Formation of Some Oxygen-Containing Heterocycles by Radical Cyclization: The Stereochemical Influence of Anomeric Effects

Article abstract of DOI:10.1021/jo980359u

The influence of the anomeric effect on radical cyclization has been examined by determining the stereochemical outcome of the ring closure of 10 suitably substituted radicals. The stereochemistry of the products formed from the acyclic precursors 4a-f indicates that for suitably constituted radicals anomeric interactions stabilize pseudoaxially substituted transition structures 14 thus affording products with stereochemistry the reverse of that normally observed.

Full text of DOI:10.1021/jo980359u

5144
J . Org. Chem. 1998, 63, 5144-5153
F or m a tion of Som e Oxygen -Con ta in in g Heter ocycles by Ra d ica l
Cycliza tion : Th e Ster eoch em ica l In flu en ce of An om er ic Effects
Athelstan L. J . Beckwith* and Dennis M. Page
Research School of Chemistry, and Department of Chemistry, Australian National University,
Canberra, ACT 0200, Australia
Received February 26, 1998
The influence of the anomeric effect on radical cyclization has been examined by determining the
stereochemical outcome of the ring closure of 10 suitably substituted radicals. The stereochemistry
of the products formed from the acyclic precursors 4a -f indicates that for suitably constituted
radicals anomeric interactions stabilize pseudoaxially substituted transition structures 14 thus
affording products with stereochemistry the reverse of that normally observed.
In recent times, the use of free radical reactions for
has received support from theoretical analyses8 and the
modeling of transition structures by a combination of
molecular orbital and molecular mechanics calculations.8
These calculations confirm the original hypothesis that
the transition structure for 1,5-exo-cyclization of hex-5-
enyl radicals and related species resembles cyclohexane
in its chair form.2-5 Hence the stereoselectivity exhibited
in the ring closure of monosubstituted hexenyl radicals
and related species is seen to arise from the propensity
of cyclizations to proceed through the more stable pseu-
doequatorially substituted cyclohexane-like conformation
of the transition structure rather than through its less
stable pseudoaxially substituted conformer.8 Since cal-
culations of the relative selectivities of the cyclized
products tend to be overestimated when only axially
substituted chairlike structures are considered, it appears
that boatlike transition structures might also be involved
in some reactions.4,8
Cyclizations of hex-5-enyl type radicals containing
heteroatoms have also been studied.1,9,10 In general such
species conform to the usual guidelines.5 For example,
most 3-substituted allylperoxyl radicals undergo stereo-
selective cyclization to afford mainly cis-substituted 1,2-
dioxolanes.10,11 Thus 1 (R ) Ph) gives 2 (R ) Ph) with
cis:trans ) 10.11 However, a notable exception occurs
when the substituent is strongly electron withdrawing.
Thus, ring closure of 1 (R ) C6F5) affords trans-2 (R )
C6F5)11 in preference to the expected cis product.5
chemical synthesis has grown steadily in popularity as
the processes which govern and control their outcomes
have become better understood.1 Intramolecular addition
reactions of suitably constituted alkenyl radicals and
related species are particularly interesting since they
readily afford a variety of ring systems commonly found
in many natural products.1 The advantages offered by
such reactions generally include high regioselectivity,
modest to good stereoselectivity,1-5 and high chemo-
selectivity.1-5
The highly regioselective exo-cyclization of the hex-5-
enyl radical has been intensively studied. The accurate
determination of its kinetics6 has underpinned its wide-
spread use as a mechanistic probe and as a radical clock.7
Most monosubstituted hex-5-enyl radicals also undergo
regioselective exo-cyclization while conforming to the
general stereochemical guideline that 1- or 3-substituted
radicals afford predominantly cis-disubstituted products,
while 2- or 4-substituted radicals give mainly trans-
disubstituted products.5
It is now generally accepted that the cyclization
behavior of simple hex-5-enyl systems reflects the ste-
reoelectronic demands of the transition structure.2-5 The
hypothesis that the strain energy engendered in attaining
the optimum orbital overlap for 1,6-ring closure out-
weighs the steric and thermodynamic factors which
would otherwise favor six-membered ring formation2-5
(1) (a) Curran, D. P. Synthesis 1988, 489. (b) J asperse, C. P.; Curran,
D. P.; Fevig, T. L. Chem. Rev. 1991, 91, 1237. (c) Ramaiah, M.
Tetrahedron 1987, 43, 3541. (d) Thebtaranonth, C.; Thebtaranonth,
Y. Tetrahedron 1990, 46, 1385. (e) Giese, B. In Radicals in Organic
Synthesis: Formation of Carbon-Carbon Bonds; Pergamon: Oxford,
1986. (f) Giese, B.; Kopping, B.; Go¨bel, T.; Dickhaut, J .; Thoma, G.;
Kulicke, K. L.; Trach, F. Org. React. 1996, 48, 301.
(2) Beckwith, A. L. J . Tetrahedron 1981, 37, 3073.
(3) Beckwith, A. L. J .; Ingold, K. U. In Rearrangements in Ground
and Excited States; de Mayo, P., Ed., Academic Press: New York, 1980;
Vol I, p 161.
When these observations were first reported, no mecha-
nistic explanation was offered for this unexpected be-
havior.11 However, it occurred to us that they might
(4) Beckwith, A. L. J . Chem. Soc. Rev. 1993, 22, 143.
(5) Beckwith, A. L. J .; Easton, C. E.; Serelis, A. K. J . Chem. Soc.,
Chem. Commun. 1980, 482. Beckwith, A. L. J .; Lawrence, T.; Serelis,
A. K. J . Chem. Soc., Chem. Commun. 1980, 484.
(6) Chatgilialoglu, C.; Ingold, K. U.; Scaiano, J . C. J . Am. Chem.
Soc. 1981, 103, 7739. Beckwith, A. L. J .; Easton, C. J .; Lawrence, T.;
Serelis, A. K. Aust. J . Chem. 1983, 36, 545.
(7) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980, 13, 317.
Newcomb, M. Tetrahedron 1993, 49, 1151.
(9) (a) Begley, M. J .; Ladlow, M.; Pattenden, G. J . Chem. Soc., Perkin
Trans. 1 1988, 1095. (b) Beckwith, A. L. J .; Phillipou, B. G. J . Am.
Chem. Soc. 1974, 96, 1613. (c) Padwa, A.; Nimmesgern, H.; Wong, G.
S. K. J . Org. Chem. 1985, 50, 5620.
(8) (a) Beckwith, A. L. J .; Schiesser, C. H. Tetrahedron 1985, 41,
3925. (b) Spellmeyer, D. C.; Houk, K. N. J . Org. Chem. 1987, 52, 959.
(c) Beckwith, A. L. J .; Zimmermann, J . J . Org. Chem. 1991, 56, 5791.
(10) Beckwith, A. L. J .; Wagner, R. D. J . Org. Chem. 1981, 46, 3638
and references therein.
(11) Feldman, K. S.; Kraebel, C. M. J . Org. Chem. 1992, 57, 4574.
S0022-3263(98)00359-4 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/01/1998

Oxygen-Containing Heterocycles by Radical Cyclization
J . Org. Chem., Vol. 63, No. 15, 1998 5145
reflect the influence of the anomeric effect12-15 on the
conformation of the cyclization transition structure.
Although there is some debate about the magnitude of
the anomeric effect,12-15 and indeed of the underlying
cause (dipole-dipole interaction,14,16 or p-σ* MO over-
lap14,17), it is clear that 2-alkoxy- and 2-(acyloxy)tetrahy-
dropyrans preferentially assume conformations in which
the substituent occupies an axial position. By analogy,
radicals of the general type 3 would be expected to
undergo cyclization via axially substituted chairlike
transition structures to afford cis-disubstituted products
in preference to their trans isomers as predicted by the
usual guidelines.5 Preliminary experiments supported
the validity of this hypothesis.18 After our work had been
commenced, two other groups independently made brief
mentions of the possibility of transition state stabilization
by anomeric effects in an attempt to account for some
unexpected stereochemical preferences in free-radical
cyclization reactions.19 However, since a detailed study
of this type of stabilization in hex-5-enyl-like ring closures
had not previously been undertaken we decided to
proceed with an extensive systematic survey.
Sch em e 1
Sch em e 2
Herein we describe an examination of the stereochem-
ical outcome of the ring closure of 10 acyclic radicals
suitably constituted to detect the stereochemical effect
of anomeric interactions. The results, which confirm the
stereoselective formation of products with stereochem-
istry the reverse of that predicted by the usual guidelines,
are indeed consistent with the hypothesis that anomeric
stabilization of suitably substituted transition structures
can be sufficiently strong to outweigh the usual steric
interactions of substituents.
Resu lts
P r ecu r sor s. Of the required acyclic bromo acetals,
4a -f were readily prepared in modest (29%) to good yield
(98%) from the appropriately substituted allyl alcohols
by an established method (Scheme 1).20,21 Cyclohexyl-
idineethanol, required for the preparation of precursor
4e, was obtained as previously described from cyclohex-
anone.22,23 Substrate 4g was obtained, albeit in very poor
yield (3%), when an attempt was made to bromo-allyl-
oxylate 1-hexene by Grady and Chokshi’s method (Scheme
2).24 However, treatment of allyltrimethylsilane with
NBS or NBA and 3-buten-1-ol successfully afforded the
required substrate 4h , although once again in very poor
yield (6%). Cleavage of 1,2-epoxyhexane25 with diphenyl
diselenide and sodium borohydride in ethanol26 gave a
hydroxyselenide (55%) which was then treated with
potassium hydride and allyl bromide to give 4i in
moderate yield (35%) after preparative TLC (Scheme 2).
The preparation of the mono-mercurial 4j (52%) followed
a modification of McNeely and Wright’s procedure,27
whereby 1,5-hexadiene in large excess (6-10 mol equiv)
was treated with an ethanolic solution of mercury(II)
acetate. Smaller quantities of diene relative to the
amount of acetate tended to promote the formation of the
diastereoisomers of the dimercurial 5.
Rin g Closu r es. The various precursors were cyclized
by treatment with tributylstannane, or, in the case of the
organomercurial 4j, with sodium trimethoxyborohydride.
The precise conditions are given in the Experimental
Section and are summarized in Table 1. Usually the
conditions were finally chosen after running a number
of test experiments designed to assess the efficiency of
the workup and analysis of the reactions mixtures, which
were often complicated by the high volatility of the cyclic
(12) Kirby, A. J . In The Anomeric Effect and Related Stereoelectronic
Effects at Oxygen; Springer-Verlag: Berlin, 1983, and references
therein.
(13) Deslongchamps, P. In Stereoelectronic Effects in Organic Chem-
istry, Pergamon: Oxford, 1983.
(14) Perrin, C. L.; Armstrong, K. B.; Fabian, M. A. J . Am. Chem.
Soc. 1994, 116, 715.
(15) The Anomeric Effect and Associated Stereoelectronic Effects;
Thatcher, G. R. J ., Ed.; ACS Symposium Series 539; American
Chemical Society: Washington, DC, 1993.
(16) Edward, J . T. Chem. Ind. 1955, 1102.
(17) Lucken, E. A. C. J . Chem. Soc. 1959, 2954. Romers, C.; Altona,
C.; Buys, H. R.; Havinga, E. Top. Stereochem. 1969, 4, 39.
(18) Beckwith A. L. J .; Page, D. In Annual Report of the Research
School of Chemistry for 1993. Australian National University: Can-
berra, 1994, p 15.
(19) Crich, D.; Sun, S.; Brunckova, J . J . Org. Chem. 1996, 61, 605.
Lopez, J . C.; Gomez, A. M.; Fraser-Reid, B. J . Chem. Soc. Chem.
Commun. 1994, 1533.
(20) Stork, G.; Sher, P. M. J . Am. Chem. Soc. 1986, 108, 303. Stork,
G.; Sher, P. M.; Chen, H.-L. J . Am. Chem. Soc. 1986, 108, 6384. Stork,
G.; Mook, R., J r.; Biller, S. A.; Rychnovsky, S. D. J . Am. Chem. Soc.
1983, 105, 3741.
(21) Uneo, Y.; Moriya, O.; Chino, K.; Watanabe, M.; Okawara, M.
J . Chem. Soc., Perkin Trans. 1 1986, 1351. Uneo, Y.; Chino, K.;
Watanabe, M.; Moriya, O.; Okawara, M. J . Am. Chem. Soc. 1982, 104,
5564.
(24) Chokshi, S. K.; Grady, G. L. Synthesis 1972, 843.
(22) Wadsworth, W. S., J r.; Emmons, W. D. In Organic Syntheses;
Wiley: New York, 1973; Baumgarten, H. E., Ed.; Coll. Vol. 5, p 547.
(23) Brink, M. Synthesis 1975, 253.
(25) Carlson, R. G.; Behn, N. S. J . Org. Chem. 1967, 32, 1363.
(26) Sharpless, K. B.; Lauer, R. F. J . Am. Chem. Soc. 1973, 95, 2696.
(27) McNeely; K. H., Wright, G. F. J . Am. Chem. Soc. 1955, 77, 2553.

5146 J . Org. Chem., Vol. 63, No. 15, 1998
Beckwith and Page
Ta ble 1. Dia ster eom er ic Ra tios of Ra d ica l Cycliza tion
P r od u cts
preparatively inseparable diastereomeric cyclized prod-
ucts. In the cases of 4b, 4f, 4h , and 4j, the products of
direct reduction, 7b, 7f, 7h , and 7j, were also detected,
but only in the case of the first was it possible for the
yield (9%) to be directly determined. In the case of 7f
and 7j, isolation was impossible because of their volatility
and the small amounts of material involved. An authen-
tic sample of 7f was obtained by reduction of 4f with Bu3-
SnH at relatively high concentration (ca. 0.2 M). The
other uncyclized products were prepared as follows: 7b
by acid-catalyzed addition of acetic acid to allyl vinyl
ether; 7h by weak acid-catalyzed addition of but-3-en-1-
ol to allyltrimethylsilane, and 7j by treatment of 1-me-
thylpent-4-en-1-ol with potassium hydride, HMPA, and
ethyl iodide.
precursor methoda temp (°C) productb cis/trans yieldc (%)
4a
4a
4b
4b
4c
4c
4d
4d
4e
4f
A
B
B
C
A
D
A
B
A
E
80
40
40
80
80
ca. 20
80
40
80
80
40
40
80
6a
6a
4.7
5.8
3.5
1.7
2.1
2.7
2.1
2.7
1.6
0.36
0.2
2.2
0.2
1.1
12
61
48
16
44
71
44
66
61
50g
72
84g
50
6bd
6bd,e
6c
6c
6d
6a
6e
6fd,f
6g
4g
4h
4i
B
B
F
6h d
6i
h
4j
G
20
6jd
-
a
A: Bu3SnH (0.2 M), benzene, AIBN; B: Bu3SnH (0.2 M),
pentane, ButONdNOBut; C: Bu3SnH (syringe pump), benzene,
AIBN; D: Bu3SnH (0.2 M), benzene, AIBN, photolysis; E: Bu3SnH
(0.07 M), benzene, AIBN; F: Bu3SnH, (0.02 M), benzene, AIBN;
b
G: Na(OMe)3BH, CH2Cl2. Unless otherwise specified the uncyc-
lized product was not detected. c Isolated. Uncyclized product also
d
formed. e This reaction also afforded the hydroxy compounds 8 (see
text). f This reaction also afforded the oxepane 9 (see text). g Total
yield of all reduced products since complete separation of cyclized
h
and uncyclized materials was not possible. Products not isolated.
Sch em e 3
The reduction of 4b at 80 °C furnished four cyclic
products (each of which was five-membered), the NMR
of which suggested that partial decomposition of the
expected products i.e., cis- and trans-6b, had occurred to
afford the corresponding hydroxy compounds 8. Analysis
of the crude reaction mixture by TLC indicated that the
decomposition process was probably not an artifact of the
workup procedure. The formation of hemiacetals 8 was
eliminated by reduction of 4b at 40 °C, but this tended
to be at the expense of the formation of more 7b. The
presence of the oxepane 9 as a minor component of the
product mixture obtained from the reduction of substrate
4f was confirmed by comparison of observed data with
those previously reported.29
The stereochemistry of both diastereomers of each of
the cyclized products 6a , 6d , 6g, and 6i was rigorously
determined by a combination of COSY, and 1D and 2D
NOE experiments. The relative stereochemistry of the
other mixtures of diastereomers was then determined by
comparison of 1H and 13C NMR spectra. An exception to
this procedure was the identification of cis- and trans-
6j. In this case the coincidence of a number of chemical
products. In general, reaction mixtures containing non-
volatile products were worked up by the DBU method28
followed by further flash chromatography. Volatile prod-
ucts were purified by careful removal of solvent by
distillation at ambient pressure, followed by short-path
(Kugelrohr) distillation under reduced pressure to remove
the material from the unwanted tributylstannane byprod-
ucts. The analysis of reaction mixtures was accomplished
by NMR spectroscopy and/or gas chromatography with
appropriate reference compounds. To ensure that no loss
of any diastereoisomer occurred, the diastereomeric ratios
of cyclized products were determined both on the crude
reaction mixtures and after workup. In general, analysis
of the crude mixtures indicated that the reactions had
proceeded cleanly and with high efficiency. The poor
yields recorded in many cases reflect the difficulty of
separating the products.
1
shifts in the H NMR spectrum of the product mixture
made unambiguous structural elucidation impossible,
and therefore cis-6j was independently prepared from the
THP ether of cis-3-hydroxycyclopentaneacetic acid (readily
prepared from norcamphor30,31) as shown in Scheme 4.
The identification of the diastereomers of the tetrahy-
dropyran 6f was accomplished by comparison of the
observed NMR spectra with those reported in the litera-
ture,32 while the diastereomers of 6h were similarly
(29) Nerdel, F.; Buddrus, J .; Brodowski, W.; Hentschel, P.; Klamann,
D.; Weyerstahl, P. J ustus Liebigs Ann. Chem. 1967, 710, 36.
(30) House, H. O.; Haack, J . L.; McDaniel, W. C.; VanDerveer, D.
J . Org. Chem. 1983, 48, 1643.
(31) Vogel, A. In Vogel’s Textbook of Practical Organic Chemistry,
4th ed.; Furniss, B. S., Hannaford, A. J ., Rogers, V., Smith, P. W. G.,
Tatchell, A. R., Eds.; Longman: London, 1978; p 475.
(32) Weber, G. F.; Hall, S. S. J . Org. Chem. 1979, 44, 364. de Hoog,
A. J . Org. Magn. Reson. 1974, 6, 233. Booth, H.; Readshaw, S. A.
Tetrahedron 1990, 46, 2097.
The structures and yields of the cyclic products ob-
tained from each of the acyclic precursors are sum-
marized in Table 1 and Scheme 3. All but two of the
substrates, 4f and 4h , furnished a mixture of two
(28) Curran, D. P.; Chang, C.-T. J . Org. Chem. 1989, 54, 3140.

Oxygen-Containing Heterocycles by Radical Cyclization
J . Org. Chem., Vol. 63, No. 15, 1998 5147
Sch em e 4
Sch em e 6
Sch em e 5
which can occur between the two oxygen atoms as
depicted in structure 14. This anomeric interaction is
not possible in the equatorially substituted conformer of
14.
Further evidence in support of this hypothesis was
obtained when 6a with an initial diastereoisomeric ratio
of cis/trans ) 4.7 was treated with trifluoromethane-
sulfonic acid. Isomerization then occurred to give pref-
erentially the trans diastereomer, the trans/cis ratio
reaching 1.5:1. This is consistent with the steric consid-
erations5,8,34 which indicate the trans isomer of 6a to be
lower in energy than the cis. Clearly the observation that
trans-6a is not formed preferentially by radical cycliza-
tion of 4a (Table 1) indicates that the radical reactions
proceed with kinetic control via a transition structure
which is not subject to the usual steric requirements. It
is also relevant that radical cyclization of 4a in acetoni-
trile-d3 gave a diastereomeric mixture of 6a with a cis/
trans ratio of 2.8 whereas the same reaction in benzene-
d6 under otherwise identical conditions gave 6a with a
cis/trans ratio of 4.2. This observation is consistent with
the generalization that anomeric effects become less
pronounced in more polar solvents.12,35
The data in Table 1 show other interesting features.
For example, the cis/trans ratio for cyclization products
decreases in the order 6a > 6c > 6d > 6e, i.e., it
decreases in magnitude with increasing bulk of the
substituent, viz., methyl < hydroxyethyl < benzyl <
cyclohexyl. This is consistent with the view that the
usual steric factors affecting the conformation of a
substituted cyclohexane-like system12-15,34 oppose the
affects of anomeric stabilization. The results in Table 1
also show that the diastereoselectivity of cyclization
increases with a decrease in reaction temperature.
Hence, both the observed preference for cis-cyclization
and the differences between various diastereomeric ratios
must arise, at least in part, from differences in activation
energies which reflect, presumably, differences in the
strain energies of the transition structures.
identified by comparison of the observed NMR data with
those reported for the structurally similar cis- and trans-
2,4-dimethyltetrahydropyrans.33 A 1D NOE experiment
conducted on trans-6h provided additional confirmation.
Discu ssion
The results presented in Table 1 show that with the
exceptions of precursor 4j, which shows virtually no
stereoselectivity on radical cyclization, and 4f and 4h ,
which give six-membered cyclic products, the remaining
ring closures fall clearly into two groups. One, compris-
ing compounds 4g and 4i, behaves similarly to other
substrates capable of affording 2-substituted hex-5-enyl
radicals and shows a preference for radical ring closure
giving trans-disubstituted cyclic products. The trans/cis
ratios of 5.0 are roughly in agreement with that reported
for cyclization of 2-methylhex-5-enyl radical.2 In accord
with widely accepted theory, the initially formed radicals
10g and 10i (Scheme 5) react preferentially through the
more stable transition structures 11g and 11i with the
substituents in the pseudoequatorial orientation to give
the trans-substituted cyclic radicals 12g and 12i. In
short, the radicals 10g and 10i show precisely the same
behavior on cyclization as many other substituted 5-hex-
enyl systems containing heteroatoms in the chain.1f
The cyclization behavior of the other major group
comprising precursors 4a -e is very different in that each
of these compounds undergoes stereoselective cyclization
to afford mainly cis-disubstituted products in contraven-
tion to the usual guidelines.2-5,8 The observed stereose-
lectivity suggests that the radicals 13a -e cyclize pref-
erentially through pseudoaxially substituted transition
structures 14a -e to give the cis-substituted cyclic radi-
cals 15a -e (Scheme 6). The clear implication is that for
these radicals the pseudoaxially substituted chairlike
transition structures are more stable than their pseu-
doequatorially substituted conformers. As adumbrated
above, we propose that the stabilization of the axially
substituted conformers reflects the anomeric interaction
If, as has been discussed above, the existence of a
stabilizing anomeric interaction can reverse the normal
diastereoselectivity for five-membered ring formation by
radical cyclization, the same should be true for formation
of six-membered rings. The data in Table 1 support this
(34) Eliel, E. L.; Wilen, S. H.; Mander, L. N. In Stereochemistry of
Organic Compounds; Wiley: New York, 1994; Chapter 11.
(35) Eliel, E. L.; Giza, C. A. J . Org. Chem. 1968, 33, 3754. Praly,
J .-P.; Lemieux, R. U. Can. J . Chem. 1987, 65, 213. Tvaroska, I.; Bleha,
T. Collect. Czech. Chem. Commun. 1980, 45, 1883. Montagnani, R.;
Tomasi, J . Int. J . Quantum Chem. 1991, 39, 851. Kysel, O.; Mach, P.
J . Mol. Struct. (THEOCHEM 73) 1991, 227, 285. Cramer, C. J . J . Org.
Chem. 1992, 57, 7034. Cramer, C. J .; Truhlar, D. G. J . Am. Chem.
Soc. 1993, 115, 5745.
(33) Eliel, E. L.; Hargrave, K. D.; Pietrusiewicz, K. M.; Manoharan,
M. J . Am. Chem. Soc. 1982, 104, 3635. Eliel, E. L.; Manoharan, M.;
Pietrusiewicz, K. M.; Hargrave, K. D. Org. Magn. Reson. 1983, 21, 94.

5148 J . Org. Chem., Vol. 63, No. 15, 1998
Beckwith and Page
Sch em e 7
reaction or the workup procedure. However, definitive
answers to these questions await further experimenta-
tion.
Con clu sion
The observations presented above show that radicals
derived from 4a -e undergo 1,5-exo-cyclization to pref-
erentially afford cis-2,4-disubstituted tetrahydrofurans.
This stereochemical outcome is the reverse of that
predicted from the usual steric guidelines for cyclization
of substituted acyclic alkenyl radicals2-5 and is consistent
with stabilization of transition state stereochemistry by
anomeric effects. Similarly the preferential formation of
the trans product by 1,6-exo-cyclization of the radical
generated from 4f is contrary to expectation. However,
the stereochemistry of cyclization of radicals generated
from 4g-j in which anomeric interactions are not pos-
sible is consistent with the usual guidelines.2-5 These
observations support the hypothesis that anomeric effects
stabilize the axially substituted conformers of appropri-
ately substituted cylization transition structures.
hypothesis. Although experimental evidence is scanty,36
it appears that the cyclization of substituted hept-6-enyl
radicals and related species is subject to the usual
thermodynamic and steric guidelines.8a,8b Thus, a 2-sub-
stituted heptenyl radical would be expected to afford a
cis-disubstituted product via a chairlike transition struc-
ture in which both substituents are pseudoequatorial.
The radical 16 (Scheme 7) derived from 4h conforms to
this hypothesis and affords mainly the cis-disubstituted
radical 18 via the transition structure 17. However, the
observation that the radical 19 derived from 4f affords
mainly the trans-cyclized product suggests that the
reaction affords the cyclic radical 21 via the pseudoaxially
substituted transition structure 20, which we expect to
be more stable than its equatorially substituted isomer
because of the anomeric interaction between the two
oxygen atoms (Scheme 7). The observation that cycliza-
tion of 4f also affords the oxepane 9 accords with previous
observations that the ring closure of hept-6-enyl radicals
and related species is less regioselective than that of
analogous hex-5-enyl systems.8a,37
Although, in general, our results fully support the
contention that anomeric stabilization can reverse the
normal diastereoselectivity of radical cyclization, there
remain two unanswered questions. First, the cyclization
of the precursor 4j is anomalous in that it shows virtually
no diastereoselectivity, although as a 2-substituted hex-
enyl system it was expected to preferentially afford trans-
6j. Perhaps, in this case the low steric bulk of the alkoxy
substituent is insufficient to outweigh the overall ther-
modynamic factors favoring the formation of the more
stable isomer cis-6j.
The second anomaly concerns the fact that cyclization
of the acetoxy-substituted precursor 4b gives a lower cis/
trans ratio of products than does the ethoxy-substituted
precursor 4a . This is contrary to expectation because the
acetoxy group, being more electronegative than the
ethoxy,12 should be subject to a greater anomeric effect,
and hence the anomeric stabilization of transition struc-
ture 14b should be greater than that of 14a . Perhaps it
is simply due to the higher steric demand of the acetoxy
groups.12,34 However, the formation during the cycliza-
tion of 4b of analogous hydroxy compounds may indicate
that cis-6a is more prone to decomposition than trans-
6b and is thus preferentially consumed during the
Exp er im en ta l Section
Gen er a l. NMR spectra were recorded as solutions in CDCl3
(unless stated otherwise) at 25 °C on 300 or 500 MHz
spectrometers with TMS or CHCl3 as reference. The samples
used for all NOE experiments were degassed by freeze-
pump-thaw methods38 prior to data accumulation, in Wilmad
Taperlok NMR tubes. Analytical thin-layer chromatography
(TLC) was conducted on glass-backed microslides coated with
0.2 mm thick Merck Silica Gel 60 GF254. Preparative TLC was
conducted by successive placements of small aliquots of
substrate onto 20 × 20 cm glass-plates loaded with Merck
Kieselgel 60 F254, and the plates were then eluted. Flash
chromatography39 was conducted over Merck Kieselgel 60
(230-400 mesh) with analytical (AR) grade solvents. Gas
chromatography (GC) was conducted with either a BP-1, BP-
5, or a BP-10 (0.25 µm, 0.22 mm × 25 m) column.
Where necessary, solvents and reagents were purified as
recommended.38 Tributylstannane was prepared as previously
described.40 Unless stated otherwise, all reactions were
conducted under an atmosphere of dry nitrogen. All radical
reaction mixtures were purged with a stream of nitrogen for
at least 10 min prior to commencement of the reaction.
Reaction temperatures refer to the external bath temperature.
All photolyses were performed at room temperature with a
quartz water-jacketed, Phillips 125 W medium-pressure mer-
cury-vapor lamp set at a distance of ca. 15 cm from the Pyrex
reaction vessel. Microanalyses were performed by the Aus-
tralian National University Analytical Services Unit. All
yields given are for isolated products unless stated otherwise.
3-(2-Br om o-1-eth oxyeth oxy)p r op -1-en e (4a ). Ethyl vi-
nyl ether (3.5 mL, 36 mmol) was added dropwise to a stirred
solution of NBS (5.2 g, 29 mmol) and allyl alcohol (1.9 mL, 28
mmol) in CH2Cl2 (20 mL) at -20 °C, and the mixture was
stirred for 2 h.20,21 The reaction mixture was then warmed to
room temperature, and any precipitated material was removed
by filtration through sintered glassware. Flash chromatog-
raphy (30:1 hexane/diethyl ether) of the concentrated crude
residue furnished the desired bromide 4a as a colorless oil41
(4.6 g, 75%); 1H NMR (CDCl3) δ 1.23 (t, 3H, J ) 6.9 Hz), 3.37
(d, 2H, J ) 5.43 Hz), 3.55-3.75 (m, 2H), 4.00-4.20 (m, 2H,),
4.71 (t, 1H, J ) 5.43 Hz), 5.17-5.34 (m, 2H), 5.84-5.98 (m,
1H); 13C NMR (CDCl3) δ 15.07, 31.57, 62.29, 67.45, 100.75,
(36) (a) Hanessian, S.; Dhanoa, D. S.; Beaulieu, P. L. Can. J . Chem.
1987, 65, 1859. (b) Fang, J . M.; Chang, H. T.; Lin, C. C. J . Chem. Soc.
Chem. Commun. 1988, 1385. (c) Ogura, K.; Kayano, A.; Fujino, T.;
Sumitani, N.; Fujita, M. Tetrahedron Lett. 1993, 34, 8313.
(37) Beckwith, A. L. J .; Moad, G. J . Chem. Soc., Chem. Commun.
1974, 472. Beckwith, A. L. J .; Roberts, D. H. J . Am. Chem. Soc. 1986,
108, 5893, for further examples see ref 1f.
(38) Armarego, W. L. F.; Perrin, D. D. Purification of Laboratory
Chemicals, 4th ed.; Butterworth Heinemann: Oxford, 1996.
(39) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 43, 2923.
(40) Szammer, J .; O¨ tvo¨s, L. Chem. Ind. 1988, 23, 764.
(41) Lidbetter, P. S.; Marples, B. A. Synth. Commun. 1986, 16, 1529.

Oxygen-Containing Heterocycles by Radical Cyclization
J . Org. Chem., Vol. 63, No. 15, 1998 5149
117.34, 133.89. Anal. Calcd for C7H13BrO2: C, 40.21; H, 6.27;
Br, 38.22. Found: C, 40.55; H, 6.27; Br, 38.48.
(CDCl3) δ 1.83 (br s, 1H), 3.45 (d, 2H, J ) 5.70 Hz), 3.63-3.85
(m, 3H), 4.05-4.23 (m, 2H), 5.20-5.38 (m, 2H), 5.87-6.01 (m,
1H); HRMS m/z calcd for C6H1079BrO, [M - OH]+ 176.9915,
found m/z 176.9914. Other spectral data were consistent with
those previously reported.43
3-(2-Br om o-1-a cetoxyeth oxy)p r op -1-en e (4b). Vinyl ac-
etate (3.4 mL, 36 mmol) was added dropwise to a stirred
solution of NBS (5.4 g, 31 mmol) and allyl alcohol (1.9 mL, 28
mmol) in CH2Cl2 (20 mL) at -20 °C, and the mixture was
stirred for 22 h. The reaction mixture was then warmed to
room temperature and stirred for a further 96 h. Workup of
the reaction mixture as described for 4a gave the bromide 4b
as a colorless oil (1.8 g, 29%); 1H NMR (CDCl3) δ 2.12 (s, 3H),
3.43 (dd, 1H, J ) 1.83, 5.07 Hz), 3.44 (dd, 1H, J ) 1.83, 5.07
Hz), 4.13 (dd, 1H, J ) 5.98, 11.96 Hz), 4.24 (dd, 1H, J ) 5.98,
11.96 Hz), 5.21-5.35 (m, 2H), 5.83-5.94 (m, 1H), 5.96 (t, 1H,
J ) 5.07 Hz); 13C NMR (CDCl3) δ 20.89, 31.17, 70.88, 95.05,
118.27, 132.97, 170.34. Anal. Calcd for C7H11BrO3: C, 37.69;
H, 4.97; Br, 35.82. Found: C, 37.63; H, 5.26; Br, 35.72.
(Z)-4-(2-Br om o-1-eth oxyeth oxy)-2-bu ten -1-ol (4c). Treat-
ment of NBS (4.7 g, 26 mmol) and (Z)-2-butene-1,4-diol (4.1
mL, 50 mmol) in CH2Cl2 (20 mL) with ethyl vinyl ether (2.4
mL, 25 mmol) as described for 4a gave the bromide 4c as a
4-[(2-Br om o-1-[(t r im et h ylsilyl)m et h yl]et h oxy]b u t -1-
en e (4h ). Allyltrimethylsilane (3.8 mL, 24 mmol) was added
to a light-protected solution of NBA (3.33 g, 24 mmol) and
3-butene-1-ol (2 mL, 23 mmol) in CH2Cl2 (20 mL), and the
mixture was stirred for 1 h and then warmed to room
temperature and stirred for a further 6 days. As the mixture
still contained starting material (TLC), further portions of
NBA (1.65 g, 12 mmol) and allyltrimethylsilane (1.9 mL, 12
mmol) were added. However, stirring of the reaction mixture
at room temperature for a further 2 days led to no further
consumption of the alcohol. The orange solution was then
poured onto brine (20 mL), and the organic layer was washed
with aqueous sodium thiosulfate (40 mL), dried (Na2SO4), and
carefully evaporated under ambient pressure. Distillation of
the residue under reduced pressure gave a lower boiling
fraction (1.7 g, bp ) 50-60 °C, 20 Torr) spectroscopically
identified as 3-buten-1-ol44 and a higher boiling fraction (612
mg, bp ) 120 °C, 20 Torr), flash chromatography of which (30:1
hexane/diethyl ether) gave the bromide 4h as a colorless oil
(191 mg, 6% based on reacted 3-buten-1-ol); 1H NMR (CDCl3)
δ 0.05 (s, 9H), 0.96 (d, 1H, J ) 1.86 Hz), 0.98 (d, 1H, J ) 3.70
Hz), 2.25-2.38 (m, 2H), 3.38-3.45 and 3.55-3.65 (m, 5H),
1
colorless oil (4.1 g, 69%); H NMR (CDCl3) δ 1.25 (t, 3H, J )
7.12 Hz), 1.81 (br s, 1H), 3.39 (d, 2H, J ) 5.50 Hz), 3.55-3.75
(m, 2H), 4.14-4.30 (m, 4H), 4.73 (t, 1H, J ) 5.5 Hz), 5.65-
5.75 and 5.82-5.93 (m, 2H); 13C NMR (CDCl3) δ 15.73, 32.05,
59.08, 62.59, 62.84, 101.40, 128.07, 133.21; HRMS m/z calcd
for C6H1081BrO2, [M - C2H5O]+ 194.9844, found m/z 194.9843.
(E)-1-P h en yl-3-(2-b r om o-1-et h oxyet h oxy)p r op -1-en e
(4d ). A solution of cinnamyl alcohol (1.7 mL, 13 mmol,
recrystallized), NBS (2.4 g, 13 mmol), and ethyl vinyl ether
(1.6 mL, 17 mmol) in CH2Cl2 (20 mL) was stirred for 6 h at
-20 °C and then for a further 16 h at room temperature. The
usual workup gave the bromide 4d as a colorless oil42 (2.1 g,
3.53-3.75 (m, 4H), 5.01-5.13 (m, 2H), 5.75-5.90 (m, 1H); 13
C
NMR (CDCl3) δ -0.89, 22.00, 34.39, 37.62, 68.54, 77.01, 116.39,
135.01; HRMS m/z calcd for C9H19OSi, [M - CH2Br]+ 171.1205,
found m/z 171.1206. The experiment was repeated several
times in order to obtain sufficient quantities of 4h for use in
subsequent reactions.
1
58%); H NMR (CDCl3) δ 1.26 (t, 3H, J ) 7.15 Hz), 3.42 (d,
2H, J ) 5.5 Hz), 3.56-3.78 (m, 2H), 4.30 (qd, 2H, J ) 13.35,
6.0 Hz), 4.79 (t, 1H, J ) 5.5 Hz), 6.29 (dt, 1H, J ) 15.9, 6.0
Hz), 6.65 (d, 1H, J ) 15.9 Hz), 7.20-7.41 (m, 5H); 13C NMR
(CDCl3) δ 15.13, 31.65, 62.34, 67.21, 100.75, 125.07, 126.41,
3-[2-(P h en ylselen yl)-1-b u t ylet h oxy]p r op -1-en e (4i).
NaBH4 (220 mg, 5.7 mmol) was added portionwise to a
vigorously stirred solution of diphenyl diselenide (890 mg, 2.9
mmol) in ethanol (12 mL) at 0 °C as previously described.26
After the evolution of gas had ceased, a solution of 1,2-
epoxyhexane (500 mg, 5 mmol) in ethanol (2 mL) was added
portionwise, and the reaction mixture was stirred at room
temperature for an additional 2 h. The solvent was then
removed under reduced pressure, and the residual oil was
poured into water and ethyl acetate. The layers were sepa-
rated, and the aqueous phase was extracted with two further
portions of ethyl acetate. The combined organic layers were
washed with brine, dried (Na2SO4), and concentrated under
reduced pressure. Flash chromatography of the residue (10:1
hexane/diethyl ether) furnished 1-(phenylselenyl)hexan-2-ol as
127.72, 128.47, 132.69, 136.36; HRMS m/z calcd for C13H17
79BrO2, [M]+• 284.0412, found m/z 284.0413. Anal. Calcd for
13H17BrO2: C, 54.75; H, 6.01; Br, 28.02. Found: C, 54.45;
-
C
H, 6.43; Br, 27.95.
1-Br om o-2-(2-cycloh exylid en eeth oxy)-2-eth oxyeth a n e
(4e). Treatment of NBS (470 mg, 4.2 mmol) and cyclohex-
ylideneethanol23 (0.5 g, 4.0 mmol) in carbon tetrachloride (20
mL) with ethyl vinyl ether (0.5 mL, 5.1 mmol) as described
for 4a gave the bromide 4e as a colorless oil (809 mg, 98%);
1H NMR (CDCl3) δ 1.24 (t, J ) 7.08 Hz, 3H), 1.56-1.70 (m,
6H), 2.17-2.30 (m, 4H), 3.38 (d, 2H, J ) 5.55 Hz), 3.45-3.70
(m, 2H), 4.15-4.32 (dq, 2H, J ) 11.63, 7.23 Hz), 4.82 (t, 1H, J
) 5.52 Hz), 5.59 (t, 1H, J ) 7.05 Hz); 13C NMR (CDCl3) δ 15.10,
26.53, 27.65, 28.22, 28.88, 31.81, 36.95, 62.06, 62.30, 100.55,
116.75, 145.61. Anal. Calcd for C12H21BrO2: C, 52.00; H, 7.64;
Br, 28.83. Found: C, 52.36; H, 7.89; Br, 28.74.
4-(2-Br om o-1-eth oxyeth oxy)bu t-1-en e (4f). Treatment
of NBS (3.0 g, 17 mmol) and 3-buten-1-ol (1.5 mL, 17 mmol)
with ethyl vinyl ether (1.8 mL, 19 mmol) in CH2Cl2 as
described above for 4a gave the bromide 4f as a colorless oil
(2.8 g, 76%); 1H NMR (CDCl3) δ 1.23 (t, 3H, J ) 7.14 Hz), 2.32-
2.40 (qt, 2H, J ) 6.71, 1.53 Hz), 3.37 (d, 2H, J ) 5.56 Hz),
3.53-3.75 (m, 4H), 4.68 (t, 1H, J ) 5.50 Hz), 5.03-5.15 (m,
2H), 5.76-5.92 (m, 1H); 13C NMR (CDCl3) δ 15.03, 31.53, 33.99,
62.36, 65.87, 101.37, 116.59, 134.72; HRMS m/z (rel intensity)
C6H1079BrO, [M - EtO]+ 176.9915, found m/z 176.9914.
3-[2-Br om o-1-(h yd r oxym eth yl)eth oxy]p r op -1-en e (4g).
An ice-chilled solution of NBS (13 g, 73 mmol) in allyl alcohol
(30 mL, 0.44 mol) was treated with 1-hexene (10 g, 73 mmol),
stirred for 2 h, and worked up as previously described.24 Flash
chromatography (3:1 hexane/diethyl ether) of the crude prod-
uct, which contained at least five components (TLC), furnished
the liquid bromo-alcohol 4g (376 mg, 3% based on total
available NBS) as the only separable compound;43 1H NMR
1
a pale yellow oil (718 mg, 56%); H NMR (CDCl3) δ 0.88 (t, J
) 7.1 Hz, 3H), 1.23-1.60 (m, 6H), 2.90 (br s, 1H), 3.15 (br s,
1H), 3.63-3.71 (m, 1H), 7.28 (m, 3H), 7.55 (m, 2H).
Potassium hydride (504 mg, 33%, 4.2 mmol) and allyl
bromide (1 g, 720 µL, 8.4 mmol) were added to a chilled
solution (0 °C) of 1-(phenylselenyl)hexan-2-ol (720 mg, 2.8
mmol) in dry THF (10 mL). The reaction mixture was then
stirred at room temperature for 19 h, further portions of
potassium hydride (1.5 equiv) and allyl bromide (3 equiv) were
added, and stirring was continued for a further 25 h after
which time all of the alcohol had been consumed (TLC).
2-Propanol was then cautiously added to destroy the excess
of potassium hydride, and the solvent was removed under
reduced pressure to give an oily residue which was worked-
up in the same manner as that described above for 1-phen-
ylselenylhexan-2-ol. However, flash chromatography failed to
give a pure product which was best obtained by preparative
TLC. A combination of four, 200-250 mg loadings on 200 ×
200 × 2 mm plates (30:1 hexane/diethyl ether), afforded the
1
selenide 4i as a yellow oil (294 mg, 35%); H NMR (CDCl3) δ
0.88 (t, J ) 7.08 Hz, 3H), 1.20-1.45 (m, 4H), 1.55-1.70 (m,
(44) The Aldrich Library of 13C and 1H FT-NMR Spectral Data;
Behnke, J ., Pouchert, C. J ., Ed.s; The Aldrich Chemical Co., Inc.:
Milwaukee, 1993; Vol. 1, p 208B. The Aldrich Library of FT-IR Spectral
Data; Pouchert, C. J . Ed.; The Aldrich Chemical Co., Inc.: Milwaukee,
1985; Vol. 1, p 139A.
(42) Iqbal, J .; Srivastava, R. R.; Gupta, K. B.; Khan, M. A. Synth.
Commun. 1989, 19, 901.
(43) Srikrishna, A.; Krishnan, K. J . Org. Chem. 1989, 54, 3981.

5150 J . Org. Chem., Vol. 63, No. 15, 1998
Beckwith and Page
2H), 2.99 (dd, 1H, J ) 12.3, 6.3 Hz), 3.11 (dd, 1H, J ) 12.3,
5.3 Hz), 3.52 (quintet, 1H, J ) 5.55 Hz), 3.91-4.06 (qd, 2H, J
) 12.48, 5.94 Hz), 5.13 (dm, 1H, J ) 10.32 Hz), 5.22 (dm, 1H,
J ) 15.63 Hz), 5.82-5.95 (m, 1H), 7.23 (m, 3H), 7.52 (m, 2H);
13C NMR (CDCl3) δ 13.93, 22.60, 27.40, 32.26, 33.87, 70.31,
78.32, 116.82, 126.68, 128.22, 128.90, 132.51, 134.93. Anal.
Calcd for C15H22OSe: C, 60.60; H, 7.46. Found: C, 60.65; H,
7.52.
DCC (1.26 g, 6.1 mmol) in the usual way,45 afforded an intense
yellow solution of the corresponding thiohydroxamic ester. This
solution was degassed and heated at reflux while tributyl-
stannane (3.3 mL, 12.3 mmol) and AIBN (10 mg, 61 µmol) in
dry benzene (15 mL) were added over a period of 1.5 h.
Refluxing of the solution was continued for 16 h after which
time the reaction was complete (TLC). The excess of stannane
was destroyed by the cautious addition of CCl4 (10 mL), and
the solution was refluxed for 1 h before being cooled and
concentrated under reduced pressure. Flash chromatography
(10:1 hexane/diethyl ether) of the crude product gave the THP-
ether of 3-hydroxy-1-methylcyclopentane in low yield (200 mg,
27%). The THP ether (200 mg, 1.1 mmol) was then added to
a stirred solution of acidified Dowex AG 50W-X2 resin (230
mg) in methanol (5 mL) and water (5 drops).46 The solution
was heated at 45 °C for 2 h, filtered through Celite, and poured
onto a water/diethyl ether bilayer. The layers were separated,
and the aqueous phase was extracted with two further portions
of diethyl ether. The combined organic layers were washed
with brine, dried (Na2SO4), and concentrated to a total volume
of ca. 5 mL under reduced pressure. The resultant solution
of 3-hydroxy-1-methylcyclopentane was subjected directly to
etherification with ethyl iodide in the manner described in the
preceding experiment to give an authentic sample of cis-6j (ca.
10 mg, 7%) with the spectral features specified above.
3-(1-Acet oxyet h oxy)p r op -1-en e (7b). Allyl vinyl ether
(1.68 g, 20 mmol) was added dropwise over ca. 5 min to a
vigorously stirred solution of p-toluenesulfonic acid (10 mg,
0.05 mmol) and acetic acid (1.14 mL, 20 mmol) in diethyl ether
(50 mL) at 5 °C. Stirring was continued at room temperature
for an additional 3.5 h and the mixture was then poured onto
a mixture of 1:1 brine and aqueous sodium bicarbonate (20
mL) and re-extracted. After washing of the ethereal solution
for a second time in with same mixture (20 mL), the organic
layer was separated, dried (Na2SO4), and concentrated by
careful distillation under ambient pressure. Short-path (Kugel-
rohr) distillation under reduced pressure (bp ) 70-80 °C, 22
Torr) afforded solely acetal 7b as a colorless oil (950 mg, 33%);
1H NMR (CDCl3) δ 1.41 (d, 3H, J ) 5.01 Hz), 2.07 (s, 3H),
3.96-4.26 (m, 2H), 5.17-5.21 (dm, 1H, J ) 10.32 Hz), 5.25-
5.32 (dm, 1H, J ) 17.28 Hz), 5.80-5.98 (m and superimposed
q, 2H, J ) 3.17 Hz); 13C NMR (CDCl3) δ 20.65, 21.13, 69.86,
95.58, 117.39, 133.55, 170.60. Anal. Calcd for C7H12O3: C,
58.32; H, 8.39. Found: C, 58.30; H, 8.57.
Ch lor o(2-eth oxyh ex-5-en yl)m er cu r y (4j). Following the
literature procedure,27 a solution of 1,5-hexadiene (600 mg, 7.3
mmol) in ethanol (1.2 mL) was added dropwise to a vigorously
stirred suspension of mercuric acetate (776 mg, 2.4 mmol) in
ethanol (7 mL) at 0 °C. The mixture was then warmed to room
temperature over a period of 20 min after which time a test
for mercuric oxide was negative.27 The reaction mixture was
then poured into aqueous sodium chloride (10%, 10 mL) and
extracted with chloroform (10 mL). The chloroform layer was
washed with water (2 × 10 mL) and evaporated under reduced
pressure to give a white solid mixture (937 mg) of the mercuric
chloride 4j and the two diastereomers of dichloro[µ-(2,5-
diethoxy-1,6-hexanediyl)]dimercury (5) in a ratio of 2:1 (1H
NMR). The solid was then added to boiling ethanol (6 mL).
Filtration of the hot suspension furnished 5 as a white powder;
1H NMR (CDCl3) δ 1.20 (dt, 6H, J ) 6.99, 1.25 Hz), 1.40-1.78
(br m, 4H), 2.13-2.19 (dt, 2H, J ) 12.30, 3.38 Hz), 2.39-2.47
(t, 1H, J ) 12.12 Hz), 2.43 (superimposed dd on t, 1H, J )
10.40, 1.83 Hz), 3.38-3.50 and 3.52-3.63 (m, 4H), 3.68-3.78
(m, 2H); 13C NMR (CDCl3) δ 15.64, 34.80, 35.55, 38.58, 38.79,
64.08, 64.20, 77.54, 77.73; IR (Polyethylene disk) 312 (Hg-
Cl) cm-1. Anal. Calcd for C10H20Cl2Hg2O2: C, 18.64; H, 3.13,
Hg, 62.26. Found: C, 18.47; H, 2.91; Hg, 62.41. Evaporation
of the filtrate under reduced pressure furnished the mono-
mercurial 4j as a pale yellow oil (90 mg, 10%).
Repetition of the above experiment with 6 mol equiv of diene
furnished the desired product 4j (52%); 1H NMR (CDCl3) δ 1.19
(t, 3H, J ) 7.05 Hz), 1.42-1.53 and 1.65-1.78 (m, 2H), 2.09-
2.19 (m and superimposed dd, 3H, J ) 12.18, 4.25 Hz), 2.42
(1H, dd, J ) 12.12, 5.16 Hz), 3.40-3.52 (m, 2H), 3.68-3.78
(m, 1H), 4.96-5.06 (m, 1H), 5.72-5.87 (m, 1H); 13C NMR
(CDCl3) δ 15.63, 29.77, 38.08, 38.75, 64.02, 77.00, 115.01,
137.81. Anal. Calcd for C8H15ClHgO: C, 26.45; H, 4.16; Cl,
9.76. Found: C, 26.13; H, 4.24; Cl, 9.56.
cis- a n d tr a n s-1-Eth oxy-3-m eth yl-cyclop en ta n e (6j).
Potassium hydride (1.2 g, 33%, 10.8 mmol), washed with
petroleum spirit (bp ) 100-120 °C), was suspended in dry
diethyl ether (10 mL) under nitrogen gas while being added
portionwise to a vigorously stirred, chilled solution (0 °C) of a
mixture of cis- and trans-3-methylcyclopentanol (830 mg, 8.3
mmol) and HMPA (800 µL, 5.4 mmol) in dry diethyl ether (10
mL). Cooling of the solution was maintained until the evolu-
tion of hydrogen gas had subsided, and freshly distilled ethyl
iodide (2 mL, 24.9 mmol) was then added while the mixture
was slowly warmed to room temperature. Additional portions
of potassium hydride (2 × 0.5 equiv) and ethyl iodide (2 × 1.5
equiv) were added after 24 and 72 h, respectively. After 96 h,
the reaction mixture was worked up as described for the
preparation of substrate 4i, and the oily residue was im-
mediately subjected to flash chromatography. Elution with
15:1 pentane/diethyl ether gave a mixture of cis-6j and trans-
4-[2-[(Tr im et h ylsilyl)m et h yl]et h oxy]b u t -1-en e (7h ).
3-Buten-1-ol (1 mL, 11.5 mmol) and concentrated hydrochloric
acid (ca. 5 drops) were dissolved in CH2Cl2 (20 mL) at 0 °C
and vigorously stirred while allyltrimethylsilane (1.85 mL, 11.5
mmol) in CH2Cl2 (20 mL) was added dropwise over 30 min
during which time the reaction mixture was slowly warmed
to room temperature. After 1.5 h, the presence of a new
compound (Rf ) 0.55, 15:1 hexane/diethyl ether) was detected
by TLC. Further catalytic quantities of concentrated hydro-
chloric acid (ca. 5 drops) were added but had no effect in the
hastening of the reaction, nor did addition of acetic acid (5
drops) and refluxing of the mixture for a further 5 days. The
reaction was therefore poured onto saturated sodium bicarbon-
ate solution (20 mL), and the organic layer was separated,
dried (Na2SO4), and carefully evaporated under reduced pres-
sure. Flash chromatography (30:1 hexane/diethyl ether) of the
crude residue gave 7h as a colorless oil (200 mg, 10%); 1H NMR
(CDCl3) δ 0.02 (s, 9H), 0.74 (dd, 1H, J ) 14.46, 7.38 Hz), 1.01
(dd, 1H, J ) 14.47, 6.65 Hz), 1.17 (d, 3H, J ) 6.04 Hz), 2.31
(q, 2H, J ) 6.84 Hz), 3.30-3.39 (dt, 1H, J ) 7.08, 7.02 Hz),
1
6j (1.8:1) as a colorless oil (1.66 g, 60%); H NMR (CDCl3) δ
0.95-1.33 (m, 2 superimposed d, and 2 superimposed t, 15H,
J ) 6.65, 6.59, 7.40, 7.40 Hz), 1.55-2.00 (m, 9H), 2.05-2.17
(quintet, 2H, J ) 6.7 Hz), 3.38-3.47 (q, 4H, J ) 7.0 Hz), 3.83-
3.95 (m, 2H); 13C NMR (CDCl3) δ cis-isomer: 15.47, 20.72,
32.08, 32.17, 32.57, 41.37, 64.08, 80.99; trans-isomer: 15.47,
20.53, 32.10, 32.50, 32.66, 40.95, 63.78, 80.93; HRMS m/z calcd
for C8H16O, [M]+• 128.1201, found m/z 128.1201. Anal. Calcd
for C8H16O: C, 74.94; H, 12.58%. Found: C, 74.63; H, 13.00%.
3.45-3.59 (m, 2H), 4.98-5.12 (m, 2H), 5.75-5.90 (m, 1H); 13
C
NMR (CDCl3) δ -0.83, 22.47, 25.71, 34.62, 67.24, 73.54, 115.98,
(45) Barton, D. H. R.; Crich, D.; Motherwell, W. B. Tetrahedron
1985, 41, 3901. Crich, D.; Quintero, L. Chem. Rev. 1989, 89, 1413.
Barton, D. H. R.; da Silva, E.; Zard, S. Z. J . Chem. Soc., Chem.
Commun. 1988, 285. Crich, D.; Ritchie, T. J . J . Chem. Soc., Chem.
Commun. 1988, 1461. Alexanian, V.; Hassner, A. Tetrahedron Lett.
1978, 46, 4475. Newcomb, M.; Glenn, A. G. J . Am. Chem. Soc. 1989,
111, 275.
cis-1-Eth oxy-3-m eth ylcyclop en ta n e (cis-6j). A degassed
solution of [3-(tetrahydropyranyl-2-oxy)cyclopent-1-yl]acetic
acid (928 mg, 4.1 mmol), prepared as previously described,30,31
in dry benzene (40 mL), when treated with N-hydroxypyridine-
2-thione (654 mg, 4.9 mmol), DMAP (745 mg, 6.1 mmol), and
(46) Beier, R.; Munday, B. P. Synth. Commun. 1979, 9, 271. Bongini,
A.; Cardillo, G.; Orena, M.; Sandri, S. Synthesis 1979, 618.

Oxygen-Containing Heterocycles by Radical Cyclization
J . Org. Chem., Vol. 63, No. 15, 1998 5151
135.52; HRMS m/z calcd for C9H19OSi, [M - Me]+ 171.1205,
found m/z 171.1206.
5-Eth oxyh ex-1-en e (7j). 1-Hexen-5-ol (1.0 g, 10 mmol),
prepared (65%) from 1-hexen-5-one as previously described47
was treated with potassium hydride (1.5 g, 33%, 13 mmol) and
ethyl iodide (2.4 mL, 30 mmol) as described above for 4i.
Concentration of the solution under reduced pressure afforded
J ) 7.10 Hz), 1.50-1.62 (m, 1H), 2.02 (dd, 1H, J ) 13.13, 7.56
Hz), 2.49 (sextet, 1H, J ) 7.24 Hz), 3.35-3.50 (m, 2H), 3.67-
3.77 (m, 1H), 4.04 (t, 1H, J ) 4.66 Hz), 5.12 (d, 1H, J ) 3.36
Hz); 13C NMR δ cis-isomer: 15.29, 17.15, 32.94, 40.82, 63.07,
73.22, 104.66; trans-isomer: 15.23, 18.38, 31.50, 41.08, 62.60,
73.95, 104.15. Anal. Calcd for C7H14O2: C, 64.58; H, 10.84.
Found: C, 64.64; H, 11.04.
a
yellow oil (733 mg, 57%), flash chromatography (15:1
(ii) In a variation of procedure A, when a solution of bromo
acetal 4a (22 mg, 0.1 mmol) in acetonitrile-d3 (1 mL) was
treated with tributylstannane (34 µL, 0.13 mmol) and heated
at reflux for 2 h, the ratio of cis- and trans-6a was 2.8:1.
An om er iza tion of cis- a n d tr a n s-2-Eth oxy-4-m eth yltet-
r a h yd r ofu r a n s (6a ). A solution of trifluoromethanesulfonic
acid in ethanol (10%, 10 µL) was added to solution of a 4.7:1
mixture of cis- and trans-6a (20 mg, 0.15 mmol) in CDCl3 (1
mL), and the mixture was kept at room temperature for 1 h
with occasional agitation. 1H NMR spectroscopy then indi-
cated the ratio of cis- and trans-6a to be 1:1.5. Experiments
in which the reaction time was 16 h or in which the mixture
was heated at 40 °C for 1 h gave the same cis:trans ratio.
cis- a n d tr a n s-4-Meth yltetr a h yd r ofu r a n -2-yl Aceta tes
(6b). Following procedure B, a solution of 4b (800 mg, 3.6
mmol) in dry pentane (36 mL) was treated with tributylstan-
nane (1.2 mL, 4.3 mmol) for 5 h. Short-path (Kugelrohr)
distillation of the concentrated residue furnished a colorless
oil (518 mg, 57%) which comprised cis- and trans-6b in a ratio
of 3.5:1. In addition, a small amount of the directly reduced
product 7b was also detected, but it could not be separated
from the cyclized material by flash chromatography. Integra-
tion of the appropriate signals (H2CdC for 7b and R1O[R]-
pentane/diethyl ether) of which gave 7j as a colorless oil (44
mg, 3%); 1H NMR (CDCl3) δ 1.13 (d, 3H, J ) 6.11 Hz), 1.19 (t,
3H, J ) 7.14 Hz), 1.40-1.55 (m, 1H), 1.55-1.68 (m, 1H), 2.03-
2.22 (m, 2H), 3.33-3.44 (m, 2H), 3.49-3.60 (m, 1H), 4.95 (dm,
1H, J ) 12.03 Hz), 5.02 (dm, 1H, J ) 17.27 Hz), 5.74-5.89
(m, 1H); 13C NMR (CDCl3) δ 15.54, 19.61, 29.75, 39.07, 63.49,
74.32, 114.28, 138.62; HRMS m/z calcd for C8H15O, [M - H]+
127.1123, found m/z 127.1123. Anal. Calcd for C8H16O: C,
74.94; H, 12.58. Found: C, 74.57; H, 13.02.
Gen er a l P r oced u r es for F r ee Ra d ica l Cycliza tion s.
General procedures for free-radical cyclizations are given
below. In each experiment the ratios of products, including
diastereoisomers, were determined by 1H NMR on both the
crude and isolated product mixtures. In some cases no
uncyclized products were detected under the conditions em-
ployed. Where no difference within experimental error was
detected, the ratio reported is that which was measured for
the purified mixture. In cases where differences occurred, both
sets of ratios are reported.
(A). AIBN (0.01 equiv) was added to a 0.2 M solution of
the halide or selenide (0.1-7.2 mmol) in dry benzene, and the
solution was degassed under a stream of dry nitrogen for 10
min. The solution was then heated at reflux, and tributyl-
stannane (1.2 equiv) was syringed into the flask in one portion.
Heating of the solution was continued until all of the starting
material was consumed (TLC). The solvent was then removed
either by distillation under ambient pressure or by evaporation
under reduced pressure for reactions that generated mixtures
containing less volatile products. If the former method was
used, further purification was effected by careful, short-path
(Kugelrohr) distillation under reduced pressure (ca. 15 Torr).
In cases where products were of low volatility, organostannane
residues were removed by a DBU workup procedure28 followed
by additional flash chromatography.
(B). This procedure is essentially the same as that specified
for procedure A, except that the solvent used was pentane.
Initiation was effected with bis-tert-butyl hyponitrite.
(C). A 0.2 M solution of the starting material (ca. 4 mmol)
in dry benzene was degassed under a stream of dry nitrogen
for 10 min and then heated at reflux while a solution of
tributylstannane (1.2 equiv) and AIBN (0.03 equiv) in dry
benzene (20 mL) was added slowly at the rate of 0.1 mL/min.
All subsequent monitoring and purification steps followed
procedure A.
1
CHOR2 for 6b) in the H NMR spectrum gave the ratio of 7b
to 6b of 1:5. This gave an adjusted yield of total cyclized
1
material (after distillation) of 250 mg (48%); H NMR (C6D6)
δ cis-isomer: 0.95 (d, 3H, J ) 6.65 Hz), 1.50-1.57 (m, 1H),
1.84 (s, 3H), 1.94-2.04 (m, 1H), 2.06-2.16 (m, 1H), 3.59 (t,
1H, J ) 8.05 Hz), 3.95 (t, 1H, J ) 8.05 Hz), 6.64-6.67 (m,
1H); trans-isomer: 0.82 (d, 3H, J ) 6.65 Hz), 1.36-1.46 (m,
1H), 1.86 (s, 3H), 1.94-2.04 (m, 1H), 2.28-2.42 (m, 1H), 3.35
(t, 1H, J ) 7.97 Hz), 4.14 (t, 1H, J ) 7.97 Hz), 6.64-6.67 (m,
1H); 13C NMR (CDCl3) δ cis-isomer: 17.51, 21.24, 32.14, 39.85,
75.12, 99.46, 170.5; trans-isomer: 17.11, 23.28, 30.82, 40.27,
75.43, 99.25, 176.91; HRMS: m/z calcd for C7H11O3, [M - H]+
143.0708, found m/z 143.0708.
Repetition of the reaction following procedure C with
heating for 24 h gave no 7b and the ratio of cis- to trans-6b
was 1.7:1. Partial decomposition of the cyclized products
occurred to form the hemiacetals 8 (1.8:1 in favor of the trans-
1
isomer); H NMR (C6D6) δ cis-isomer: 1.04 (d, 3H, J ) 6.47
Hz), 1.48-1.63 (m, 1H), 2.02-2.18 (m, 2H), 3.69 (t, 1H, J )
8.18 Hz), 3.96 (t, 1H, J ) 7.88 Hz), 5.59-5.61 (m, 1H); trans-
isomer: 0.89 (d, 3H, J ) 6.68 Hz), 1.37-1.46 (m, 1H), 2.02-
2.18 (m, 1H), 2.42-2.57 (m, 1H), 3.87 (t, 1H, J ) 7.48 Hz),
4.23 (t, 1H, J ) 7.63 Hz), 5.59-5.61 (m, 1H); 13C NMR (C6D6)
δ cis-isomer: 17.92, 33.73, 42.24, 73.81, 99.98. trans-isomer:
18.16, 31.92, 42.35, 74.45, 99.52.
(D). This procedure is essentially the same as A except that
the reaction mixture was irradiated at room temperature with
a quartz water-jacketed, 125 W medium-pressure mercury-
vapor lamp.
cis- a n d tr a n s-2-Eth oxy-4-(2-h yd r oxyeth yl)tetr a h yd r o-
fu r a n (6c). Following procedure D, a solution of bromo acetal
4c (1.0 g, 4.2 mmol) in dry benzene (21 mL) was treated with
tributylstannane (1.4 mL, 5.0 mmol) for 3 h. The concentrated
residue was then subjected to DBU workup28 and flash
chromatography (Rf ) 0.18, 1:1 hexane/diethyl ether) to afford
a pale yellow oil (473 mg, 71%) comprising the two tetrahy-
drofurans cis- and trans-6c in a ratio of 5.0:1. Repetition of
the experiment with the same quantities of reagents, but
following procedure A, afforded after 3.5 h, the same products
(296 mg, 44%) in a ratio of 3.3:1; 1H NMR (CDCl3) δ cis-
isomer: 1.20 (t, 3H, J ) 7.15 Hz), 1.53 (br s, 1H), 1.53-1.59
(m, 1H), 1.73 (q, 2H, J ) 6.59 Hz), 2.23-2.37 (m, 2H), 3.39-
3.49 (m, 1H), 3.54 (t, 1H, J ) 8.13 Hz), 3.63-3.70 (m, 2H),
3.69-3.79 (m, 1H), 3.99 (t, 1H, J ) 8.13 Hz), 5.11-5.14 (m,
1H); trans-isomer: 1.18 (t, 3H, J ) 7.14 Hz), 1.53 (br s, 1H),
1.60-1.69 (m, 1H), 1.73 (q, 2H, J ) 6.59 Hz), 2.02-2.10 (dd,
1H, J ) 12.7, 7.5 Hz), 2.48-2.58 (quintet, 1H, J ) 8.00 Hz),
3.39-3.49 (m, 1H), 3.54 (t, 1H, J ) 8.13 Hz), 3.63-3.70 (m,
2H), 3.69-3.79 (m, 1H), 4.08 (t, 1H, J ) 8.13 Hz), 5.11-5.14
cis- an d tr a n s-2-Eth oxy-4-m eth yltetr ah ydr ofu r an s (6a).
(i) Following procedure A, a solution of 4a 41 (1.5 g, 7.2 mmol)
in dry benzene (36 mL) was treated with tributylstannane (2.3
mL, 8.6 mmol) for 2 h. Short-path (Kugelrohr) distillation of
the concentrated residue furnished a colorless oil (114 mg,
12%) which comprised the two tetrahydrofurans cis- and trans-
6a 48 in a ratio of 4.7:1. An analogous experiment in benzene-
d6 gave a ratio between cis- and trans-6a of 4.2:1. Repetition
of the experiment with the same quantities of reagents but
following procedure B afforded, after 4 h, the same products
(570 mg, 61%) in
a
ratio of 5.8:1. 1H NMR (CDCl3) δ
cis-isomer: 1.07 (d, 3H, J ) 6.47 Hz), 1.19 (t, 3H, J ) 7.10
Hz), 1.40-1.50 (m, 1H), 2.16-2.33 (m, 2H), 3.35-3.50 (m, 2H),
3.67-3.77 (m, 1H), 3.92 (t, 1H, J ) 4.66 Hz), 5.11 (d, 1H, J )
3.24 Hz); trans-isomer: 1.02 (d, 3H, J ) 6.76 Hz), 1.17 (t, 3H,
(47) Luche, J . L. J . Am. Chem. Soc. 1978, 100, 2226.
(48) Parham, W. E.; Holmquist, H. E. J . Am. Chem. Soc. 1951, 73,
913.

5152 J . Org. Chem., Vol. 63, No. 15, 1998
Beckwith and Page
(m, 1H); 13C NMR (CDCl3) δ cis-isomer: 15.13, 35.06, 35.61,
38.60, 61.52, 62.93, 71.50, 104.04; trans-isomer: 15.09, 35.84,
36.42, 39.00, 61.52, 62.49, 71.50, 103.59. HRMS m/z calcd for
C8H15O3, [M - H]+ 159.1021, found m/z 159.1021.
(see below); 1H NMR (CDCl3) δ trans-6f: 0.95 (d, 3H, J ) 6.16
Hz), 1.29-1.50 (m, 3H), 1.35 (t, 3H, J ) 7.4 Hz), 1.87-1.95
(dm, 1H, J ) 13.1 Hz), 2.15-2.28 (m, 1H), 3.45-3.80 (m, 2H),
3.90-4.05 (m, 2H), 5.03 (d, 1H, J ) 4.40 Hz). Additional
spectral data were similar to those reported previously.32
The second fraction contained 2-ethoxyoxepane (9)29 as a
cis- an d tr a n s-2-Eth oxy-4-ben zyltetr ah ydr ofu r an s (6d).
Following procedure A, a solution of bromo acetal 4d 42 (710
mg, 2.5 mmol) in dry benzene (12.5 mL) was treated with
tributylstannane (0.74 mL, 2.7 mmol) for 55 h. Two further
aliquots of stannane (1.15 equiv and 0.73 equiv) together with
small amounts of AIBN (0.01 equiv) were added at 6.5 and 30
h, respectively. The concentrated residue was worked up with
DBU28 and flash chromatography (10:1:0.5 hexane/diethyl
ether/chloroform) to afford cis- and trans-6d 49 in a ratio of 2.1:1
as a colorless oil (512 mg, 44%); 1H NMR (CDCl3) δ cis-
isomer: 1.23 (t, 3H, J ) 7.1 Hz), 1.58-1.66 (ddd, 1H, J ) 13.3,
6.6, 3.0 Hz), 2.16-2.25 (ddd, 1H, J ) 13.3, 11.8, 5.6 Hz), 2.48
(septet, 1H, J ) 7.38 Hz), 2.77 (dd, 2H, J ) 8.1, 2.4 Hz), 3.38-
3.50 (m, 1H), 3.61 (t, 1H, J ) 8.3 Hz), 3.68-3.81 (m, 1H), 3.90
(t, 1H, J ) 8.4 Hz), 5.10-5.16 (m, 1H), 7.13-7.33 (m, 5H);
trans-isomer: 1.18 (t, 3H, J ) 7.1 Hz), 1.67-1.78 (m, 1H),
1.96-2.03 (ddd, 1H, J ) 12.9, 6.9, 1.1 Hz), 2.65-2.84 (m, 3H),
3.38-3.50 (m, 1H), 3.54-3.59 (dd, 1H, J ) 8.2, 5.8 Hz), 3.68-
3.81 (m, 1H), 3.95-4.01 (dd, 1H, J ) 8.3, 7.0 Hz), 5.10-5.16
(m, 1H), 7.13-7.33 (m, 5H); 13C NMR (CDCl3) δ cis-isomer:
15.23, 38.65, 39.07, 40.03, 62.98, 71.56, 104.27, 125.91, 125.99,
128.29, 128.47, 128.55, 140.69; trans-isomer: 15.05, 38.57,
38.98, 39.74, 62.60, 71.64, 103.80, 125.91, 125.99, 128.29,
128.47, 128.55, 140.50. Anal. Calcd for C13H18O2: C, 75.69;
H, 8.79. Found: C, 75.42; H, 9.11.
1
colorless oil (10 mg); H NMR (CDCl3) δ 1.30-1.37 (d, 3H, J
) 6.8 Hz), 1.40-1.96 (m, 7H), 2.11-2.20 (m, 1H), 3.48-3.58
(m, 1H), 3.64-3.69 (1H, dm, J ) 12.61 Hz, H-7a), 3.93-4.02
(t, 1H J ) 10.99 Hz), 4.00-4.02 (m, 2H), 4.89-4.93 (dd, 1H, J
) 8.55, 5.37 Hz); 13C NMR (CDCl3) δ 15.98, 23.49, 30.39, 31.55,
36.01, 61.77, 62.97, 102.22.
The third fraction contained exclusively cis-2-ethoxy-4-
methyltetrahydropyran (cis-6f)32 as a colorless oil (21 mg); 1H
NMR (CDCl3) δ 0.95 (d, 3H, J ) 6.16 Hz), 1.06-1.62 (m, 4H),
1.39 (t, 3H, J ) 6.96 Hz), 1.88-2.00 (1H, qt, J ) 8.55, 1.93
Hz), 3.36 (dt, 1H, J ) 11.48, 2.99 Hz), 3.60-3.70 (m, 1H), 4.05-
4.11 (ddd, 1H, J ) 11.60, 4.33, 2.19 Hz), 4.20-4.30 (m, 1H),
4.44 (dd, 1H, J ) 8.60, 2.32 Hz). Additional spectral data were
consistent with those reported previously.32
4-(1-Eth oxyeth oxy)bu t-1-en e (7f). Following procedure
B a solution of 4f (500 mg, 2.2 mmol) in dry pentane (11 mL)
was treated with tributylstannane (0.71 mL, 2.6 mmol) for 5
h. A DBU workup28 and flash chromatography (15:1 pentane/
diethyl ether) gave solely 7f as a colorless oil; 1H NMR (CDCl3)
δ 1.20 (t, 3H, J ) 5.28 Hz), 1.31 (d, 3H, J ) 6.80 Hz), 2.34 (qt,
2H, J ) 6.97, 2.09 Hz), 3.42-3.72 (m, 4H), 4.71 (q, 1H, J )
5.58 Hz), 5.02-5.15 (m, 2H,), 5.77-5.91 (m, 1H); 13C NMR
(CDCl3) δ 15.16, 19.66, 34.21, 60.63, 64.14, 99.34, 116.23,
135.21. HRMS m/z calcd for C8H15O2, [M - H]+ 143.1072,
found m/z 143.1073.
Repetition of the experiment with 4d (274 mg, 0.96 mmol)
by a variation of procedure B afforded, after 5 days, cis- and
trans-6d (131 mg, 66%) in a ratio of 2.7:1. In this experiment,
four further aliquots of stannane (0.35 equiv) together with
small amounts of AIBN (0.01 equiv) were added at 4.3, 6, 72,
and 96 h, respectively.
cis- a n d tr a n s-2-(H yd r oxym et h yl)-4-m et h ylt et r a h y-
d r ofu r a n (6g). Following procedure B, a solution of 4g (304
mg, 1.6 mmol) in dry pentane (10 mL) was treated with
tributylstannane (0.52 mL, 1.9 mmol) for 5 h. A DBU
workup28 and flash chromatography (Rf ) 0.50, diethyl ether)
gave cis and trans-4g in a ratio of 1:5.0 as a pale yellow oil
cis- a n d tr a n s-2-Eth oxy-4-cycloh exyltetr a h yd r ofu r a n s
(6e). Following procedure A, a solution of bromo acetal 4e
(150 mg, 0.5 mmol) in dry benzene (2.5 mL) was heated with
tributylstannane (161 µL, 0.6 mmol) for 2 h. A DBU workup28
and flash chromatography (Rf ) 0.31, 15:1 hexane/diethyl
ether) gave cis- and trans-6e in a ratio of 1.7:1 as a pale yellow
oil (65 mg, 61%); 1H NMR (CDCl3) δ cis-isomer: 0.85-0.98
(m, 2H), 1.16-1.31 (m and superimposed t, 4 + 3H, J ) 7.08
Hz), 1.44-1.70 (m, 6H), 1.80-1.95 (sextet, 1H, J ) 7.9 Hz),
2.20-2.30 (m, 1H), 3.35-3.55 (m and t, 2H, J ) 8.06 Hz),
3.55-3.78 (m, 1H), 3.92 (t, 1H, J ) 7.76 Hz), 5.07-5.13 (m,
1H); trans-isomer: 0.85-0.98 (m, 2H), 1.16-1.31 (m and
superimposed t, 4 + 3H, J ) 7.08 Hz), 1.48-1.70 (m, 6H),
1.95-2.03 (dd, 1H, J ) 12.5, 7.2 Hz), 2.15-2.30 (m, 1H), 3.35-
3.55 (m and t, 2H, J ) 8.12 Hz), 3.55-3.78 (m, 1H), 4.05 (t,
1H, J ) 8.12 Hz), 5.07-5.13 (m, 1H); 13C NMR (CDCl3) δ cis-
isomer: 16.00, 27.13, 38.21, 63.51, 70.65, 105.11; trans-
isomer: 15.96, 27.08, 38.08, 62.92, 71.90, 104.77. The remain-
ing signals could not be assigned to a particular isomer: 26.71,
26.77, 26.81, 32.28, 32.62, 32.79, 32.87, 41.88, 42.64, 43.74,
45.99. Anal. Calcd for C12H22O2: C, 72.68; H, 11.18. Found:
C, 72.67; H, 11.50.
Redu ction of 4-(2-Br om o-1-eth oxyeth oxy)bu t-1-en e (4f).
In a variation of procedure A, a solution of 4f (980 mg, 4.4
mmol) in dry benzene (72 mL) when heated with tributylstan-
nane (1.4 mL, 5.3 mmol) for 18 h gave a complex mixture of
products (1H NMR), two of which were tentatively identified
as being the diastereomeric 2-ethoxy-4-methyltetrahydropy-
rans 6f,32 while the detection of olefinic resonances suggested
that the uncyclized product 7f had also been formed. The
residue was then subjected to a DBU workup procedure28
followed by repetitive flash chromatography (15:1 pentane/
diethyl ether).
1
(131 mg, 72%); H NMR (CDCl3) δ cis-isomer: 1.05 (d, 3H, J
) 6.74 Hz), 1.21-1.32 (m, 1H), 2.02-2.11 (quintet, 1H, J )
6.80 Hz), 2.24-2.41 (m, 1H), 3.30-3.35 (dd, 1H, J ) 8.28, 7.12
Hz), 3.47-3.55 (dd, 1H, J ) 11.53, 6.22 Hz), 3.66-3.71 (dd,
1H, J ) 11.50, 3.29 Hz), 3.90-3.96 (t, 1H, J ) 8.24 Hz), 3.96-
4.10 (m, 1H); trans-isomer: 1.04 (d, 3H, J ) 6.74 Hz), 1.53-
1.61 (dd, 1H, J ) 12.47, 6.66 Hz), 1.83-1.90 (dd, 1H, J ) 12.49,
6.07 Hz), 2.24-2.41 (octet, 1H, J ) 6.80 Hz), 3.31-3.36 (dd,
1H, J ) 8.28, 7.12 Hz), 3.43-3.49 (dd, 1H, J ) 11.53, 6.22
Hz), 3.61-3.67 (dd, 1H, J ) 11.50, 3.29 Hz), 3.96-4.01 (dd,
1H, J ) 8.24, 6.65 Hz), 4.08-4.17 (m, 1H); 13C NMR (CDCl3)
δ 17.59, 33.64, 35.34, 65.07, 75.11, 78.81; HRMS m/z calcd for
C6H11O2, [M - H]+ 115.0759, found m/z 115.0754.
R ed u ct ion of 4-[2-Br om o-1-[(t r im et h ylsilyl)m et h yl-
]eth oxy]bu t-1-en e (4h ). In a variation of procedure A,
treatment of 4h (430 mg, 1.6 mmol) with tributylstannane
(0.52 mL, 1.9 mmol) in dry benzene (81 mL) for 3 h gave a
complex mixture of products (1H NMR). Two of these were
tentatively identified as cis- and trans-6h , while the presence
of olefinic signals suggested that 7h had also been formed.
DBU workup28 followed by repetitive flash chromatography
(30:1 pentane/diethyl ether) gave as the first fraction a mixture
of cis-6h (Rf ) 0.40) with 7h (Rf ) 0.50) which was identified
by comparison with an authentic specimen (see above). The
second fraction gave exclusively cis-2-[(trimethylsilyl)methyl]-
4-methyltetrahydropyran (cis-6h ) as a colorless oil (144 mg);
1H NMR (CDCl3) δ 0.02 (s, 9H), 0.74 (dd, 1H, J ) 14.56, 7.08
Hz), 0.90 (dd, 1H, J ) 13.52, 6.11 Hz), 0.91 (d, 3H, J ) 6.32
Hz), 1.00-1.85 (m, 5H), 3.29-3.42 (m, 2H), 3.94 (ddd, 1H, J
) 11.26 Hz, J ) 4.40, 1.16 Hz); 13C NMR (CDCl3) δ -0.90,
22.27, 25.23, 30.45, 34.32, 43.62, 67.86, 75.68; HRMS m/z calcd
for C9H19OSi, [M - Me]+ 171.1205, found m/z 171.1206.
The third fraction gave exclusively trans-2-[(trimethylsilyl)-
methyl]-4-methyltetrahydropyran (trans-6h ) as a colorless oil
(19 mg); 1H NMR (CDCl3) δ 0.02 (s, 9H), 0.74 (dd, 1H, J )
14.40, 7.30 Hz), 0.97 (dd, 1H, J ) 14.40, 7.30 Hz), 1.03 (d, 3H,
J ) 7.08 Hz), 1.16-1.27 (m, 1H), 1.38-1.44 (dm, 1H, J ) 13.30
Hz), 1.68-1.80 (octet, 1H, J ) 4.54 Hz), 1.93-2.04 (m, 1H),
The first fraction, a colorless oil (112 mg), contained a 1:2
mixture of trans-2-ethoxy-4-methyltetrahydropyran 6f6 and 7f
which was identified by comparison with an authentic sample
(49) Srikrishna, A.; Nagaraju, S.; Sharma, G. V. R. Synth. Commun.
1992, 22, 1127.

Oxygen-Containing Heterocycles by Radical Cyclization
J . Org. Chem., Vol. 63, No. 15, 1998 5153
3.58-3.82 (m, 3H); 13C NMR (CDCl3) δ -0.92, 19.00, 23.41,
24.85, 32.29, 40.37, 61.97, 70.17.
hydride (150 mg each, 1.2 mmol) were added at 6 and 48 h,
respectively, in an attempt to hasten the consumption of the
starting material. Stirring of the mixture was continued at
room temperature for 3 days during which time metallic
mercury formed. The mixture was then filtered through
Celite, and the solvent was removed by spinning band distil-
lation to give a mixture of cis-6j, trans-6j, and 7j (1.1:1.0:1.3)
as a pale yellow oil, the components of which were identified
by comparison with authentic samples (see above). Repetition
of the experiment in CDCl3 confirmed that ratios above were
unchanged by distillation of the crude product mixture.
cis- a n d tr a n s-2-Bu tyl-4-m eth yltetr a h yd r ofu r a n (6i).
Following procedure B, 4i (520 mg, 1.8 mmol) in dry pentane
(8.7 mL) was treated with tributylstannane (0.58 mL, 2.2
mmol) for 5 h. Short-path (Kugelrohr) distillation (bp ) 107
°C, 15 Torr) of the concentrated residue furnished a mixture
of cis- and trans-6i50 (1:5) as a colorless oil (124 mg, 50%); 1H
NMR (CDCl3) δ cis-isomer: 0.90 (t, 3H, J ) 6.7 Hz), 1.03 (d,
3H, J ) 6.7 Hz), 1.07 (m, 1H), 1.21-1.72 (m, 6H), 2.08-2.17
(m, 1H), 2.22-2.36 (m, 1H), 3.30 (t, 1H, J ) 7.95 Hz), 3.78 (t,
1H, J ) 7.95 Hz), 3.74-3.83 (m, 1H); trans-isomer: 0.90 (t,
3H, J ) 6.7 Hz), 1.01 (d, 3H, J ) 6.7 Hz), 1.21-1.72 (m, 8H),
2.22-2.36 (m, 1H), 3.17 (dd, 1H, J ) 8.3, 6.9 Hz), 3.80-3.87
(m, 1H), 3.90 (dd, 1H, J ) 8.3, 6.9 Hz); 13C NMR (CDCl3) δ
cis-isomer: 13.96, 17.86, 22.68, 28.44, 35.73, 34.21, 40.82,
74.26, 80.18; trans-isomer: 13.90, 18.01, 22.68, 28.37, 35.73,
33.08, 39.53, 74.72, 78.71. Anal. Calcd for C9H18O: C, 76.00;
H, 12.75. Found: C, 75.99; H, 12.58.
Red u ction of Ch lor o(2-eth oxyh ex-5-en yl)m er cu r y (4j)
with Sodiu m Tr im eth oxybor oh ydr ide. Sodium trimethoxy-
borohydride (200 mg, 1.55 mmol) was suspended in dry CH2-
Cl2 (15 mL), and the mixture was degassed under nitrogen
for ca. 10 min. In a separate flask, the mono-mercurial 4j (470
mg, 1.3 mmol) was dissolved in dry CH2Cl2 (1 mL), and the
resulting solution was similarly degassed. The mercurial
solution was then added via a syringe to the vigorously stirred
borohydride suspension. Two additional portions of the boro-
Ack n ow led gm en t. We thank Mr. R. Longmore for
the preparation of tributylstannane and for other tech-
nical assistance. One of us (D.M.P.) gratefully acknowl-
edges the award of a Ph.D. Scholarship from the
Australian National University. This paper was written
at the University of Freiburg during the tenure of a von
Humboldt Senior Award to A.L.J .B., the receipt of which
is gratefully acknowledged.
Su p p or tin g In for m a tion Ava ila ble: IR and MS data for
compounds 4a -j, 5, 6a -j, 7b, 7f, 7h , 7j, 8, and 9. Copies of
1H and 13C NMR spectra for compounds 4c, 4f, 4g, 4h , 6b, 6c,
6f, 6g, 6h , 7f, 7h , 8, and 9 (33 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
(50) Bel’skii, I. F.; Minashkina, Z. K. Bull. Acad. Sci. USSR (Engl.).
1970, 10, 2195.
J O980359U