Stereoselective Cyclizations and Rearrangements in Vinyl Radicals Promoted by Regioselective Sulfanyl Radical Addition to Enynes

Article abstract of DOI:10.1021/jo9701142

Regioselective radical addition of 4-cyanotoluenethiol (1a) to enynes 3-6 leads to vinyl radicals 7-10 that can undergo five- or six-membered cyclization onto styrene or terminal double bonds in competition with 5-exo cyclization onto the aryl ring. The latter affords spiro-cyclohexadienyl radical intermediates which can either be trapped by 2-cyanoisopropyl radicals or give 1,4-aryl migration products. Regioselective radical addition of phenethanethiol (1b) to enynes 3-6 gives radicals 11-14 which undergo five- or six-membered cyclization onto the alkene double bond; the stereoelectronically disfavored 6-exo cyclization can compete with intermolecular hydrogen abstraction and (to a small extent) with 1,5-hydrogen migration. The 5- and 6-(π-exo)exo cyclization of vinyl radicals 7, 9, 11-13 is highly stereoselective and exclusively or predominately affords products deriving from the (Z)-isomers. Stereochemical evidence indicates that the six-membered endo-cyclization products 30 and 50 could derive from a direct 6-endo cyclization of (E)-radicals (E)-10 and (E)-14 rather than from a 5-exo cyclization/ring expansion process. Sulfanyl radical addition to enynes 3-5 is highly regioselective to the terminal triple bond. In contrast, reaction of thiols 1a,b with enyne 6 leads to products deriving from sulfanyl radical addition to both the CC triple and double bond. This behavior is accounted for by assuming that sulfanyl radical addition to the alkyne triple bond is not reversible, while the addition to the alkene double bond is. Vinyl radical 10 affords product 33 by a rare 1,4-hydrogen migration/fragmentation process.

Full text of DOI:10.1021/jo9701142

5600
J . Org. Chem. 1997, 62, 5600-5607
Ster eoselective Cycliza tion s a n d Rea r r a n gem en ts in Vin yl
Ra d ica ls P r om oted by Regioselective Su lfa n yl Ra d ica l Ad d ition to
En yn es
Pier Carlo Montevecchi* and Maria Luisa Navacchia
Dipartimento di Chimica Organica “A. Mangini”, Universita` di Bologna, Viale Risorgimento 4,
I-40136 Bologna, Italy
Received J anuary 22, 1997X
Regioselective radical addition of 4-cyanotoluenethiol (1a ) to enynes 3-6 leads to vinyl radicals
7-10 that can undergo five- or six-membered cyclization onto styrene or terminal double bonds in
competition with 5-exo cyclization onto the aryl ring. The latter affords spiro-cyclohexadienyl radical
intermediates which can either be trapped by 2-cyanoisopropyl radicals or give 1,4-aryl migration
products. Regioselective radical addition of phenethanethiol (1b) to enynes 3-6 gives radicals 11-
14 which undergo five- or six-membered cyclization onto the alkene double bond; the stereoelec-
tronically disfavored 6-exo cyclization can compete with intermolecular hydrogen abstraction and
(to a small extent) with 1,5-hydrogen migration. The 5- and 6-(π-exo)exo cyclization of vinyl radicals
7, 9, 11-13 is highly stereoselective and exclusively or predominately affords products deriving
from the (Z)-isomers. Stereochemical evidence indicates that the six-membered endo-cyclization
products 30 and 50 could derive from a direct 6-endo cyclization of (E)-radicals (E)-10 and (E)-14
rather than from a 5-exo cyclization/ring expansion process. Sulfanyl radical addition to enynes
3-5 is highly regioselective to the terminal triple bond. In contrast, reaction of thiols 1a ,b with
enyne 6 leads to products deriving from sulfanyl radical addition to both the CC triple and double
bond. This behavior is accounted for by assuming that sulfanyl radical addition to the alkyne
triple bond is not reversible, while the addition to the alkene double bond is. Vinyl radical 10
affords product 33 by a rare 1,4-hydrogen migration/fragmentation process.
Vinyl radicals can be readily generated in several ways,
In the last few years we have been particularly
interested in the chemistry of vinyl radicals generated
by sulfanyl radical addition to alkynes,8 which we have
considered from both a synthetic and a mechanistic
standpoint. Vinyl radicals display a large variety of
reactivity: they can readily undergo 5- or 6-exo cycliza-
tion onto carbon-carbon double and triple bonds,9 and
aromatic2f-k,8b,c,10 and heteroaromatic8e rings, while the
5-endo cyclization onto alkene double bonds8f and aro-
matic rings8b is preferred with linear R-phenyl-substi-
tuted radicals. Moreover, vinyl radicals can undergo 1,5-
radical translocation toward an activated methylene by
intramolecular hydrogen migration.8f,11 We have recently
reported a sequential radical process involving a 1,5-
radical translocation of vinyl radicals as the key step.8f
It appears that no data are presently available con-
cerning the competition between the different cyclization
most of these involving addition to the alkyne triple bond1
of carbon-centered2 and heteroatom-centered radicals
(including sulfanyl,3 stannyl,4 silyl,5 germyl,6 or selenyl7
radicals).
X Abstract published in Advance ACS Abstracts, J uly 1, 1997.
(1) For reviews on radical addition to the alkyne triple bonds see:
(a) Chatgilialoglu, C.; Ferreri, C. Radical Addition involving CC Triple
Bonds. In The Chemistry of Triple-bonded Functional Groups; Patai,
S., Ed.; J . Wiley: New York, 1994; Suppl. C2, Vol. 2, Chapter 16. (b)
Curran, D. P. Radical Addition Reactions. Comprehensive Organic
Synthesis; Pergamon Press: Oxford, U.K., 1991; Vol. 4, Chapter 4.1.
(2) (a) Barton, D. H. R.; Crich, D.; Kretzschmar, K. J . Chem. Soc.,
Perkin Trans. 1 1986, 39; Giese, B.; Lachhein, S. Angew. Chem., Int.
Ed. Engl. 1982, 21, 768. (b) Giese, B.; Gonzales-Gomez, J . A.; Lachhein,
S.; Metzger, O. Angew. Chem., Int. Ed. Engl. 1987, 26, 479. (c) Curran,
D. P.; Kim, D. Tetrahedron 1991, 32, 6171. (d) Curran, D. P.; Kim, D.;
Ziegler, C. Tetrahedron 1991, 32, 6189. (e) Ichinose, Y.; Matsunaga,
S.; Fugami, K.; Oshima, K.; Utimoto, K. Tetrahedron Lett. 1989, 30,
3155. (f) Citterio, A.; Sebastiano, R.; Maronati, A.; Santi, R.; Bergamini,
F. J . Chem. Soc., Chem. Commun. 1994, 1517. (g) Citterio, A.; Cardani,
S.; Sebastiano, R.; Nicolini, M. Gazz. Chim. Ital. 1993, 123, 1035. (h)
Santi, R.; Bergamini, F.; Citterio, A.; Sebastiano, R.; Nicolini, M. J .
Org. Chem. 1992, 57, 4250. (j) Citterio, A.; Sebastiano, R.; Carvayal,
M. C. J . Org. Chem. 1991, 56, 5335.
(3) For a review on sulfanyl radical addition to the alkyne triple
bond see: Chatgilialoglu, C.; Guerra, M. Supplement S: The Chemistry
of Sulfur-containing Functional Groups; Patai, S.; Rappoport, Z., Eds.;
J . Wiley: Chichester, 1993, Chapter 8. See also ref 8 and references
cited therein.
(4) (a) Capella, L.; Montevecchi, P. C.; Nanni, D. J . Org. Chem. 1994,
59, 2818. (b) Pereyre, M.; Quintard, J .-P.; Rahm, A. Tin in Organic
Synthesis; Butterworths: London, 1987. (c) J ung, M. F.; Light, L. A.
Tetrahedron Lett. 1982, 23, 3851. (d) Nozaki, K.; Oshima, K.; Utimoto,
K Tetrahedron 1989, 45, 923. (e) Nozaki, K.; Oshima, K.; Utimoto, K
J . Am. Chem. Soc. 1987, 109, 2547. (f) Konoike, T.; Araki, Y.
Tetrahedron Lett. 1992, 33, 5093. Neumann, W. P. Synthesis, 1987,
665.
(7) (a) Ogawa, A.; Obayashi, R.; Sekiguchi, M.; Masawaki, T.;
Kambe, N.; Sonoda, N. Tetrahedron Lett. 1992, 33, 1329. (b) Kataoka,
T.; Yoshimatsu, M.; Shimizu, H.; Hori, M. Tetrahedron Lett. 1990, 31,
5927. (c) Kataoka, T.; Yoshimatsu, M.; Noda, Y.; Sato, T.; Shimizu,
H.; Hori, M. J . Chem. Soc., Perkin Trans. 1 1993, 121. (d) Ogawa, A.;
Yokoyama, H.; Yokoyama, T.; Masawaki, T.; Kambe, T.; Sonoda, N. J .
Org. Chem. 1991, 56, 5721.
(8) (a) Benati, L.; Montevecchi, P. C.; Spagnolo, P. J . Chem. Soc.,
Perkin Trans. 1 1991, 2103. (b) Benati, L.; Montevecchi, P. C.;
Spagnolo, P. J . Chem. Soc., Perkin Trans. 1 1992, 1659. (c) Benati, L.;
Montevecchi, P. C.; Capella, L.; Spagnolo, P. J . Org. Chem. 1994, 59,
3368. (d) Benati, L.; Capella, L.; Montevecchi, P. C.; Spagnolo, P. J .
Chem. Soc., Perkin Trans. 1 1995, 1035. (e) Benati, L.; Capella, L.;
Montevecchi, P. C.; Spagnolo P. J . Org. Chem. 1995, 60, 7941. (f)
Capella, L.; Montevecchi, P. C.; Navacchia, M. L. J . Org. Chem. 1996,
61, 6783.
(9) (a) Curran, D. P. Radical Addition Reactions. Comprehensive
Organic Synthesis; Pergamon Press: Oxford, U.K., 1991; Vol. 4,
Chapter 4.2. (b) Capella, L.; Montevecchi, P. C.; Navacchia, M. L. J .
Org. Chem. 1995, 60, 7424.
(10) Capella, L.; Montevecchi, P. C.; Nanni, D. J . Org. Chem. 1994,
59, 3368.
(5) (a) Ojima, I. The Chemistry of Organic Silicon Compounds; Patai,
S., Rappoport, Z., Eds.; J . Wiley: Chichester, 1989; Vol. 2, Chapter
25. (b) Chatgilialoglu, C. Acc. Chem. Res. 1992, 25, 188. (c) Kopping,
B.; Chatgilialoglu, C.; Zehnder, M.; Giese, B. J . Org. Chem. 1992, 57,
5147.
(6) Nozaki, K.; Ichinose, Y.; Wakamatsu, K.; Oshima, K., Utimoto,
K. Bull. Chem. Soc. J pn. 1990, 63, 2268.
(11) Curran, D. P.; Shen, W. J . Am. Chem. Soc. 1993, 115, 6051.
Gimisis, T.; Chatgilialoglu, C. J . Org. Chem 1996, 61, 1908.
S0022-3263(97)00114-X CCC: $14.00 © 1997 American Chemical Society

Regioselective Sulfanyl Radical Addition to Enynes
J . Org. Chem., Vol. 62, No. 16, 1997 5601
Sch em e 1
Sch em e 2
modes (five- or six-membered, exo or endo) onto different
radicophilic moieties and the intramolecular hydrogen
migration. With the aim of reaching this goal we planned
a study of the fate of vinyl radicals generated by addition
of methanesulfanyl radicals 2 (having suitable R-radico-
philic substituents) to alkynes carrying radicophilic
moieties in 5- or 6-position. We report herein results
obtained from the radical addition of thiols 1a ,b to the
alkynyl ethers 3-6. We reasoned that the regioselective
addition12 of AIBN-generated sulfanyl radicals 2a ,b to
the alkyne triple bond of 3-6 would lead to radicals
7-14, from which 5- (or 6)-exo cyclization onto the
terminal (or the styrene) double bond might occur in
competition with the 5-exo cyclization onto the aryl ring
(see radicals 7-10) and the radical migration toward the
R-oxy allylic methylene and the benzylic methylene (for
radicals 11-14)13 (Scheme 1).
intermediates by 2-cyanopropyl radicals owing to the
high AIBN concentration employed. However, in these
cases more complex reaction mixtures were generally
obtained.
Reaction of thiol 1a with butynyl ether 3 was carried
out under method B conditions. Chromatography of the
resulting reaction mixture allowed for the separation of
the pyran 17 (28%), the methyl sulfide 20 (6%), and the
spiro-compound 18 (11%) (Scheme 2). The pyran 17 and
the sulfide 20 were obtained as pure stereoisomers. The
Z configuration for 20 was postulated on the basis of our
previous evidence on the 1,4-aryl migration toward vinyl
radicals.8c The E configuration for the pyran 17 was
established by NOE measurements. Compounds 17, 18,
and 20 were derived from vinyl radical intermediate 7
through two competitive processes. 6-Exo cyclization of
radical (Z)-7 onto the styrene double bond led to benzyl
radical 15, and then to the pyran 17 by hydrogen
abstraction, while 5-exo cyclization of radical (E)-7 onto
the aromatic ring led to the spiro-cyclohexadienyl radical
16. This radical could either be trapped by 2-cyanopropyl
radicals, leading to the spiro-compound 18, or undergo
fragmentation of the spiro-ring leading to the thiomethyl
radical 19 and then to the 1,4-aryl migration product 20
by subsequent hydrogen abstraction. We have previously
reported examples of 1,4-aryl migrations toward vinyl
radicals,8c although at that time the intermediacy of
cyclohexadienyl radicals was only postulated. The present
formation of the coupling product 18 strongly supports
our previous claim.
Resu lts a n d Discu ssion
Reactions of thiols 1a ,b with the alkynyl ethers 3-6
were generally carried out under conditions of method A
and/or method B, as described in the Experimental
Section. According to method A, a benzene solution of
thiol 1a ,b containing a 2-fold excess of ether 3-6 and
0.1 mol equiv of AIBN was refluxed for 3 h. Following
method B, a benzene solution of thiol 1a ,b was slowly
added over 3 h with a syringe pump to a boiling benzene
solution containing a 2-fold excess of the appopriate
alkyne 3-6 and equimolar amounts of AIBN. In both
cases the reaction mixtures were washed with 10%
aqueous NaOH to separate the unreacted thiol 1a ,b and
then chromatographed on silica gel column. In all cases
thiol 1a ,b was generally recovered in 10-40% yield after
acidification of the aqueous layer.
In principle the conditions of method A should favor
intermolecular hydrogen abstraction reactions, since
thiols are good hydrogen donors. Vice versa, the condi-
tions of method B should be preferred for trapping radical
(12) It is well known that radical species, including arenesulfanyl
(Broka, C. A.; Reichert, D. E. C. Tetrahedron Lett. 1987, 28, 1503),
selenyl (ref 7d), and stannyl radicals (ref 16a) promote the cyclization
of enynes by regioselective addition to the CC triple bond. These
radicals facilely add to both CC double and triple bonds of enynes
leading to alkyl and vinyl radical intermediates. The reason for the
regioselective addition is the greater reversibility of radical addition
to alkenes compared with alkynes (see ref 1a).
Although the overall yields from the reaction of 1a with
3 are not quite satisfactory, it appears that the 5-exo
cyclization onto the aromatic ring and the 6-exo cycliza-
tion onto the styrene double bond occur at comparable
rates. These two cyclization modes compete favorably
(13) Preliminary results have been previously reported; see: Mon-
tevecchi, P. C.; Navacchia, M. L. Tetrahedron Lett. 1996, 37, 6583.

5602 J . Org. Chem., Vol. 62, No. 16, 1997
Montevecchi and Navacchia
Sch em e 5
Sch em e 3
Sch em e 4
with both the intermolecular hydrogen abstraction and
the 1,5-hydrogen migration from the R-oxy allylic meth-
ylene.
incapable of undergoing competitive 5-exo cyclization onto
the aromatic ring.
Thiol 1a was let to react with ether 4 under conditions
of both methods A and B to give exclusively products 21
and 23 deriving from radical 8 through 5-exo cyclization
onto the aromatic ring (Scheme 3). In this case as well
no products deriving from radical 8 by intermolecular
hydrogen abstraction or hydrogen migration were ob-
tained. Moreover, we did not detect any product derived
from radical 8 by 6-exo cyclization onto the terminal CC
double bond. The scarce tendency of the terminal double
bond to undergo 6-exo cyclization can be ascribed to both
stereoelectronic effects, which favor the 5-exo cycliza-
tion,15 and to thermodynamic effects, which favor the
cyclization onto the styrene double bond with formation
of a stable benzyl radical.
The importance of stereoelectronic effects is clearly
evidenced by the reaction of thiol 1a with ether 5, which
gave only the tetrahydrofuran 24 under method A and
B conditions (64% and 50% yields, respectively) (Scheme
4). Compound 24 (as pyran 17) was obtained as pure
(E) isomer through 5-exo cyclization of radical intermedi-
ate (Z)-9. The E configuration was established by NOE
measurements. Vinyl radical intermediate 9 seems
Thiol 1a reacted with ether 6 to give products 29, 30,
33, and 35 (Scheme 5). Reaction products 29 and 30
arose from vinyl radical 10 by competitive cyclization onto
the aromatic ring and the terminal double bond, respec-
tively. Methyl sulfide 29 was expected to derive from
5-exo cyclization onto the aromatic ring and subsequent
fragmentation of intermediate spiro-radical 27, while the
pyran 30 might be derived from either 6-endo and/or
5-exo cyclization onto the terminal double bond. In this
latter case, the exo radical 26 would afford the endo
radical 28 through
a well known ring-expansion
process.8c,15c,d,16 Pyran 30 was obtained in both (E) and
(Z) configurations, with a ca. 2:1 (E)/(Z) ratio. The actual
configurational structure was established for the (E)
isomer by NOE measurements.
The rearranged sulfide 33 was initially believed13 to
derive from radical 10 through initial 1,4-hydrogen
migration leading to the translocated radical 31. Sub-
sequent carbon-oxygen bond cleavage (with loss of
acrolein) would afford the allyl radical 32 and then 33
by hydrogen abstraction. Analogues 1,4-hydrogen migra-
tion/C-O fragmentation processes have been recently
reported.17 However, we observed that the homologous
(14) Compound (Z)-23 was contaminated by minor amounts of its
(E)-isomer [Z/E ratio ca. 9:1] probably deriving from a postisomeriza-
tion process.
(15) (a) Baldwin, J . E. J . Chem. Soc., Chem. Commun. 1976, 734.
(b) Beckwith, A. L. J .; Easton, C. J .; Serelis, A. K. J . Chem. Soc., Chem.
Commun. 1980, 482. (c) Crich, D.; Fortt, S. H. Tetrahedron Lett., 1987,
28, 2895. (d) Stork, G.; Mook, R Tetrahedron Lett., 1986, 27, 4529.
(16) (a) Stork, G.; Mook, R. J . Am. Chem. Soc. 1987, 109, 2829. (b)
Beckwith, A. L. J .; O’Shea, D. M. Tetrahedron Lett. 1986, 27, 4525.
(17) Crich, D.; Sun, S.; Brunckova, J . J . Org. Chem. 1996, 61, 605;
Hart, D. J .; Kuzmich, D J . Chin. Chem. Soc. 1995, 42, 873.

Regioselective Sulfanyl Radical Addition to Enynes
J . Org. Chem., Vol. 62, No. 16, 1997 5603
Sch em e 6
Sch em e 7
products 40 and 41 in a 59:41 ratio (75% overall yield).
Both products 40 and 41 were obtained in a stereoselctive
fashion as pure (E) stereoisomers; the actual configura-
tions were assigned by NOE measurements. This finding
led us to suggest that the methyl ether (E)-41 was derived
from stereoselective π-exo cyclization of radical (Z)-37
followed by scission of the spiro-ring of radical 39, while
the methyl sulfide (E)-40 was the expected product from
π-endo cyclization of radical (E)-37 (Scheme 6). Since
R-alkyl-substituted vinyl radicals exist as rapidly inter-
converting (E) and (Z) isomers,1a the [41]/[40] ratio
depends on kexo and kendo values as well as on the E/Z
equilibrium constant, Ke, through the equation [41]/[40]
) Kekexo/kendo
. The observed slight preference of the
π-endo cyclization mode (Kekexo/kendo ) 0.69) might result
from the fact that (E) radicals could be preferred at the
equilibrium (Ke < 1) rather than from more favorable
stereoelectronic conditions.
In all cases examined, we did not find any products
deriving from radicals 7-10 by hydrogen abstraction nor
any products deriving from radicals 7, 8 by 1,5-hydrogen
migration, even though the 1,5-hydrogen migration from
an activated methylene toward vinyl radicals is a feasible
process. In particular, we have previously reported8f that
â-ethanesulfanyl-substituted vinyl radicals carrying a
suitable activating group in δ position smoothly undergo
a 1,5-hydrogen migration/fragmentation process. This
reaction leads to alkenesulfanyl radicals and alkene by
C-S bond scission of the new translocated radicals
(Scheme 7). Thus, in order to study the competition
between the 1,5-hydrogen migration and the five- and
six-membered cyclization onto the CC double bond, the
fate of radicals 11-14, which were generated by reacting
thiol 1b with alkynyl ethers 3-6 under method A
conditions, was considered.
The reaction of thiol 1b with ether 3 gave products (E)-
42 (42%) and 45 (8%). Product 45 is expected from vinyl
radical 11 through hydrogen abstraction. This reaction
can compete, although to a minor extent, with the 6-exo
cyclization onto the styrene double bond leading to pyran
42 (Scheme 8). On the other hand, the reaction with
ether 4 afforded major amounts of the hydrogen abstrac-
tion product 46 (44%) in addition to minor amounts of
the 6-exo cyclization product 43 (21%), as a result of the
lesser reactivity of the terminal double bond with respect
to the styrene double bond (Scheme 8). Pyran 42,
analogous to the cyclization products 17 and 24, was
obtained in a high stereoselective fashion as an (E)
stereoisomer,19 while radical 12 gave a 80/20 mixture of
isomeric (E)- and (Z)-43. The actual configurational
structure for pyran 43 was established for the (E) isomer
by NOE measurements.
vinyl radical 8 did not undergo any 1,5-hydrogen migra-
tion (see Scheme 3), even though this process (much more
feasible than the corresponding 1,4-hydrogen migration18)
would lead to an R-oxy allyl radical analogous to the
translocated radical 31. This observation led us to claim
that the loss of acrolein, with concomitant formation of
the stable R-thio-substituted allyl radical 32, might be
the driving force of the entire process. On this basis we
might consider the possibility that a one-step, concerted
mechanism takes place leading to 32 from 10; that is,
the 1,4-hydrogen migration might be favored by the
concomitant carbon-oxygen bond cleavage.
Compound (Z)-30 was contaminated by a minor prod-
uct which was likely the tetrahydrofuran 35 [80:20 (Z)-
30/35 ratio](Scheme 5). Compound 35 could be derived
from competitive sulfanyl radical 2a addition to the
carbon-carbon double bond of ether 6 followed by 5-exo
cyclization onto the CC triple bond of the resulting alkyl
radical 34. The nonregioselective sulfanyl radical pro-
moted cyclization of the enyne 6 will be discussed later.
The above results indicate that the cyclization onto the
styrene double bond predominates over the cyclization
onto both the terminal double bond and the aromatic
ring; these latter reactions occur at comparable rates.
However, it should be noted that the cyclization onto the
aromatic ring is π-endo, while the cyclization onto the
CC double bond is π-exo; these cyclization modes might
have different stereoelectronic demands, also in light of
the fact that different heteroatoms (oxygen and sulfur)
are present in the exo and endo chains of radicals 7-10.
To evaluate this point we investigated the fate of
radical 37 which was expected to cyclize onto both aryl
rings to an extent only depending on the different
stereoelectronic demands (Scheme 6). The reaction of
thiol 1a with alkynyl ether 36 carried out under condi-
tions of method A furnished a mixture of the rearranged
No product deriving from radicals 11 or 12 through
1,5-hydrogen migration could be separated by column
(18) 1,4-Hydrogen migrations are very rare. For examples see:
Brunton, G.; Griller, D.; Barclay, L. R. C.; Ingold, K. U. J . Am. Chem.
Soc. 1976, 98, 6803. J ournet, M.; Malacria, M. J . Org. Chem. 1992,
57, 3085. Wallace, T. J .; Gritter, R. J . J . Org. Chem., 1961, 26, 5256.
(19) The (E) configuration of the 6-(π-exo)exo cyclization product 42
and the 5-(π-exo)exo product 25 was attributed on the basis of the
general evidence provided by related cyclization products 17, 24, and
43, whose (E) configurations were assigned by NOE measurements.

5604 J . Org. Chem., Vol. 62, No. 16, 1997
Montevecchi and Navacchia
Sch em e 9
Sch em e 8
been reported on alkanesulfanyl radical addition to
alkynes, and little is known about alkanesulfanyl radical
addition to alkenes.20 However, even though no com-
parative data are available for alkanesulfanyl radicals,
results reported for arenesulfanyl radical addition to
1-pentyne and 1-hexene indicate that alkylacetylenes and
terminal alkenes react with sulfanyl radicals nearly at
the same rate.21 Both sulfanyl radical addition to alkenes
and alkynes are reversible, but the reversibility is less
important for alkynes. Ito and co-workers21 estimated
the absolute rate constants k1 and k-1 by measuring the
arenesulfanyl radical decay in the presence of alkenes
(and alkynes) and oxygen. The k-1 values for the addition
to terminal alkenes not having electron acceptor substit-
uents was found to be ca. 100 times greater than k1. In
contrast, the addition to terminal alkynes was not found
to be reversible under the conditions employed. The
importance of the C-S bond cleavage should strongly
diminish on passing from arenethiyl to alkanethiyl
radicals as a result of the increased C-S bond strength
(the C-SPh bond is 10 kcal/mol weaker than the C-SAlk
bond22). On this basis, it can be assumed that alkane-
sulfanyl radical addition to enynes occurs in a nonrevers-
ible fashion to the CC triple bond, and in a reversible
fashion to the CC double bond. Thus, products deriving
from the latter reaction are expected only if the resulting
alkyl radical can react before equilibrating. This is the
case for radicals 34 and 48 (see Schemes 5 and 9), which
can undergo a facile 5-exo cyclization onto the CC triple
bond.
St er eoselect ivit y of t h e (π-exo) Vin yl R a d ica l
Cycliza tion .23 As mentioned above, the (π-exo)exo five-
membered cyclization products 24 and 25 (see Scheme
4), as well as the (π-exo)exo six-membered products 17
and 42 (see Schemes 2 and 8), were obtained as pure (E)
stereoisomers. These products were derived from ste-
reoselective cyclization of vinyl radicals (Z)-7, -9, -11, -13.
Moreover, the rearranged compound 41 was obtained as
pure (E) isomer which was derived from radical (Z)-37
by stereoselective 5-(π-exo)exo cyclization onto the aryl
ring (see Scheme 6). The apparent incapability of
radicals (E)-7, -9, -11, -13, -37 to undergo (π-exo)exo
cyclization can be ascribable to steric repulsion which
chromatography. However, GC-MS analysis of the reac-
tion mixture of 1b with 4 detected formation of styrene
which was expected from the translocated radical 44
through C-S bond cleavage (Scheme 8).
However, it would appear that the hydrogen abstrac-
tion reaction and, to a lesser extent, the 1,5-hydrogen
migration can only compete with the stereoelectronically
disfavored 6-exo cyclization. In fact, thiol 1b reacted with
ether 5 to give compound (E)-25 (53%) as the only
identifiable reaction product (Scheme 4). Compound (E)-
25 was obtained in a highly stereoselective fashion by
radical (Z)-13 through 5-exo cyclization onto the styrene
double bond.19
Similarly, reaction of thiol 1b with the ether 6 gave
the pyran 50 (67%) as the only product deriving from
radical 14. The pyran 50 resulted from cyclization onto
the terminal double bond through intermediacy of the
six-membered endo-radical 49. The latter, analogous to
the endo-radical 28 (see Scheme 5), might be formed
either by a 5-exo cyclization followed by ring expansion
of resulting exo-radical 47 or by a 6-endo cyclization.
Pyran 50 was obtained in both (E) and (Z) configurational
forms with a ca. 2:1 (E)/(Z) isomeric ratio. The actual
configurational structure was established for the (E)
isomer by NOE measurements.
The reaction of 1b with 6 furnished product 52 (22%)
in addition to pyran 50. Product 52, analogous to 35,
could be the result of sulfanyl radical addition to the CC
double bond of ether 6 and subsequent 5-exo cyclization
of the resulting alkyl radical 48 onto the terminal alkyne
triple bond (Scheme 9).
Regioselectivity of Su lfa n yl Ra d ica l Ad d ition to
En yn es. Sulfanyl radical addition to the employed
enynes was generally found to be regioselective to the
CC triple bond, but with ether 6 a marked nonregiose-
lectivity was observed. This behavior can be accounted
for on the basis of the general evidence reported in the
literature for the sulfanyl radical addition to CC double
and triple bonds. To our knowledge no kinetic data have
(21) Ito, O.; Omori, R.; Matsuda, M. J . Am. Chem. Soc. 1982, 104,
3934.
(22) Benson, S. W. Chem. Rev. 1978, 78, 23.
(23) For a recent monograph of stereoselective radical reactions
see: Curran, D. P.; Porter, N. A.; Giese, B. Stereochemistry of Radical
Reactions; VCH Publishers: New York, 1996.
(20) In particular, the absolute rate of addition of butanesulfanyl
radicals to cyclopropylethylene and 1-octene (McPhee, D. J .; Campre-
don, M.; Lesage, M.; Griller, D J . Am. Chem. Soc. 1989, 111, 7563)
and 1-pentene (Sivertz, C. J . Phys. Chem. 1959, 63, 34) have been
reported.

Regioselective Sulfanyl Radical Addition to Enynes
J . Org. Chem., Vol. 62, No. 16, 1997 5605
occurs in the transition state between the radicophilic
moiety (styrene double bond or aromatic ring) and the
â-substituent. Support for our claim arose from the
evidence that vinyl radical 12 can undergo 6-(π-exo)exo
cyclization onto the terminal double bond in both (Z) and
(E) configurations leading to (E)- and (Z)-43, respectively,
in a 80:20 ratio (see Scheme 8). The formation of minor
amounts of the (Z) isomer (Z)-43 is consistent with the
minor steric demand of the terminal CC double bond with
respect to the styrene double bond (and the aromatic
ring), which allows radical 12 to cyclize in the (E)
configuration as well, although to a minor extent.
In contrast, the formal six-membered endo-cyclization
onto the terminal CC double bond products 30 and 50
(see Schemes 5 and 9) were obtained in a stereoselective
mode as 2:1 (E)/(Z) mixtures. In light of the evidence
provided above which indicates that the (π-exo)exo cy-
clization is largely preferred for (Z)-vinyl radicals, this
finding seems inconsistent with a mechanism leading to
30 and 50 by initial 5-(π-exo)exo cyclization followed by
ring-expansion (see Schemes 5 and 9). In fact, in this
case we would expect either complete nonstereoselectivity
or (if transannular bond scission is fast enough to
compete with rotation around the CC single bond)
predominance of the (Z) isomer. Vice versa, the observed
stereoselectivity might be accounted for by assuming that
a 6-(π-exo)endo cyclization occurs. In our opinion no
steric hindrance should be expected between the radico-
philic double bond and the â-substituents for this cy-
clization mode; thus, the cyclization rates of the (E) and
(Z) radicals should be equalized. In this assumption the
observed predominance of the (E) isomers (E)-30 and (E)-
50 would parallel the predominance at the equilibrium
of the more stable (E)-vinyl radicals (E)-10 and (E)-14
with respect to their (Z)-isomers. However, our sugges-
tion deserves further attention, since there is abundant
evidence in the literature that vinyl radical cyclizations
lead to thermodynamically more stable 6-endo radicals
through ring-expansion of kinetically preferred 5-exo
radicals.8c,15d,16
Results obtained from radicals 11-14 indicate that the
intermolecular hydrogen abstraction can only compete
with the stereoelectronically disfavored 6-exo cyclization
onto the styrene bond (to a minor extent) and onto the
terminal double bond (to a major extent). Some evidence
has been provided which indicates that the 1,5-hydrogen
migration from a benzylic group can compete with the
reaction onto the terminal double bond.
The six-membered endo-cyclization products 30 and 50,
deriving from radicals 10 and 14 (Schemes 5 and 9),
might be formed either by a 5-exo cyclization/ring expan-
sion process or by a direct 6-endo cyclization. However,
stereochemical evidence appears to be inconsistent with
the 5-exo cyclization/ring expansion process. In fact,
pyrans 30 and 50 were obtained predominantly in the
(E) configuration, which arose by cyclization of (E)-
radicals (E)-10 and (E)-14. A preferred 5-exo cyclization
of (E)-radicals (E)-10 and (E)-14 is not consistent with
the general behavior exhibited by vinyl radicals 7, 9, 11-
13, and 37 which undergo exclusive (or predominant) 5-
or 6-(π-exo)exo cyclization in the (Z)-configuration due to
steric repulsion occurring in the transition state between
the exo double bond (or the aryl ring) and the â-substitu-
ent.
Sulfanyl radical addition to enynes 3-5 is highly
regioselective to the terminal triple bond. In contrast,
reaction of thiols 1a ,b with enyne 6 led to products
deriving from sulfanyl radical addition to both the CC
triple and double bond. This peculiar behavior is ac-
counted for by assuming that sulfanyl radical addition
to the alkyne triple bond is not reversible, while the
addition to the alkene double bond is. The only route
opened for alkyl radicals deriving from addition to the
alkene double bond of 3-5 appears to be â-scission
leading to starting enyne and sulfanyl radical. In
contrast, sulfanyl radical addition to enyne 6 gives alkyl
radicals 34 and 48 which can undergo a facile 5-exo
cyclization onto the alkyne triple bond.
Exp er im en ta l Section
Sta r tin g Ma ter ia ls. Thiol 1a 8c was prepared as previously
described. Thiol 1b is commercially available. Propynyl
ethers 5,24 6,25 and 36 were obtained in ca. 80% yields by
heating in a sealed tube a THF solution (50 mL) of sodium
propargylate (50 mmol) (prepared from equimolar amounts of
propargyl alcohol and sodium hydride in anhydrous THF) and
the appropriate commercially available bromide (trans-cin-
namyl bromide, allyl bromide, and 4-cyanobenzyl bromide,
respectively) at 80 °C for 8 h. Butynyl ethers 3 and 4 were
similarly prepared from sodium but-3-yn-1-oate. [3: 1H NMR
δ (200 MHz) 2.0 (1H, t, J ) 2.6 Hz), 2.50 (2H, dt, J d ) 2.6 Hz,
J t ) 7.0 Hz), 3.60 (2H, t, J ) 7.0 Hz), 4.20 (2H, br d, J ) 6.0
Hz), 6.30 (1H, dt, J d ) 16 Hz, J t ) 6.0 Hz), 6.65 (1H, br d, J d
) 16.0 Hz), 7.2-7.45 (5H, m); MS m/z (rel inten) 186 (M+, 10),
185 (10), 117 (100), 115 (45), 105 (100). Anal. Calcd for
C13H14O: C, 83.83; H, 7.58; O, 8.60. Found: C, 83.95; H, 7.55.
4: 1H NMR δ (200 MHz) 1.96 (1H, t, J ) 2.6 Hz), 2.50 (2H, dt,
J t ) 7.0 Hz, J d ) 2.6 Hz), 3.60 (2H, t, J ) 7.0 Hz), 4.05 (2H,
br d, J ) 6 Hz), 5.20 (1H, br d, J ) 10 Hz), 5.30 (1H, br d, J
) 15 Hz), 5.92 (1H, ddt, J 1 ) 15 Hz, J 2 ) 10 Hz, J t ) 6 Hz);
MS m/z (rel inten) 110 (M+, 2), 79 (20), 71 (20), 41 (100), 39
(60). Anal. Calcd for C7H10O: C, 76.32; H, 9.15; O, 14.58.
Found: C, 76.5; H, 9.20. 36: 1H NMR δ (200 MHz) 2.48 (1H,
t, J ) 2 Hz), 4.20 (2H, d, J ) 2 Hz), 4.60 (2H, s), 7.40 (2H, d,
J ) 8.5 Hz), 7.60 (2H, d, J ) 8.5 Hz); MS m/z (rel inten) 171
Con clu sion s
Regioselective radical addition of thiols 1a ,b to enynes
3-6 leads to vinyl radicals 7-14, suitable for the study
of the competition between cyclization onto different
radicophilic groups and inter- or intramolecular hydrogen
abstraction reaction. The results obtained from vinyl
radicals 7-10 indicate that the five-membered cyclization
onto the styrene double bond predominates over the six-
membered endo-cyclization onto the terminal double bond
as well as the 5-exo cyclization onto the 4-cyanophenyl
ring. This last reaction affords spiro-cyclohexadienyl
radical intermediates which can either be trapped by
2-cyanoisopropyl radicals or give 1,4-aryl migration
products by fragmentation of the spiro-ring. Moreover,
5-exo cyclization onto the 4-cyanophenyl ring and 6-exo
cyclization onto the styrene double bond occur at com-
parable rates; both largely predominate over the 6-exo
cyclization onto the terminal double bond.
Cyclizations onto the aromatic ring and the double
bond occur in a (π-endo) and a (π-exo) mode, respectively.
Evidence has been provided that these two cyclization
modes have comparable stereoelectronic demands, as
shown by radical 37 (see Scheme 6) which furnished the
5-(π-endo)exo and the 5-(π-exo)exo products 40 and 41 in
a 59:41 ratio.
(24) Trost, B. M.; Edstrom, E. D.; Carter-Petillo, M. B. J . Org. Chem.
1989, 54, 4489.
(25) Fellous, R.; Rabin, J . P.; Lizzani-Cuvelier, L.; Luft, R. Bull.
Chem. Soc. Fr. 1974, 923.

5606 J . Org. Chem., Vol. 62, No. 16, 1997
Montevecchi and Navacchia
(M+, 10), 170 (20), 132 (95), 130 (60), 117 (60), 116 (95), 104
(75), 39 (100). Anal. Calcd for C11H9NO: C, 77.17; H, 5.30;
N, 8.18; O, 9.35. Found: C, 77.3; H, 5.32; N, 8.15].
12.40.]; (Z)-3-(4-cya n op h en yl)-4-(m et h ylt h io)b u t -3-en -1-
yl a llyl eth er (Z)-23 (305 mg, 70% (method A); 70 mg, 18%
(method B)] [1H NMR δ 2.30 (3H, s), 2.72 (2H, t, J ) 7 Hz),
3.40 (2H, t, J ) 7 Hz), 3.90 (2H, br d, J ) 6 Hz), 5.12 (1H, br
Rea ction P r od u cts. Reaction products were separated by
Merck silica gel column chromatography (0.040-0.063 particle
size) of the reaction mixtures by gradual elution with light
petroleum ether (bp 40-70 °C)-diethyl ether. Yields were
based on reacted thiol 1a ,b. Structural assignments generally
came from 1H NMR and MS spectral data in addition to
elemental analysis. Elemental analysis was not performed for
the spiro compound 18, owing to the impossibility of obtaining
a pure sample, nor for compounds 35 and 52 which were
obtained as inseparable mixtures with (Z)-30 and (Z)-50,
respectively. Their identification arose from careful MS and
1H NMR spectral analysis. 1H NMR spectra were recorded at
200 (or 300 MHz) with Me4Si as internal standard. Mass
spectra were recorded with the electronic impact method.
Rea ction s of Th iols 1a ,b w ith Alk yn yl Eth er s 3-6, 36.
Meth od A. A benzene solution (25 mL) of the appropriate
thiol 1a ,b (2 mmol), the appropriate ether 3-6, 36 (4 mmol),
and AIBN (0.2 mmol) was refluxed for 3 h. The reaction
mixture was washed twice with NaOH 10% and once with
water, the organic layer was dried over Na2SO4, and the
solvent was evaporated off. The unreacted thiol 1a ,b was
recovered by acidification of the aqueous layer and extraction
with diethyl ether.
d, J ) 10 Hz), 5.20 (1H, br d, J ) 16 Hz), 5.82 (1H, ddt, J 1
)
16 Hz, J 2 ) 10 Hz, J t ) 6 Hz), 6.20 (1H, s), 7.40-7.70 (4H,
A2B2 system); MS m/z (rel inten) 259 (M+, 25), 202 (30), 188
(60), 41 (100). Anal. Calcd for C15H17NOS: C, 69.46; H, 6.61;
N, 5.40; O, 6.17; S, 12.36. Found: C, 69.35; H, 6.58; N, 5.42
; S, 12.30.]; sp ir o-cycloh exa d ien e 21 [45 mg, 9% (method
B)] [1H NMR δ 1.55 (6H, s), 2.20 (2H, dt, J d ) 1.5 Hz, J t ) 6.5
Hz), 3.20 (2H, s), 3.55 (2H, t, J ) 6.5 Hz), 4.00 (2H, br d, J )
6 Hz), 5.20 (1H, br d, J ) 10 Hz), 5.30 (1H, br d, J ) 15 Hz),
5.92 (2H, d, J ) 9.5 Hz; superimposed to 1H, ddt, J 1 ) 10 Hz,
J 2 ) 15 Hz, J t ) 6 Hz), 6.10 (1H, t, J ) 1.5 Hz), 6.20 (2H, d,
J ) 9.5 Hz); MS m/z (rel inten) 326 (M+, 20), 200 (40), 116
(60), 41 (100). Anal. Calcd for C19H22N2OS: C, 69.90; H, 6.79;
N, 8.58; O, 4.90; S, 9.82. Found: C, 69.65; H, 6.75; N, 8.61; S,
9.78.]. Thiol 1a was recovered in ca. 15% yield (method A)
and 25% (method B).
F r om 4-Cya n o-r-tolu en eth iol (1a ) a n d P r op yn yl P h en -
ylp r op en yl Eth er (5). Chromatography gave (E)-3-ben zyl-
4-[(4-cyan o-r-tolylth io)m eth yliden e]tetr ah ydr ofu r an (E)-
24 [220 mg, 64% (method A); 250 mg, 50% (method B)] [1H
NMR δ 2.68 (ABX system, J AB ) 12 Hz, J AX ) 8.5 Hz, J BX
)
7 Hz; inner lines separation 6 Hz), 2.95 (1H, m), 3.6 (1H, dd,
J 1 ) 8.5 Hz, J 2 ) 5.4 Hz), 3.8 (2H, s), 3.85 (1H, dd, J 1 ) 8.5
Hz, J 2 ) 7 Hz), 4.30 (2H, ABX system, J AB ) 14 Hz, J AX ) J BX
) 2 Hz; inner lines separation 2 Hz), 5.50 (1H, q, J ) 2 Hz,
collapsing to triplet upon irradiation at δ 2.95), 7.10-7.30 (5H,
m), 7.35 (2H, d, J ) 8.5 Hz), 7.60 (2H, d, J ) 8.5 Hz).
Irradiation at δ 5.50 caused an enhancement at δ 2.95 (allylic
proton) and 2.68 (benzylic protons). MS m/z (rel inten) 321
(M+, 10), 230 (25), 205 (10), 116 (100), 91 (40). Anal. Calcd
for C20H19NOS: C, 74.73; H, 5.96; N, 4.36; O, 4.98; S, 9.97.
Found: C, 74.50; H, 5.93; N, 4.34; S, 9.93.]. Thiol 1a was
recovered in ca. 45% yield (method A) and 20% (method B).
F r om 4-Cya n o-r-tolu en eth iol (1a ) a n d P r op yn yl P r o-
p en yl Eth er (6). Chromatography gave 4-cya n oben zyl a llyl
su lfid e 33 [40 mg, 15% (method A)] [1H NMR δ 3.00 (2H, br
d, J ) 7 Hz), 3.70 (2H, s), 5.05 (1H, br d, J ) 16.5 Hz), 5.15
(1H, br d, J ) 10 Hz), 5.77 (1H, ddt, J 1 ) 16.5 Hz, J 2 ) 10 Hz,
J t ) 7 Hz), 7.40 (2H, d, J ) 8.5 Hz), 7.60 (2H, d, J ) 8.5 Hz);
MS m/z (rel inten) 189 (M+, 30), 147 (45), 116 (100), 73 (50).
Anal. Calcd for C11H11NS: C, 69.80; H, 5.86; N, 7.40; S, 16.94.
Found: C, 69.60; H, 5.88; N, 7.37; S, 17.00.]; (Z)-2-(4-
cya n op h en yl)-3-(m eth ylth io)p r op -2-en -1-yl a llyl eth er 29
[26 mg, 7% (method A)] [1H NMR δ 2.32 (3H, s), 4.00 (2H, dt,
J d ) 6 Hz, J t ) 1.5 Hz), 4.22 (2H, s), 5.20 (1H, dq, J d ) 10 Hz,
J q ) 1.5 Hz), 5.26 (1H, dq, J d ) 15 Hz, J q ) 1.5 Hz), 5.90 (1H,
ddt, J 1 ) 15 Hz, J 2 ) 10 Hz, J t ) 6 Hz), 6.45 (1H, s), 7.50-
7.70 (4H, A2B2 system, J ) 8.5 Hz). Minor peaks at δ 2.45
(s), 4.42 (s) and 6.80 (s) were assigned to the E isomer (Z/E
ratio ) 9:1); MS m/z (rel inten) 245 (M+, 20), 188 (30), 71(60),
41(100). Anal. Calcd for C14H15NOS: C, 68.54; H, 6.16; N,
5.71; O, 6.52; S, 13.07. Found: C, 68.75; H, 6.18; N, 5.69; S,
13.10.]; a 4:1 mixture of (Z)-3-[(4-cya n o-r-tolylth io)m eth -
ylid en e]p yr a n (Z)-30 and (possible) 3-methylidene-4-[(4-
cyano-R-tolylthio)methyl]tetrahydrofuran 35 [73 mg, 19%
overall yield] [1H NMR δ [(Z)-30] 1.60-1.70 (2H, m); 2.30 (2H,
br t, J ) 7 Hz; collapsing to broad singlet upon irradiation at
δ 1.65), 3.70 (2H, t, J ) 7 Hz, collapsing to singlet upon
irradiation at δ 1.65), 3.82 (2H, s), 4.15 (2H, s), 5.62 (1H, br
s), 7.40-7.60 (4H, A2B2 system, J ) 9 Hz); 1H NMR spectrum
showed peaks at δ 4.20 (2H, br s, OCH2Cd), 3.95 (2H, m,
OCH2), 4.90 (1H, br s, CdCH2), and 4.95 (2H, br s, CdCH2)
ascribable to product 35; MS m/z (rel inten) 245 (M+, 10), 129
(30), 116 (100)]; (E)-3-[(4-cyan o-r-tolylth io)m eth yliden e]py-
r a n (E)-30 [135 mg, 35% (method A)]: 1H NMR δ 1.60-1.70
(2H, m); 2.33 (2H, br t, J ) 7 Hz, collapsing to broad singlet
upon irradiation at δ 1.65), 3.70 (2H, t, J ) 7 Hz, collapsing
to singlet upon irradiation at δ 1.65), 3.86 (2H, s), 3.95 (2H,
s), 5.70 (1H, s), 7.40-7.60 (4H, A2B2 system, J ) 9 Hz).
Irradiation at δ 5.70 caused en enhancement at δ 3.95
(OCH2Cd). MS m/z (rel inten) 245 (M+, 10), 129 (30), 116
(100). Anal. Calcd for C14H15NOS: C, 68.54; H, 6.16; N, 5.71;
Meth od B. A benzene solution (5 mL) of thiol 1a (2 mmol)
was added over 4 h with a syringe pump to a boiling benzene
solution (20 mL) of the appropriate alkynyl ether 3-6 (4 mmol)
and AIBN (55 mg, 0.33 mmol). Further portions of AIBN (55
mg) were added after 1.5 and 3 h. The resulting reaction
mixture was refluxed for an additional 1 h and then treated
as described in method A.
F r om 4-Cya n o-r-tolu en eth iol (1a ) a n d Bu tyn yl P h en -
ylp r op en yl Eth er 3. Chromatography gave: (E)-3-ben zyl-
4-[(4-cya n o-r-t olylt h io)m et h ylid en e]p yr a n (E)-17 [135
mg, 28% (method B)] [1H NMR δ 2.4 (3H, m), 2.80 (2H, ABX
system, J AB ) 12 Hz, J AX ) J BX ) 7 Hz; inner line separation
8 Hz), 3.4 (1H, m), 3.52 (2H, ABX system, J AB ) 11 Hz, J AX
)
J BX ) 4 Hz; inner line separation 5 Hz), 3.73 (2H, s), 3.8 (1H,
m), 5.50 (1H, s), 7.00-7.35 (7H, m), 7.55 (2H, d, J ) 8.5 Hz).
Irradiation at δ 5.50 caused an enhancement at δ 2.4 (allylic
protons) and 2.8 (benzylic protons). MS m/z (rel inten) 335
(M+, 20), 254 (25), 244 (100), 219 (20), 128 (95) 116 (95), 91
(55). Anal. Calcd for C21H21NOS: C, 75.19; H, 6.31; N, 4.18;
O, 4.77; S, 9.56. Found: C, 75.40; H, 6.34; N, 4.16; S, 9.50.],
(Z)-3-(4-cya n op h en yl)-4-(m eth ylth io)bu t-3-en -1-yl (E)-3-
p h en ylp r op -2-en -1-yl eth er (Z,E)-20 [27 mg, 6% (method
B)] [1H NMR δ 2.30 (3H, s), 2.75 (2H, t, J ) 6.9 Hz), 3.45 (2H,
t, J ) 6.9 Hz), 4.10 (2H, br d, J ) 5.5 Hz), 6.18 (1H, dt, J d
)
16 Hz, J t ) 5.5 Hz), 6.20 (1H, s), 6.54 (1H, br d, J ) 16 Hz),
7.20-7.70 (9H, m); MS m/z (rel inten) 335 (M+, 10), 188 (30),
137 (40), 117 (100). Anal. Calcd for C21H21NOS: C, 75.19; H,
6.31; N, 4.18; O, 4.77; S, 9.56. Found: C, 75.45; H, 6.28; N,
4.20; S, 9.60.] and sp ir o-cycloh exa d ien e 18 [64 mg, 11%
(method B)] [1H NMR δ 1.50 (6H, s), 2.20 (2H, br t, collapsing
to br s upon irradiation at δ 3.60), 3.15 (2H, s), 3.58 (2H, t, J
) 7 Hz), 4.15 (2H, br d, J ) 6.0 Hz), 5.90 (2H, d, J ) 10 Hz),
6.08 (1H, br s), 6.18 (2H, d, J ) 10 Hz), 6.27 (1H, dt, J d ) 15.9
Hz, J t ) 6.0 Hz; collapsing to doublet upon irradiation at δ
4.15), 6.60 (1H, br d, J ) 15.9 Hz; collapsing to doublet upon
irradiation at δ 4.15), 7.20-7.40 (5H, m); MS m/z (rel inten)
402 (M+, 2), 335 (4), 296 (10), 232 (15), 219 (20), 116 (100).
Thiol 1a was recovered in ca. 30% yield.
F r om 4-Cya n o-r-tolu en eth iol (1a ) a n d Bu tyn yl P r o-
pen yl Eth er (4). Chromatography gave: (E)-3-(4-cyan oph en -
yl)-4-(m eth ylth io)bu t-3-en -1-yl a llyl eth er (E)-23 (35 mg,
8% (method A); 5 mg, 2% (method B)] [1H NMR δ 2.40 (3H, s),
2.88 (2H, t, J ) 7 Hz), 3.52 (2H, t, J ) 7 Hz), 3.95 (2H, br d,
J ) 6 Hz), 5.12 (1H, br d, J ) 10 Hz), 5.20 (1H, br d, J ) 16
Hz), 5.85 (1H, m, collapsing to dd, J 1 ) 10 Hz, J 2 ) 16 Hz
upon irradiation at δ 3.95), 6.50 (1H, s), 7.40-7.70 (4H, A2B2
system); MS m/z (rel inten) 259 (M+, 30), 202 (35), 188 (50),
41 (100). Anal. Calcd for C15H17NOS: C, 69.46; H, 6.61; N,
5.40; O, 6.17; S, 12.36. Found: C, 69.60; H, 6.60; N, 5.37; S,

Regioselective Sulfanyl Radical Addition to Enynes
J . Org. Chem., Vol. 62, No. 16, 1997 5607
O, 6.52; S, 13.07. Found: C, 68.75; H, 6.14; N, 5.73; S, 13.02.].
Thiol 1a was recovered in 22% yield.
Bis(phenethylthio)but-1-yl allyl ether (25 mg, 9%). Thiol 1b
was recovered in ca. 30% yield.
F r om P h en eth a n eth iol (1b) a n d Bu tyn yl P h en ylp r o-
p en yl Eth er (3). Chromatography gave (Z)-4-(p h en eth -
ylth io)bu t-3-en -1-yl (E)-3-ph en ylpr op-2-en -1-yl eth er (Z,E)-
45 (30 mg; 8%) [1H NMR δ 2.45 (2H, m), 2.90 (4H, s), 3.50
(2H, t, J ) 6.5 Hz), 4.15 (2H, m), 5.60-5.76 (1H, m; collapsing
to A part of an AB system upon irradiation at δ 2.40, J AB ) 11
Hz), 6.05 (1H, br d, collapsing to B part of an AB system upon
irradiation at δ 2.45, J AB ) 11 Hz), 6.30 (1H, m, collapsing to
A part of an AB system upon irradiation at δ 4.15, J ) 15
Hz), 6.70 (1H, br d, collapsing to B part of an AB system upon
irradiation at δ 4.15, J ) 15 Hz), 7.10-7.40 (10H, m); MS m/z
(rel inten) 324 (M+, 10), 233 (30), 227 (40), 117 (50), 105 (100).
Anal. Calcd for C21H24OS: C, 77.73; H, 7.46; O, 4.93; S, 9.88.
Found: C, 77.90; H, 7.50; S, 9,84.] and (E)-3-ben zyl-4-
[(p h en eth ylth io)m eth ylid en e]p yr a n (E)-42 (160 mg, 42%)
[1H NMR δ 2.35-2.45 (3H, m), 2.8 (4H, br s), 2.9 (2H, m), 3.6
(3H, m), 3.90 (1H, dt, J d ) 10 Hz, J t ) 6.50 Hz), 5.60 (1H, s),
7.20-7.40 (10H, m); MS m/z (rel inten) 324 (M+, 10), 233 (55),
105 (100), 91 (30). Anal. Calcd for C21H24OS: C, 77.73; H,
7.46; O, 4.93; S, 9.88. Found: C, 77.54; H, 7.42; S, 9.92.]. Thiol
1b was recovered in ca. 40% yield.
F r om P h en eth a n eth iol (1b) a n d Bu tyn yl P r op en yl
Eth er (4). Chromatography gave a ca. 1:1 mixture of (E)-
and (Z)-4-(p h en eth ylth io)bu t-3-en -1-yl a llyl eth er (E)-
a n d (Z)-46 (150 mg, 44%) [1H NMR δ 2.40 (2H, q, J ) 7 Hz),
2.45 (2H, q, J ) 7 Hz), 2.90 (8H, br s), 3.42 (2H, t, J ) 7 Hz),
3.48 (2H, t, J ) 7 Hz), 3.95 (4H, br d, J ) 6 Hz), 5.16 (2H, br
d, J ) 10.5 Hz), 5.25 (2H, br d, J ) 16 Hz), 5.64 (2H, m,
collapsing to 1H, d, J ) 9.5 Hz and 1H, d, J ) 15 Hz, upon
irradiation at δ 2.42), 5.90 (2H, m), 6.00 (1H, br d; collapsing
to doublet, J ) 9.5 Hz, upon irradiation at δ 2.42) and 6.00
(1H, br d, collapsing to doublet, J ) 15 Hz, upon irradiation
at 2.42), 7.10-7.40 (10H, m); MS m/z (rel inten) 248 (M+, 5),
143 (30), 105 (100), 41 (50). Anal. Calcd for C15H20OS: C,
72.54; H, 8.12; O, 6.44; S, 12.91. Found: C, 72.75; H, 8.15; S,
12.85.]; a ca. 80:20 mixture of (E)- (Z)-3-m eth yl-4-[(p h en -
eth ylth io)m eth ylid en e]p yr a n (E)- a n d (Z)-43 (75 mg; 21%)
[MS m/z (rel inten) 248 (M+, 90), 219 (20), 105 (100). Anal.
Calcd for C15H20OS: C, 72.53; H, 8.12; O, 6.44; S, 12.91.
Found: C, 72.45; H, 8.08; S, 12.95.]. Repeated column
chromatography allowed for the separation of a fraction
composed mainly of the (E)-isomer and a fraction containing
a ca. 1:1 isomeric mixture. 1H NMR (E-isomer) δ 1.0 (3H, d,
J ) 6.5 Hz, collapsing to singlet upon irradiation at δ 2.35),
2.23 (1H, ddd, J 1 ) 13.5 Hz, J 2 ) 9.5 Hz, J 3 ) 4.8 Hz,
collapsing to d1d2 upon irradiation at δ 3.5 and to d1d3 upon
irradiation at δ 3.8), 2.35 (1H, m); 2.60 (1H, ddd, J 1 ) 13.5
Hz, J 2 ) J 3 ) 4 Hz, collapsing to dd upon irradiation at δ 3.5
or 3.8), 2.90 (4H, s), 3.15 (1H, dd, J 1 ) 10.5 Hz, J 2 ) 8.3 Hz,
collapsing to d, J ) 10.5 Hz upon irradiation at δ 2.35), 3.50
(1H, ddd, J 1 ) 12.5 Hz, J 2 ) 9.5 Hz, J 3 ) 4 Hz), 3.8 (2H, m)
5.64 (1H, s), 7.15-7.35 (5H, m). Irradiation at δ 5.64 caused
an enhancement at δ 2.90 (PhCH2CH2), 2.35 (dCCH ) and 1.0
(exo methyl group). 1H NMR (Z-isomer) δ 1.20 (3H, d, J ) 6.5
Hz), 2.0 (1H, br d), 2.6 (1H, m), 2.8 (1H, m), 2.88 (4H, s), 3.3
(1H, m), 3.5 (1H, m), 3.75 (1H, m), 4.0 (1H, m), 5.65 (1H, s),
7.15-7.35 (5H, m)]; a product which was probably 4,4-
F r om P h en eth a n eth iol (1b) a n d P r op yn yl P h en ylp r o-
p en yl Eth er (5). Chromatography gave (E)-3-ben zyl-4-
[(p h en et h ylt h io)m et h ylid en e]t et r a h yd r ofu r a n (E)-(25)
(200 mg, 53%) [1H NMR δ 2.65 (1H, A part of an ABX system,
J AB ) 13 Hz, J AX ) 8.5 Hz), 2.85 (1H, B part of an ABX system,
J AB ) 13 Hz, J BX ) 6 Hz), 2.88 (4H, br s), 3.00 (1H, m), 3.65
(1H, A part of an ABX system, J AB ) 8.5 Hz, J AX ) 5 Hz), 3.90
(1H, B part of an ABX system, J AB ) 8.5 Hz, J BX ) 6 Hz), 4.35
(2H, ABX system, J AB ) 13 Hz, J AX ) J BX ) 2 Hz, inner line
separation 2 Hz), 5.60 (1H, q, J ) 2 Hz), 7.1-7.4 (10H, m);
MS m/z (rel inten) 310 (M+, 15), 219 (40), 185 (20), 105 (100),
91 (40). Anal. Calcd for C20H22OS: C, 77.38; H, 7.14; O, 5.15;
S, 10.33. Found: C, 77.60; H, 7.17; S, 10.30.]. Thiol 1b was
recovered in ca. 30% yield.
F r om P h en eth a n eth iol (1b) a n d P r op yn yl P r op en yl
Eth er (6). Chromatography gave a 50:50 mixture of (Z)-3-
[(ph en eth ylth io)m eth yliden e]pyr an (Z)-50 and 3-r-[(ph en -
eth ylth io)m eth yl]-4-m eth ylid en etetr a h yd r ofu r a n 52 (145
mg, 44% overall yield) [1H NMR δ[(Z)-50)] 1.70 (2H, m); 2.34
(2H, dt, J t ) 6.5 Hz, J d ) 1.2 Hz), 2.90 (4H, br s), 3.72 (2H, t,
J ) 6 Hz, collapsing to singlet upon irradiation at δ 1.70), 4.22
(2H, s), 5.75 (1H, s), 7.20-7.40 (5H, m); δ(52) 2.70-2.90 (7H,
m), 3.74 (1H, dd, J 1 ) 8.8 Hz, J 2 ) 5.6 Hz), 4.03 (1H, dd, J 1
)
8.8 Hz, J 2 ) 7.0 Hz), 4.32 (2H, br s), 4.95 (2H, br s), 7.20-
7.40 (5H, m); MS m/z (rel inten) 234 (M+, 30), 152 (35), 129
(75), 105 (60), 97 (90), 83 (100)]; (E)-3-[(p h en et h ylt h io)-
m eth ylid en e]p yr a n (E)-50 (150 mg, 45%) [1H NMRδ 1.70
(2H, m); 2.40 (2H, dt, J t ) 6 Hz, J d ) 1.2 Hz), 2.90 (4H, br s),
3.75 (2H, t, J ) 6 Hz), 4.00 (2H, s), 5.80 (1H, s), 7.15 -7.40
(5H, m). Irradiation at δ 4.0 caused an enhancement at δ 5.8
(vinylic proton). MS m/z (rel inten) 234 (M+, 25), 129 (70),
105 (40), 97 (100). Anal. Calcd for C14H18OS: C, 71.75; H,
7.74; O, 6.83; S, 13.68. Found: C, 71.55; H, 7.71; S, 13.73.].
Thiol 1b was recovered in ca. 30% yield.
F r om 4-Cya n o-r-tolu en eth iol (1a ) a n d P r op yn yl 4-Cy-
a n oben zyl Eth er (36). Column chromatography gave a 59:
41 mixture of 4-cya n oben zyl 3-(4-cya n op h en yl)-4-(m eth -
ylt h io)p r op -2-en -1-yl et h er 40 and 4-cya n ob en zyl
3-m et h oxy-2-(4-cya n op h en yl)p r op -1-en -1-yl su lfid e 41
(310 mg, 75%) [1H NMR δ(40) 2.35 (3H, s), 4.32 (2H, s), 4.58
(2H, s), 6.50 (1H, s), 7.30-7.70 (8H, m). Irradiation at δ 6.50
caused a 2% enhancement at δ 2.35 (CH3S) and a 5%
enhancement at δ 4.32 (dCHCH2O). 1H NMR δ(41) 3.30 (3H,
s), 4.05 (2H, s), 4.38 (2H, s), 6.60 (1H, s), 7.40-7.70 (8H, m).
Irradiation at δ 6.60 caused a 2% enhancement at δ 4.05
(CH2S). MS m/z (rel inten) 320 (M+, 30), 204 (45), 116 (100).
Anal. Calcd for C19H16N2OS: C, 71.22; H, 5.03; N, 8.74; O,
4.99; S, 10.01. Found: C, 71.50; H, 5.05; N, 8.70; S, 10.05.].
Thiol 1b was recovered in ca. 35% yield.
Ack n ow led gm en t. This work has been carried out
under the “Progetto di Finanziamento Triennale Ateneo
di Bologna”. We also acknowledge financial support
from the Ministero dell’Universita` e della Ricerca Sci-
entifica e Tecnologica (MURST) and CNR (Rome).
J O9701142