Allylstannylation of carbon-carbon and carbon-oxygen unsaturated bonds via a radical chain process

Article abstract of DOI:10.1021/jo0057295

In the presence of a radical initiator, allyltributylstannanes bearing an electron-withdrawing group at the Β-position smoothly reacted with electron-deficient terminal alkenes to give allylstannylated products in good yields. The stannyl group was introduced into the terminal carbon with high regioselectivity. The allylstannylation of homochiral 8-phenylmenthyl acrylate proceeded with moderate to good diastereoselectivity. Terminal and electron-deficient internal alkynes as well efficiently underwent the radical-initiated allylstannylation in an anti addition mode. The reaction of terminal alkynes showed the same regioselectivity as that of terminal alkenes. The present radical reaction was applicable to allylation of aromatic aldehydes and ketones.

Full text of DOI:10.1021/jo0057295

3348
J . Org. Chem. 2001, 66, 3348-3355
Allylsta n n yla tion of Ca r bon -Ca r bon a n d Ca r bon -Oxygen
Un sa tu r a ted Bon d s via a Ra d ica l Ch a in P r ocess1
Katsukiyo Miura, Hiroshi Saito, Daisuke Itoh, Toshie Matsuda, Naoki Fujisawa,
Di Wang, and Akira Hosomi*
Department of Chemistry, Graduate School of Pure and Applied Sciences, University of Tsukuba
Tsukuba, Ibaraki 305-8571, J apan
hosomi@chem.tsukuba.ac.jp
Received November 10, 2000
In the presence of a radical initiator, allyltributylstannanes bearing an electron-withdrawing group
at the â-position smoothly reacted with electron-deficient terminal alkenes to give allylstannylated
products in good yields. The stannyl group was introduced into the terminal carbon with high
regioselectivity. The allylstannylation of homochiral 8-phenylmenthyl acrylate proceeded with
moderate to good diastereoselectivity. Terminal and electron-deficient internal alkynes as well
efficiently underwent the radical-initiated allylstannylation in an anti addition mode. The reaction
of terminal alkynes showed the same regioselectivity as that of terminal alkenes. The present radical
reaction was applicable to allylation of aromatic aldehydes and ketones.
In tr od u ction
allylmetalation, which proceeds via a radical chain
mechanism.6-8
The carbometalation of alkenes and alkynes is one of
the most useful reactions for the stereocontrolled con-
struction of organic molecules because it usually proceeds
with high regio- and stereoselectivity, and the resultant
organometallics react with various electrophiles with
retention of the stereochemical integrity.2 A variety of
organometallics are known to add to the carbon-carbon
unsaturated bond spontaneously or by the aid of a
catalyst. In particular, carbometalation with allylmetals,
that is allylmetalation, has been extensively studied in
view of their unique reactivities, the ease of preparation,
and the synthetically useful products.2-5 The known
allylmetalation reactions can be classified into the fol-
lowing three types: uncatalyzed concerted reactions
using reactive allylmetals, transition metal-catalyzed re-
actions, and Lewis acid-catalyzed reactions through ionic
intermediates. In contrast, we reported a novel type of
Allylstannanes are valuable reagents for allylation of
various carbon radical precursors such as alkyl halides,
dithiocarbonates, sulfides, and selenides.9-11 Acyclic ste-
reocontrol of the homolytic allylation has been the focus
of intensive investigation in recent years.12 In the radical
chain process, an allylstannane plays two important roles
as a radical transfer agent. One is to generate a stannyl
radical, which abstracts an atom or group from a sub-
strate to provide an alkyl radical, and the other is to
allylate the alkyl radical with regeneration of the stannyl
radical. It is also well-established that a stannyl radical
easily adds to an alkene to form a â-stannylalkyl radical.11a
From this knowledge, it occurred to us that a radical-
initiated reaction of an alkene with an allylstannane
could realize homolytic allylstannylation by the stannyl
radical addition to the alkene and the subsequent ally-
lation with the allylstannane.13 Under this premise, we
initially found that electron-deficient alkenes actually
underwent the allylstannylation, and further investi-
(1) Free Radical Chemistry. 36. For part 35, see: Miura, K.; Saito,
H.; Fujisawa, N.; Hosomi, A. J . Org. Chem. 2000, 65, 8119.
(2) (a) Knochel, P. Comprehensive Organometallic Chemistry II;
Abel, E. W., Stone, F. G. A., Wilkinson, G., Eds.; Pergamon Press:
Oxford, 1995; Vol. 11, p 159. (b) Knochel, P. Comprehensive Organic
Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford,
1991; Vol. 4, p 865. (c) Oppolzer, W. Angew. Chem., Int. Ed. Engl. 1989,
28, 38. (d) Negishi, E. Pure Appl. Chem. 1981, 53, 2333. (e) Normant,
J . F.; Alexakis, A. Synthesis 1981, 841.
(6) For preliminary papers on this work, see: (a) Miura, K.; Itoh,
D.; Hondo, T.; Saito, H.; Ito, H.; Hosomi, A. Tetrahedron Lett. 1996,
37, 8539. (b) Miura, K.; Matsuda, T.; Hondo, T.; Ito, H.; Hosomi, A.
Synlett 1996, 555.
(7) Miura, K.; Saito, H.; Nakagawa, T.; Hondo, T.; Tateiwa, J .;
Sonoda, M.; Hosomi, A. J . Org. Chem. 1998, 63, 5740.
(8) The carbometalations of alkylcobalts(III) and acylgermanes are
known to proceed via a radical chain mechanism. (a) Samsel, E. G.;
Kochi, J . K. J . Am. Chem. Soc. 1986, 108, 4790 (C-Co). (b) Curran,
D. P. Synthesis, 1988, 489 and references therein (C-Co). (c) Curran,
D. P.; Liu, H. J . Org. Chem. 1991, 56, 3463 (C-Ge).
(9) (a) Kosugi, M.; Kurino, K.; Takayama, K.; Migita, T. J . Orga-
nomet. Chem. 1973, 56, C11. (b) Grignon, J .; Pereyre, M. J . Organomet.
Chem. 1973, 61, C33.
(10) Keck, G. E.; Yates, J . B. J . Am. Chem. Soc. 1982, 104, 5829.
(11) Reviews: (a) Davies, A. G. Organotin Chemistry; VCH: Wein-
heim, 1997. (b) Curran, D. P. In Comprehensive Organic Synthesis;
Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford, 1991; Vol. 4,
p 715. (c) Pereyre, M.; Quintard, J .-P.; Rahm, A. Tin in Organic
Synthesis; Butterworth: London, 1987. (d) Giese, B. Radicals in
Organic Synthesis: Formation of Carbon-Carbon Bonds; Pergamon
Press: Oxford, 1986.
(3) Yamamoto, Y.; Asao, N. Chem. Rev. 1993, 93, 2207.
(4) For recent studies on allylmetalation of alkynes, see: (a)
Yoshikawa, E.; Gevorgyan, V.; Asao, N.; Yamamoto, Y. J . Am. Chem.
Soc. 1997, 119, 6781 (allyl-Si). (b) Yeon, S. H.; Han, J . S.; Hong, E.;
Do, Y.; J ung, I. N. J . Organomet. Chem. 1995, 499, 159 (allyl-Si). (c)
Asao, N.; Matsukawa, Y.; Yamamoto, Y. Chem. Commun. 1996, 1513
(allyl-Sn). (d) Shirakawa, E.; Yamasaki, K.; Yoshida, H.; Hiyama, T.
J . Am. Chem. Soc. 1999, 121, 10221 (allyl-Sn). (e) Araki, S.; Imai, A.;
Shimizu, K.; Yamada, M.; Mori, A.; Butsugan, Y. J . Org. Chem. 1995,
60, 1841 (allyl-In). (f) Fujiwara, N.; Yamamoto, Y. J . Org. Chem. 1997,
62, 2318 (allyl-In). (g) Okada, K.; Oshima, K.; Utimoto, K. J . Am. Chem.
Soc. 1996, 118, 6076 (allyl-Mg).
(5) For recent studies on allylmetalation of alkenes and allenes,
see: (a) Yeon, S. H.; Lee, B. W.; Yoo, B. R.; Suk, M.-Y.; J ung, I. N.
Organometallics 1995, 14, 2361 (allyl-Si). (b) Nakamura, M.; Arai, M.;
Nakamura, E. J . Am. Chem. Soc. 1995, 117, 1179 (allyl-Zn). (c) Araki,
S.; Usui, H.; Kato, M.; Butsugan, Y. J . Am. Chem. Soc. 1996, 118, 4699
(allyl-In).
(12) Renaud, P.; Gerster, M. Angew. Chem., Int. Ed. Engl. 1998,
37, 2562.
10.1021/jo0057295 CCC: $20.00 © 2001 American Chemical Society
Published on Web 04/18/2001

Allylstannylation via a Radical Chain Process
J . Org. Chem., Vol. 66, No. 10, 2001 3349
Ta ble 1. Allylsta n n yla tion of Meth yl Acr yla tea
Ta ble 2. Allylsta n n yla tion of Mon o- a n d
1,1-Disu bstitu ted Alk en es w ith 2ea
alkene
entry
R1
R2
time/h
product
isolated yield/%b
1
2
3
4
5
6
CN
Ph
C8H17
CO2Me
CN
H
H
H
Me
Me
Me
4
2
24
2
4
2
6a
6b
6c
6d
6e
6f
83
40
no reaction
allylstannane
isolated yield/%
63
62
7
entry
R
time/h
3
4
Ph
1
2
3
4
5
6
H
2a
4
2
4
2
1
2
13
21
48
40b
47b
33b
<3c
Me
SiMe3
Ph
CO2Me
CN
2b
2c
2d
2e
2f
a
b
See footnote a in Table 1. In all entries, the formation of 5e
was observed.
69
88d,e
81d
Ta ble 3. Allyla tion of In ter n a l Alk en es w ith 2ea
a
All reactions were performed with a substrate (0.50 or 1.00
mmol), 2 (4.0 equiv), and AIBN (5 mol %) in benzene (5 mL per 1
b
mmol of the substrate) at 80 °C. Diesters 4a -c were obtained
as a ca. 3:2 mixture of diastereomers. c Including unidentified
impurities. One diastereomer could be characterized by 1H NMR
d
analysis. Allylstannane 2e or 2f was recovered in 72% or 73%
alkene
isolated
R3
geometry
timeb/h
product
yieldc/%
yield based on the initial amount, respectively. In addition, the
formation of a trace amount of 5e or 5f, dimer of the allylstannane,
was observed. e The reaction with 2.0 equiv of 2e gave 3e in 77%
yield.
CO2Me
CO2Me
CN
Z
E
Z
2
2
4
7a
7a
7b
67
40
76
a
After the reaction of an alkene (1.00 mmol) with 2e was
performed under the conditions described in Table 1, the resultant
mixture was evaporated and treated with concentrated HCl (0.5
gated its applicability to other unsaturated bonds. Herein
we wish to disclose the scope and limitations of the
homolytic allylstannylation of alkenes, alkynes, and
carbonyl compounds.6
b
mL) in CH3CN (3.5 mL) for 20 min. The reaction time of the
allylstannylation. c See footnote b in Table 2.
Resu lts a n d Discu ssion
sponding 1,1-disubstituted ones. Electron-deficient al-
kenes readily underwent the allylstannylation to give
adducts 6 in good yields (Table 2, entries 1, 4, and 5),
while styrene and R-methylstyrene exhibited much lower
reactivity (Table 2, entries 2 and 6). The use of 1-decene
resulted in no adduct even after prolonged reaction time
(Table 2, entry 3).
In marked contrast with methyl acrylate, methyl cro-
tonate did not react with 2e under the same conditions.
However, highly electron-deficient internal alkenes such
as dimethyl maleate, dimethyl fumarate, and fumaroni-
trile were reactive to 2e although the allylstannylated
products could not be isolated because of their partial
destannylation in purification by silica gel column chro-
matography. After the crude products were treated with
hydrochloric acid for complete destannylation, allylated
products 7a ,b were obtained in moderate to good yields
(Table 3).
A plausible mechanism for the formation of allylstan-
nylated products 3 or 6 and 2:1 adducts 4 is shown in
Scheme 1. In the initiation step, the radical generated
from AIBN, that is Me2(NC)C•, reacts with an allylstan-
nane 2 to generate tributylstannyl radical. Reversible
addition of the stannyl radical to an alkene gives a
â-stannylalkyl radical 8,14 which leads to 3 or 6 by ho-
molytic allylation with 2. Further addition of 8 to the
alkene and the subsequent allylation gives 4. The origin
of the regioselective allylstannylation is that terminal
addition of the stannyl radical is favored over internal
Allylsta n n yla tion of Alk en es. First, methyl acrylate
(1), a good radical acceptor, was selected as a substrate
to evaluate the reactivities of some allylstannanes (Table
1). Treatment of 1 with 4 equiv of allyltributylstannane
(2a ) in the presence of AIBN at 80 °C gave R-allyl-â-
stannylester 3a in only 13% yield (Table 1, entry 1). A
diastereomeric mixture of diester 4a consisting of two
molecules of 1 and one molecule of 2a was obtained as a
major product. The reactivity of methallylstannane 2b
to 1 was similar to that of 2a (Table 1, entry 2), while
the radical-initiated reaction with 2c formed 3c in
preference to 4c (Table 1, entry 3). Introduction of a
phenyl group as R effectively suppressed the formation
of 2:1 adduct 4d to improve the efficiency of allylstan-
nylation (Table 1, entry 4). Allylstannanes 2e,f bearing
an electron-withdrawing group as R smoothly reacted
with 1 to afford allylstannylated products 3e,f in high
yields without 2:1 adducts 4e,f (Table 1, entries 5 and
6). In these reactions, dimerization of 2e,f to 5e,f was
observed; however, 2e,f were recovered in high yields.
The decreased amount (2.0 equiv) of 2e lowered the yield
of 3e to 77%.
The allylstannylation of several mono- and 1,1-disub-
stituted alkenes with 2e was next attempted to examine
its applicability to alkenes. As shown in Table 2, mono-
substituted alkenes were more reactive than the corre-
(13) The allylsulfonation of alkenes with allyl sulfone via a similar
radical chain process has been reported. (a) Harvey, I. W.; Phillips, E.
D.; Whitham, G. H. J . Chem. Soc., Chem. Commun. 1990, 481. (b)
Chuang, C.-P. Synlett 1990, 527.
(14) Kuivila, H. G.; Sommer, R. J . Am. Chem. Soc. 1967, 89,
5616.

3350 J . Org. Chem., Vol. 66, No. 10, 2001
Miura et al.
Ta ble 4. Dia ster eoselective Allylsta n n yla tion w ith 2fa
Sch em e 1
isolated yield/%
entry substrate time/h product
major
minor
dr
addition in view of the low steric hindrance and the high
stability of the resultant radical 8.
1
2
3
4
5
6
11a
11b
11c
11d
11e
11f
24
1
1
1
4
12a
12b
12c
12d
12e
12f
0
49
56
38
79b
67
0
22
22
23
69:31
72:28
62:38
63:37b
82:18
For successful allylstannylation, the allylation of 8
must be faster than the reversion (â-elimination) and the
further addition to the alkene. The high efficiency of the
allylstannylation with 2e,f in Table 1 is probably because
the electron-withdrawing â-substituent R (CO2Me or CN)
accelerates the allylation of 8a (R1 ) CO2Me, R2 ) H).
The selective formation of 4a ,b with 2a ,b indicates that
the reaction of 8a with methyl acrylate (1) is faster than
that with 2a ,b. The high reactivity of 8a to 1 and 2e,f
supports the assumption that 8a has a nucleophilic
character although the substituent on the radical center
is an electron-withdrawing group. However, the compari-
son between the results with 2a ,b implies that 8a has
slightly higher reactivity to more electron-donating al-
lylstannane 2b than to 2a .15 Accordingly, in a strict sense,
8a seemed to be an ambiphilic radical rather than a nu-
cleophilic radical.11b,16 The reason for the diminished reac-
tivity of 1,1-disubstituted alkenes is unclear (Table 2),
but it may be due to deceleration of the allylation of 8 by
the increased steric bulkiness around the radical center.
The insensitivity of 1-decene to 2e is attributable to the
instability of the corresponding â-stannylalkyl radical
intermediate 8 (R1 ) C8H17, R2 ) H), which would induce
the reversion to prevent the subsequent allylation.14,17
It is known that â-stannyl esters can be converted into
titanium homoenolates by treatment with TiCl4 and the
homoenolates smoothly add to carbonyl compounds.18
Therefore, we attempted the TiCl4-mediated reactions of
3e,f with benzaldehyde to show the synthetic utility of
the allylstannylated products. As expected, the reactions
afforded adducts 10 in moderate yields (eq 1). These
results disclosed that the present products served as
homoenolate equivalents.19
4
15c
a
All reactions were performed with 11 (0.40 mmol), 2f (1.60
b
mmol), and AIBN (0.02 mmol) in benzene (2.5 mL) at 80 °C. The
product was obtained as a diastereomeric mixture. The diastere-
omeric ratio was determined by 1H NMR analysis. c The product
was contaminated by 11f. The yield was determined by 1H NMR
analysis.
carbon bond formation in radical processes as well as
ionic processes.12,20 In particular, allylation of R-carbonyl
carbon radicals bearing a chiral auxiliary with allylstan-
nanes promises high levels of diastereoselectivity.21 This
fact induced us to investigate asymmetric homolytic
allylstannylation using acrylic acid derivatives with a
chiral auxiliary.
Initially, we prepared six homochiral acrylic acid
derivatives 11 according to the reported procedure.22
Then, they were subjected to the AIBN-initiated reaction
with 2f in benzene at 80 °C (Table 4). N-Acryloylcamphor
sultam 11a was insensitive to 2f, contrary to our expec-
tation (Table 4, entry 1),21a while N-acryloyloxazolidino-
nes 11b-d and menthyl acrylates 11e,f smoothly reacted
with 2f to give the corresponding allylstannylated prod-
ucts 12 in good yields (Table 4, entries 2-6). 8-Phenyl-
menthyl acrylate 11f showed the best diastereoselectivity
of the five.23
(17) In hydrostannylation with trialkylstannanes, electron-deficient
alkenes show higher reactivity than simple alkenes such as 1-octene.
Kupchik, E. J . Organotin Compounds; Sawyer, A. K., Ed.; Dekker:
New York, 1971; Vol. 1, p 7.
(18) Goswami, R. J . Org. Chem. 1985, 50, 5907.
(19) Kuwajima, I.; Nakamura, E. In Comprehensive Organic Syn-
thesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford, 1991;
Vol. 2, p 441.
(20) (a) Sibi, M. P.; Porter, N. A. Acc. Chem. Res. 1999, 32, 163. (b)
Curran, D. P.; Porter, N. A.; Giese, B. Stereochemistry of Radical
Reactions; VCH: Weinheim, 1995. (c) Smadja, W. Synlett 1994, 1.
(21) (a) Curran, D. P.; Shen, W.; Zhang, J .; Heffner, T. A. J . Am.
Chem. Soc. 1990, 112, 6738. (b) Porter, N. A.; Rosenstein, I. J .; Breyer,
R. A.; Bruhnke, J . D.; Wu, W.-X.; McPhail, A. T. J . Am. Chem. Soc.
1992, 114, 7664. (c) Yamamoto, Y.; Onuki, S.; Yumoto, M.; Asao, N. J .
Am. Chem. Soc. 1994, 116, 421. (d) Sibi, M. P.; J i, J . Angew. Chem.,
Int. Ed. Engl. 1996, 35, 190.
(22) For 11a : (a) Oppolzer, W.; Chapuis, C.; Bernardinelli, G. Helv.
Chim. Acta 1984, 67, 1397. For 11b,c: (b) Gage, J . R.; Evans, D. A
Organic Syntheses; Wiley: New York, 1993; Collect. Vol. VIII, p. 528.
(c) Evans, D. A.; Chapman, K. T.; Bisaha, J . J . Am. Chem. Soc. 1988,
110, 1238. For 11d : (d) Farmer, R. F.; Hamer, J . J . Org. Chem. 1966,
31, 2418. For 11e: (e) Sibi, M. P.; J i, J . J . Org. Chem. 1996, 61, 6090.
For 11f: (f) Ort, O. Organic Syntheses; Wiley: New York, 1993; Collect.
Vol. VIII, p 522. (g) Corey, E. J .; Ensley, H. E. J . Am. Chem. Soc. 1975,
97, 6908.
Asym m etr ic Hom olytic Allylsta n n yla tion . Asym-
metric induction by chiral auxiliary control is now known
to be a reliable method for the enantioselective carbon-
(15) Hagen, G.; Mayr, H. J . Am. Chem. Soc. 1991, 113, 4954.
(16) (a) Beranek, I.; Fischer, H. Free Radicals in Synthesis and
Biology; Minisci, F., Ed.; Kluwer: Dordrecht, 1989; p 303. (b) Giese,
B.; He, J .; Mehl, W. Chem. Ber. 1988, 121, 2063.

Allylstannylation via a Radical Chain Process
J . Org. Chem., Vol. 66, No. 10, 2001 3351
Ta ble 5. Allylsta n n yla tion of Acr yla te 11fa
Sch em e 2
allylstannane
entry
R
solvent T/°C time/h product yieldb/% drb
1
2
3
4
5
6
7
8
9
Me
Ph
CO2Me
CN
CN
CN
CN
CN
Ph
2b PhH
2d PhH
2e PhH
2f AcOEt
2f THF
2f hexane
2f hexane
2f hexane -78
2d hexane
2e hexane
80
80
80
77
67
69
25
4
4
4
4
4
12g
12h
12i
12f
12f
12f
12f
12f
12h
12i
39 (50) 63:37
81 (0) 84:16
68 (20) 70:30
82 (12) 83:17
47 (49) 83:17
69 (23) 86:14
4
24
10
24
24
72 (6)
90:10
38 (51) 91:9
39 (43) 87:13
68 (16) 87:13
25
25
10 CO2Me
Sch em e 3
a
All reactions were performed with 11f (0.40 mmol), 2 (1.60
mmol), and AIBN (0.02 mmol, entries 1-6) or Et3B (0.40 mmol,
b
entries 7-10) in solvent (2.5 mL). The yield and diastereomeric
ratio were determined by 1H NMR analysis of a mixture of 11f
and 12 obtained after purification by column chromatography. The
recovery (%) of 11f is shown in parentheses.
The AIBN-initiated allylstannylation of 11f with 2b,e
in benzene resulted in lower diastereoselectivity, while
the reaction with 2d proceeded with good selectivity
(entries 1-3 in Table 5). The use of 2d in the allylstan-
nylation of 11b was not effective in improving the
selectivity (eq 2). To achieve higher diastereoselectivity,
the reaction of 11f with 2f was carried out under several
reaction conditions (Table 5, entries 4-8). As a result,
the Et3B-initiated reaction in hexane at 25 °C improved
the selectivity to 90% (Table 5, entry 7).24 The allylstan-
nylation with 2d ,e showed good selectivity under these
conditions (Table 5, entries 9 and 10).
rotation was opposite to that of (R)-14.27 Accordingly, the
major isomers of 12h and other adducts derived from 11f
were determined to possess S-configuration at the carbon
R to the carbonyl group.
The allylstannylation of 11f with 2 would proceed
through radical intermediate 17 having a planar struc-
ture by conjugation with the carbonyl group (Scheme
3).20b J udging from the previous studies on the reactions
of R-carbonyl carbon radicals bearing an 8-phenylmenthyl
group,23 17 presumably takes conformation A, in which
the stannylmethyl group is anti to the alkoxy group and
the phenyl group shields one face of the prostereogenic
radical center.28 Attack of 2 to A on the opposite side to
the phenyl group well agrees with the present stereo-
chemical outcome.
The LAH-reduction of the major isomer of 12j was
found to give the same enantiomer of 13 ((S)-13) as
derived from the major isomer of 12h by measurement
of optical rotation (eq 3). Thus, the major isomers of 12b,j
Lewis acids are now routinely used for stereoselectivity
control of radical reactions as well as ionic and concerted
reactions.12 However, our efforts to attain high levels of
diastereoselectivity in the Lewis acid-controlled homolytic
allylstannylation of 11b,d ,f resulted in failure.
To determine the absolute configuration of 12f-i, the
major isomer of adduct 12h (X* ) 8-phenylmenthyl, R
) Ph) was converted into alcohol 14 by reduction with
LAH followed by destannylation with BuLi (Scheme 2).25
In addition, the authentic sample of (R)-14 was prepared
from (S)-propanoyloxazolidinone 15 by the stereodefined
method reported by Evans et al.26 The sign of 14 in optical
(23) For diastereoselective reactions of R-carbonyl radicals generated
from 8-phenylmenthyl esters, see: (a) Crich, D.; Davies, J . W.
Tetrahedron Lett. 1987, 28, 4205. (b) Hamon, D. P. G.; Razzino, P.;
Massy-Westropp, R. A. J . Chem. Soc., Chem. Commun. 1991, 332. (c)
Hamon, D. P. G.; Massy-Westropp, R. A.; Razzino, P. Tetrahedron 1993,
49, 6419. (d) Chen, M.-Y.; Fang, J .-M.; Tsai, Y.-M.; Yeh R.-L. J . Chem.
Soc., Chem. Commun. 1991, 1603.
(24) Nozaki, K.; Oshima, K.; Utimoto, K. J . Am. Chem. Soc. 1987,
109, 2547.
(25) (a) Meyer, N.; Seebach, D. Chem. Ber. 1980, 113, 1290. (b)
Brieden, W.; Ostwald, R.; Knochel, P. Angew. Chem., Int. Ed. Engl.
1993, 32, 582.
were determined to have (4S,2′S)-configuration. This
(27) Alcohol 14 and the authentic sample of (R)-14 were proved to
be enatiomerically pure by their conversion into the diastereomeric
MTPA esters with (+)-MTPA-Cl. See the Supporting Information
and: Dale, J . A.; Mosher, H. S. J . Am. Chem. Soc. 1973, 95, 512.
(28) The intramolecular π-stacking interaction between the phenyl
group and the R-carbonyl radical is likely to effect the conformation.
8-Phenylmenthyl crotonate is known to have such an interaction
between the phenyl group and the carbon-carbon double bond.
Mezrhab, B.; Dumas, F.; d’Angelo, J .; Riche, C. J . Org. Chem. 1994,
59, 500.
(26) Evans, D. A.; Ennis, M. D.; Mathre, D. J . J . Am. Chem. Soc.
1982, 104, 1737.

3352 J . Org. Chem., Vol. 66, No. 10, 2001
Miura et al.
Ta ble 6. Allylsta n n yla tion of Eth yl P r op yn oa tea
Sch em e 4
allylstannane
R
yield/%
20b,c 21 + 22c,d
entry
time/h
21:22e
1
2
3
4
5
6
H
Me
SiMe3
Ph
CO2Me
CN
2a f
2b
4
2
2
1
1
1
12g
27g
38
61
70
<1h
<2h
<2
66:34
75:25
50:50
34:66
73:27
100:0i
2cf
2d
<5
2ef
2ff
<15
66
11i
diastereoselectivity in the allylstannylation of 11b is
opposite to that in the Lewis acid-controlled allylation
of R-carbonyl carbon radicals bearing a chiral oxazolidi-
none auxiliary.29 Under the present conditions, the
radical intermediate 18 generated from 11b would favor
conformation C over B because B has a dipole-dipole
repulsion between the two carbonyl groups (Scheme 4).
The observed selectivity may be due to the reaction of
conformer C with 2 although the origin of the diastereo-
face-selectivity is not clear.
a
b
See footnote a in Table 1. Isolated yield of a pure product
except for entries 1 and 2. c The configurations of 20-22 were
assigned by NOE experiments, chemical shifts of the olefinic
protons, and coupling constants between the olefinic proton and
119Sn or 117Sn. See the Supporting Information. The mixture of
d
vinylstannanes 21 and 22 was contaminated by unidentified
impurities except for entry 6. e Determined by 1H NMR analysis.
f The allylstannane was recovered in 62-74% yield based on the
initial amount. Including (Z)-23. The yield was estimated by H
NMR analysis. Including (E)-23. Only 21f was obtained.
g
1
h
i
Ta ble 7. Allylsta n n yla tion of Alk yn es 24a
Allylsta n n yla tion of Alk yn es. We next directed our
efforts to homolytic allylstannylation of alkynes to de-
velop a new stereoselective route to functionalized vinyl-
stannanes. Ethyl propynoate (19) was first elected as a
substrate for the reaction with allylstannanes 2. The
AIBN-initiated reaction of 19 with the parent allylstan-
nane 2a gave a mixture of R-allyl-â-stannyl-substituted
acrylate 20a (12%) and (Z)-â-stannylacrylate 23 (1.4%)
after purification by silica gel column chromatography
(Table 6). A mixture of 21a , 22a (stereo- and regioisomers
of 20a ), and (E)-23 was also obtained although each yield
was fairly low (<0.5%). On the other hand, â-substituted
allylstannanes exhibited higher reactivity than 2a as
shown in the allylstannylation of methyl acrylate. In
particular, 2e,f efficiently reacted with 19 to give trans
adducts 20e,f selectively.
alkyne
R2
isolated yield/%b
entry
R1
time/h
25
27
1
2
3
4
5
6
7
8
9
H
H
H
H
H
Bu
Ph
Ph
n-C10H21
(CH2)2OH
CH(OH)Me 24d
CH(OAc)Me 24e
24a
24b
24c
1
6
2
2
2
2
1
1
6
94
44c
69
34d,e
62
5d
To investigate the scope and limitations of the allyl-
stannylation, a variety of alkynes were subjected to the
reaction with 2e (Table 7). In these cases, the formation
of cis-adducts 26 was not observed except for entry 8.
Phenylacetylene (24a ) showed high reactivity to 2e,
leading to vinylstannane 25a in a high yield without
regioisomer 27a (Table 7, entry 1). Although the reaction
of 1-dodecyne (24b) also gave only 25b, isolation of 25b
from the reaction mixture including 2e and its dimer 5e
was a laborious process (Table 7, entry 2). The yield of
CO2Me
CO2Me
24f
24g
24h
24i
24j
15f
2
38
84
MeO2C CO2Me
Bu Ph
n-C5H11 n-C5H11
85g
19
10
24
no reaction
a
b
See footnote a in Table 1. The configuration of product was
assigned mainly by NOE experiments and/or the coupling con-
stant between 1H and 119Sn or 117Sn. See the Supporting In-
formation. c Treatment of the crude product with HCl-CH3CN
d
gave 28 in 70% yield. A mixture of 25d and 27d was obtained.
The yields were determined by 1H NMR analysis. e Lactone 29
was also obtained (<4%). f Allylvinylstannane 30 was also ob-
1
25b was estimated to be 63% by the H NMR analysis of
the crude product, but it was isolated in only 44% yield
(>98% pure) by distillation. Protonolysis of the crude
product with HCl-CH3CN provided 1,4-diene 28 in 70%
isolated yield. The reactivity of 24b is in sharp contrast
to that of 1-decene, which was insensitive to 2e under
the same reaction conditions. 3-Butyn-1-ol (24c) as well
as 24a ,b smoothly underwent the allylstannylation to
g
tained (7%). The formation of cis adduct 26h was observed
(<1%).
form only 25c (Table 7, entry 3). Thus, the present
reaction is tolerable to a hydroxy group unlike most of
the known allylmetalation reactions. The use of 3-butyn-
2-ol (24d ) as a substrate resulted in three adducts, both
regioisomers 25d , 27d , and δ-lactone 29 (Table 7, entry
4). Protection of the hydroxy group of 24d was fairly
effective to suppress the formation of 27e as well as
lactonization (Table 7, entry 5).
(29) Sibi et al. have reported that the addition of multipoint binding
Lewis acids (e.g., Yb(OTf)3, MgBr2) changes the diastereoselectivity
in the allylation of such carbon radicals. See ref 21d and: Sibi, M. P.;
Rheault, T. R. J . Am. Chem. Soc. 2000, 122, 8873.

Allylstannylation via a Radical Chain Process
J . Org. Chem., Vol. 66, No. 10, 2001 3353
Sch em e 5
Internal alkynes conjugated with an ester group were
also available for the allylstannylation. The reaction of
methyl 2-heptynoate (24f) gave trans-adducts 25f and
27f with a 1:2 regioselectivity along with allylvinylstan-
nane 30 (Table 7, entry 6). When methyl 3-phenyl-2-
propynoate (24g) was employed, the regioselectivity
increased to more than 40:1 (Table 7, entry 7). Dimethyl
acetylenedicarboxylate (24h ), a highly electron-deficient
alkyne, also underwent the trans-allylstannylation in
high efficiency (Table 7, entry 8). In contrast, internal
alkynes with no electron-withdrawing group were much
less reactive to 2e than electron-deficient alkynes (Table
7, entries 9 and 10).
Sch em e 6
A plausible mechanism for the homolytic allylstanny-
lation of alkynes is illustrated in Scheme 5. First,
tributylstannyl radical generated from an allylstannane
2 by the action of AIBN adds to an alkyne reversibly.
Then, the resulting vinyl radical 31 or 32 reacts with 2
to afford the corresponding allylstannylated product and
regenerate the stannyl radical. The formation of 23 in
the reaction of 19 with 2a ,b is probably due to low
radical-allylating ability of 2a ,b, which allows hydrogen
abstraction of the radical intermediate 31 (R1 ) H, R2 )
CO2Et). As mentioned above, the high reactivity of 2e,f
toward the homolytic allylstannylation would be im-
parted by the â-substituent, thus accelerating the ally-
lation step.
of products would also affect the stereoselectivity.14,34
Indeed, (E)-vinylstannanes 20 were partly isomerized to
the Z-isomers 21 in the presence of Bu3SnH and Et3B.35
The difference between 1-decene and 1-dodecyne in
reactivity to 2e is attributable to the lifetimes of the
â-stannyl carbon radical intermediates arising from these
substrates. â-Stannylalkyl radicals are believed to revert
to stannyl radicals faster than â-stannylvinyl radicals.36
The high reactivity of 1-dodecyne is probably because the
corresponding â-stannylvinyl radical intermediate has
enough lifetime to react with 2e unlike the â-stannylalkyl
one from 1-decene.
In the case of terminal alkynes (R1 ) H), the regiose-
lective introduction of the stannyl group can be rational-
ized by avoidance of steric repulsion from the substituent
R2 and the formation of more stabilized radical interme-
diate 31. The formation of 22 and 27d from terminal
alkynes 19 and 24d indicates that the ester and hydroxy
groups facilitate addition of the stannyl radical to the
internal acetylenic carbon close to them. This directing
effect was distinctly observed in the reaction of internal
alkyne 24f. It is known that homolytic hydrostannyla-
tions of 2-alkynoates and propargyl alcohols with
Bu3SnH show a similar trend in regioselectivity.30,31 How-
ever, the origin of the directing effect remains obscure
at present. The results with phenyl-substituted alkynes
24g,i disclose that the phenyl group effectively controls
the regioselectivity. The regiocontrol would be caused by
strong stabilization of the corresponding radical inter-
mediates by the phenyl group.
The formation of 30 can be explained by the stepwise
mechanism shown in Scheme 6, which consists of (1)
addition of tributylstannyl radical to the R-carbon of ester
24f, (2) 1,5-hydrogen transfer from sp3-carbon to sp2-
carbon, (3) elimination of a dibutylvinylstannyl radical
from the rearranged radical, and (4) allylation of the
stannyl radical by SH2′ process. There are two factors that
assist the 1,5-hydrogen transfer. One is that the vinyl
radical intermediate 32 (R1 ) Bu, R2 ) CO2Me), which
is not stabilized by conjugation, has enough reactivity to
cause the 1,5-hydrogen transfer.37 The other is that the
steric bulkiness around the radical center decelerates the
intermolecular allylation with 2e.
Allyla tion of Ca r bon yl Com p ou n d s. The allylation
of carbonyl compounds with allylstannanes is signifi-
cantly valuable for regio- and stereoselective carbon-
carbon bond formation.3,11a,c Allylstannanes react with
aldehydes and ketones spontaneously or in the presence
of a Lewis acid to give homoallyl alcohols. The allylation
reactions are generally believed to proceed via a concerted
or ionic process. There is no example of the radical-
initiated allylation of carbonyl carbons except for our
previous work.38,39 In contrast, hydrostannanes are known
to reduce aldehydes and ketones to alcohols by a radical
The reason for the preferred formation of trans-adducts
in the present allylstannylation is that 2 approaches the
radical center of 31 or 32 from the opposite side to the
stannyl group to avoid its steric hindrance.32,33 In the
reaction of 19, the stannyl radical-mediated isomerization
(30) (a) Konoike, T.; Araki, Y. Tetrahedron Lett. 1992, 33, 5093. (b)
Ensley, H. E.; Buescher, R. R.; Lee, K. J . Org. Chem. 1982, 47, 404.
(31) (a) Leusink, A. J .; Budding, H. A. J . Organomet. Chem. 1968,
11, 533. (b) Leusink, A. J .; Budding, H. A.; Marsman, J . W. J .
Organomet. Chem. 1967, 9, 285.
(32) (a) Giese, B.; Gonza´lez-Go´mez, J . A.; Lachhein, S.; Metzger, J .
O. Angew. Chem., Int. Ed. Engl. 1987, 26, 479. (b) J ournet, M.;
Malacria, M. J . Org. Chem. 1992, 57, 3085.
(34) Leusink, A. J .; Budding, H. A.; Drenth, W. J . Organomet. Chem.
1968, 11, 541.
(35) Taniguchi, M.; Nozaki, K.; Miura, K.; Oshima, K.; Utimoto, K.
Bull. Chem. Soc. J pn. 1992, 65, 349.
(36) Stork, G.; Mook, R., J r. J . Am. Chem. Soc. 1987, 109, 2829.
(37) (a) Curran, D. P.; Shen, W. J . Am. Chem. Soc. 1993, 115, 6051.
(b) Bogen, S.; Malacria, M. J . Am. Chem. Soc. 1996, 118, 3992.
(38) We have recently reported the radical-initiated allylsilylation
of aldehydes and ketones. See ref 7.
(33) (a) Miura, K.; Itoh, D.; Hondo, T.; Hosomi, A. Tetrahedron Lett.
1994, 35, 9605. (b) Miura, K.; Saito, H.; Itoh, D.; Hosomi, A. Tetrahe-
dron Lett. 1999, 40, 8841.

3354 J . Org. Chem., Vol. 66, No. 10, 2001
Miura et al.
Ta ble 8. Allyla tion of Ald eh yd es a n d Keton esa
lation step to attain high efficiency. Electron-deficient
alkenes and various alkynes smoothly undergo the ho-
molytic allylstannylation. The high reactivity of these
substrates can be rationalized by the stability of radical
intermediates arising from them. Namely, â-stannylalkyl
radicals stabilized by an electron-withdrawing group or
â-stannylvinyl radicals have enough lifetimes to react
with allylstannanes before their reversion to the sub-
strates and tributylstannyl radical. The present allyl-
stannylation provides a new method for the synthesis of
functionalized â-stannylesters and vinylstannanes, which
are used as potent carbon nucleophiles for carbon-carbon
bond formation.11a,c,18 The use of acrylates bearing a chiral
auxiliary enables asymmetric allylstannylation reaction.
In addition, the radical-initiated reaction with allylstan-
nanes can be utilized for the allylation of aromatic
carbonyl compounds. In conclusion, we have developed
a novel type of carbometalation reaction with wide
applicability, which proceeds via a radical chain mech-
anism unlike the known carbometalation reactions.
carbonyl compound
entry
R1
R2
product
isolated yield/%
1
2
3
4
5
Ph
H
H
H
H
33a
33b
33c
33d
33e
59
35
53
0
4-MeOC6H4
2-furyl
Ph(CH2)2
Ph
Me
29
a
See footnote a in Table 1. The resulting reaction mixture was
treated with DBU and passed through a short silica gel column.
Sch em e 7
Exp er im en ta l Section
Gen er a l Meth od s. Unless otherwise noted, all reactions
and distillation of solvents were carried out under N2. Solvents
were dried by distillation from sodium metal/benzophenone
ketyl (THF, Et2O) and CaH2 (benzene, hexane, CH2Cl2).
Bu3SnCl was simply distilled in vacuo. All other commercial
reagents were used as received.
Syn th esis of Su bstr a tes. Homochiral acrylic acid deriva-
tives 11 were prepared by the reported procedure.22 1-Methyl-
2-propynyl acetate (24e) was obtained from 3-butyn-2-ol by
acetylation with AcCl-Et3N in Et2O.40 The reaction of methyl
chloroformate with lithium acetylide derived from 1-hexyne
was carried out for the synthesis of methyl 2-heptynoate
(24f).41 Other substrates were purchased.
Syn th esis of Allylsta n n a n es. Allylstannanes 2a -e were
prepared by reductive coupling between the corresponding allyl
bromides and Bu3SnCl using Mg and a catalytic amount of
PbBr2.42 2-Trimethylsilyl-3-bromo-1-propene,43 3-bromo-2-
phenyl-1-propene,44 and methyl 2-(bromomethyl)acrylate45
were prepared by the reported procedures. Allylstannane 2f
was synthesized from the corresponding allyl sulfonates by
homolytic substitution with Bu3SnH.46
AIBN-In it ia t ed Allylst a n n yla t ion (Gen er a l P r oce-
d u r e). Allylstannane 2 (4.0 equiv) was added to a solution of
a substrate (1.0 equiv, 0.50 or 1.00 mmol) and AIBN (5 mol
%) in benzene (5.0 mL per 1.00 mmol of the substrate). The
mixture was stirred at reflux. After completion of the reaction,
the resultant mixture was evaporated and purified by silica
gel column chromatography. In the reaction of 11, 0.40 mmol
of 11 and 2.5 mL of benzene or other solvent were used.
Allyla tion of In ter n a l Alk en es a n d 1-Dod ecyn e. The
AIBN-initiated reaction of an electron-deficient internal alkene
(0.5 mmol) with 2e was performed under the same conditions
as the above general procedure. After concentration of the
reaction mixture, it was diluted with CH3CN (3.5 mL) and
treated with concentrated HCl (0.5 mL) for 10 min. The
resultant mixture was neutralized with saturated aqueous
chain mechanism as well as a concerted or ionic
process.11c,17 The homolytic reduction involves addition
of a stannyl radical to the carbon-oxygen bond as a prop-
agation step. Therefore, our attention was next focused
on the homolytic allylation of carbonyl compounds.
Treatment of benzaldehyde with 2f in the presence of
AIBN and the subsequent destannylative workup gave
homoallyl alcohol 33a in a moderate yield (entry 1 in
Table 8). The reaction without AIBN resulted in no
adduct. This observation supports our assumption that
the present allylation proceeds via a radical process. As
expected from the above results, 2a hardly reacted with
benzaldehyde even in the presence of AIBN. The use of
2e gave lactone 34 exclusively, which would be formed
by the allylstannylation-lactonization-allylstannylation
process shown in Scheme 7. While other aromatic alde-
hydes and acetophenone also underwent the AIBN-
initiated allylation with 2f (Table 8, entries 2, 3, and 5),
3-phenylpropanal was quite insensitive to 2f (Table 8,
entry 4).
Con clu sion
We have found that, in the presence of a radical
initiator, allyltributylstannanes react with alkenes and
alkynes to give the corresponding allylstannylated prod-
ucts. This homolytic allylstannylation involves two propa-
gation steps: reversible addition of tributylstannyl radi-
cal to the unsaturated bonds and allylation of the
resultant â-stannyl carbon radicals with allylstannanes.
Introduction of an electron-withdrawing group into the
â-position of allylstannanes accelerates the latter ally-
(40) Nakamura, K.; Takenaka, K.; Ohno, A. Tetrahedron: Asymmetry
1998, 9, 4429.
(41) Zhou, Z.; Larouche, D.; Bennett, S. M. Tetrahedron 1995, 51,
11623.
(42) Tanaka, H.; Abdul Hai, A. K. M.; Ogawa, H.; Torii, S. Synlett
1993, 835.
(43) Lee, E.; Yu, S.-G.; Hur, C.-U.; Yang, S.-M. Tetrahedron Lett.
1988, 29, 6969.
(44) Hatch, L. F.; Patton, T. L. J . Am. Chem. Soc. 1954, 76, 2705.
(45) Byun, H.-S.; Reddy, K. C.; Bittman, R. Tetrahedron Lett. 1994,
35, 1371.
(46) Baldwin, J . E.; Adlington, R. M.; Lowe, C.; O’Neil, I. A.; Sanders,
G. L.; Schofield, C. J .; Sweeney, J . B. J . Chem. Soc., Chem. Commun.
1988, 1030.
(39) Enholm et al. have reported that the AIBN-initiated reactions
of cyclopropyl and R,â-unsaturated ketones with allylstannanes give
γ- and â-allylated ketones, respectively. (a) Enholm, E. J .; J ia, Z. J . J .
Org. Chem. 1997, 62, 9159. (b) Enholm, E. J .; Moran, K. M.; Whitley,
P. E. Tetrahedron Lett. 1998, 39, 971.

Allylstannylation via a Radical Chain Process
J . Org. Chem., Vol. 66, No. 10, 2001 3355
NaHCO3 (20 mL) and extracted with Et2O (2 × 15 mL). The
extract was treated with DBU (0.5 mL) for 5 min and passed
through a short silica gel column.47 The filtrate was evaporated
and purified by silica gel column chromatography. This method
was also used for allylstannylation of 1-dodecyne (24b) with
2e followed by protonolysis (footnote c in Table 7).
mmol). The mixture was stirred for 12 h and concentrated in
vacuo. Purification of the residual oil by silica gel column
chromatography (hexane-CH2Cl2 2:1) gave 21a (4.3 mg, 0.010
mmol) in 5% yield with 62% recovery of 20a . Other (Z)-
vinylstannanes 20 as well were partially isomerized to 21
under these conditions.
TiCl4-Med ia t ed R ea ct ion of â-St a n n ylest er 3e w it h
Ben za ld eh yd e.18 To a solution of 3e (475 mg, 1.00 mmol) in
CH2Cl2 (3.5 mL) at room temperature was added TiCl4 (1.0 M
in CH2Cl2, 0.80 mL, 0.80 mmol). After 30 min, benzaldehyde
(74 mg, 0.70 mmol) was added to the mixture. After 7 h, the
resultant mixture was poured into saturated aqueous NaHCO3
(20 mL) and extracted with Et2O (2 × 15 mL). A halostannane
byproduct was removed from the extract by the above DBU-
silica gel method.47 Purification of the crude product by silica
gel column chromatography gave 10a (144 mg, 0.46 mmol) in
66% yield.
Et3B-In itia ted Allylsta n n yla tion of Acr yla te 11f. Et3B
(1.0 M in hexane, 0.40 mL, 0.40 mmol) was added to a solution
of 11f (0.40 mmol) and 2 (1.6 mmol) in hexane (2.5 mL) at 25
or -78 °C. Then, dry air (5 mL) was introduced into the
mixture. After completion of the reaction, the resultant
mixture was evaporated and purified by silica gel column
chromatography.
Allyla tion of Ca r bon yl Com p ou n d s. The AIBN-initiated
reaction of a carbonyl compound (0.5 mmol) with 2f was
performed under the same conditions as the above general
procedure. After concentration of the reaction mixture, it was
diluted with Et2O (10 mL), treated with DBU (0.3 mL) for 5
min, and passed through a short silica gel column.47 The
filtrate was evaporated and purified by silica gel column
chromatography.
Ack n ow led gm en t. This work was partly supported
by Grants-in-Aid for Scientific Research and Grants-
in-Aid for Scientific Research on Priority Areas (No.
706: Dynamic Control of Stereochemistry) from the
Ministry of Education, J apan, and Pfizer Inc.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedure for the synthesis of allylstannanes 2 and alcohol 14,
characterization data for allylstannanes and products, and
discussion on configurational assignment of vinylstannane
products. This material is available free of charge via the
Internet at http://pubs.acs.org.
Sta n n yl Ra d ica l-In d u ced Isom er iza tion of 20 (Typ ica l
P r oced u r e). To a solution of 20a (86 mg, 0.20 mmol) in
benzene (0.4 mL) at room temperature were added Bu3SnH
(22 µL, 0.08 mmol) and Et3B (1.0 M in hexane, 0.08 mL, 0.08
(47) Curran, D. P.; Chang, C.-T. J . Org. Chem. 1989, 54, 3140.
J O0057295