8-Endo cyclization of (alkoxycarbonyl)methyl radicals: Radical ways for preparation of eight-membered-ring lactones

Article abstract of DOI:10.1021/ja980908d

Cyclization of (alkoxycarbonyl)methyl radicals generated from bromoacetates proceeds in the 8-endo made to generate heptanolactones. Three distinct types of 8-endo/5-exo tandem radical cyclizations produce different bicyclic heptanolactones. In certain cases, intramolecular free-radical attack on the heptanolactone carbonyl group initiates further skeletal rearrangement. Ab initio calculations indicate that the preference of the 8-endo cyclization over the 5-exo mode originates from the conformational bias of (alkoxycarbonyl)methyl radicals favoring the Z- over the E-conformation.

Full text of DOI:10.1021/ja980908d

J. Am. Chem. Soc. 1998, 120, 7469-7478
7469
8-Endo Cyclization of (Alkoxycarbonyl)methyl Radicals: Radical
Ways for Preparation of Eight-Membered-Ring Lactones
Eun Lee,* Cheol Hwan Yoon, Tae Hee Lee, Sun Young Kim, Tae Joon Ha,
Yong-suk Sung, Sang-Hyun Park, and Sangyoub Lee
Contribution from the Department of Chemistry, College of Natural Sciences, Seoul National UniVersity,
Seoul 151-742, Korea
ReceiVed March 17, 1998
Abstract: Cyclization of (alkoxycarbonyl)methyl radicals generated from bromoacetates proceeds in the 8-endo
mode to generate heptanolactones. Three distinct types of 8-endo/5-exo tandem radical cyclizations produce
different bicyclic heptanolactones. In certain cases, intramolecular free-radical attack on the heptanolactone
carbonyl group initiates further skeletal rearrangement. Ab initio calculations indicate that the preference of
the 8-endo cyclization over the 5-exo mode originates from the conformational bias of (alkoxycarbonyl)-
methyl radicals favoring the Z- over the E-conformation.
Introduction
Scheme 1
Radical cyclization reactions developed in the past decade
are now firmly established as indispensable tools in synthetic
chemistry. In forming carbo- and heterocyclic compounds, these
reactions exhibit useful regio- and stereoselectivity employing
a variety of functional groups as radical acceptors, and efficient
synthetic schemes for a plethora of complex natural products
have been formulated based on key radical cyclization reactions.1
Lactones were among the primary targets in the early stage
of development. Initial attempts to cyclize (alkoxycarbonyl)-
alkyl radical species from R-haloalkanoate esters under standard
radical-generating conditions with tributylstannane and AIBN
led only to simple reduction products, and the results were
attributed to the unfavorable conformational bias of these
carbonyl conjugated radicals. Stork2 and Ueno3 solved this
problem by developing tin hydride-mediated cyclization of
R-haloacetals as an indirect route to γ- and δ-lactones (Scheme
1). Another practical solution to this problem was devised by
Curran4 based on the halogen atom-transfer reactions (Scheme
2). More recently, however, direct syntheses of γ- and
δ-lactones via 5-exo or 6-exo radical cyclizations using R-ha-
loalkanoates as substrates were reported5 (Scheme 3).
Scheme 2
Scheme 3
In our continuing efforts to synthesize hydroazulenic ses-
quiterpenes, we had occasion to examine the tributylstannane-
(1) For recent reviews, see: (a) Giese, B.; Kopping, B.; Go¨bel, T.;
Dickhaut, J.; Thoma, G.; Kulicke, K. J.; Trach, F. Org. React. 1996, 48,
301. (b) Jasperse, C. P.; Curran, D. P.; Fevig, T. L. Chem. ReV. 1991, 91,
1237. (c) Curran, D. P. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 4, pp 715, 779.
(2) (a) Stork, G.; Mook, R., Jr.; Biller, S. A.; Rychnovsky, S. D. J. Am.
Chem. Soc. 1983, 105, 3741. (b) Stork, G.; Sher, P. M. J. Am. Chem. Soc.
1983, 105, 6765. (c) Stork, G.; Sher, P. M. J. Am. Chem. Soc. 1986, 108,
303. (d) Stork, G.; Sher, P. M.; Chen, H. L. J. Am. Chem. Soc. 1986, 108,
6384. (e) Stork, G. Bull. Chem. Soc. Jpn. 1988, 61, 149.
(3) (a) Ueno, Y.; Chino, K.; Watanabe, M.; Moriya, O.; Okawara, M. J.
Am. Chem. Soc. 1982, 104, 5564. (b) Ueno, Y.; Moriya, O.; Chino, K.;
Watanabe, M.; Okawara, M. J. Chem. Soc., Perkin Trans. 1 1986, 1351.
(4) (a) Curran, D. P.; Chang, C. T. Tetrahedron Lett. 1987, 28, 2477.
(b) Curran, D. P.; Chang, C. T. J. Org. Chem. 1989, 54, 3140. (c) Curran,
D. P.; Chang, C. T. Tetrahedron Lett. 1990, 31, 933.
(5) (a) Belletire, J. L.; Mahmoodi, N. O. Tetrahedron Lett. 1989, 30,
4363. (b) Clough, J. M.; Pattenden, G.; Wight, P. G. Tetrahedron Lett.
1989, 30, 7469. (c) Hannesian, S.; Di Fabio, R.; Marcoux, J. F.;
Prud’homme, M. J. Org. Chem. 1990, 55, 3436.
mediated radical cyclization reaction of the bromoacetate 1. In
light of the findings cited above, we hoped to obtain a tricyclic
γ-lactone via 5-exo/7-endo tandem radical cyclizations, which
would be used as a pivotal intermediate en route to guaianolide
natural products.6 The reaction proceeded smoothly under the
standard high-dilution radical-generating conditions, and a
product was obtained in 80% isolated yield. Upon careful
S0002-7863(98)00908-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/18/1998

7470 J. Am. Chem. Soc., Vol. 120, No. 30, 1998
Lee et al.
Table 1
Scheme 4
spectroscopic analysis, we were quite surprised to find that what
we had was the tricyclic heptanolactone 2, a product of 8-endo/
5-exo tandem radical cyclizations (Scheme 4). For the (alkoxy-
carbonyl)methyl radical generated from the bromoactate 1,
8-endo cyclization was preferred over the usual 5-exo cycliza-
tion. In other words, eight-membered heptanolactone ring
formation was kinetically much faster than five-membered
γ-lactone formation!
It is well-known that the eight-membered-ring lactones are
the least accessible ones via traditional lactone-forming reactions
starting from ω-halo- and ω-hydroxycarboxylic acids and
ω-alkenoic acids.7-9 Aside from a few scattered examples,10,11
syntheses of heptanolactones frequently employ indirect schemes
which may involve Baeyer-Villiger oxidation of cyclohep-
tanones or sigmatropic rearrangements among others.12 In this
context, the preferential formation of a eight-membered hep-
tanolactone via 8-endo cyclization reaction of an (alkoxycar-
bonyl)methyl radical was truly remarkable, especially when an
alternative 5-exo mode of cyclization was also possible. In this
paper, we report results of further experimental examples of
8-endo cyclization of (alkoxycarbonyl)methyl radicals.13 We
also provide a theoretical basis for the preferential 8-endo
cyclization by presenting results of ab initio calculations.
(6) For the guaianolide synthesis via 5-exo/7endo tandem radical
cyclization reactions, see: Lee, E.; Lim, J. W.; Yoon, C. H.; Sung, Y.-s.;
Kim, Y. K.; Yun, M.; Kim, S. J. Am. Chem. Soc. 1997, 119, 8391.
(7) For reviews, see: (a) Back, T. G. Tetrahedron 1977, 33, 3041. (b)
Rousseau, G. Tetrahedron 1995, 51, 2777.
(8) The difficulty in the formation of heptanolactones is discussed in
the following examples: (a) Mukaiyama, T.; Usui, M.; Saigo, K. Chem.
Lett. 1976, 49. (b) Galli, C.; Illuminati, G.; Mandolini, L. J. Am. Chem.
Soc. 1973, 95, 8374. (c) Galli, C.; Illuminati, G.; Mandolini, L.; Tamborra,
P. J. Am. Chem. Soc. 1977, 99, 2591. (d) Galli, C.; Mandolini, L. J. Chem.
Soc., Chem. Commun. 1982, 251. (e) Simonot, B.; Rousseau, G. J. Org.
Chem. 1994, 59, 5912.
(9) In many other examples, heptanolactones are omitted as targets but
it is obvious that they are the most difficult ones: (a) Corey, E. J.; Nicolaou,
K. C. J. Am. Chem. Soc. 1974, 96, 5614. (b) Kurihara, T.; Nakajima, Y.;
Mitsunobu, O. Tetrahedron Lett. 1976, 2455. (c) Vorbru¨ggen, H.; Krolik-
iewicz, K. Angew. Chem., Int. Ed. Engl. 1977, 16, 876. (d) Kruizinga, W.
H.; Kellogg, R. M. J. Chem. Soc., Chem. Commun. 1979, 286. (e) Steliou,
K.; Szczygielska-Nowosielska, A.; Favre, A.; Poupart, M. A.; Hanessian,
S. J. Am. Chem. Soc. 1980, 102, 7578. (f) Kruizinga, W. H.; Kellogg, R.
M. J. Am. Chem. Soc. 1981, 103, 5183. (g) Regen, S. L.; Kimura, Y. J.
Am. Chem. Soc. 1982, 104, 2064. (h) Steliou, K.; Poupart, M. A. J. Am.
Chem. Soc. 1983, 105, 7130. (i) Kimura, Y.; Regen, S. L. J. Org. Chem.
1983, 48, 1533.
(10) There are a few recent examples of successful formation of
heptanolactones from substituted 7-hydroxyheptanoic acids. In these cases,
the conformational constraint imposed by the substituents is considered
important: (a) Funk, R. L.; Abelman, M. M. J. Org. Chem. 1986, 51, 3247.
(b) Buszek, K. R.; Sato, N.; Jeong, Y. J. Am. Chem. Soc. 1994, 116, 5511.
(c) Andrus, M. B.; Argade, A. B. Tetrahedron Lett. 1996, 37, 5049.
(11) There are more successful examples for synthesis of unsaturated
heptanolactones: Nicolaou, K. C.; McGarry, D. G.; Somers, P. K.; Kim,
B. H.; Ogilvie, W. W.; Yiannikouros, G.; Prasad, C. V. C.; Veale, C. A.;
Hark, R. R. J. Am. Chem. Soc. 1990, 112, 6263.
(12) (a) McWilliams, J. C.; Clardy, J. J. Am. Chem. Soc. 1994, 116,
8378. (b) Robinson, R. A.; Clark, J. S.; Holmes, A. B. J. Am. Chem. Soc.
1993, 115, 10400. (c) Curtis, N. R.; Holmes, A. B.; Looney, M. G.
Tetrahedron 1991, 47, 7171.
Results
The bromoacetates 3a-20a were obtained via reaction of the
corresponding alkenols and alkadienols with bromoacetic acid
in the presence of dicyclohexylcarbodiimide (DCC) and 4-(di-
methylamino)pyridine (DMAP). Most alkenols and alkadienols
used are known, and they were either purchased or synthesized
by employing procedures in the literature. Other alkenols and
alkadienols were obtained by following procedures described
in the Experimental Section.
Slow addition (via a syringe pump) of a solution of tribu-
tylstannane (1.4 equiv) in benzene containing azobisisobutyro-
nitrile (AIBN) (0.1 equiv) over 5 h into a benzene solution
(0.015 M) of 4-pentenyl bromoacetate (3a) under reflux gave
4-pentenyl acetate (3b, 31%) and heptanolactone (3c, 38%).14
The presence of the 4-methyl substituent did not increase the
yield of the lactone 4c (38%), but substitution with the
4-trimethylsilyl group facilitated the cyclization considerably
in forming lactone 5c (54%). Further examples of the simple
8-endo cyclization are presented in Table 1. 1-Phenyl and 2,2-
dimethyl substitutents increased the yield of the lactones 6c
(52%) and 7c (53%) as expected. The lower yield for 8c (25%)
(14) (a) Matsubara, S.; Takai, K.; Nozaki, H. Bull. Chem. Soc. Jpn. 1983,
56, 2029. (b) Wiberg, K. B.; Waldron, R. F. J. Am. Chem. Soc. 1991, 113,
7697.
(13) Part of this work has appeared in communication form: Lee, E.;
Yoon, C. H.; Lee, T. H. J. Am. Chem. Soc. 1992, 114, 10981.

Eight-Membered-Ring Lactone Preparation
J. Am. Chem. Soc., Vol. 120, No. 30, 1998 7471
Table 2
Table 3
Scheme 5
may be attributed to the steric crowding in the transition-state
conformations. (Alkoxycarbonyl)ethyl radicals are presumably
much less active in this type of cyclization as shown by the
low yield of the lactone 9c (13%) from the R-bromopropionate
9a. From the results shown in Table 1, it was concluded that
8-endo cyclization of (alkoxycarbonyl)methyl radicals was
indeed generally applicable.
On the contrary, reactions of the lower and higher homologues
10a and 11a did not yield cyclization products (6-exo or 7-endo
mode for 10a and 8-exo or 9-endo mode for 11a) (Table 2).
The reaction of 10a produced a complex product mixture from
which 10b was isolated in 42% yield. The conversion of 11a
into the acetate 11b (79%) was accompanied by one major
byproduct, which appeared to be a dimeric species formed by
the intermolecular attack of the (alkoxycarbonyl)methyl radical
and subsequent reduction. Formation of larger lactone rings
was also unfavorable, as the acetates 12b (57%) and 13b (90%)
were the only products isolated after reaction of the substrates
12a and 13a. These results indicate that 8-endo mode of
cyclization is the intrinsically favored pathway for (alkoxycar-
bonyl)methyl radicals.
Next, substrates 14a-16a were reacted under the same
conditions for 5-exo/8-endo competition experiments (Table 3).
Radical cyclization of the bromoacetate 14a proceeded smoothly
to yield the heptanolactone 14b (31%) and the bicyclic
heptanolactone 14c (53%). The result of the reaction of 15a
was more interesting. Thin-layer chromatographic analysis
(silica gel, 7:1 hexane/ethyl acetate) of the crude reaction
mixture revealed four clean spots for the simple acetate 15b
(25%), the heptanolactone 15c (25%), the bicyclic butyrolactone
15e (25%), and the bridged bicyclic lactone 15d (23%). The
reaction of 16a yielded mainly three products: the acetate 16b
(44%), the butyrolactone 16d (28%), and the bicyclic lactone
16c (14%). The heptanolactones 14b and 15c were clearly
obtained via 8-endo cyclization, and the bicyclic heptanolactones
14c, 15d, and 16c are products of 8-endo/5-exo tandem radical
cyclizations. No simple 5-exo mode cyclization products were
obtained. It is now quite clear that 8-endo mode cyclization is
much faster than 5-exo cyclization for (alkoxycarbonyl)methyl
radicals. It is to be noted that the combined yield (84%) of
14b and 14c from 14a for 8-endo cyclization is substantially
higher than the yield (38%) of 3c from 3a. This could be the
effect of an extra substituent on the reactive conformation.
Comments on the structural elucidation of the bicyclic
heptanolactones are warranted. The secondary methyl group
in 14c was determined to be exo oriented whereas the secondary
methyl groups in 15d and 16c were assigned to be endo oriented.
The NOE difference spectrum of 16c showed that the intensity
of the signals (δ 1.57 and 1.69, ABq, J ) 14.4 Hz) from the
protons of the isolated methylene group of the carbocyclic ring
did not change upon irradiation of the secondary methyl group
signals (δ 0.94, d, J ) 7.2 Hz) but the signals (δ 1.67, dd, J )
4.5, 13.5 Hz) from one of the methylene protons â to the
carbonyl group were enhanced.
In the formation of the bicyclic heptanolactone 14c, the
conformation 14d may be important, in which the pendant vinyl
group is directed away from the heptanolactone center (Scheme
5). Apparently, the transition-state conformation for the 5-exo
cyclization of the heptanolactone radical formed by 8-endo
cyclization of the initial (alkoxycarbonyl)methyl radical from
15a (and 16a) is much influenced by the presence of the gem
dimethyl groups so that the conformation 15f may become more
important (Scheme 6). In the ensuing methyl radical 15g, the
radical center is directed toward the center of the lactone ring,
enticing further rearrangement. It may be assumed that the
reaction proceeds from the methyl radical 15g via transannular
attack to the lactone carbonyl group and fragmentation of the
oxy radical 15h to the ethyl radical 15i. The mechanism
delineated in Scheme 6 calls for the cis relationship of the
lactone ring and the ethyl group in 16d. This was convincingly
shown in the NOE difference spectrum: irradiation of the
lactone methylene proton signals (δ 2.49, d, J ) 7.6 Hz) resulted
in the enhancement of the signals (δ 1.29, q, J ) 7.4 Hz) from
the methylene protons of the ethyl group. We believe this is

7472 J. Am. Chem. Soc., Vol. 120, No. 30, 1998
Lee et al.
Scheme 6
Scheme 7
Table 4
bromoacetates 19a and 20a, almost identical yields of 19c and
20b were obtained, and it may be argued that the initial 8-endo
radical cyclization proceeded under irreversible conditions. The
efficiency of 8-endo radical cyclizations in general may thus
have to be explained by assuming that the reactions proceeded
irreversibly.
In the reaction of 19a and 20a, products 19d and 20c were
isolated, and their structures were confirmed by independent
syntheses. One obvious (but unlikely) explanation was the vinyl
hydrogen abstraction by the initial (alkoxycarbonyl)methyl
radicals followed by the loss of acetylene. When 20a was
reacted in the presence of deuteriotributylstannane instead of
tributylstannane, deuterium-labeled 20b (45%) and 20c (36%)
were isolated. Spectroscopic analysis of 20c easily located the
deuterium atom at the acetate methyl carbon. Obviously, the
rationalization given above is not operational and an alternative
explanation is needed.
Discussion
There are several known examples of 8-endo radical cycliza-
tion. Some of them involve relatively rigid templates, and it is
difficult to correlate these results with the present one.17 For
flexible carbocycle synthesis, it is predicted that 8-endo mode
of cyclization of a 7-octenyl radical is preferred over the
alternative 7-exo mode of cyclization,18 and examples of
cyclooctane synthesis were indeed reported.19 But the prefer-
ential formation of heptanolactones cannot be explained on the
same grounds as the formation of cyclooctanes. Contemporary
to our initial reports,13 Speckamp and co-workers reported
medium-sized lactone synthesis via copper(I)-catalyzed atom-
transfer cyclizations of dichloroacetates and trichloroacetates,20
which is probably more closely related to the present study.
More recently, 8-endo cyclization of unsaturated acrylates upon
reaction with t-BuHgI/KI has been reported by Russell and Li.21
The preferred 8-endo mode of cyclization of (alkoxycarbo-
nyl)methyl radicals reflects the conformational bias of these
the first clear-cut example of the radical rearrangements
involving ester or lactone carbonyl functionalities.15,16
Further examples of 8-endo radical cyclization were collected
(Table 4). The reaction of the bromoacetate 17a yielded the
acetate 17b (12%) and the heptanolactone 17c (58%). Obvi-
ously, transannular 6-exo cyclization was not feasible after initial
8-endo cyclization. However, the substrate 18a was transformed
into a mixture (27:32) of the bicyclic lactones 18c (59%) and
the heptanolactone 18b (18%). The formation of 18c may be
explained by a second type of 8-endo/5-exo tandem radical
cyclization. A third type of 8-endo/5-exo tandem radical
cyclization is also possible: the reaction of the bromoacetate
19a afforded the acetate 19b (24%), 6-heptenyl acetate (19d,
13%), and the bicyclic heptanolactone 19c (38%). The straight-
chain acetate 20c (39%) and the bicyclic heptanolactone 20b
(39%) were also isolated from the reaction of the bromoacetate
20a.
(17) (a) Sato, T.; Ishida, S.; Ishibashi, H.; Ikeda, M. J. Chem. Soc., Perkin
Trans. 1 1991, 353. (b) Crich, D.; Chen, C.; Hwang, J. T.; Yuan, H.;
Papadatos, A.; Walter, R. I. J. Am. Chem. Soc. 1994, 116, 8937.
(18) Beckwith, A. L. J.; Schiesser, C. H. Tetrahedron 1985, 41, 3925.
(19) (a) Merritt, J. E.; Sasson, M.; Kates, S. A.; Snider, B. B. Tetrahedron
Lett. 1988, 29, 5209. (b) Snider, B. B.; Merritt, J. E. Tetrahedron 1991,
47, 8663. (c) Molander, G. A.; McKie, J. A. J. Org. Chem. 1994, 59, 3186.
(20) (a) Pirrung, F. O. H.; Steeman, W. J. M.; Hiemstra, H.; Speckamp,
W. N.; Kaptein, B.; Boesten, W. H. J.; Schoemaker, H. E.; Kamphuis, J.
Tetraheron Lett. 1992, 33, 5141. (b) Pirrung, F. O. H.; Hiemstra, H.; Kaptein,
B.; Sobrino, M. E. M.; Petra, D. G. I.; Schoemaker, H. E.; Speckamp, W.
N. Synlett 1993, 739. (c) Pirrung, F. O. H.; Hiemstra, H.; Speckamp, W.
N.; Kaptein, B.; Schoemaker, H. E. Tetrahedron 1994, 50, 12415. (d)
Pirrung, F. O. H.; Hiemstra, H.; Speckamp, W. N.; Kaptein, B.; Shoemaker,
H. E. Synthesis 1995, 458.
The stereoselectivity in the formation of 20b (and 19c) may
be explained as shown in Scheme 7. From two different
(15) For a review for radical-mediated rearrangements, see: Dowd, P.;
Zhang, W. Chem. ReV. 1993, 93, 2091.
(16) For some examples of intramolecular addition of alkyl radicals to
aldehyde and ketone carbonyl groups, see: (a) Tsang, R.; Fraser-Reid, B.
J. Am. Chem. Soc. 1986, 108, 2116. (b) Tsang, R.; Fraser-Reid, B. J. Am.
Chem. Soc. 1986, 108, 8102. (c) Tsang, R.; Dickson, J. K. Jr.; Park, H.;
Walton, R.; Fraser-Reid, B. J. Am. Chem. Soc. 1987, 109, 3484. (d) Dowd,
P.; Choi, S. C. J. Am. Chem. Soc. 1987, 109, 3493. (e) Dowd, P.; Choi, S.
C. J. Am. Chem. Soc. 1987, 109, 6548.
(21) Russell, G. A.; Li, C. Tetrahedron Lett. 1996, 37, 2557.

Eight-Membered-Ring Lactone Preparation
J. Am. Chem. Soc., Vol. 120, No. 30, 1998 7473
Scheme 8
Figure 1. Schematic representation of possible reaction modes of
101: (a) 8-endo cyclization of the Z-conformation; (b) 8-endo cycliza-
tion of the E-conformation; (c) 7-exo cyclization of the Z-conformation;
(d) 7-exo cyclization of the E-conformation.
radicals favoring Z-conformation (s-trans) over E-conformation
(s-cis). In line with many theoretical and experimental studies
corroborating the relative stability of Z-ester conformations, the
Z-conformation of the (alkoxycarbonyl)alkyl radicals was also
judged to be more stable than the E-conformation.22 Atom-
transfer cyclization of allyl iodoacetates is much more efficient
at 80 °C than at 25 °C. This beneficial effect of temperature
arises because, at higher temperature, there is relatively larger
population of the less stable E-conformer (which can cyclize
in the 5-exo mode) of (allyloxycarbonyl)methyl radicals than
at lower temperature.23 The selective formation of eight-
membered-ring heptanolactones in these cyclization reactions
is probably also connected with the finding that heptanolactone
is the smallest lactone for which Z-conformers are found as low-
energy conformers.14b
For a better understanding of the selectivity shown in the
above experimental results, ab initio calculations were performed
on a number of cyclizations of (alkoxycarbonyl)methyl radicals.
All calculations were carried out with the GAMESS24 series of
programs using the ROHF25 method. All structures reported
were fully optimized with the 3-21G basis sets26 and were
characterized by harmonic frequency analysis. Energies are
obtained with second-order Møller-Plesset perturbation theory
(MP2).27 The intrinsic reaction coordinate calculations (IRC)28
were performed to confirm the connectivity between the
respective reactant and product via a proposed transition
structure.
Table 5. Energies and Geometry Data for Cyclization of
(Alkoxycarbonyl)methyl Radicals
(ROHF/MP2/3-21G//ROHF/3-21G)
structurea
energy (hartree)
distance (Å)
101 Z
E
102 Z
E
103 Z
E
104 Z
E
105 Z
E
106 Z
E
107 Z
E
108 Z
E
109 (E)
110 (E)
111 (E)
112 (E)
113 (E)
114 (E)
115 (E)
-419.653 628
-419.639 786
-419.647 271
-419.632 564
-419.679 941
-419.669 688
-419.632 049
-419.627 706
-419.667 150
-419.670 360
-380.743 785
-380.727 826
-380.725 982
-380.719 304
-380.753 889
-380.760 319
-380.719 333
-380.760 378
-341.817 171
-341.798 573
-341.851 156
-341.807 211
-341.853 688
d(C1-C8) ) 4.19
d(C1-C8) ) 4.94
d(C1-C8) ) 2.12
d(C1-C8) ) 2.12
d(C1-C8) ) 1.57
d(C1-C8) ) 1.56
d(C1-C7) ) 2.14
d(C1-C7) ) 2.13
d(C1-C7) ) 1.55
d(C1-C7) ) 1.55
d(C1-C7) ) 4.59
d(C1-C7) ) 3.95
d(C1-C7) ) 2.12
d(C1-C7) ) 2.13
d(C1-C7) ) 1.55
d(C1-C7) ) 1.56
d(C1-C6) ) 2.12
d(C1-C6) ) 1.55
d(C1-C6) ) 3.29
d(C1-C6) ) 2.13
d(C1-C6) ) 1.54
d(C1-C5) ) 2.09
d(C1-C5) ) 1.54
a Z or E means Z- and E-conformation.
Results from ab initio calculations on the cyclizations of
homologues of (alkoxycarbonyl)methyl radicals (Scheme 8) are
given in Figure 1 and Table 5. The Z-conformations of
(alkoxycarbonyl)methyl radicals are calculated to be more stable
than the E-conformations along the reaction pathway up to the
transition state.
For (4-pentenyloxycarbonyl)methyl radical 101, the Z-con-
formation (101Z) is more stable than the E-conformation (101E)
Figure 2. Calculated structure of the transition state for 8-endo
cyclization (102Z).
(22) (a) Fisher, H.; Wu, L. M. HelV. Chim. Acta 1983, 66, 138. (b)
Beckwith, A. L. J.; Glover, S. A. Aust. J. Chem. 1987, 40, 157.
(23) Curran, D. P.; Tamine, J. J. Org. Chem. 1991, 56, 2746.
(24) GAMESS version 11/96: Schmidt, M. W.; Baldridge, K. K.; Boatz,
J. A.; Elbert, S. T.; Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga,
N.; Nguyen, K. A.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J.
Comput. Chem. 1993, 14, 1347.
(25) (a) McWeeny, R.; Diercksen, G. J. Chem. Phys. 1968, 49, 4852.
(b) Faegri, K.; Manne, R. Mol. Phys. 1976, 31, 1037. (c) Hsu, H.; Davidson,
E. R.; Pitzer, R. M. J. Chem. Phys. 1976, 65, 609.
(26) Bikley, J. S.; Pople, J. A.; Hehre, W. J. J. Am. Chem. Soc. 1980,
102, 939.
(27) (a) Knowles, P. J.; Andrews, J. S.; Amos, R. D.; Handy, N. C.;
Pople, J. A. Chem. Phys. Lett. 1991, 186, 130. (b) Lauderdale, W. J.; Stanton,
J. F.; Gauss, J.; Watts, J. D.; Bartlett, R. J. Chem. Phys. Lett. 1991, 187,
21.
by 8.7 kcal/mol. The activation energies for the 8-endo and
7-exo cyclizations of 101E are 4.5 and 7.6 kcal/mol, respectively
(Figure 1b and d). The activation energy (4.0 kcal/mol) of the
8-endo cyclization of 101Z is much smaller than that (13.5 kcal/
mol) of the 7-exo cyclization (Figure 1a and c). Therefore,
ROHF/MP2/3-21G//ROHF/3-21G ab initio calculations predict
that 8-endo cyclization of the Z-conformation is the most
preferred mode of reaction for 101. Figure 2 shows the
calculated structure of the transition-state 102Z for 8-endo
cyclization. In the transition-state 102Z, the C1-C8 bond
length is 2.12 Å, and the radical approach angle C1-C8-C7
is 106.6°. The pyramidalization at C8 is 158.3°. These values
(28) (a) Schmidt, M. W.; Gordon, M. S.; Dupuis, M. J. Am. Chem. Soc.
1985, 107, 2585. (b) Gonzales, C.; Schlegel, H. B. J. Phys. Chem. 1990,
94, 5523.

7474 J. Am. Chem. Soc., Vol. 120, No. 30, 1998
Lee et al.
Scheme 9
Table 6. Energies and Geometry Data for the 8-Endo/5-Exo
Tandem Cyclization of (Alkoxycarbonyl)methyl Radicals
(ROHF/MP2/3-21G//ROHF/3-21G)
Figure 4. Calculated structures of the transition states 202, 204, and
206.
structurea energy (hartree)
distance (Å)
201 Z
E
-496.280 597 d(C1-C8) ) 3.72
of 201 over the 5-exo mode originates from the conformational
bias of this radical favoring the Z- over the E-conformation.
The 7-exo mode of cyclization is also possible for 201Z, but
the activation energy for this mode of reaction should be much
higher than that of the 8-endo cyclization (cf. Figure 1a and
c). Figure 4 shows the calculated transition-state structures 202,
204, and 206.
In conclusion, 8-endo cyclization is the fundamentally
preferred mode of reaction for (alkoxycarbonyl)methyl radicals
and eight-membered heptanolactones are obtained in reasonably
good yields from bromoacetates. Further mechanistic details
and synthetic utility of these unique reactions30 will be reported
in due course.
-496.262 789 d(C1-C8) ) 3.95
202
203
204
205
206
207
-496.272 977 d(C1-C8) ) 2.12
-496.309 176 d(C1-C8) ) 1.56
-496.281 915 d(C7-C9) ) 2.10, d(C9-C10) ) 1.35
-496.324 056 d(C7-C9) ) 1.55, d(C9-C10) ) 1.51
-496.256 771 d(C1-C9) ) 2.09, d(C9-C10) ) 1.35
-496.304 592 d(C1-C9) ) 1.55, d(C9-C10) ) 1.51
a Z or E means Z- and E-conformation.
Experimental Section
NMR spectra were obtained on Varian EM-360A (60 MHz),
Bruker AC-80 (80 MHz), Bruker AW-80 (80 MHz), Varian
VXR-200 (200 MHz), Varian Gemini-300BB (300 MHz), and
Bruker AMX 500 (500 MHz) instruments. Chemical shifts are
reported as δ values relative to internal tetramethylsilane. Mass
spectra were recorded on a VG-Trio 2 spectrometer using
electron impact (EI) or chemical ionization (CI) method, and
high-resolution mass spectra (HRMS) were recorded on a JEOL
JMS GSX-300 spectrometer. Infrared spectra were taken on a
Perkin-Elmer model 782 spectrometer or a Bruker IFS48 FT-
IR spectrophotometer as neat oil. Optical rotations were
measured on a Jasco DIP-360 digital polarimeter. GC chro-
matograms were recorded on a Hewlett-Packard model HP
5880A gas chromatograph using nitrogen as carrier gas.
TLC was performed on Merck precoated silica gel plates (no.
5554), and the TLC spots were visualized under 254-nm UV
light and/or by charring after dropping the plate into vanillin
solution in 5% sulfuric acid/methanol. Purification of products
was accomplished via Merck silica gel (no. 7734 and no. 9385)
flash chromatography. Hexane and ethyl acetate were simple
distilled and used in column chromatography.
Figure 3. Schematic representation of 8-endo/5-exo tandem radical
cyclizations of 201.
are comparable to those calculated for intermolecular addition
of various radicals to different alkenes.29
Results from ab initio calculations on the 8-endo/5-exo
tandem radical cyclizations of the radical 201 (Scheme 9) are
summarized in Figure 3 and Table 6. The Z-conformation
(201Z) of the radical 201 is more stable than the E-conformation
(201E) by 11.3 kcal/mol. The activation barrier (3.8 kcal/mol)
for the 5-exo cyclization of 201E is comparable to that (4.8
kcal/mol) of the 8-endo cyclization of 201Z. These calculations
therefore indicate that the preference of the 8-endo cyclization
Unless otherwise specified, all reactions were conducted under
a slight positive pressure of dry nitrogen and the usual workup
(29) (a) Zipse, H.; He, J.; Houk, K. N.; Giese, B. J. Am. Chem. Soc.
1991, 113, 4324. (b) Wong, M. W.; Pross, A.; Radom, L. J. Am. Chem.
Soc. 1994, 116, 6284. (c) Wong, M. W.; Pross, A.; Radom, L. J. Am. Chem.
Soc. 1994, 116, 11938.
(30) For an example in the synthesis of (-)-clavukerin A, see: Lee, E.;
Yoon, C. H. Tetrahedron Lett. 1996, 37, 5929.

Eight-Membered-Ring Lactone Preparation
J. Am. Chem. Soc., Vol. 120, No. 30, 1998 7475
refers to washing of the quenched reaction mixture with
saturated sodium chloride solution, drying over anhydrous
MgSO4, and evaporating under reduced pressure using a rotary
evaporator.
Scheme 10
All solvents used in reactions were dried under nitrogen or
argon atmosphere. THF was distilled from Na-benzophenone.
Dichloromethane and benzene were washed with concentrated
H2SO4 and distilled from P2O5 and stored over 4-Å molecular
sieves. Ethyl ether was distilled from lithium aluminum hydride
(LAH). Pyridine was dried over KOH and stored over 4-Å
molecular sieves.
Bromoacetate 1. To a stirred solution of DCC (280 mg,
1.31 mmol) in 10 mL of dry dichloromethane was added
bromoacetic acid (210 mg, 1.48 mmol) at 0 °C, and the reaction
mixture was stirred for 30 min at that temperature. To this
mixture was added the solution of the allylic alcohol (210 mg,
0.87 mmol) in 5 mL of dichloromethane followed by a catalytic
amount (40 mg) of DMAP, and the reaction mixture was stirred
for 30 min at room temperature. After filtration on a silica gel
pad, the filtrate was concentrated under reduced pressure. Flash
chromatography (silica gel, 1.5 × 15 cm, 12:1 hexane/ethyl
acetate) afforded the bromoacetate 1 (300 mg, 95%): 1H NMR
(200 MHz, CDCl3) δ 1.07 (d, 3 H, J ) 7.1 Hz), 1.79 (br s, 3
H), 2.34-2.44 (m, 1 H), 2.91-2.97 (m, 1 H), 3.35 (s, 3 H),
3.98 (s, 2 H), 4.08 (t, 1 H, J ) 4.6 Hz), 4.58 and 4.65 (ABq, 2
H, J ) 6.8 Hz), 4.82 (br, 1 H), 5.07-5.29 (m, 4 H), 5.77 (ddd,
1 H, J ) 5.4, 11.0, 17.0 Hz).
Heptanolactone 2. To a stirred solution of the bromoacetate
1 (310 mg, 0.86 mmol) in dry benzene (34 mL, 0.025 M) under
reflux was added a mixture of tributylstannane (0.28 mL, 1.2
equiv) and a catalytic amount of AIBN in 5 mL of benzene via
a syringe pump for 4 h. After complete addition, the reaction
mixture was heated under reflux for 1 h. After evaporation of
solvent, flash chromatography (silica gel, 1.5 × 15 cm, 10:1
hexane/ethyl acetate) afforded the heptanolactone 2 (194 mg,
octadien-4-ol,39 and 2-(2′-propenyl)-4-penten-1-ol40 were syn-
thesized according to the literature procedures. 4,4,6-Trimethyl-
1,6-heptadien-3-ol (21), 3-ethenyl-6-hepten-1-ol (22), and ethyl
6-(2′-hydroxyethyl)-2,7-octadienoate (23) were prepared as
described below (Scheme 10).
4,4,6-Trimethyl-1,6-heptadien-3-ol (21). Isobutyraldehyde
(24; 10 g, 138 mmol) and methallyl alcohol (25; 15 g, 208
mmol) were dissolved in o-xylene (50 mL) containing p-TsOH
(0.1 g). The solution was heated 48 h under reflux with
continuous removal of water on a Dean-Stark trap. The
product aldehyde 26 (7.8 g, 45%) was obtained via fractional
distillation through a short packed column. A sample of the
aldehyde 26 (5.0 g, 41 mmol) in dry THF (50 mL) was added
dropwise at 0 °C to the THF solution (50 mL) of vinylmagne-
sium bromide obtained from magnesium (1.25 g, 51 mmol) and
vinyl bromide (6.6 g, 62 mmol). The reaction mixture was
stirred 30 min at room temperature and poured into ice/2 N
HCl. The resulting mixture was extracted three times with ether.
The combined ether extracts were washed with water, saturated
NaHCO3 solution, and brine and dried over anhydrous MgSO4.
Concentration and flash chromatography afforded 4,4,6-tri-
30
1
80%): [R]D -86.3° (c 2.35, CCl4); H NMR (200 MHz,
CDCl3) δ 0.929 (d, 3 H, J ) 7.1 Hz), 0.936 (s, 3 H), 1.162 (d,
3 H, J ) 6.8 Hz), 1.20-2.00 (m, 6 H), 2.28-2.64 (m, 4 H),
3.38 (s, 3 H), 3.99 (br t, 1 H, J ) 3.2 Hz), 4.14 (d, 1 H, J ) 5.6
Hz), 4.58 and 4.69 (ABq, 2 H, J ) 6.7 Hz); 13C NMR (50.3
MHz, CDCl3) δ 174.8, 95.02, 88.89, 81.32, 55.41, 55.31, 50.78,
46.69, 44.66, 43.07, 34.08, 32.96, 30.57, 22.70, 13.83, 7.68;
IR (neat, cm-1) 2932, 1777, 1725, 1446, 1403, 1281, 1245,
1194, 1149, 1091, 1038; MS (EI) m/z (relative intensity) 282
(M+, 7), 264 (6), 250 (25), 237 (59), 219 (64), 209 (21), 193
(32), 178 (29), 165 (100), 147 (66), 133 (33), 121 (47), 111
(42), 107 (37), 81 (78), 69 (30), 55 (31); HRMS m/z calcd for
C16H26O4 282.1832, found 282.1839.
1
Preparation of Alkenols and Alkadienols. 4-Penten-1-ol,
3-methyl-3-buten-1-ol, and 5-hexen-1-ol were purchased from
Aldrich. 4-Methyl-4-penten-1-ol,31 4-(trimethylsilyl)-4-penten-
1-ol,32 1-phenyl-4-penten-1-ol,33 2,2-dimethyl-4-penten-1-ol,20c
2,2,4-trimethyl-4-penten-1-ol,34 6-hepten-1-ol,35 12-tridecen-1-
ol,36 1,6-heptadien-3-ol,37 4,4-dimethyl-1,6-heptadien-3-ol,38 1,7-
methyl-1,6-heptadien-3-ol (21, 4.8 g, 78%): H NMR (80 MHz,
CDCl3) δ 0.92 (s, 6 H), 1.59 (s, 1 H), 1.80-1.82 (m, 3 H),
1.99 and 2.16 (ABq, 2 H, J ) 12.9 Hz), 3.84 (br d, 1 H, J )
7.5 Hz), 4.70-4.74 (m, 1 H), 4.84-4.93 (m, 1 H), 5.10-5.36
(m, 2 H), 5.77-6.19 (m, 1 H).
3-Ethenyl-6-hepten-1-ol (22). DMSO (3.56 mL, 41.8 mmol)
was added to a dichloromethane solution (60 mL) of oxalyl
chloride (1.84 mL, 18.9 mmol) under a nitrogen atmosphere at
-78 °C. After 5 min, 4-penten-1-ol (27; 1.5 g, 17.4 mmol)
was added dropwise to this solution, followed by triethylamine
(31 mL, 120 mmol) after another 15 min at the same temper-
ature. This reaction mixture was stirred for 30 min at -78 °C
and allowed to warm to room temperature. To this solution,
(31) Kauffmann, T.; Nienaber, H.; Stach, D. Inorg. Chim. Acta 1994,
220, 85.
(32) Overman, L. E.; Thompson, A. S. J. Am. Chem. Soc. 1988, 110,
2248.
(33) Hartung, J.; Schwarz, M. Synlett 1997, 1116.
(34) Shishido, K.; Umimoto, K.; Shibuya, M. Heterocycles 1994, 38,
641.
(35) Diedrich, M. K.; Klaerner, F.-G.; Beno, B. R.; Houk, K. N.;
Senderowitz, H.; Still, W. C. J. Am. Chem. Soc. 1997, 119, 10255.
(36) Shono, T.; Matsumura, Y.; Kashimura, S. Chem. Lett. 1978, 69.
(37) Chen, L.; Wang, D.; Sun, C.; Zhu, Y. Trans. Tianjin UniV. 1996,
2, 81.
(39) Wilson, S. E. Tetrahedron Lett. 1975, 4651.
(40) Yamamoto, Y.; Iwasa, M.; Sawada, S.; Oda, J. Agric. Biol. Chem.
1990, 54, 3269.
(38) Salomon, R. G.; Ghosh, S. Org. Synth. 1984, 62, 125.

7476 J. Am. Chem. Soc., Vol. 120, No. 30, 1998
Lee et al.
(carbethoxymethylene)triphenylphosphorane (6.7 g, 41.8 mmol)
in dry dichloromethane (10 mL) was added. After being stirred
for 2 h at room temperature, the reaction was quenched by
adding water and the aqueous layer was extracted with dichlo-
romethane three times. The combined organic layer was washed
with brine, dried over anhydrous MgSO4, and concentrated to
afford 2.0 g of the ester 28. The product was dissolved in
anhydrous ether (10 mL), and this solution was added dropwise
to a suspension of LAH (661 mg, 17.3 mmol) in anhydrous
ether at 0 °C. After stirring 30 min at room temperature, water
(0.8 mL), 15% NaOH solution (0.8 mL), and water (2.1 mL)
were added successively. This mixture was filtered on a silica
gel pad, and the filtrate was concentrated. Flash chromatog-
raphy on silica gel gave 2,6-hepatadien-1-ol (29; 750 mg, 60%
orthoacetate (34.1 g, 210 mmol) and a catalytic amount (0.2 g)
of propionic acid at 160-170 °C for 12 h with continuous
removal of ethanol. The reaction mixture was worked up as
described above. Flash chromatography afforded the ester 33
(4.06 g, 50%): 1H NMR (80 MHz, CDCl3) δ 1.24 (t, 3 H, J )
7.1 Hz), 1.20-1.80 (m, 10 H), 2.30-2.70 (m, 3 H), 3.24-3.87
(m, 4 H), 4.12 (q, 2 H, J ) 7.1 Hz), 4.55 (br s, 1 H), 4.90-
5.88 (m, 3 H). This sample of the ester 33 was dissolved in
anhydrous ether (30 mL), and the solution was added to a
suspension of LAH (573 mg, 15 mmol) in anhydrous ether (40
mL) under reflux. After being stirred for 30 min at room
temperature, the reaction mixture was worked up as described
above. The alcohol (3.40 g, 100%) was obtained via flash
chromatography. This sample of the alcohol was treated with
pyridine (1.25 mL) and acetic anhydride (1.53 g, 15 mmol) at
0 °C. After stirring 1 h at 0 °C, the reaction mixture was
warmed to room temperature and stirred for 12 h. The reaction
mixture was washed with 2 N HCl and brine, dried over
anhydrous MgSO4, and concentrated. Flash chromatography
1
yield from 4-penten-1-ol): H NMR (80 MHz, CDCl3) δ 1.95-
2.20 (m, 4 H), 3.57-3.80 (m, 2 H), 4.87-6.05 (m, 5 H). A
sample (1.56 g, 13.9 mmol) of 29 was heated with triethyl
orthoacetate (15.8 g, 97.4 mmol) in the presence of propionic
acid (0.15 g) at 130-140 °C for 20 h with continuous removal
of ethanol. The reaction was quenched with 2 N HCl, and the
reaction mixture was extracted with ether three times. The
combined ether extracts were washed with saturated NaHCO3
solution and brine, dried over anhydrous MgSO4, and concen-
trated. Flash chromatography afforded the product ester 30
(1.55 g, 60%). A sample (822 mg, 4.4 mmol) of the ester 30
was dissolved in anhydrous ether (10 mL), and this solution
was added dropwise into a suspension of LAH (168 mg, 4.4
mmol) in ether (10 mL) at 0 °C. After being stirred for 30 min
at room temperature, this reaction mixture was worked up as
described above. Flash chromatography afforded 3-ethenyl-6-
hepten-1-ol (22; 430 mg, 70%): 1H NMR (80 MHz, CDCl3) δ
1.24-1.73 (m, 5 H), 1.91-2.30 (m, 3 H), 3.64 (t, 2 H, J ) 6.5
Hz)), 4.86-6.06 (m, 6 H).
1
afforded the protected diol 34 (4.0 g, 98%): H NMR (80 MHz,
CDCl3) δ 1.30-1.80 (m, 13 H). 2.04 (s, 3 H), 3.23-3.87 (m,
4 H), 3.96-4.15 (m, 2 H), 4.55 (br s, 1 H), 4.85-5.78 (m, 3
H). This sample of the protected diol 34 was dissolved in
methanol (40 mL) containing a catalytic amount (0.2 g) of
p-TsOH, and the solution was stirred for 90 min at room
temperature. The reaction mixture was concentrated, and the
residue was dissolved in ether. The ether solution was washed
with brine, dried over anhydrous MgSO4, and concentrated to
give the alcohol (2.64 g, 95%). Oxalyl chloride (1.40 mL, 14.5
mmol) in dry dichloromethane (50 mL) was treated with DMSO
(2.67 mL, 31.3 mmol) at -78 °C under a nitrogen atmosphere.
After 5 min, the sample of the alcohol (2.64 g, 14.1 mmol) in
dry dichloromethane (15 mL) was added dropwise to this
solution, and the reaction mixture was further stirred for 15 min
at -78 °C. Triethylamine (23 mL, 89.9 mmol) was added to
the reaction mixture, which was stirred for 30 min at -78 °C
before warming to 0 °C. The reaction mixture was quenched
with water, and the aqueous layer was extracted three times
with dichloromethane. The combined organic extracts were
washed with brine, dried over anhydrous MgSO4, and concen-
trated to give the aldehyde (2.34 g, 90%). This sample of the
aldehyde was dissolved in dry THF (10 mL) and the solution
was treated with (carbethoxymethylene)triphenylphosphorane
(2.20 g, 12.8 mmol) dissolved in dry THF (20 mL) at 0 °C.
The reaction mixture was stirred for 2 h at room temperature.
Concentration and flash chromatography afforded the acetoxy
ester (3.07 g, 95%). This sample of the acetoxy ester was
dissolved in ethanol (50 mL) with K2CO3 (0.3 g). The solution
was stirred for 1 day at room temperature. Concentration and
flash chromatography afforded ethyl 6-(2′-hydroxyethyl)-2,7-
octadienoate (23, 2.17 g, 85%): 1H NMR (80 MHz, CDCl3) δ
1.28 (t, 3 H, J ) 7.1 Hz), 1.35-1.79 (m, 5 H), 1.97-2.34 (m,
3 H), 3.65 (t, 2 H, J ) 6.5 Hz), 4.18 (q, 2 H, J ) 7.1 Hz),
4.89-7.14 (m, 5 H).
Ethyl 6-(2′-Hydroxyethyl)-2,7-octadienoate (23). To a
solution of 1,4-butanediol (31; 10.2 g, 113 mmol) and a catalytic
amount (0.2 g) of p-TsOH in dichloromethane (150 mL) was
added DHP (10.3 mL, 113 mmol) in dichloromethane (40 mL)
via a syringe pump for 5 h. After complete addition, the reaction
mixture was stirred for 2 h at room temperature. The reaction
was quenched with water, and the reaction mixture was extracted
three times with dichloromethane. The combined organic
extracts were washed with brine, dried over anhydrous MgSO4,
and concentrated. Flash chromatography yielded the THP-
protected diol (10.4 g, 60%). This sample of the alcohol was
reacted with PCC (25.4 g, 118 mmol) in dichloromethane (100
mL) in the presence of NaOAc and 4-Å molecular sieves. The
reaction mixture was filtered through a silica gel pad, and the
filtrate was concentrated to yield the aldehyde (7.11 g, 70%).
The aldehyde sample was dissolved in dry THF (20 mL), and
this solution was added to (carbethoxymethylene)triphen-
ylphosphorane (2.58 g, 15 mmol) in dry THF (80 mL) at 0 °C.
The reaction mixture was stirred for 2 h at room temperature.
Concentration and flash chromatography afforded the ester 32
(8.51 g, 85%): 1H NMR (80 MHz, CDCl3) δ 1.28 (t, 3 H, J )
7.1 Hz), 1.43-1.91 (m, 8 H), 2.18-2.37 (m, 2 H), 3.26-3.91
(m, 4 H), 4.18 (q, 2 H, J ) 7.1 Hz), 4.56 (br s, 1 H), 5.83 (dt,
1 H, Jt ) 1.5 Hz, Jd ) 15.6 Hz), 7.00 (dt, 1 H, Jt ) 6.8 Hz, Jd
) 15.6 Hz). The ester 32 (8.0 g, 33 mmol) was dissolved in
anhydrous ether (75 mL), and this solution was added dropwise
into a solution of LAH (1.26 g, 33 mmol) in anhydrous ether
(75 mL) at 0 °C. After being stirred 1 h at room temperature,
this reaction mixture was worked up as described above. Flash
chromatography afforded the allylic alcohol (5.95 g, 30 mmol).
This sample of the allylic alcohol was heated with triethyl
General Procedure for Preparation of Bromoacetates. To
a stirred solution of DCC (1.2 equiv) in dry dichloromethane
was added bromoacetic acid (1.4 equiv) at room temperature,
and the reaction mixture was stirred for 30 min. To this mixture
was added an alcohol solution in dichloromethane followed by
a catalytic amount (0.1 equiv) of DMAP, and the reaction
mixture was stirred for 30 min at room temperature. After
filtration on a silica gel pad, the filtrate was concentrated under
reduced pressure. Flash chromatography afforded the corre-
sponding bromoacetate in good yield.

Eight-Membered-Ring Lactone Preparation
J. Am. Chem. Soc., Vol. 120, No. 30, 1998 7477
General Procedure for Radical Cyclizations. To a stirred
solution of a bromoacetate in benzene (0.015 M) under reflux
was added a mixture of tributylstannane (1.4 equiv) and AIBN
(0.1 equiv) in benzene via a syringe pump for 5 h (typical
scale: 1.0 mmol of bromoacetate in 67 mL of benzene). The
reaction mixture was further heated under reflux for 1 h and
concentrated under reduced pressure. The products were
separated by flash chromatography.
Oxocan-2-one (3c). From 207 mg of 3a, 48 mg (38%) of
3c was obtained after chromatographic separation (10:1-3:1
hexane/ether): 1H NMR (80 MHz, CDCl3) δ 1.56-2.05 (m, 8
H), 2.52 (t, 2 H, J ) 5.6 Hz), 4.32 (t, 2 H, J ) 5.6 Hz); 13C
NMR (50.3 MHz, CDCl3) δ 176.7, 67.83, 31.24, 30.90, 28.31,
25.79, 23.91; IR (neat, cm-1) 2920, 2860, 1725, 1450, 1232,
1130, 1097; MS (EI) m/z (relative intensity) 128 (M+, 0.1), 110
(6), 100 (19), 98 (16), 70 (27), 69 (48), 55 (100), 42 (70); HRMS
m/z calcd for C7H12O2 128.0837, found 128.0844.
5-Methyloxocan-2-one (4c). From 221 mg of 4a, 54 mg
(38%) of 4c was obtained after chromatographic separation (4:1
hexane/ethyl acetate): 1H NMR (80 MHz, CDCl3) δ 0.98 (d, 3
H, J ) 5.8 Hz), 1.25-2.00 (m, 7 H), 2.48-2.79 (m, 2 H), 4.32
(t, 2 H, J ) 5.6 Hz); 13C NMR (50.3 MHz, CDCl3) δ 176.4,
68.19, 36.12, 32.34, 32.29, 30.36, 30.09, 24.68; MS (EI) m/z
(relative intensity) 142 (M+, 0.2), 124 (4), 113 (19), 112 (11),
101 (10), 94 (21), 83 (39), 70 (38), 69 (51), 55 (100); HRMS
m/z calcd for C8H14O2 142.0995, found 142.0988.
5-Trimethylsilyloxocan-2-one (5c). From 150 mg of 5a,
54 mg (54%) of 5c was obtained after chromatographic
separation (5:1 hexane/ethyl acetate): 1H NMR (300 MHz,
CDCl3) δ -0.04 (s, 9 H), 0.45-0.55 (m, 1 H), 1.06-1.20 (m,
1 H), 1.46-1.61 (m, 2 H), 1.77-1.85 (m, 1 H), 1.96-2.09 (m,
2 H), 2.49 (ddd, 1 H, J ) 4.3, 9.0, 12.3 Hz), 2.70 (ddd, 1 H, J
) 4.1, 8.0, 12.3 Hz), 4.29-4.46 (m, 2 H); 13C NMR (50.3 MHz,
CDCl3) δ 176.6, 67.73, 32.58, 31.79, 29.08, 25.15, 24.41,
-3.777; IR (neat, cm-1) 2955, 2858, 2254, 1722, 1448, 1249,
1124, 837; MS (CI) m/z (relative intensity) 241 (M + 41, 1),
201 (M + 1, 40), 185 (38), 172 (7), 157 (3), 143 (10), 129 (3),
111 (53), 103 (3), 83 (100), 73 (55), 69 (21); HRMS m/z calcd
for C10H20O2Si 200.1233, found 200.1206.
2960, 2925, 1725, 1455, 1375, 1340, 1165, 1120, 1075; MS
(EI) m/z (relative intensity) 170 (M+, 1), 155 (1), 96 (100), 83
(45), 69 (28), 55 (45); HRMS m/z calcd for C10H18O2 170.1307,
found 170.1313.
8-Ethenyloxocan-2-one (14b). From 233 mg of 14a, 48 mg
(31%) of 14b and 82 mg (53%) of 14c were obtained after
chromatographic separation (8:1 hexane/ether): 1H NMR (80
MHz, CDCl3) δ 1.36-2.03 (m, 8 H), 2.43-2.59 (m, 2 H), 5.08-
5.43 (m, 3 H), 5.93 (ddd, 1 H, J ) 5.5, 10.0, 17.0 Hz); 13C
NMR (125.7 MHz, CDCl3) δ 176.4, 136.7, 115.8, 78.55, 37.33,
32.99, 29.14, 26.59, 24.02; IR (neat, cm-1) 2920, 2850, 1725,
1445, 1360, 1240; MS (EI) m/z (relative intensity) 154 (M+,
0.2), 126 (3), 111 (8), 98 (52), 80 (14), 69 (32), 55 (100); HRMS
m/z calcd for C9H14O2 154.0995, found 154.0975.
9-Methyl-2-oxabicyclo[4.2.1]nonan-3-one (14c). 1H NMR
(200 MHz, CDCl3) δ 1.06 (d, 3 H, J ) 7.1 Hz), 1.20-2.75 (m,
10 H), 4.61 (dd, 1 H, J ) 5.4, 7.7 Hz); 13C NMR (50.3 MHz,
CDCl3) δ 175.3, 83.55, 42.77, 40.20, 33.45, 32.25, 28.97, 22.36,
10.21; IR (neat, cm-1) 2940, 2860, 1730, 1720, 1430, 1350,
1250, 1190, 1170, 1105, 1040; MS (EI) m/z (relative intensity)
154 (M+, 3), 125 (2), 110 (89), 97 (72), 82 (27), 67 (64), 55
(100); HRMS m/z calcd for C9H14O2 154.0995, found 154.0999.
7,7-Dimethyl-8-ethenyloxocan-2-one (15c). From 340 mg
of 15a, 60 mg (25%) of 15c, 59 mg (25%) of 15e, and 55 mg
(23%) of 15d were obtained after chromatographic separation
(30:1-5:1 hexane/ether): 1H NMR (80 MHz, CDCl3) δ 0.89
(s, 3 H), 0.93 (s, 3 H), 1.10-2.00 (m, 6 H), 2.35-2.59 (m, 2
H), 4.76 (d, 1 H, J ) 6.0 Hz), 5.16-5.43 (m, 2 H), 5.89 (ddd,
1 H, J ) 5.9, 9.7, 17.3 Hz); 13C NMR (50.3 MHz, CDCl3) δ
175.4, 133.0, 117.8, 84.18, 39.56, 32.93, 29.12, 24.75, 21.97,
21.63; IR (neat, cm-1) 2960, 2920, 2855, 2830, 1720, 1460,
1350, 1325, 1270, 1205, 1140, 1105; MS (EI) m/z (relative
intensity) 182 (M+, 0.1), 126 (24), 111 (4), 82 (100), 69 (22),
55 (44); HRMS m/z calcd for C11H18O2 182.1307, found
182.1279.
8,8,9-Trimethyl-2-oxabicyclo[4.2.1]nonan-3-one (15d). 1H
NMR (300 MHz, CDCl3) δ 1.025 (d, 3 H, J ) 7.1 Hz), 1.138
(s, 3 H), 1.178 (s, 3 H), 1.29-1.39 (m, 1 H), 1.49 (d, 1 H, J )
14.3 Hz), 1.75-1.85 (m, 1 H), 1.89 (dd, 1 H, J ) 8.3, 14.3
Hz). 2.40-2.78 (m, 4 H), 3.97 (d, 1 H, J ) 4.7 Hz); 13C NMR
(125.7 MHz, CDCl3) δ 175.4, 93.18, 43.54, 42.13, 40.87, 40.33,
34.04, 33.70, 23.42, 22.29, 10.83; IR (neat, cm-1) 3010, 2920,
1705, 1460, 1215, 1160, 1140; MS (EI) m/z (relative intensity)
182 (M+, 1), 167 (1), 154 (3), 138 (19), 125 (42), 110 (16), 97
(32), 83 (16), 81 (37), 69 (40), 55 (100); HRMS m/z calcd for
C11H18O2 182.1307, found 182.1316.
8-Phenyloxocan-2-one (6c). From 283 mg of 6a, 107 mg
(52%) of 6c was obtained after chromatographic separation (8:
1
1-4:1 hexane/ether): H NMR (80 MHz, CDCl3) δ 1.60-2.20
(m, 8 H), 2.58 (t, 2 H), 5.69 (t, 1 H, J ) 6.9 Hz), 7.29-7.44
(m, 5 H); 13C NMR (50.3 MHz, CDCl3) δ 176.2, 140.2, 128.3,
127.7, 125.8, 79.63, 39.71, 32.93, 29.27, 26.53, 24.31; MS (EI)
m/z (relative intensity) 204 (M+, 0.5), 144 (17), 117 (25), 105
(26), 99 (59), 77 (36), 69 (36), 55 (100); HRMS m/z calcd for
C13H16O2 204.1150, found 204.1180.
6-Ethyl-8,8-dimethyl-2-oxabicyclo[3.3.0]octan-3-one
(15e). 1H NMR (300 MHz, CDCl3) δ 0.902 (t, 3 H, J ) 7.3
Hz), 0.985 (s, 3 H), 1.106 (s, 3 H), 1.21 (t, 1 H, J ) 12.7 Hz),
1.27-1.41 (m, 2 H), 1.58 (dd, 1 H, J ) 5.3, 12.7 Hz), 2.05-
2.20 (m, 1 H), 2.50 (d, 1 H, J ) 1.6 Hz), 2.52 (d, 1 H, J ) 5.0
Hz), 2.98-3.07 (m, 1 H), 4.39 (d, 1 H, J ) 6.3 Hz); 13C NMR
(50.3 MHz, CDCl3) δ 177.7, 93.20, 43.49, 41.79, 40.49, 39.98,
29.36, 25.61, 24.25, 23.39, 12.92; MS (EI) m/z (relative
intensity) 182 (M+, 2), 153 (12), 140 (2), 125 (17), 111 (7), 97
(100), 85 (13), 69 (45), 55 (59); HRMS m/z calcd for C11H18O2
182.1307, found 182.1271.
7,7-Dimethyloxocan-2-one (7c). From 235 mg of 7a, 83
mg (53%) of 7c was obtained after chromatographic separation
(7:1 hexane/ethyl acetate): 1H NMR (80 MHz, CDCl3) δ 0.96
(s, 6 H), 1.20-2.00 (m, 6 H), 2.45-2.62 (m, 2 H), 3.94 (s, 2
H); 13C NMR (50.3 MHz, CDCl3) δ 176.1, 75.36, 37.27, 36.51,
31.32, 28.74, 25.12, 22.01; MS (EI) m/z (relative intensity) 156
(M+, 0.5), 126 (5), 124 (11), 111 (3), 95 (3), 83 (15), 82 (100),
69 (19), 55 (37); HRMS m/z calcd for C9H16O2 156.1150, found
156.1178.
5,7,7-Trimethyloxocan-2-one (8c). From 210 mg of 8a, 35
mg (25%) of 8c was obtained after chromatographic separation
(8:1-4:1 hexane/ether): 1H NMR (300 MHz, CDCl3) δ 0.927
(s, 3 H), 0.969 (d, 3 H, J ) 6.8 Hz), 0.994 (s, 3 H), 1.14-1.95
(m, 5 H), 2.55-2.60 (m, 2 H), 3.80 and 4.08 (ABq, 2 H, J )
12.1 Hz); 13C NMR (50.3 MHz, CDCl3) δ 175.9, 75.53, 46.02,
6,8,8,9-Tetramethyl-2-oxabicyclo[4.2.1]nonan-3-one (16c).
From 275 mg of 16a, 54 mg (28%) of 16d and 30 mg (14%) of
16c were obtained after chromatographic separation (10:1-4:1
hexane/ether): 1H NMR (300 MHz, CDCl3) δ 0.935 (d, 3 H, J
) 7.2 Hz), 1.081 (s, 3 H), 1.09-1.18 (m, 1H), 1.146 (s, 3 H),
1.160 (s, 3 H), 1.568 and 1.692 (ABq, 2 H, J ) 14.4 Hz), 1.671
(dd, 1 H, J ) 4.5, 13.5 Hz), 2.18-2.24 (m, 1 H), 2.63 (ddd, 1
36.79, 36.18, 30.86, 28.68, 27.36, 25.59, 22.70; IR (neat, cm-1
)

7478 J. Am. Chem. Soc., Vol. 120, No. 30, 1998
Lee et al.
H, J ) 5.4, 13.4, 16.5 Hz), 2.75 (dt, 1 H, Jt ) 4.3 Hz, Jd )
16.5 Hz), 4.03 (d, 1 H, J ) 5.0 Hz); 13C NMR (50.3 MHz,
CDCl3) δ 175.0, 94.07, 51.75, 46.13, 44.98, 40.95, 34.61, 29.60,
28.88, 23.50, 8.659; IR (neat, cm-1) 2960, 2920, 1720, 1715,
1460, 1390, 1340, 1285, 1240, 1190, 1120; MS (EI) m/z (relative
intensity) 196 (M+, 2), 153 (3), 139 (100), 121 (15), 111 (68),
109 (29), 96 (45), 83 (53), 69 (49), 55 (72); HRMS m/z calcd
for C12H20O2 196.1465, found 196.1451.
6-Ethyl-6,8,8-trimethyl-2-oxabicyclo[3.3.0]octan-3-one
(16d). 1H NMR (300 MHz, CDCl3) δ 0.864 (t, 3 H, J ) 7.4
Hz), 1.067 (s, 3 H) 1.098 (s, 3 H), 1.171 (s, 3 H), 1.285 (q, 2
H, J ) 7.4 Hz), 1.495 and 1.674 (ABq, 2 H, J ) 13.7 Hz),
2.490 (d, 2 H, J ) 7.6 Hz), 2.60-2.75 (m, 1 H), 4.530 (d, 1 H,
J ) 6.0 Hz); 13C NMR (50.3 MHz, CDCl3) δ 177.7, 94.35,
52.39, 51.02, 42.22, 41.20, 32.78, 31.11, 29.52, 27.53, 26.02,
9.203; MS (EI) m/z (relative intensity) 196 (M+, 0.1), 181 (2),
167 (23), 153 (8), 139 (36), 126 (25), 111 (100), 97 (15), 83
(70), 71 (27), 69 (39), 55 (74); HRMS m/z calcd for C12H20O2
196.1465, found 196.1434.
180.3, 75.19, 40.10, 39.35, 38.88, 36.75, 34.19, 32.42, 24.84,
15.56; IR (neat, cm-1) 3020, 2958, 2931, 2874, 1732, 1215;
MS (EI) m/z (relative intensity) 168 (M+, 0.5), 138 (36), 120
(5), 109 (8), 94 (100), 81 (63), 67 (70); HRMS m/z calcd for
C10H16O2 168.1151, found 168.1139.
The major isomer with a lower Rf value on TLC: 1H NMR
(300 MHz, CDCl3) δ 1.030 (d, 3 H, J ) 6.7 Hz), 1.31-1.41
(m, 1 H), 1.50-1.55 (m, 1 H), 1.94-2.03 (m, 5 H), 2.13-2.27
(m, 1 H), 2.30-2.40 (m, 2 H), 2.50-2.60 (m, 1 H), 3.97 (dd,
1 H, J ) 5.4, 12.1 Hz), 4.58 (dd, 1 H, J ) 3.5, 12.1 Hz); 13C
NMR (125.7 MHz, CDCl3) δ 178.4, 73.13, 43.95, 41.98, 40.88,
36.45, 34.31, 31.29, 28.82, 22.03; IR (neat, cm-1) 3020, 2957,
2930, 2870, 1730, 1217; MS (EI) m/z (relative intensity) 168
(M+, 5), 150 (17), 138 (14), 124 (49), 106 (16), 94 (98), 81
(95), 67 (74), 55 (79), 41 (100); HRMS m/z calcd for C10H16O2
168.1151, found 168.1173.
9-Methyl-4-oxabicyclo[6.3.0]undecan-5-one (19c). From
261 mg of 19a, 70 mg (38%) of 19c was obtained after
1
chromatographic separation (3:1 hexane/ethyl acetate): H NMR
8-(2′-Propenyl)oxocan-2-one (17c). From 247 mg of 17a,
98 mg (58%) of 17c was obtained after chromatographic
separation (14:1-5:1 hexane/ether): 1H NMR (300 MHz,
CDCl3) δ 1.47-1.96 (m, 8 H), 2.27-2.37 (m, 1 H), 2.41-2.59
(m, 3 H), 4.57-4.66 (m, 1 H), 5.08-5.17 (m, 2 H), 5.76-5.90
(m, 1 H); 13C NMR (50.3 MHz, CDCl3) δ 176.5, 133.6, 117.7,
78.01, 39.79, 36.70, 32.28, 28.72, 26.20, 23.74; MS (EI) m/z
(relative intensity) 168 (M+, 0.1), 128 (7), 127 (90), 99 (14),
81 (100), 69 (15), 55 (60); HRMS m/z calcd for C10H16O2
168.1150, found 168.1144.
(300 MHz, CDCl3) δ 0.875 (d, 3 H, J ) 7.1 Hz), 1.18-1.78
(m, 7 H), 1.89-2.19 (m, 4 H), 2.51 (ddd, 1 H, J ) 4.6, 8.7,
12.6 Hz), 2.68 (ddd, 1 H, J ) 4.4, 8.3, 12.6 Hz), 4.36 (t, 2 H,
J ) 5.5 Hz); 13C NMR (125.7 MHz, CDCl3) δ 176.8, 68.51,
45.54, 42.03, 39.04, 37.74, 33.94, 33.58, 30.79, 29.24, 15.13;
IR (neat, cm-1) 2932, 2861, 1732, 1248, 1038, 910; MS (EI)
m/z (relative intensity) 182 (M+, 1), 164 (1), 153 (13), 138 (24),
127 (10), 109 (28), 95 (48), 85 (49), 81 (52), 67 (51), 55 (100),
41 (69); HRMS m/z calcd for C11H18O2 182.1307, found
182.1311.
7-(2′-Propenyl)oxocan-2-one (18b). From 247 mg of 18a,
45 mg (27%) of the minor isomer of 18c, 53 mg (32%) of the
major isomer of 18c, and 30 mg (18%) of 18b were obtained
after chromatographic separation (5:1 hexane/ethyl acetate): 1H
NMR (300 MHz, CDCl3) δ 1.19-2.20 (m, 9 H), 2.52-2.60
(m, 2 H), 4.15 (dd, 1 H, J ) 6.3, 12.2 Hz), 4.34 (dd, 1 H, J )
3.7, 12.2 Hz), 5.03-5.10 (m, 2 H), 5.68-5.82 (m, 1 H); 13C
NMR (125.7 MHz, CDCl3) δ 176.5, 136.0, 116.9, 70.19, 40.97,
36.37, 31.58, 29.67, 28.61, 24.21; IR (neat, cm-1) 3020, 2930,
1732, 1248, 1217, 1047, 910; MS (CI) m/z (relative intensity)
209 (M + 41, 0.5), 169 (M + 1, 34), 151 (89), 133 (70), 123
(33), 109 (100), 81 (14), 69 (23); HRMS m/z calcd for C10H16O2
168.1151, found 168.1180.
8-Methyl-3-oxabicyclo[5.2.1]decan-4-one (18c). The minor
isomer with a higher Rf value on TLC: 1H NMR (300 MHz,
CDCl3) δ 1.022 (d, 3 H, J ) 6.3 Hz), 1.36-1.46 (m, 1 H),
1.63-1.86 (m, 3 H), 1.89-2.14 (m, 4 H), 2.34-2.44 (m, 2 H),
2.48-2.56 (m, 1 H), 3.96 (dd, 1 H, J ) 3.4, 11.8 Hz), 4.59
(dd, 1 H, J ) 4.0, 11.8 Hz); 13C NMR (125.7 MHz, CDCl3) δ
9-(Ethoxycarbonyl)methyl-4-oxabicyclo[6.3.0]undecan-5-
one (20b). From 334 mg of 20a, 100 mg (39%) of 20b was
obtained after chromatographic separation (3:1 hexane/ethyl
acetate): 1H NMR (300 MHz, CDCl3) δ 1.25 (t, 3 H, J ) 7.1
Hz), 1.30-1.80 (m, 6 H), 1.82-1.96 (m, 3 H), 1.98-2.10 (m,
1 H), 2.15-2.41 (m, 2 H), 2.43-2.57 (m, 2 H), 2.63-2.71 (m,
1 H), 4.14 (q, 2 H, J ) 7.1 Hz), 4.30-4.41 (m, 2 H); 13C NMR
(75.4 MHz, CDCl3) δ 177.1, 173.9, 68.48, 60.66, 44.11, 42.56,
40.76, 37.18, 35.10, 33.60, 31.68, 30.56, 29.10, 14.34; IR (neat,
cm-1) 2939, 2864, 1726, 1278, 1194, 912; MS (EI) m/z (relative
intensity) 254 (M+, 2), 236 (4), 209 (19), 190 (6), 180 (59),
167 (84), 149 (18), 135 (17), 121 (39), 107 (92), 93 (78), 79
(73), 67 (61), 55 (100), 41 (81); HRMS m/z calcd for C14H22O4
254.1518, found 254.1527.
Acknowledgment. Authors thank the Organic Chemistry
Research Center (KOSEF) and Ministry of Education (BSRI-
94-3416) for financial support.
JA980908D