The Scope and Limitations of Intramolecular Radical Cyclizations of Acylsilanes with Alkyl, Aryl, and Vinyl Radicals

Article abstract of DOI:10.1021/jo9711302

5-Exo cyclizations of primary and secondary radicals with acylsilanes successfully give cyclopentyl silyl ethers. The corresponding 6-exo cyclizations are sensitive to changes of the size of silyl groups. Secondary radicals undergo 6-exo cyclizations with

Full text of DOI:10.1021/jo9711302

J . Org. Chem. 1997, 62, 9089-9098
9089
Th e Scop e a n d Lim ita tion s of In tr a m olecu la r Ra d ica l Cycliza tion s
of Acylsila n es w ith Alk yl, Ar yl, a n d Vin yl Ra d ica ls
Sheng-Yueh Chang, Weir-Torn J iaang, Chaur-Donp Cherng, Kuo-Hsiang Tang,
Chih-Hao Huang, and Yeun-Min Tsai*
Department of Chemistry, National Taiwan University, Taipei, Taiwan 106, Republic of China
Received J une 20, 1997X
5-Exo cyclizations of primary and secondary radicals with acylsilanes successfully give cyclopentyl
silyl ethers. The corresponding 6-exo cyclizations are sensitive to changes of the size of silyl groups.
Secondary radicals undergo 6-exo cyclizations with acylsilanes more slowly. Reaction of aryl radical
with acylsilane proceeds well for 5-exo cyclization but not for 6-exo cyclization. Vinyl radicals give
best results in 5-exo cyclizations with acylsilanes but give low yields (∼30%) in 6-exo cyclizations.
Intramolecular cyclizations of vinyl radicals with acylsilanes give enol silyl ethers regiospecifically.
Sch em e 1
In tr od u ction
Radical addition to carbon-carbon multiple bonds is
an important strategy in the construction of carbocycles
or heterocycles.1 A carbon-carbon double bond is the
most often used radical acceptor in this type of strategy.
When a carbon-carbon double bond is used for cycliza-
tion, the cyclic structure obtained is always accompanied
by an additional carbon appendage. This may sometimes
be an unwanted feature. Although the carbonyl chem-
istry constitutes a major part of the organic chemistry,
it was not until recently that the radical addition to
carbonyls attracted the attention of synthetic organic
chemists.2-4 When the carbonyl is used as the radical
acceptor, a cyclic alcohol is obtained. This alcohol does
not have the extra carbon appendage on the ring as in
the olefin system mentioned above. In addition, the
resulting hydroxyl group can be transformed to other
useful functional groups very easily. Therefore, the
radical carbonyl cyclization reaction is a potentially
useful process.2
X Abstract published in Advance ACS Abstracts, December 1, 1997.
(1) (a) Fossey, J .; Lefort, D.; Sorba, J . Free Radicals in Organic
Chemistry; Wiley: New York, 1995. (b) Giese, B.; Kopping, B.; Go¨bel,
T.; Dickhaut, J .; Thoma, G.; Kulicke, K. J .; Trach, F. Org. React. 1996,
48, 303.
As shown in Scheme 1,5 Beckwith’s kinetic data reveal
that the k1 value for 5-exo ring closure of 4-formylbutyl
radical (1) is 8.7 × 105 s-1 at 80 °C. This value is
comparable with the value of 1.4 × 106 s-1 (80 °C) for
5-exo ring closure of 5-hexenyl radical.5b,6,7 In the case
of 5-formylpentyl radical (3), k2 is 1.0 × 106 s-1 (80 °C),
which is faster than that for 6-exo ring closure of
6-heptenyl radical (4.3 × 104 s-1 at 80 °C).5b,6 The
problem of the two cyclization systems shown in Scheme
(2) (a) Tsang, R.; Fraser-Reid, B. J . Am. Chem. Soc. 1986, 108, 2116,
8102. (b) Tsang, R.; Dickson, J r., J . K.; Pak, H.; Walton, R.; Fraser-
Reid, B. J . Am. Chem. Soc. 1987, 109, 3484. (c) Dickson, J . K., J r.;
Tsang, R.; Llera, J . M.; Fraser-Reid, B. J . Org. Chem. 1989, 54, 5350.
(d) Walton, R.; Fraser-Reid, B. J . Am. Chem. Soc. 1991, 113, 5791. (e)
Paquette, L. A.; Ra, C. S.; Silvestri, T. W. Tetrahedron 1989, 45, 3099.
(f) Knapp, S.; Gibson, F. S.; Choe, Y. H. Tetrahedron Lett. 1990, 38,
5397. (g) Knapp, S.; Gibson, F. S. J . Org. Chem. 1992, 57, 4802. (h)
Grissom, J . W.; Klingberg, D. J . Org. Chem. 1993, 58, 6559. (i) Grissom,
J . W.; Klingberg, D.; Meyenburg, S.; Stallman, B. L. J . Org. Chem.
1994, 59, 7876. (j) Clive, D. L. J .; Postema, M. H. D. J . Chem. Soc.,
Chem. Commun. 1993, 429. (k) J ung, M. E.; Choe, S. W. T. Tetrahedron
Lett. 1993, 34, 6247. (l) Hays, D. S.; Fu, G. C. J . Am. Chem. Soc. 1995,
117, 7283.
(3) For the use of radical addition to carbonyl in ring expansion
reactions, see: (a) Beckwith, A. L. J .; O’Shea, D. M.; Gerba, S.;
Westwood, S. W. J . Chem. Soc., Chem. Commun. 1987, 666. (b)
Beckwith, A. L. J .; O’Shea, D. M.; Westwood, S. W. J . Am. Chem. Soc.
1988, 110, 2565. (c) Dowd, P.; Choi, S.-C. J . Am. Chem. Soc. 1987,
109, 3493, 6548. (d) Dowd, P.; Zhang, W. Chem. Rev. (Washington, D.C.)
1993, 93, 2091. (e) Zhang, W.; Dowd, P. Tetrahedron Lett. 1996, 37,
957. (f) Nishida, A.; Takahashi, H.; Takeda, H.; Takada, N.; Yonemitsu,
O. J . Am. Chem. Soc. 1990, 112, 902. (g) Boger, D. L.; Mathvink, R. J .
J . Org. Chem. 1990, 55, 5442. (h) Baldwin, J . E.; Adlington, R. M.;
Singh, R. Tetrahedron 1992, 48, 3385. (i) Crimmins, M. T.; Dudek, C.
M.; Cheung, A. W.-H. Tetrahedron Lett. 1992, 33, 181. (j) Bowman,
W. R.; Westlake, P. J . Tetrahedron 1992, 48, 4027. (k) Dygutsch, D.
P.; Newmann, W. P.; Peterseim, M. Synlett 1994, 363. (l) Nemoto, H.;
Shiraki, M.; Yamada, N.; Raku, N.; Fukumoto, K. Tetrahedron Lett.
1996, 37, 6355. (m) Nemoto, H.; Shiraki, M.; Yamada, N.; Raku, N.;
Fukumoto, K. Tetrahedron 1996, 52, 13339.
1 is that the ring openings of alkoxy radicals 2 (k-1
)
4.7 × 108 s-1 at 80 °C) and 4 (k-2 ) 1.1 × 107 s-1 at 80
°C) are faster than the cyclizations. In fact, the most
useful application of radical addition to carbonyl com-
pounds relies on the controlled ring openings of the
cyclized alkoxy radicals to prepare ring expansion prod-
ucts.3
To drive these kind of neutral carbonyl addition
reactions to the right-hand side, one methodology is to
trap or convert the cyclized alkoxy radical irreversibly.
A successful demonstration of this idea is the acylsilane
system developed by us.8 Using radical 5 derived from
5-bromoacylsilane for illustration (Scheme 2), the 5-exo
ring closure product 6 undergoes a facile radical-Brook
rearrangement9 to give R-silyloxy radical 7. Due to the
strong bonding energy of O-Si,10 the silyl migration is
(4) For acyl migration, see: (a) Giese, B.; Heinrich, N.; Horler, H.;
Koch, W.; Schwarz, H. Chem. Ber. 1986, 119, 3528. (b) Wollowitz, S.;
Halpern, J . J . Am. Chem. Soc. 1988, 110, 3112. (c) Renaud, P.; Vionnet,
J .-P. J . Org. Chem. 1993, 58, 5895. (d) Leardini, R.; Lucarini, M.;
Nanni, A.; Nanni, D.; Pedulli, G. F.; Tundo, A.; Zanardi, G. J . Org.
Chem. 1993, 58, 2419. (e) Chen, S.-H.; Huang, S.; Gao, Q.; Golik, J .;
Farina, V. J . Org. Chem. 1994, 59, 1475. (f) Kaliappan, K.; Rao, G. S.
R. S. J . Chem. Soc., Chem. Commun. 1996, 2331.
(5) (a) Beckwith, A. L. J .; Hay, B. P. J . Am. Chem. Soc. 1989, 111,
230, 2674. (b) Beckwith, A. L. J .; Raner, K. D. J . Org. Chem. 1992, 57,
4954.
(6) Beckwith, A. L. J . Tetrahedron 1981, 37, 3073.
(7) Lusztyk, J .; Maillard, B.; Deycard, S.; Lindsay, D. A.; Ingold, K.
U. J . Org. Chem. 1987, 52, 3509.
S0022-3263(97)01130-4 CCC: $14.00 © 1997 American Chemical Society

9090 J . Org. Chem., Vol. 62, No. 26, 1997
Chang et al.
Sch em e 2
Sch em e 4
Sch em e 3
In the case of 5-hexenyl radical cyclization, it is well-
known that an alkyl substituent at C(5) severely reduces
the rate of 5-exo cyclization.6,14 In comparison, a recent
report by Curran, Diederichsen, and Palovich12c on radi-
cal cyclizations of acylgermanes showed that although
the cyclization rate is dependent on the germanium
ligand, the steric effect of the germyl group is not obvious
for 5-exo cyclization. This may be attributed to the longer
carbon-germanium bond. Therefore, it is important to
determine the size effect of the silyl group of acylsilane
toward the cyclization. The feasibility of the acylsilane
cyclization using primary, secondary, aryl, and vinyl
radicals is also studied and described in this full account.
most likely irreversible. Therefore, the cyclization step
is essentially driven toward the right-hand side by this
design. The cyclization of 4-silylcarbonylbutyl radical 5
can be considered as an equivalent of the cyclization of
4-formylbutyl radical 1. In addition, through the radi-
cal-Brook rearrangement, the radical is relocated to the
initial site of attack. Therefore, acylsilane can be re-
garded as a reagent equivalent of the geminal radical
acceptor/radical donor synthon.11
Parallel to our study of acylsilanes, Curran developed
the radical chemistry of acylgermanes (Scheme 3),12 In
the acylgermane system, the cyclized R-germyl alkoxy
radical undergoes a â-scission reaction to obtain cyclic
ketone with concomitant formation of a germyl radical
which carries on the chain reaction. Reminiscent to the
acylgermane system, Kim and J on13 recently reported the
radical cyclization of thioesters and selenoesters (Scheme
4). Instead of a catalytic cycle as in the acylgermane
system, the thioester and selenoester systems require the
use of 1.1 equiv of hexabutylditin. The â-scission process
involved in these reactions serves to drive the carbonyl
cyclizations to the right-hand side. The added advantage
of the acylgermane system is that it is tin free and the
removal of the germanium byproduct is easy.
Resu lts a n d Discu ssion
Syn th esis of Acylsila n e Su bstr a tes. Retrosyntheti-
cally, radicals such as acylsilane 5 were generated from
the corresponding haloacylsilanes using the tributyltin
hydride methodology. The haloacylsilanes were synthe-
sized using method developed by Brook and Corey.15,16
As shown in Table 1, 2-silyl-substituted 1,3-dithianes 8
were alkylated with suitable halides. The resulting
aliphatic bromo 1,3-dithianes could not be stored for a
long period due to the presence of a nucleophilic sulfur
and electrophilic halide in the same molecule.17 There-
fore, it is best to hydrolyze the crude alkylation product
directly without delay. The use of red mercuric oxide and
boron trifluoride etherate in wet THF for the hydrolysis
of 1,3-dithiane was generally successful.18 However,
when olefin was present as in entries 21-24, ceric
ammonium nitrate (CAN) in wet acetonitrile19 was used
instead to avoid complication due to the participation of
the olefin in the reaction.20 In the cases with hindered
silyl group as in dithianes 8d 15a and 8e21 (entries 4, 5,
(8) (a) Tsai, Y.-M.; Cherng, C.-D. Tetrahedron Lett. 1991, 32, 3515.
(b) Tsai, Y.-M.; Tang, K.-H.; J iaang, W.-T. Tetrahedron Lett. 1993, 34,
1303. (c) Curran, D. P.; J iaang, W.-T.; Palovich, M.; Tsai, Y.-M. Synlett
1993, 403. (d) Tsai, Y.-M.; Chang, S.-Y. J . Chem. Soc., Chem. Commun.
1995, 981. (e) Chuang, T.-H.; Fang, J .-M.; J iaang, W.-T.; Tsai, Y.-M.
J . Org. Chem. 1996, 61, 1794. (f) Tsai, Y.-M.; Tang, K.-H.; J iaang, W.-
T. Tetrahedron Lett. 1996, 37, 7767. (g) Tsai, Y.-M.; Nieh, H.-C.; Pan,
J .-S.; Hsiao, D.-D. J . Chem. Soc., Chem. Commun. 1996, 2469.
(9) (a) Dalton, J . C.; Borque, R. A. J . Am. Chem. Soc. 1981, 103,
699. (b) Harris, J . M.; MacInnes, I.; Walton, J . C.; Maillard, B. J .
Organomet. Chem. 1991, 403, C25. (c) Tsai, Y.-M.; Ke, B.-W. J . Chin.
Chem. Soc. (Taipei) 1993, 40, 641. (d) Robertson, J .; Burrows, J . N.
Tetrahedron Lett. 1994, 35, 3777.
(14) Beckwith, A. L. J .; Blair, I. A.; Phillipou, G. Tetrahedron Lett.
1974, 2251.
(15) (a) Brook, A. G.; Duff, J . M.; J ones, P. F.; Davis, N. R. J . Am.
Chem. Soc. 1967, 89, 431. (b) Corey, E. J .; Seebach, D.; Freedman, R.
J . Am. Chem. Soc. 1967, 89, 434.
(16) For reviews about acylsilanes, see: (a) Ricci, A.; Degl’Innocenti,
A. Synthesis 1989, 647. (b) Page, P. C. B.; Klair, S. S.; Rosenthal, S.
Chem. Soc. Rev. 1990, 19, 147. (c) Cirillo, P. F.; Panek, J . S. Org. Prep.
Proc. Int. 1992, 24, 553.
(17) (a) Davey, A. E.; Parsons, A. F.; Taylor, R. J . K. J . Chem. Soc.,
Perkin Trans. 1 1989, 1853. (b) Sui, Z.; Furth, P. S.; De Voss, J . J . J .
Org. Chem. 1992, 57, 6658.
(18) Vedejs, E.; Fuchs, P. L. J . Org. Chem. 1971, 36, 366.
(19) (a) Ho, T.-L.; Ho, H. C.; Wong, C. M. J . Chem. Soc., Chem.
Commun. 1972, 791. (b) Ho, H. C.; Ho, T.-L.; Wong, C. M. Can. J .
Chem. 1972, 50, 2718.
(10) J ackson, R. A. J . Organomet. Chem. 1979, 166, 17.
(11) Ryu, I.; Sonoda, N.; Curran, D. P. Chem. Rev. (Washington,
D.C.) 1996, 96, 177.
(12) (a) Curran, D. P.; Liu, H. J . Org. Chem. 1991, 56, 3463. (b)
Curran, D. P.; Palovich, M. Synlett 1992, 631. (c) Curran, D. P.;
Diederichsen, U.; Palovich, M. J . Am. Chem. Soc. 1997, 119, 4797.
(13) Kim, S.; J on, S. Y. J . Chem. Soc., Chem. Commun. 1996, 1335.
(20) Tsai, Y.-M.; Nieh, H.-C.; Cherng, C.-D. J . Org. Chem. 1992, 57,
7010.
(21) Yoshida, J .; Itoh, M.; Matsunaga, S.; Isoe, S. J . Org. Chem.
1992, 57, 4877.

Intramolecular Radical Cyclizations of Acylsilanes
J . Org. Chem., Vol. 62, No. 26, 1997 9091
Ta ble 1. P r ep a r a tion of (Ha loa cyl)sila n es fr om 2-Silyl-1,3-d ith ia n es
a
Method A: red HgO (2 equiv), BF3‚OEt2 (2 equiv), H2O/THF (15/85), rt, 1 h. Method B: CAN (3 equiv), NaHCO3 (1.5 equiv), H2O/
CH3CN (1/8), -30 °C, 5 min.
10, and 11), the alkylations required the use of HMPA.
28 resulted from the elimination side reactions of dibro-
mides 23 and 27 during alkylations.
In entries 19 and 21, low yields of acylsilanes 248e and

9092 J . Org. Chem., Vol. 62, No. 26, 1997
Chang et al.
Ta ble 2. Ra d ica l Cycliza tion s of Acylsila n es In volvin g
P r im a r y Ra d ica lsa
tive to the change of the silyl group. Similar to that
mentioned above, the 6-exo cyclization of acyltrimethyl-
silane 12a gave the volatile cyclization product which was
converted directly to cyclohexyl benzoate in 64% yield
(entry 6). Therefore, the yield of cyclized silyl ether 36a
should be at least of the same value. Careful analysis of
the crude reaction mixture by GC showed the presence
of reduction product 37a which was identified by com-
parison with an authentic sample prepared alterna-
tively.22,24 The ratio of 36a /37a was determined to be
5.4/1 by GC. In the cases of acyltriphenylsilane 12d
(entry 7) and acyl(tert-butyl)dimethylsilane 12b (entry
8), substantial amounts of straight reduction products
37d (26%) and 37b (38%) were also obtained. With a
very bulky tert-butyldiphenylsilyl group (entry 9), acyl-
silane 12e gave only 30% of cyclization product 36e, and
the major product was debrominated acylsilane 37e
(63%). Both acylsilanes 12d and 12b contain large silyl
groups; however, the silyl groups are smaller than the
tert-butyldiphenylsilyl group.23 On the basis of these
observations, it is clear that for the 6-exo cyclizations of
acylsilanes 12, as the size of the silyl group increases,
the yield of the cyclization product decreases. Together,
these results indicated that the rate of 6-exo cyclization
of acylsilane is slower than that of the 5-exo cyclization.
The best two cases of this type are the cyclizations of
acylsilanes 12c20 (entry 10) and 12f (entry 11). Both gave
∼80% of cyclized products (36c,f) with ∼5% of reduction
products (37c,f). As we have mentioned above the larger
the size of the silyl group, the lower the yield of silyl ether
36. Assuming that the dimethylphenylsilyl and meth-
yldiphenylsilyl groups are larger than the trimethylsilyl
group, it is surprising that both acylmethyldiphenylsilane
12c (entry 10) and acyldimethylphenylsilane 12f (entry
11) gave more cyclization products than acyltrimethyl-
silane 12a (entry 6). We speculate that the unexpected
cyclization efficiency of 12c and 12f is due to the presence
of the phenyl group on silicon. However, with the
presence of three phenyl groups on silicon as in acyltri-
phenylsilane 12d (entry 7), the cyclization is less efficient
due to the enhanced steric effect.
a
The cyclization was performed by slow addition (6 h) of a
benzene solution of tributyltin hydride (1.2-1.5 equiv) and AIBN
(5 mol %) to a refluxing solution of the substrate. Equal amounts
of benzene were used to prepare the two solutions, and the final
concentration relative to the substrate was 0.05 M. b The yield was
extrapolated from the yield of the corresponding benzoate derived
from the silyl ether. c The ratio of 36a /37a was determined to be
d
5.4/1 by GC. The final concentration relative to 12b was 0.025
M.
In the case of acyltriphenylgermanes, Curran, Dieder-
ichsen, and Palovich12c found that the 5-exo primary
radical cyclization (kc ) 6.4 × 106 s-1 at 80 °C) is faster
than the 6-exo cyclization (kc ) 1.3 × 106 s-1 at 80 °C). It
was also reported that acylgermanes with triphenylger-
myl group exhibit higher reactivity than those with
nonaromatic germanium ligands. This was attributed to
a lower carbonyl LUMO energy of acyltriphenylger-
manes.12c The results of the cyclizations of acylsilanes
shown in Table 2 are consistent with the acylgermane
chemistry. However, acyltriphenylsilane 12d (entry 7)
does not cyclize as efficiently as the corresponding
acyltriphenylgermane12c (87% cyclization).
When the phenyl group on silicon contains an ortho
substituent as in acylsilane 12g (entry 12), the cyclization
also is less efficient. An appreciable amount of straight
reduction product 37g (32%) was obtained along with
52% of cyclized product 36g. In the case of acylsilane
12h (entry 13) with a 4-methoxy group attached on the
phenyl group, the cyclization gave silyl ether 36h (81%)
exclusively. This result was similar to the cyclization of
acyldimethylphenylsilane 12f (entry 11). Therefore, the
Cycliza tion s In volvin g Alk yl Ra d ica ls. The radical
cyclization studies were generally performed by slow
addition (6 h) of a benzene solution of tributyltin hydride
(1.2-1.5 equiv) and AIBN (0.05 equiv) to a refluxing
benzene solution of bromoacylsilane. Our results involv-
ing primary radicals are shown in Table 2. The results
show that primary radicals undergo 5-exo cyclization very
efficiently (entries 1-5). Variation of the size of the silyl
group did not have a significant effect on this type of
cyclizations. High yields were obtained, and we did not
isolate any uncyclized reduction product. In the case of
entry 1, we had difficulty isolating trimethylsilyl ether
35a because of its volatility. To overcome this problem,
we directly added benzoyl chloride (5 equiv), TBAF/THF
(2.5 equiv), and triethylamine (4 equiv) at the end of
cyclization and heated the reaction mixture for 3 h at 80
°C. In this way, we were able to isolate cyclopentyl
benzoate in 68% yield. Therefore, the yield of 35a should
be at least 68%. Gas chromatographic analysis of the
cyclization mixture derived from acyltrimethylsilane 10a
also showed that there was no uncyclized product.22
Contrary to 5-exo cyclizations, 6-exo cyclizations of
primary radicals (entries 6-11, Table 2) are more sensi-
(23) Hwu, J . R.; Wang, N. Chem. Rev. (Washington, D.C.) 1989, 89,
1599.
(24) Miller, J . A.; Zweifel, G. Synthesis 1981, 288.
(22) Authentic straight reduction product was obtained by alkylation
of 8a with the corresponding alkyl bromide followed by hydrolysis.

Intramolecular Radical Cyclizations of Acylsilanes
J . Org. Chem., Vol. 62, No. 26, 1997 9093
Ta ble 3. Ra d ica l Cycliza tion s of Acylsila n es In volvin g Secon d a r y Ra d ica lsa
a
The cyclization was performed by slow addition of a benzene solution of tributyltin hydride (1.2-1.3 equiv) and AIBN (5 mol %) to a
refluxing benzene solution of the substrate. The final concentration relative to the substrate. c A separable mixture of cis and trans
isomers, cis/ trans ) 55/45. The reaction was performed by directly mixing the substrate, stannane, and AIBN in benzene and heated
to reflux for 0.5 h. A separable mixture of cis and trans isomers, cis/ trans ) 3/7.
b
d
e
sluggish cyclization of 12g is more likely due to the steric
effect of the ortho substituent and not to the electronic
effect.
radical cyclization system which is less sensitive to the
substituents at C(1).6,14
To determine the stereochemistry of the cyclized five-
membered ring product 38, we started from the com-
mercially available trans-alcohol 41 and prepared an
authentic sample of silyl ether trans-38 via a well-known
method25 (eq 1). Therefore, the stereochemical structure
5-Exo radical cyclizations of acylsilanes with secondary
radicals were also studied (Table 3). Under the same
condition as used in Table 2, acylsilane 14 (entry 1, Table
3) gave silyl ether 38 in 76% yield. Compared with the
cyclization of the primary substrate 10c (Table 2, entry
3), no apparent difference between the primary and
secondary substrate is observed. We found that this
reaction could be performed in a concentrated condition
(0.5 M in benzene). When the tributyltin hydride solu-
tion was quickly added over 0.5 h, acylsilane 14 still gave
silyl ether 38 (80%) as the only product (entry 2). Even
when we mixed 14 (0.5 M in benzene), tributyltin
hydride, and AIBN (5 mol %) directly and heated for 0.5
h (entry 3), we were able to isolate 52% of cis-38, 42% of
trans-38, and only 5% of straight reduction product 37c.
These experiments demonstrate that the 5-exo cyclization
of acylsilane can be performed very practically with a
fairly concentrated condition in a short time.
In contrast to the 5-exo cyclization of acylsilane 14
(entry 2), the homologous acylsilane 16 (entry 4) gave
only straight reduction product 39 (76%). Note that the
conditions used in entries 2 and 4 were the same. We
had to dilute the concentration and lengthen the stan-
nane addition time (entries 5 and 6) to obtain the
cyclization product. Thus, in entry 6 we were able to
isolate 17% of silyl ether cis-40, 40% of trans-40, and 39%
of reduction product 39. Compared with the 6-exo
cyclization of the primary substrate 12c (entry 10, Table
2), the cyclization of the secondary substrate 16 is less
efficient. Therefore, the 6-exo cyclization of acylsilane
is sensitive not only to the size of the silyl group but also
to the substituent at the carbon bearing the initial
radical. A similar trend was also observed in the
acylgermane system.12c This is contrary to the 5-hexenyl
of trans-38 was identified. 13C NMR (CDCl3) absorption
of the methyl carbon in trans-38 appeared at δ 18.1 and
that of cis-38 appeared at δ 14.5. The higher field
absorption of the methyl carbon of cis-38 indicated that
the methyl group was located at a sterically more
encumbered position.26
1
In the H NMR (CDCl3) spectra of six-membered ring
silyl ether 40, the hydrogen absorption at C-1 appeared
at δ 3.75-3.90 (m, OCH) for cis-40 and δ 3.24 (td, J )
10.0, 4.0 Hz, OCH) for trans-40. These values correlate
reasonably well with those of cis- and trans-2-methylcy-
clohexanol (δ 3.82 for the cis-isomer and δ 3.10 for the
trans-isomer).27 In particular, the coupling pattern for
C(1)-H of trans-40 indicated that it belongs to the trans-
diequatorial conformer. In the 13C NMR spectra of 40,
the methyl carbon of cis-40 appeared at δ 17.6 and that
of trans-40 appeared at δ 19.3. Similar to silyl ether 38,
the methyl carbon of cis-40 appeared at higher field.
(25) Corey, E. J .; Venkateswarlu, A. J . Am. Chem. Soc. 1972, 94,
6190.
(26) Breitmaier, E.; Voelter, W. Carbon-13 NMR Spectroscopy; 3rd
ed.; VCH: Weinhein, 1990; p 115.
(27) Iguchi, S.; Nakai, H.; Hayashi, M.; Yamamoto, H.; Maruoka,
K. Bull. Chem. Soc. J pn. 1981, 54, 3033.

9094 J . Org. Chem., Vol. 62, No. 26, 1997
Chang et al.
Ta ble 4. Ra d ica l Cycliza tion s of Acylsila n es in th e
P r esen ce of Et3Ba
comparison of the reactions at room temperature (entry
5) and 80 °C (entry 3) showed that without triethylborane
the cyclization process is preferred at higher tempera-
ture. However, when we performed the cyclization of 12b
(entry 5) at 80 °C using triethylborane-air for initiation,
we found only a slight increase in the ratio of 36b/37b
(cf. entry 3). Therefore, the beneficial effect of trieth-
ylborane seems to operate only at lower temperature. The
influence of triethylborane is small at higher tempera-
ture.
amount of
Et3B (equiv)
temp
(°C)
products
(ratio)d
entry
substrates
1b
2
12a
12a
12b
12b
12b
12b
12c
0c
80
rt
80
rt
rt
80
rt
36a /37a (5.4/1)
36a /37a (36/1)
36b/37b (1.6/1)
36b/37b (4.4/1)
36b/37b (0.85/1)f
36b/37b (2.2/1)
36cg
1.3
3e
4
0c
1.2
0c
1.5
0.18
5e
6
7
Attem p ted F or m a tion of Oth er Rin g Sizes a n d
In ter m olecu la r P r ocess. To test if this method could
be used to construct rings other than five- and six-
membered, we examined the cyclization of acylsilanes 18,
20, and 22 (entries 16-18, Table 1). The chloride 20 and
selenide 22 were used because we were not able to
synthesize the corresponding bromides using the 1,3-
dithiane strategy. Treatment of the 7-exo substrate 18
with tributyltin hydride (1.6 equiv) gave 68% of reduction
product 39 and 23% of R-silyl alcohol 43. Alcohol 43 was
presumably derived from further reduction of acylsilane
39 with excess stannane.31 The reaction of the 4-exo
substrate 20 with tributyltin hydride (1.65 equiv) gave
34% of reduction product 44, 13% of chloride 45, 5% of
silyl ether 46, and 30% of recovered 20. The presence of
chloride 45 indicated that the rate of acylsilane carbonyl
reduction by stannane is competitive to the reduction of
the chloride moiety.31,32 We do not know the origin of
the silyl ether 46. The reaction of the 3-exo substrate
22 with tributyltin hydride (1.3 equiv) gave 79% of
reduction product 47. Therefore, the formation of un-
substituted seven-, four- and three-membered rings is not
feasible using this method.
a
A benzene solution of tributyltin hydride (1.3 equiv) was added
over 2 h to a benzene solution of the substrate and triethylborane.
The final concentration relative to the substrate was 0.05 M. Dry
air was slowly bubbled through the reaction mixture via a syringe.
The reaction condition described in Table 2 was used. c AIBN
b
was used for initiation. The ratio was determined by GC. e A
d
benzene solution of tributyltin hydride (1.2 equiv) and AIBN (5
mol %) was slowly added over 2 h to a benzene solution of the
substrate. The final concentration relative to the substrate was
0.05 M. The reaction mixture was irradiated with two 100 W
tungsten lamps. f Unreacted 12b (32%) was observed. Straight
g
reduction product was not observed.
Sch em e 5
The stereochemical outcome of the cyclizations of
acylsilanes 14 and 16 is determined at the hydrogen
abstraction step. As shown in Scheme 5, the cyclization
of 16 went through R-silyloxy radical 42 in which the
methyl group preferred to adopt an equatorial position.
It is well-known that the cyclohexyl radical prefers to
abstract hydrogen from the axial site.28 Therefore, axial
hydrogen abstraction of radical 42 should give the trans-
product preferentially as observed.
Effect of Tr ieth ylbor a n e. Clive and Postema2j re-
cently reported that triethylborane-stannane-air sys-
tem29 could improve the cyclization of 5-formylpentyl
radical (3) system. On the basis of their finding, we
performed radical cyclization of acylsilanes using trieth-
ylborane-air for initiation. As shown in Table 4, without
triethylborane the cyclization of the primary 6-exo sub-
strate 12a at 80 °C gave cyclization product 36a /reduc-
tion product 37a 24 in a ratio of 5.4/1 by GC analysis (entry
1). When the triethylborane method (rt) was used (entry
2), the ratio of 36a /37a improved to 36/1. Likewise, in
the case with a tert-butyldimethylsilyl group the ratio of
36b/37b30 obtained from acylsilane 12b (entries 3 and
4) also increased. When acylmethyldiphenylsilane 12c
was cyclized using the triethylborane method at rt (entry
6), we did not observe reduction product 37c by GC.
To determine whether this improvement was due to
the temperature or not, we performed the cyclization of
12b at room temperature without triethylborane (entry
5). This reaction was initiated by the irradiation of two
100 W tungsten lamps in the presence of AIBN (5 mol
%). The ratio of cyclization product 36b/reduction prod-
uct 37b was 0.85/1. This result indicated that the use
of triethylborane is helpful for cyclization. Interestingly,
We also examined the possibility of intermolecular
coupling of alkyl radical with acylsilane. This was
performed by slow addition (0.5 h) of a benzene solution
of tributyltin hydride (1.2 equiv) and AIBN (5 mol %) to
a refluxing benzene solution of 5 equiv of acetyltrimeth-
ylsilane (48) and hexadecyl bromide. The final concen-
tration relative to the bromide was 0.5 M. Unfortunately,
1
we did not observe any coupling product. The H NMR
of the crude concentrate of the reaction mixture showed
that the hexadecyl bromide was completely reduced with
the acylsilane 48 left intact.
Cycliza tion s w ith Ar yl Ra d ica ls. An aryl radical
is potentially more reactive than an alkyl radical. How-
ever, for intramolecular cyclization involving an aryl
radical, the two sp2 centers of the aryl group incorporated
in the forming ring may increase the strain energy for
cyclization.33 Therefore, it is essential to test the acyl-
silane system with an aryl radical.
(28) (a) Curran, D. P.; Porter, N. A.; Giese, B. Stereochemistry of
Radical Reactions; VCH: Weinheim, 1996; Chapter 3. (b) Giese, B.
Angew. Chem., Int. Ed. Engl. 1989, 28, 969.
(29) (a) Nozaki, K.; Oshima, K.; Utimoto, K. J . Am. Chem. Soc. 1987,
109, 2547. (b) Miura, K.; Ichinose, Y.; Nozaki, K.; Fugami, K.; Oshima,
K.; Utimoto, K. Bull. Chem. Soc. J pn. 1989, 62, 143.
(31) Ingold, K. U.; Lusztyk, J .; Scaiano, J . C. J . Am. Chem. Soc.
1984, 106, 343.
(32) Beckwith, A. L. J .; Pigou, P. E. Aust. J . Chem. 1986, 39, 77.
(33) Beckwith, A. L. J .; Gara, W. B. J . Chem. Soc., Perkin Trans. 2
1975, 795.
(30) Dondy, B.; Doussot, P.; Portella, C. Synthesis 1992, 995.

Intramolecular Radical Cyclizations of Acylsilanes
J . Org. Chem., Vol. 62, No. 26, 1997 9095
Sch em e 6
Sch em e 7
Sch em e 8
As shown in Scheme 6, the 5-exo substrate 248e cyclized
effectively when treated with tributyltin hydride and
gave silyl ether 49 in 83% yield. On the contrary, under
the same condition (condition 1) the homologous acylsi-
lane 26 gave only reduction product 50 (86%). Even
when we lengthened the stannane addition rate to 4 h
(condition 2) and made the concentration more dilute
(0.05 M), we still obtained acylsilane 50 in 81% yield
without the formation of 6-exo cyclization product. Previ-
ously, we noted that 1,5-hydrogen transfer involving the
carbonyl R-hydrogen is a potential side reaction in our
6-exo cyclization system.8b The 1,5-hydrogen transfer
process in ω-formylalkyl radical cyclizations5b and acylger-
mane cyclizations12c has also been reported. We believed
that this might be the major process for the attempted
cyclization of 26. Grissom et al.2i recently reported the
cyclizations of aldehydes 51. The 5-exo cyclization prod-
uct 52 was obtained in 23% yield from aldehyde 51a . The
acylsilane system provides an attractive alternative for
the 5-exo cyclization. However, the acylsilane system is
worse for 6-exo cyclization involving an aryl radical
because the corresponding aldehyde 51b gave 34% of
6-exo cyclization product 53 whereas acylsilane 26 gave
only reduction product.
Cycliza tion s w ith Vin yl Ra d ica ls. Acylsilane cy-
clization involving a vinyl radical34 provides an intriguing
situation. First, there are two types of possible system.
The exocyclic type (Scheme 7) having the vinyl group
exocyclic to the forming ring involves one sp2 center of
the vinyl group in the forming ring. The endocyclic type
having the vinyl group endocyclic to the forming ring
incorporates two sp2 centers of the vinyl group in the
forming ring. Apparently, the degree of strain involved
in these two systems is different. Second, cyclization
followed by silyl shift provides an allylic radical as in 54
and 57. Hydrogen abstractions of these allylic radicals
are confronted with regioselectivity options. Either enol
silyl ethers (55 and 58) or allyl silyl ethers (56 and 59)
may be obtained.
Our cyclization results of vinyl radicals are sum-
marized in Scheme 8. Initially, we had problems initiat-
ing the reaction by just heating. Therefore, we increased
the amount of AIBN to 15 mol % and irradiated the
reaction mixture with two 100 W tungsten lamps.34 The
two 5-exo cyclization systems 28 and 30 gave only
cyclization products. More importantly, both cases gave
regioselectively the enol silyl ethers 6035 (62%) and 61
(83%). Therefore, the intermediate allyl radicals prefer
to abstract hydrogen at the less substituted carbon. This
approach provides a unique way to prepare five-mem-
bered cyclic silyl enol ether regiospecifically.8d
The two 6-exo cyclization systems 32 and 34 behave
similarly and gave low yields of enol silyl ethers 6236
(23%) and 64 (35%). Reduction products 63 (45%) and
65 (48%) were obtained as the major products. Presum-
ably, 1,5-hydrogen transfer is a major competing process.
Con clu sion s
Acylsilanes that bear a bromo substituent at the δ and
ꢀ position could be easily prepared. Intramolecular
(35) Leigh, W. J .; Bradaric, C. J .; Sluggett, G. W. J . Am. Chem. Soc.
1993, 115, 5332.
(34) (a) Stork, G.; Baine, N. H. J . Am. Chem. Soc. 1982, 104, 2321.
(b) Stork, G.; Mook, R., J r. J . Am. Chem. Soc. 1983, 105, 3720.
(36) Denmark, S. E.; Hammer, R. P.; Weber, E. J .; Habermas, K. L.
J . Org. Chem. 1987, 52, 165.

9096 J . Org. Chem., Vol. 62, No. 26, 1997
Chang et al.
(200 MHz) δ 0.73 (s, 3 H), 1.91-2.19 (m, 2 H), 2.69 (dt, J )
14.0, 4.0 Hz, 2 H), 2.89 (td, J ) 14.0, 4.0 Hz, 2 H), 4.20 (s, 1
H), 7.28-7.49 (m, 6 H), 7.55-7.71 (m, 4 H); 13C NMR (50 MHz)
δ -5.5, 26.0, 31.4, 32.8, 127.8, 129.9, 133.3, 135.1. Anal. Calcd
for C17H20S2Si: C, 64.50; H, 6.37. Found: C, 64.34; H, 6.57.
2-(ter t-Bu tyld ip h en ylsilyl)-1,3-d ith ia n e (8e). To a solu-
tion of 600 mg (5.00 mmol) of 1,3-dithiane in 5 mL of dry THF
cooled in a dry ice-acetone bath was added dropwise over 20
min a solution of 1.50 M of n-butyllithium in hexane (3.60 mL,
5.40 mmol). The resulting solution was stirred at the same
temperature for 20 min followed by the addition of 1.75 mL
(10.0 mmol) of HMPA. The resulting solution was stirred for
another 30 min at the same temperature followed by slow
addition of 1.44 mL (5.50 mmol) of tert-butylchlorodiphenyl-
silane over 20 min. The resulting mixture was warmed slowly
to room temperature and then stirred overnight. The reaction
mixture was partitioned between 100 mL of ether and 100 mL
of water. The organic layer was washed with water (100 mL)
and brine (50 mL), dried (MgSO4), and concentrated in vacuo.
The residual oil was chromatographed over silica gel (eluted
with hexane/ethyl acetate, 98/2) to give 1.35 g (75%) of 8e as
a white solid: mp 113.5-115.0 °C (EtOH); 1H NMR (200 MHz)
δ 1.18 (s, 9 H), 2.00-2.15 (m, 2 H), 2.70 (dt, J ) 14.0, 3.6 Hz,
2 H), 2.78-3.05 (m, 2 H), 4.32 (s, 1 H), 7.35-7.48 (m, 6 H),
7.68-7.85 (m, 4 H); 13C NMR (75 MHz) δ 19.2, 26.0, 28.5, 32.0,
32.7, 127.3, 129.4, 132.0, 136.2; HRMS calcd for C20H26S2Si
m/z 358.1245, found 358.1256.
2-[(2-Meth oxyp h en yl)d im eth ylsilyl]-1,3-d ith ia n e (8g).
To a mixture of 290 mg (12.0 mmol) of magnesium turnings
and 2 mL of THF was added a small amount of iodine and
2-bromoanisole. When the reaction was initiated, an ad-
ditional portion of THF (8 mL) was added. A solution of 1.25
mL (10.0 mmol) of 2-bromoanisole in 10 mL of THF was added
slowly to the above mixture in a rate as to keep a gentle reflux.
The resulting mixture was stirred for another hour at room
temperature and then added slowly to a solution of 6.3 mL
(50 mmol) of dichlorodimethylsilane in 20 mL of THF. The
reaction mixture was stirred at room temperature for 1 h,
diluted with 150 mL of hexane, and filtered. The filtrate was
concentrated in vacuo to remove the excess dichlorosilane, and
the resulting crude chlorosilane was dissolved in 10 mL of dry
THF and then cooled in an ice-water bath. To another
solution of 1.20 g (10.0 mmol) of 1,3-dithiane in 10 mL of THF
cooled at 0 °C was added 7.1 mL (11 mmol) of n-butyllithium
(1.55 M in hexane). The resulting solution was added slowly
to the above-mentioned crude chlorosilane solution. The
resulting mixture was stirred for another 30 min and then
partitioned between 100 mL of ether and 60 mL of water. The
organic layer was washed with brine (20 mL), dried (MgSO4),
and concentrated. The residual oil was chromatographed over
silica gel (eluted with hexane/dichloromethane, 9/1, 8/2, and
7/3 in sequence) to give 1.5 g (53%) of 8g as a colorless liquid:
1H NMR (300 MHz) δ 0.46 (s, 6 H), 1.90-2.16 (m, 2 H), 2.68
(dt, J ) 14.0, 3.0 Hz, 2 H), 2.86 (ddd, J ) 14.0, 11.0, 3.0 Hz,
2 H), 3.83 (s, 3 H), 4.10 (s, 1 H), 6.85 (d, J ) 8.0 Hz, 1 H), 6.98
(t, J ) 8.0 Hz, 1 H), 7.39 (t, J ) 8.0 Hz, 1 H), 7.46 (d, J ) 8.0
Hz, 1 H); 13C NMR (50 MHz) δ -4.4, 26.1, 31.1, 33.4, 54.9,
109.6, 120.4, 122.8, 131.5, 135.7, 164.2; HRMS calcd for
C13H20OS2Si m/z 284.0725, found 284.0729.
radical cyclizations of these bromoacylsilanes were most
effective for 5-exo cyclizations. For 6-exo cyclizations, the
system was sensitive to the changes of the size of the
silyl group and to the type of the initial radical center.
6-Exo cyclizations involving primary radicals were most
effective when the silyl group was a methyldiphenylsilyl
group. However, when a secondary radical was involved
in 6-exo cyclization, only a moderate yield of cyclization
product was obtained. Using triethylborane-air for
initiation at room temperature also improved the cycliza-
tion efficiency. An aryl radical underwent 5-exo cycliza-
tion with acylsilane successfully. In contrast, the 6-exo
cyclization did not occur for an aryl radical. Vinyl
radicals underwent 5-exo cyclizations with acylsilanes
successfully and afforded enol silyl ethers as the products.
The corresponding 6-exo cyclizations only gave low yields
of cyclization products.
The radical-Brook rearrangement involved in the
acylsilane cyclizations not only serves to drive the reac-
tion toward the cyclization side but also converts the
cyclized oxygen radical to a carbon radical. This unique
feature makes the acylsilane systems different from the
ω-formylalkyl radical cyclization systems. In the latter
systems cyclic alcohols are the expected products, while
in the former systems silyl group protected cyclic alcohols
are obtained. In more functionalized cases such as the
vinyl radical cyclizations of acylsilanes, regiospecific enol
silyl ethers are accessible. This type of product cannot
be obtained easily via the ω-formylalkyl radical cycliza-
tions. The cyclizations of acylsilanes are similar to
acylgermanes in many respects. However, the two
processes lead to different end products and are comple-
mentary. The acylgermanes cyclize to give cyclic ketones,
and the reaction stops at the ketone stage.12 The acyl-
silane cyclizations lead to a cyclized R-silyloxy radical and
create opportunities for further manipulations.8
Exp er im en ta l Section
Melting points are uncorrected. 1H NMR spectra were
recorded at 200 or 300 MHz; 13C NMR spectra were recorded
at 50 or 75 MHz. Tetramethysilane (δ ) 0 ppm) or CHCl3 (δ
) 7.24 ppm) were used as internal standards, and CDCl3 was
used as the solvent. Benzene and THF were distilled from
sodium benzophenone ketyl under N2. HMPA was distilled
over calcium hydride. The benzene used for cyclization
reactions was deoxygenated by passing a gentle stream of
argon through for 0.5 h before use. All reactions were
performed under a blanket of N2 or Ar. Dibromides 23,37 25,38
27,39 29,40 31,39 and 3340 were synthesized according to
literature methods.
2-(Meth yld ip h en ylsilyl)-1,3-d ith ia n e (8c). To a solution
of 2.00 g (16.7 mmol) of 1,3-dithiane in 16 mL of dry THF
cooled at 0 °C was added dropwise over 20 min a solution of
1.60 M of n-butyllithium in hexane (11.0 mL, 17.6 mmol). The
resulting solution was stirred at 0 °C for 20 min followed by
the addition of 3.51 mL (16.7 mmol) of chlorodimethylphenyl-
silane in one portion. The resulting mixture was stirred at 0
°C for 10 min and then at rt for 30 min and partitioned
between 70 mL of ether and 25 mL of water. The ether phase
was washed with brine (25 mL), dried (MgSO4), and concen-
trated in vacuo. The residual liquid was chromatographed
over silica gel (eluted with hexane/ethyl acetate, 98/2) to give
2-[(4-Meth oxyp h en yl)d im eth ylsilyl]-1,3-d ith ia n e (8h ).
According to the same procedure for the synthesis of 8g, from
0.50 g (4.2 mmol) of 1,3-dithiane we obtained 1.0 g (85%) of
8h as a colorless liquid: 1H NMR (200 MHz) δ 0.42 (s, 6 H),
1.90-2.15 (m, 2 H), 2.67 (dt, J ) 14.0, 3.0 Hz, 2 H), 2.83 (ddd,
J ) 14.0, 11.0, 3.0 Hz, 2 H), 3.79 (s, 3 H), 3.83 (s, 1 H), 6.92
(br d, J ) 9.0 Hz, 2 H), 7.51 (br d, J ) 9.0 Hz, 2 H); 13C NMR
(50 MHz) δ -4.7, 25.9, 31.0, 33.9, 54.8, 113.5, 125.8, 135.5,
160.8; HRMS calcd for C13H20OS2Si m/z 284.0725, found
284.0719.
1
3.70 g (70%) of 8c as a white solid: mp 44-46 °C; H NMR
Gen er a l P r oced u r e for th e Alk yla tion of 2-Silyl-1,3-
d ith ia n es w ith Alk yl Dibr om id es follow ed by Hyd r olysis
w ith Red Mer cu r ic Oxid e a n d Bor on Tr iflu or id e Eth er -
a te. To a solution of the 2-silyl-1,3-dithiane (3.0 mmol) in 6
mL of THF cooled in an ice-water bath was added dropwise
over 20 min a solution of n-butyllithium in hexane (1.1 equiv).
After another 30 min of stirring at 0 °C, the resulting solution
(37) Ponton, J .; Helquist, P.; Conrad, P. C.; Fuchs, P. L. J . Org.
Chem. 1981, 46, 118.
(38) Rowley, L. E.; Swan, J . M. Aust. J . Chem. 1974, 27, 801.
(39) Bestmann, H. J .; Zeibig, T.; Vostrowsky, O. Synthesis 1990,
1039.
(40) Shi, L.; Narula, C. K.; Mak, K. T.; Kao, L.; Xu, Y.; Heck, R. F.
J . Org. Chem. 1983, 48, 3894.

Intramolecular Radical Cyclizations of Acylsilanes
J . Org. Chem., Vol. 62, No. 26, 1997 9097
was added over 45 min to a solution of the dibromide (2 equiv)
in 5 mL of THF cooled at -30 °C using a dry ice-acetonitrile
bath. The resulting mixture was stirred at the same temper-
ature for 1.5 h and then partitioned between 80 mL of ether
and 60 mL of water. The organic layer was washed with 50
mL of brine, dried (MgSO4), and concentrated in vacuo. The
residual oil was dissolved in 15 mL of wet THF (15%) followed
by the addition of 1.29 g (6.0 mmol) of red mercuric oxide, 1.20
g of Celite, and 0.73 mL (6.0 mmol) of boron trifluoride
etherate. The resulting mixture was stirred at room temper-
ature for 1 h, diluted with ether (80 mL), and filtered. The
filtrate was washed with brine (30 mL), dried (MgSO4), and
concentrated in vacuo. The residual oil was chromatographed
over silica gel using hexane/ethyl acetate as eluent.
6-Br om o-1-[(2-m eth oxyp h en yl)d im eth ylsilyl]h exa n -1-
on e (12g). To a solution of 852 mg (3.0 mmol) of 8g in 5 mL
of THF cooled in an ice-water bath was added dropwise over
20 min a 1.55 M solution of n-butyllithium in hexane (2.3 mL,
3.6 mmol). After another 20 min of stirring at 0 °C, the
resulting solution was added over 30 min to a solution of 2.0
mL (15 mmol) of 11 in 10 mL of THF cooled at -30 °C using
a dry ice-acetonitrile bath. The resulting mixture was stirred
at the same temperature for 30 min and then partitioned
between 100 mL of ether and 80 mL of water. The organic
layer was washed with 50 mL of brine, dried (MgSO4), and
concentrated in vacuo. The residual oil and 1.5 g (18 mmol)
of sodium bicarbonate was mixed with 8 mL of acetonitrile
and 5 mL of dichloromethane and then cooled in a dry ice-
acetonitrile bath. A mixture of 4.11 g (7.50 mmol) of CAN in
8 mL of acetonitrile and 1 mL of water was added over 1 min
to the above mixture. The resulting mixture was stirred for
10 min at the same temperature and then diluted with 80 mL
of ether and filtered. The filtrate was washed with water (80
mL) and brine (50 mL), dried (MgSO4), and concentrated. The
residual oil was chromatographed over silica gel (eluted with
hexane/ethyl acetate, 97/3) to give 12g as a pale yellow oil:
IR (neat) 1633 cm-1; 1H NMR (200 MHz) δ 0.42 (s, 6 H), 1.20-
1.35 (m, 2 H), 1.35-1.50 (m, 2 H), 1.76 (quintet, J ) 7.0 Hz,
2 H), 2.55 (t, J ) 7.0 Hz, 2 H), 3.32 (t, J ) 7.0 Hz, 2 H), 3.75
(s, 3 H), 6.83 (br d, J ) 8.0 Hz, 1 H), 6.98 (ddd, J ) 8.0, 7.0,
1.0 Hz, 1 H), 7.35-7.45 (m, 2 H); 13C NMR (75 MHz) δ -4.7,
21.2, 27.8, 32.5, 33.6, 47.5, 54.9, 109.5, 120.9, 123.2, 131.8,
135.4, 163.9, 246.1; HRMS calcd for C15H23BrO2Si m/z 342.0651,
found 342.0650.
Gen er a l P r oced u r e for Ra d ica l Cycliza tion s of Br o-
m oa cylsila n es 10 a n d 12. To a refluxing benzene (10 mL)
solution of the bromoacylsilane (1 mmol) was added via syringe
pump a benzene (10 mL) solution of tributyltin hydride (1.2-
1.5 mmol) and AIBN (0.05 mmol) over 6 h. The resulting
solution was heated at 80 °C for another hour and concentrated
in vacuo. To the residual liquid was added a few drops of wet
triethylamine,41 and the resulting mixture was chromato-
graphed over silica gel (eluted with hexane/ethyl acetate) to
isolate the products.
Cyclop en tyl ter t-bu tyld im eth ylsilyl eth er (35b): 1H
NMR (200 MHz) δ 0.02 (s, 6 H), 0.86 (s, 9 H), 1.40-1.55 (m, 4
H), 1.60-1.75 (m, 4 H), 4.21 (quintet, J ) 5 Hz, 1 H); 13C NMR
(50 MHz) δ -4.7, 18.2, 23.1, 25.9, 35.7, 74.4; HRMS calcd for
C11H24OSi m/z 200.11596, found 200.1601.
Ra d ica l Cycliza tion of 14. cis- a n d tr a n s-2-Meth ylcy-
clop en tyl Meth yld ip h en ylsilyl Eth er (cis- a n d tr a n s-38).
To a refluxing solution of 150 mg (0.400 mmol) of 14 in 0.4
mL of benzene was added via syringe pump over 30 min a
solution of 0.129 mL (0.48 mmol) of tributyltin hydride and
3.5 mg (0.022 mmol) of AIBN in 0.4 mL of benzene. The
resulting solution was stirred for 2 h at 80 °C and then
concentrated in vacuo. To the residual liquid was added a few
drops of wet triethylamine,41 and the resulting mixture was
chromatographed over silica gel (eluted with hexane) to give
50 mg (42%) of cis-38 as a colorless liquid: 1H NMR (200 MHz)
δ 0.65 (s, 3 H), 1.01 (d, J ) 6.5 Hz, 3 H), 1.30-1.94 (m, 7 H),
4.16 (q, J ) 4.0 Hz, 1 H), 7.26-7.46 (m, 6 H), 7.48-7.70 (m, 4
H); 13C NMR (50 MHz) δ -2.4, 14.5, 21.8, 30.8, 34.7, 39.8,
127.7, 129.5, 134.4, 137.1. Anal. Calcd for C19H24OSi: C,
76.97; H, 8.16. Found: C, 76.48; H, 8.12. Further elution gave
45 mg (38%) of trans-38 as a colorless liquid: 1H NMR (200
MHz) δ 0.66 (s, 3 H), 0.89 (d, J ) 6.0 Hz, 3 H), 0.99-1.21 (m,
1 H), 1.51-2.03 (m, 6 H), 3.77 (q, J ) 6.0 Hz, 1 H), 7.26-7.43
(m, 6 H), 7.50-7.68 (m, 4 H); 13C NMR (50 MHz) δ -2.3, 18.1,
21.4, 31.2, 34.2, 42.5, 81.3, 127.7, 129.6, 134.3, 136.9; HRMS
calcd for C19H24OSi m/z 296.1597, found 296.1597.
5-Br om o-1-(tr im eth ylsilyl)pen tan -1-on e (10a): IR (neat)
1640 cm-1; 1H NMR (200 MHz) δ 0.21 (s, 9 H), 1.58-1.92 (m,
4 H), 2.65 (t, J ) 6.0 Hz, 2 H), 3.39 (t, J ) 6.0 Hz, 2 H); 13C
NMR (50 MHz) δ -3.4, 20.5, 32.1, 33.2, 47.0, 247.1. Anal.
Calcd for C8H17BrOSi: C, 40.49; H, 7.22. Found: C, 40.12;
H, 7.12.
5-Br om o-1-(tr ip h en ylsilyl)p en ta n -1-on e (10d ). To a
solution of 2.0 g (5.3 mmol) of 8d in 10 mL of THF cooled in
an ice-water bath was added dropwise over 10 min a 1.50 M
solution of n-butyllithium in hexane (4.0 mL, 6.0 mmol). After
another 20 min of stirring, the resulting mixture was added
slowly over 30 min to a solution of 1.20 mL (10.0 mmol) of 9
in 5 mL of dried HMPA cooled in an ice-water bath. The
reaction mixture was stirred at 0 °C for another 1 h and then
partitioned between 150 mL of ether and 100 mL of water.
The organic layer was washed with water (100 mL) and brine
(100 mL), dried (MgSO4), and concentrated. The residual oil
was dissolved in 20 mL of wet THF (15%) followed by the
addition of 2.30 g (10.0 mmol) of red mercuric oxide and 2.3 g
of Celite. The resulting mixture was cooled in an ice-water
bath followed by the addition of 1.65 mL (13.3 mmol) of boron
trifluoride etherate over 5 min. The resulting mixture was
stirred at room temperature for 1 h, diluted with 40 mL of
hexane/ethyl acetate (9/1), and filtered over a short pad of silica
gel. The filtrate was concentrated, and the residual crude
material was recrystallized from a mixture of 15 mL of hexane
and 5 mL of ethyl acetate to give 1.69 g (73%) of 10d as a
white crystal: mp 133.5-134.5 °C; IR (CH2Cl2) 1635 cm-1; 1H
NMR (200 MHz) δ 1.51-1.71 (m, 4 H), 2.73 (t, J ) 6.7 Hz, 2
H), 3.27 (t, J ) 6.5 Hz, 2 H), 7.30-7.50 (m, 9 H), 7.51-7.63
(m, 6 H); 13C NMR (50 MHz) δ 20.8, 32.1, 33.4, 49.3, 128.2,
130.3, 131.1, 136.1, 242.2; HRMS calcd for C23H23BrOSi m/z
422.0702, found 422.0707.
5-Br om o-1-(ter t-bu tyld ip h en ylsilyl)p en ta n -1-on e (10e).
To a solution of 535 mg (1.5 mmol) of 8e in 5 mL of THF cooled
in an dry ice-acetone bath was added dropwise over 10 min
a 1.40 M solution of n-butyllithium in hexane (1.2 mL, 1.7
mmol). The resulting solution was stirred for another 20 min
at the same temperature followed by the addition of 1.05 mL
(6.0 mmol) of HMPA and 0.20 mL (1.7 mmol) of 9. The
resulting mixture was warmed slowly to room temperature
and then stirred for 1 h. The reaction mixture was partitioned
between 100 mL of ether and 50 mL of water. The organic
layer was washed with water (50 mL) and brine (100 mL),
dried, and concentrated. The residual oil was dissolved in 15
mL of wet THF (15%) followed by the addition of 480 mg (2.4
mmol) of red mercuric oxide and 480 mg of Celite. The
resulting mixture was cooled in an ice-water bath followed
by the addition of 0.50 mL (4.0 mmol) of boron trifluoride
etherate over 5 min. The resulting mixture was stirred at
room temperature for 1 h, diluted with 50 mL of ether and
filtered. The filtrate was washed with brine (20 mL), dried
(MgSO4), and concentrated in vacuo. The residual oil was
chromatographed over silica gel (eluted with hexane/ethyl
acetate, 98/2) to give 184 mg (31%) of 10e as a pale yellow oil:
IR (neat) 1635 cm-1; 1H NMR (300 MHz) δ 1.11 (s, 9 H), 1.48-
1.58 (m, 2 H), 1.58-1.71 (m, 2 H), 2.50 (t, J ) 6.7 Hz, 2 H),
3.25 (t, J ) 6.7 Hz, 2 H), 7.32-7.45 (m, 6 H), 7.57-7.66 (m, 4
H); 13C NMR (75 MHz) δ 18.4, 20.7, 27.5, 32.0, 33.2, 49.8, 128.0,
129.9, 131.7, 136.1, 244.4; HRMS calcd for C21H27BrOSi m/z
402.1015, found 402.1019.
Ra d ica l Cycliza tion of 16. 1-(Meth yld ip h en ylsilyl)-
h ep ta n -1-on e (39), cis- a n d tr a n s-2-Meth ylcycloh exyl
Meth yld ip h en ylsilyl Eth er (cis- a n d tr a n s-40). To a
refluxing solution of 100 mg (0.257 mmol) of 16 in 22 mL of
benzene was added via syringe pump over 2 h a solution of
0.090 mL (0.33 mmol) of tributyltin hydride and 3.0 mg (0.019
(41) Curran, D. P.; Chang, C.-T. J . Org. Chem. 1989, 54, 3140.

9098 J . Org. Chem., Vol. 62, No. 26, 1997
Chang et al.
mmol) of AIBN in 7.1 mL of benzene. The resulting solution
was stirred for 2 h at 80 °C and then concentrated in vacuo.
To the residual liquid was added a few drops of wet triethyl-
amine,41 and the resulting mixture was chromatographed over
silica gel (eluted with hexane followed by hexane/ethyl acetate,
99/1) to give 13 mg (17%) of cis-40 as a colorless liquid: 1H
NMR (200 MHz) δ 0.63 (s, 3 H), 0.87 (d, J ) 6.0 Hz, 3 H),
1.08-1.82 (m, 9 H), 3.75-3.90 (m, 1 H), 7.26-7.46 (m, 6 H),
7.50-7.70 (m, 4 H); 13C NMR (50 MHz) δ -2.3, 17.6, 20.9, 24.7,
29.1, 33.0, 36.7, 72.5, 127.7, 129.5, 134.4, 137.2. Anal. Calcd
for C20H26OSi: C, 77.37; H, 8.44. Found: C, 77.37; H, 8.19.
Further elution gave 32 mg (40%) of trans-40 as a colorless
liquid: 1H NMR (200 MHz) δ 0.65 (s, 3 H), 0.93 (d, J ) 6.5
Hz, 3 H), 1.00-1.87 (m, 9 H), 3.24 (td, J ) 10.0, 4.0 Hz, 1 H),
7.28-7.45 (m, 6 H), 7.52-7.68 (m, 4 H); 13C NMR (50 MHz) δ
-2.1, 19.3, 25.1, 25.6, 33.6, 35.8, 40.2, 78.0, 127.7, 129.6, 134.4,
137.1. Anal. Calcd for C20H26OSi: C, 77.37; H, 8.44. Found:
C, 77.17; H, 8.19. Continued elution gave 31 mg (39%) of 39
drops of wet triethylamine,41 and the resulting mixture was
chromatographed over silica gel (eluted with hexane/ethyl
acetate, 98/2) to give 82 mg (83%) of 49 as a colorless liquid:
1H NMR (300 MHz) δ 0.79 (s, 3 H), 2.09 (dtd, J ) 12.7, 8.0,
6.6 Hz, when irradiated at δ 5.42 collapsed into a dt, J ) 12.7,
8.0 Hz, 1 H), 2.30-2.41 (m, when irradiated at δ 5.42 collapsed
into a ddd, J ) 12.7, 8.0, 3.7 Hz, 1 H), 2.76 (dt, J ) 15.8, 8.0
Hz, 1 H), 3.04 (ddd, J ) 15.8, 8.0, 3.7 Hz, 1 H), 5.42 (t, J ) 6.6
Hz, 1 H), 7.15-7.25 (m, 1 H), 7.26-7.35 (m, 1 H), 7.36-7.51
(m, 8 H), 7.70-7.75 (m, 4 H); 13C NMR (75 MHz) δ -2.2, 29.7,
36.2, 77.0, 124.4, 124.6, 126.3, 127.7, 127.8, 129.7, 134.4, 136.4,
142.7, 145.0. Anal. Calcd for C22H22OSi: C, 79.95; H, 6.71.
Found: C, 79.84; H, 6.72.
Ra d ica l Cycliza tion of 28. 1-Cyclop en ten -1-yl Meth -
yld ip h en ylsilyl Eth er (60). To a refluxing benzene (10 mL)
solution of 360 mg (1.0 mmol) of 28 was added via syringe
pump a benzene (10 mL) solution of 0.35 mL (1.3 mmol) of
tributyltin hydride and 25 mg (0.15 mmol) of AIBN over 2 h.
The resulting solution was heated at 80 °C for another 1 h.
Throughout the reaction period, the reaction mixture was
irradiated with two 100 W tungsten lamps. The resulting
solution was concentrated in vacuo. To the residual liquid was
added a few drops of wet triethylamine,41 and the resulting
mixture was chromatographed over silica gel (eluted with
hexane/ethyl acetate, 99/1) to give 173 mg (62%) of 60 as a
1
as a pale yellow liquid: IR (neat) 1679, 1427, 1112 cm-1; H
NMR (200 MHz) δ 0.75 (s, 3 H), 0.82 (t, J ) 7.0 Hz, 3 H), 1.05-
1.34 (m, 6 H), 1.45 (quintet, J ) 7.0 Hz, 2 H), 2.62 (t, J ) 7.0
Hz, 2 H), 7.30-7.49 (m, 6 H), 7.49-7.66 (m, 4 H); 13C NMR
(50 MHz) δ -5.3, 14.0, 22.1, 22.4, 28.8, 31.6, 49.7, 128.1, 130.0,
132.8, 134.9, 244.7. Anal. Calcd for C20H26OSi: C, 77.36; H,
8.44. Found: C, 76.98; H, 8.37.
Gen er a l P r oced u r e for Ra d ica l Cycliza tion s of 12a -c
u sin g Et3B a n d Air for In itia tion . To a solution of the
bromide (1 mmol) and triethylborane (1.3 mmol; 1 M in
hexane) in 5 mL of benzene was added via syringe pump over
2 h a solution of tributyltin hydride (1.3 equiv). During the
same period, a slow stream of dry air was passed through the
solution via syringe pump. The resulting solution was stirred
at room temperature for another hour. The reaction mixture
was then analyzed with gas chromatography using a 3.3 mm
× 3 m column (10% SE30 on chromosorb W, 80-100 mesh)
with a flow rate of 30 mL/min: tR (36a ) ) 4.42 min (130 °C);
tR (37a ) ) 6.19 min (130 °C); tR (36b) ) 6.90 min (170 °C); tR
(37b) ) 9.26 min (170 °C); tR (36c) ) 13.84 min (230 °C); tR
(37c) ) 18.27 min (230 °C).
1
colorless liquid: IR (neat) 1637 cm-1; H NMR (300 MHz) δ
0.72 (s, 3 H), 1.80 (quintet, J ) 7.2 Hz, 2 H), 2.13-2.31 (m, 4
H), 4.57 (t, J ) 2.0 Hz, 1 H), 7.29-7.44 (m, 6 H), 7.54-7.68
(m, 4 H); 13C NMR (75 MHz) δ -2.7, 21.2, 28.6, 33.4, 103.4,
127.8, 129.9, 134.3, 135.7, 154.7; HRMS calcd for C18H20OSi
m/z 280.1284, found 280.1295.
Ack n ow led gm en t. Financial support by the Na-
tional Science Council of the Republic of China is
gratefully acknowledged.
Su p p or tin g In for m a tion Ava ila ble: Details of compound
characterization of 10b,c, 12a ,b,f, 14, 16, 12d ,e,h , 26, 28, 30,
32, 34, 35c-e, 36c-h , 37c-g, 50, 61, 62, 64, and 65; 1H and
13C NMR spectra of new compounds lacking analysis (34
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.
Ra d ica l Cycliza tion of 24. Meth yld ip h en ylsilyl 1-In -
d a n yl Eth er (49). To a refluxing benzene (1.5 mL) solution
of 124 mg (0.302 mmol) of 24 was added via syringe pump a
benzene (1.5 mL) solution of 0.094 mL (0.36 mmol) of tribu-
tyltin hydride and 2.5 mg (0.015 mmol) of AIBN over 2 h. The
resulting solution was heated at 80 °C for another 2 h and
concentrated in vacuo. To the residual liquid was added a few
J O9711302