt, Jd ϭ 15 Hz, Jt ϭ 7 Hz, 0.4 H), 5.90 (br d, J ϭ 9.5 Hz, 0.6 H),
5.92 (br d, J ϭ 15 Hz, 0.4 H), 7.4 and 7.65 (A and B parts of an
A2B2 system, J ϭ 8.5 Hz, 4 H). Ϫ 5b: δ ϭ 1.6 (m, 2 H), 2.3 (s, 3
H), 2.55 (br t, J ϭ 7 Hz, 2 H), 3.25 (t, J ϭ 7 Hz, 2 H), 6.15 (br s,
1 H), 7.4 and 7.65 (A and B parts of an A2B2 system, J ϭ 8.5 Hz,
4 H). Ϫ MS; m/z (%): 258 [Mϩ] (7), 230 (15), 215 (65), 183 (65),
171 (100), 116 (90). Ϫ MS (HR); m/z: 258.0935 (C13H14N4S, [Mϩ],
From 2-Azidodiphenylacetylene 21a: Chromatography gave unre-
acted azide (25%) and 2-phenyl-3-(phenylthio)indole (32a) (85%).
Ϫ
1H NMR: δ ϭ 7.0Ϫ7.8 (aromatic protons), 8.6 (br s, NH). Ϫ
MS; m/z (%): 301 [Mϩ] (100), 223 (25). Ϫ MS (HR); m/z: 301.0921
(C20H15NS, [Mϩ], calcd. 301.0925). Ϫ IR: ν˜max ϭ 3480 cmϪ1
.
From (2-Azidophenyl)(trimethylsilyl)acetylene 21b: Chromatog-
raphy gave unreacted azide (55%), 3-(phenylthio)-2-(trimethyl-
silyl)indole (32b) (42%), and an unknown, green product (20%) [the
1H-NMR spectrum showed singlets at δ ϭ 1.2Ϫ1.5 and aromatic
protons at δ ϭ 7.3Ϫ7.7. Ϫ MS; m/z (%): 374 [Mϩ] (15), 373 (35),
359 (20), 301 (25), 286 (40), 117 (30), 73 (100). Ϫ IR: ν˜max ϭ 3400
(very broad), 2160 (m, sharp)].
calcd. 258.0939). Ϫ IR: ν˜max ϭ 2260 (m, sh) and 2150 cmϪ1
.
From Phenylbutynyl Azide 14 and Ͱ-Toluenethiol 1a: Chromatog-
raphy gave unreacted azide (45%), unreacted thiol 1a, and an insep-
32b: 1H NMR: δ ϭ 0.5 (s, 9 H), 7.0Ϫ7.8 (m, 9 H), 8.5 (br s,
arable mixture of 4-azido-2-(methylthio)-1,1-diphenylbut-1-ene (19), NH). Irradiation at δ ϭ 0.5 caused a 5% enhancement of the broad
1224 Eur. J. Org. Chem. 1998, 1219Ϫ1226

Study of Vinyl Radical Cyclization
FULL PAPER
singlet at δ ϭ 8.5. Ϫ MS; m/z (%): 297 [Mϩ] (100), 282 (20), 73
(Z)-16: 1H NMR: δ ϭ 2.5 (t, J ϭ 7 Hz, 2 H; enhanced upon
(100). Ϫ MS (HR); m/z: 297.1003 (C17H19NSSi, [Mϩ], calcd. irradiation at δ ϭ 6.9), 3.4 (t, J ϭ 7 Hz, 2 H), 6.9 (s, 1 H), 7.2Ϫ7.5
297.1007). Ϫ IR: ν˜max ϭ 3500 cmϪ1
.
(m, 10 H). MS; m/z (%): 281 [Mϩ] (15), 253 (10), 252 (20), 115
(100). Ϫ MS (HR); m/z: 281.0981 (C16H15N3S, [Mϩ], calcd.
From 1-(2-Azidophenyl)hex-1-yne 21c: Chromatography gave
unreacted azide (45%) and 1-(2-azidophenyl)-2-(phenylthio)hex-1-
281.0987). Ϫ IR: ν˜max ϭ 2150 cmϪ1
.
1
1
ene (31c) (80%) (Z/E ratio, 70:30). 31c: H NMR: (Z) isomer: δ ϭ
(E)-16: H NMR: δ ϭ 2.7 (t, J ϭ 7 Hz, 2 H), 3.5 (t, J ϭ 7 Hz,
2 H), 6.87 (s, 1 H), 7.2Ϫ7.5 (m, 10 H). MS; m/z (%): 281 [Mϩ] (15),
253 (10), 252 (20), 115 (100). Ϫ MS (HR): 281.0982 (C16H15N3S,
0.85 (t, J ϭ 7.5 Hz, 3 H), 1.2Ϫ1.7 (m, 4 H), 2.3 (t, J ϭ 7.5 Hz, 2
H), 7.8 (s, 1 H), 7.0Ϫ7.8 (m, 9 H); E-isomer: δ ϭ 0.85 (t, J ϭ 7.5
Hz, 3 H), 1.2Ϫ1.7 (m, 4 H), 2.45 (t, J ϭ 7.5 Hz, 2 H), 6.5 (s, 1 H),
7.0Ϫ7.8 (m, 9 H). Ϫ MS; m/z (%): 309 [Mϩ] (25), 281 (15), 238
(20), 204 (100), 130 (90). Ϫ MS (HR); m/z: 309.1304 (C18H19N3S,
[Mϩ], calcd. 281.0987). Ϫ IR: ν˜max ϭ 2150 cmϪ1
.
Reactions of Alkynyl Azides 14 and 21a with Tributyltin Hydride.
Ϫ General Procedure: A solution of tributyltin hydride (2 mmol),
the appropriate alkyne (2 mmol), and AIBN (0.4 mmol) was re-
fluxed for 3 h. The solvent was then removed, the residue was ana-
[Mϩ], calcd. 309.1300). Ϫ IR: ν
ϭ 2150 cmϪ1
.
˜
max
From Hexynyl Azide 8: Chromatography gave unreacted azide
(23%) and a fraction consisting of a 50:50 mixture of the fused
triazole 11 and sulfide 10, n ϭ 2 (90% overall yield). The reaction
was repeated at room temperature in the presence of oxygen. Col-
umn chromatography gave unreacted azide (50%) and 6-azido-1-
(phenylthio)-hex-1-ene (10), n ϭ 2 (90%). 10:1H NMR: δ ϭ
1.5Ϫ1.8 (m, 4 H), 2.25 (m, 2 H), 3.30 (m, 2 H), 5.58 (dt, Jd ϭ 9
Hz, Jt ϭ 7 Hz, 0.6 H), 5.95 (dt, Jd ϭ 14.5 Hz, Jt ϭ 7 Hz, 0.4 H),
6.18 (dt, Jd ϭ 9 Hz, Jt ϭ 1 Hz, 0.6 H), 6.25 (dt, Jd ϭ 14.5 Hz,
Jt ϭ 1 Hz, 0.4 H), 7.1Ϫ7.4 (m, 5 H). Ϫ MS; m/z (%): 204 (20), 172
1
lysed by H-NMR, and chromatographed.
From 2-Azidodiphenylacetylene 21a: Chromatography gave unre-
acted azide (20%) and 2-aminodiphenylacetylene 42 (93%).
From Phenylbutynyl Azide 14: Chromatography gave unreacted
azide (30%) and 4-phenylbut-3-ynylamine (41) (85%). 41: 1H
NMR: δ ϭ 2.55 (t, J ϭ 7 Hz, 2 H), 2.9 (t, J ϭ 7 Hz, 2 H), 2.0 (br
s, 2 H), 7.2Ϫ7.5 (m, 5 H). Ϫ MS; m/z (%): 145 [Mϩ] (10), 144 (90),
115 (100). Ϫ MS (HR) 145.0895 (C10H11N, [Mϩ], calcd. 145.0891).
(45), 84 (100). Ϫ IR: ν
ϭ 2150 cmϪ1. Thermal decomposition
˜
max
Reactions of Alkynyl Azides 14 and 21a with TMSS. Ϫ General
Procedure: A solution of TMSS (0.2 mmol), the appropriate alkyne
(2 mmol), and AIBN (0.4 mmol) in fluorobenzene was refluxed for
3 h.
of the azide in boiling fluorobenzene for 3 h gave 1,8,9-triazabicy-
clo[4.3.0]nona-6,8-diene (11) in ca. 50% yield (based on starting
1
azide). 11: H NMR: δ ϭ 1.8 (m, 2 H), 2.0 (m, 2 H), 2.75 (t, J ϭ
7 Hz, 2 H), 4.30 (t, J ϭ 7 Hz, 2 H), 7.4 (s, 1 H). Ϫ MS; m/z (%):
123 [Mϩ] (30), 94 (15), 95 (10), 69 (100). Ϫ MS (HR); m/z: 123.0793
(C6H9N3, [Mϩ], calcd. 123.0796)].
1
From Phenylbutynyl Azide 14: H-NMR analysis of the reaction
mixture showed the presence of starting azide (55%), amine 41
(52%) and silylated pyrrole 46 (15%) as the only products. Column
chromatography resulted in the isolation of starting azide, amine
From Pentynyl Azide 2: 1H-NMR analysis of the reaction mix-
ture showed the presence of unreacted azide and adduct 10, n ϭ 1
in a 50:50 ratio. Chromatography gave unreacted azide and a frac-
tion (90%) consisting of the sulfide 10, n ϭ 1, and, possibly, benzo-
thiophene 12 and pyrrole 13 in a 95:5:5 ratio. Repeated chromatog-
raphy led to separation of some pure 5-azido-1-(phenylthio)pent-1-
41, and
a
2:1 mixture (14%) of 2,3-dihydro-5-phenyl-4-
[tris(trimethylsilyl)silyl]-1H-pyrrole (46), and 3,4-dihydro-5-
phenyl-2H-pyrrole (47) .
46: 1H NMR: δ ϭ 0.1 (s, 27 H), 2.75 (t, J ϭ 7 Hz, 2 H; collapsing
to a singlet upon irradiation at δ ϭ 3.8), 3.8 (t, J ϭ 7 Hz, 2 H),
7.2Ϫ7.4 (m, 5 H). Ϫ GC-MS; m/z (%): 391 [Mϩ] (10), 318 (20),
1
ene (10), n ϭ 1 (1:1 Z/E ratio). 10: H NMR: δ ϭ 1.6Ϫ1.8 (m, 4
H), 2.2Ϫ2.4 (m, 4 H; collapsing to dd, J ϭ 6.0 and 1.0 Hz, 2 H,
at δ ϭ 2.27 and dd, J ϭ 6.0 and 1.0 Hz, 2 H, at δ ϭ 2.37 upon
irradiation at δ ϭ 1.71), 3.30 (t, J ϭ 7 Hz, 2 H; collapsing to singlet
upon irradiation at δ ϭ 1.71), 3.34 (t, J ϭ 7 Hz, 2 H; collapsing to
singlet upon irradiation at δ ϭ 1.71), 5.78 (dt, Jd ϭ 10 Hz, Jt ϭ 6
Hz, 1 H), 5.9 (dt, Jd ϭ 15 Hz, Jt ϭ 6 Hz, 1 H), 6.2 (dt, Jd ϭ 15
Hz, Jt ϭ 1 Hz, 1 H), 6.26 (dt, Jd ϭ 10 Hz, Jt ϭ 1 Hz, 1 H), 7.1Ϫ7.4
(m, 5 H). Ϫ MS; m/z (%): 191 [Mϩ Ϫ 28] (80), 158 (100), 130 (60),
109 (40). Ϫ IR: ν˜max ϭ 2150 cmϪ1. Products 12 and 13 could not
be isolated as pure compounds. Their structural assignments are
based only on 1H-NMR and GC-MS analysis of the above mixture.
202 (25), 173 (30), 73 (100).
Ϫ MS (HR); m/z: 391.2008
(C19H37NSi4, [Mϩ], calcd. 391.2003).
47: 1H NMR: δ ϭ 2.0 (quintuplet, J ϭ 7.5 Hz, 2 H), 2.95 (tt,
J1 ϭ 7.5 Hz, J2 ϭ 2.0 Hz, 2 H; collapsing to t, J ϭ 2.0 Hz upon
irradiation at δ ϭ 2.0; collapsing to t, J ϭ 7.5 Hz upon irradiation
at δ ϭ 4.05), 4.05 (tt, J1 ϭ 7.5 Hz, J2 ϭ 2.0 Hz, 2 H; collapsing to
t, J ϭ 2.0 Hz upon irradiation at δ ϭ 2.0), 7.2Ϫ7.4 (m, 5 H). Ϫ
GC-MS; m/z (%): 145 [Mϩ] (30), 144 (20), 117 (100).
From 2-Azidodiphenylacetylene 21a: Chromatography of the re-
action mixture gave unreacted azide (24%), the amine 42 (70%)
contaminated with trace amounts of silylated indole 43 (as sug-
gested by TLC and 1H-NMR), and minor amounts of an unknown
compound [MS; m/z (%): 412 [Mϩ] (10%)]. Repeated chromatogra-
phy of the above amine separated few milligrams of a ca. 5:1 mix-
ture of 42 and 43. This mixture was treated for 10 min. at room
temperature with an excess of anhydrous tetrabutylammonium flu-
oride in THF to give unchanged amine and 2-phenylindole 44
(identified by GC-MS spectral comparison with an authentic speci-
men).[24]
1
Benzothiophene 12: H NMR: δ ϭ 2.0 (quintuplet, J ϭ 7 Hz, 2
H), 2.94 (dt, Jt ϭ 7 Hz, Jd ϭ 1 Hz, 2 H; collapsing to doublet upon
irradiation at δ ϭ 2.0), 3.37 (t, J ϭ 7 Hz, 2 H), 7.1Ϫ7.4 (aromatic
protons). Ϫ GC-MS; m/z (%): 217 [Mϩ] (2), 189 (60), 188 (80), 160
1
(80), 147 (100), 134 (90), 115 (50). Pyrrole 13: H NMR: δ ϭ 2.85
(t, J ϭ 7 Hz, ϪNHϪCH2Ϫ) and 6.55 (s, ϭCHϪ). Ϫ GC-MS; m/z
(%): 191 [Mϩ] (100), 158 (60), 82 (70).
From Phenylbutynyl Azide 14: Chromatography gave unreacted
azide (45%) and a mixture of the sulfide 16 (90:10 Z/E ratio) and
an unknown compound in ca. 90:10 ratio (80% overall yield). Re-
peated chromatography gave pure (Z)-4-azido-1-phenyl-2-
Thermal Decomposition of Azidophenylpropene 37: A solution of
(phenylthio)-but-1-ene [(Z)-16], the pure (E) isomer (E)-16, and a propene 37 (1 mmol) in fluorobenzene (10 ml) was refluxed for 3
1
mixture of (E)-16 and the unknown (1H NMR: δ ϭ 3.1 (t, J ϭ 7 h. GC-MS and H-NMR analysis of the reaction mixture showed
Hz, 2 H), 3.5 (t, J ϭ 7 Hz, 2 H), 7.2Ϫ7.5 (aromatic protons). Ϫ the presence of a 70:30 mixture of starting azide and 2-methylpyr-
MS; m/z: 279 [Mϩ]).
role 40.[20]
Eur. J. Org. Chem. 1998, 1219Ϫ1226
1225

P. C. Montevecchi, M. L. Navacchia, P. Spagnolo
FULL PAPER
Ƞ
Dedicated to the memory of Professor Antonino Fava
nicello, G. Mauriello, P. Scafato, P. Spagnolo, J. Org. Chem.
1995, 60, 2254Ϫ2256; A. DeglЈinnocenti, M. Funicello, P.
Scafato, P. Spagnolo, Chem. Lett. 1994, 1873Ϫ1876; M. Fun-
icello, P. Spagnolo, P. Zanirato, Acta Chem. Scand. 1993, 47,
231Ϫ243.
For a preliminary report of this work see: P. C. Montevecchi,
M. L. Navacchia, P. Spagnolo, Tetrahedron Lett. 1997, 38,
7913Ϫ7916.
(1923Ϫ1997).
[1]
D. P. Curran, N. A. Porter, B. Giese, in Stereochemistry of Rad-
ical Reactions, VCH Publishers, New York, 1996.
[2]
D. P. Curran, “Radical Cyclizations and Sequential Radical Re-
[11]
[12]
actions” in Comprehensive Organic Synthesis, Pergamon Press,
Oxford, U.K., 1991, vol. 4, Chapter 4.2.
[3] [3a]
J. M. Surzur, L. Stella, P. Tordo, Tetrahedron Lett. 1970,
3107Ϫ3108. Ϫ [3b] M. Newcomb, T. M. Deed, D. J. Marquardt,
For previous reports on thiol-mediated radical cyclizations of
alkynes see refs.(([9b][9c][9g][9h])). See also: C. A. Broka, D. E.
C. Reichert, Tetrahedron Lett. 1987, 28, 1503Ϫ1506 and L. El
Kaim, A. Gacon, A. Perroux, Tetrahedron Lett. 1998, 39,
371Ϫ374.
[3c]
Tetrahedron 1990, 46, 2317Ϫ2328. Ϫ
A. L. J. Beckwith, B.
J. Maxwell, J. Tsanaktsidis, Aust. J. Chem. 1991, 44, 1809Ϫ1812.
Ϫ
[3d] S. Kim, G. H. Joe, J. Y. Do, J. Am. Chem. Soc. 1993, 115,
3328Ϫ3329. Ϫ [3e] W. R. Bowmann, D. N. Clark, R. J. Marmon,
Tetrahedron 1994, 50, 1275Ϫ1294. Ϫ [3f] S. Kim, K. S. Yoon, S.
C. Chatgilialoglu, C. Ferreri, “Radical Addition Involving
[13] [13a]
[3g]
S. Kim, H. S. Seo, Tetrahedron 1995, 51, 8437Ϫ8446. Ϫ
S.
CC Triple Bonds” in The Chemistry of Triple-Bonded Functional
Kim, J. Y. Do, J. Chem. Soc., Chem. Commun. 1995,
Groups (Ed.: S. Patai), J. Wiley, New York, 1994, Suppl. C2,
[3h]
[13b]
1607Ϫ1608. Ϫ
B. J. Maxwell, J. Tsanaktsidis, J. Am. Chem.
Vol. 2, Chapter 16. Ϫ
D. P. Curran, “Radical Addition Re-
Soc. 1996, 118, 4276Ϫ4283.
actions” in Comprehensive Organic Synthesis, Pergamon Press,
Oxford, U.K., 1991, Vol. 4, Chapter 4.1.
[4]
[5]
S. Atmaran, A. E. Forrester, M. Gill, R. H. Thomson, J. Chem.
Soc., Perkin Trans 1 1981, 1721Ϫ1724; J. Biovin, E. Fouquet,
S. Z. Zard, Tetrahedron 1994, 50, 1745Ϫ1756; L. E. Kaim, C.
Meyer, J. Org. Chem. 1996, 61, 1556Ϫ1557.
[14]
[15]
P. C. Montevecchi, M. L. Navacchia, P. Spagnolo, J. Org. Chem.
1997, 62, 5846Ϫ5848.
H. S. Dang, B. P. Roberts, J. Chem. Soc., Perkin Trans. 1 1996,
W. R. Bowmann, P. T. Stephenson, N. K. Terrett, A. R. Young,
Tetrahedron 1995, 51, 7959Ϫ7980; M. J. Tomaszewski, J. War-
kentin, N. H. Werstiuk, Aust. J. Chem. 1995, 48, 291Ϫ321; S.
Takano, M. Suzuki, K. Ogasawara, Heterocycles 1994, 37,
149Ϫ153; M. Gioanola, R. Leardini, D. Nanni, P. Pareschi, G.
Zanardi, Tetrahedron 1995, 51, 2039Ϫ2054.
1493Ϫ1497, and references cited therein; see also refs.[3d]
and[3g]
.
[16]
[17]
B. P. Roberts, J. N. Winter, J. Chem. Soc., Perkin Trans. 2
1979, 1353Ϫ1361.
Arenethiols are more powerful hydrogen donors than alkanethi-
ols as a result of their lower SϪH bond strength (80 and 88
kcal/mol, respectively). The syringe pump technique allowed the
use of a very low concentration of benzenethiol 1c.
L. Benati, P. C. Montevecchi, P. Spagnolo, J. Chem. Soc., Perkin
Trans. 1, 1992, 1659Ϫ1664.
[6]
[7]
L. Horner, G. Bauer, Tetrahedron Lett. 1966, 3573Ϫ3577.
[7a] L. Benati, P. C. Montevecchi, P. Spagnolo, Tetrahedron Lett.
[7b]
[18]
[19]
1978, 815Ϫ818. Ϫ
Chem. 1981, 46, 4570Ϫ4573.
L. Benati, P. C. Montevecchi, J. Org.
[8]
S. Kim, G. H. Joe, J. Y. Do, J. Am. Chem. Soc. 1994, 116,
5521Ϫ5522; M. Kizil, J. A. Murphy, J. Chem. Soc., Chem. Com-
mun. 1995, 1409Ϫ1410.
See ref.[10g]; see also: O. Ito, R. Omori, M. Matsuda, J. Am.
Chem. Soc. 1982, 104, 3934Ϫ3937, and S. W. Benson, Chem.
Rev. 1978, 78, 23Ϫ35.
[9] [9a]
[20]
[21]
L. Capella, P. C. Montevecchi, D. Nanni, J. Org. Chem.
P. A. Smith, S. P. Chou, J. Org. Chem. 1981, 46, 3970Ϫ3977.
It is known that thiols are better H-atom donors towards alkyl
radicals than Bu3SnH or TMSS even though sulfurϪhydrogen
bonds in the alkanethiols are relatively strong (see: M. New-
comb, A. G. Glenn, M. B. Maneck, J. Org. Chem. 1989, 54,
4603Ϫ4606; C. Chatigilialoglu, Acc. Chem. Res. 1992, 25,
188Ϫ194; M. Ballestri, C. Chatgilialoglu, P. Dembech, A. Guer-
rini, G. Seconi in Sulfur-Centered Reactive Intermediates in
Chemistry and Biology (Eds.: C. Chatgilialoglu, K.-D. Asmus),
Plenum Press, New York, 1990, pp. 319Ϫ325.
1994, 59, 3368Ϫ3374. Ϫ [9b] L. Benati, L. Capella, P. C. Monte-
[9c]
vecchi, P. Spagnolo, J. Org. Chem. 1994, 59, 2818Ϫ2823. Ϫ
L. Benati, L. Capella, P. C. Montevecchi, P. Spagnolo, J. Org.
Chem. 1995, 60, 7941Ϫ7946. Ϫ [9d] L. Capella, P. C. Montevec-
chi, M. L. Navacchia, J. Org. Chem. 1995, 60, 7424Ϫ7432. Ϫ
[9e]
L. Capella, P. C. Montevecchi, M. L. Navacchia, J. Org.
[9f]
Chem. 1996, 61, 6783Ϫ6789. Ϫ
P. C. Montevecchi, M. L.
Navacchia, P. Spagnolo, Tetrahedron 1997, 53, 7929Ϫ7936. Ϫ
[9g]
P. C. Montevecchi, M. L. Navacchia, J. Org. Chem. 1997,
[9h]
[22]
62, 5600Ϫ5607. Ϫ
Org. Chem. 1998, 63, 537Ϫ542.
P. C. Montevecchi, M. L. Navacchia, J.
M. Alami, F. Ferri, G. Linstrumelle, Tetrahedron Lett. 1993,
34, 6403Ϫ6406.
[10]
[23]
[24]
L. Benati, G. Calestani, D. Nanni, P. Spagnolo, M. Volta, Tetra-
hedron 1997, 53, 9269Ϫ9278; A. DeglЈinnocenti, M. Funicello,
P. Scafato, P. Spagnolo, P. Zanirato, J. Chem. Soc., Perkin Trans.
1 1996, 2561Ϫ2563; A. Capperucci, A. DeglЈinnocenti, M. Fu-
D. R. Coulson, Inorg. Synth. 1972, 13, 121Ϫ124.
R. D. Stephens, C. E. Castro, J. Org. Chem. 1963, 28,
3313Ϫ3315.
[98027]
1226
Eur. J. Org. Chem. 1998, 1219Ϫ1226