A tandem radical cyclization route to tricyclo[4.3.n.01,5] alkanes

Article abstract of DOI:10.1016/j.tetlet.2004.08.031

A facile route to tricyclo[4.3.n.0^1,5]alkane skeletons from conjugated cyclic enones was developed through tandem free radical cyclization reaction sequence involving the cyclopropylmethyl radical mediated rearrangement. The scope and limitation of the reaction was investigated. A facile route to tricyclo[4.3.n.0^1,5]alkane skeletons from conjugated cyclic enones was developed through tandem free radical cyclization reaction sequence involving the cyclopropylmethyl radical mediated rearrangement. The scope and limitation of the reaction was investigated.

Full text of DOI:10.1016/j.tetlet.2004.08.031

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 7225–7229
A tandem radical cyclization route to tricyclo[4.3.n.01,5]alkanes
Hee-Yoon Lee,* Sejin Lee, Byung Gyu Kim and Jong Soo Bahnz
Center for Molecular Design and Synthesis, Department of Chemistry and BK21 School of Molecular Science,
Korea Advanced Institute of Science and Technology, Daejeon, 305–701, Korea
Received 3 December 2003; revised 2 August 2004; accepted 6 August 2004
Available online 24 August 2004
Abstract—A facile route to tricyclo[4.3.n.01,5]alkane skeletons from conjugated cyclic enones was developed through tandem free
radical cyclization reaction sequence involving the cyclopropylmethyl radical mediated rearrangement. The scope and limitation
of the reaction was investigated.
Ó 2004 Elsevier Ltd. All rights reserved.
Since the introduction of trialkyltin hydride mediated
radical cyclization reaction,1 free radical cyclization
reaction has been one of the most widely used synthetic
methodologies in organic synthesis.2 Vinyl radical cycli-
zation reaction has been particularly interesting to us
since the cyclization reaction could produce the 6-endo
product as well as the 5-exo product through cyclopro-
pyl methyl radical mediated rearrangement.3 Based on
the observations of the rearrangement in organic synthe-
sis,4 we envisioned a tandem radical cyclization reaction
to construct tricyclo[4.3.2.01,5]undecane structure from
cyclopentane ring and successfully executed the idea in
the total synthesis of suberosenone (Scheme 1).5 It was
believed that the trialkyltin hydride mediated radical
cyclization reaction of 1 produced tricyclic compound
2 via intermediate A and the rearranged intermediate
B as the direct radical cyclization of A would be less
favorable due to the strain energy associated in the
transition state to form the bicyclic structure. With the
successful construction of the tricyclic structure and
the completion of the total synthesis of suberosenone,
we became interested in generalization of the cyclization
reaction.
Herein, we report the scope and limitation of this cycli-
zation reaction.
First, several substrates with various substitution pat-
terns were prepared from cyclopentenone or 2-methyl
cyclopentenone by addition of butynyl Grignard reagent
or butenyl Grignard reagent followed by allylation or
propargylation to furnish the enyne compounds. The
two-step process offered a quick entry into various sub-
strates. The substrates were subjected to the standard
radical cyclization reaction conditions for enynes6 and
the products were treated with SiO2 or 1% HCl(aq)/
MeOH for the destannylation reaction. The result was
summarized in Table 1.
i) Bu3SnH,
TMSO
AIBN(cat.)
ii) HCl
OH
OH
2
OTMS
1
OTMS
OTMS
SnBu3
SnBu3
TMSO
TMSO
The cyclization reaction appeared to be quite general as
all the substrates underwent tandem cyclization reaction
when W was the methyl group that was believed to drive
the rearrangement toward the more stable tertiary radi-
cal intermediates like B in Scheme 1 (entries 1–5). When
W was the hydrogen that could not provide extra stabi-
lization after the rearrangement, neither rearrangement
nor tandem cyclization reaction was observed (entry 6).
A
B
Scheme 1. Tandem cyclization through rearrangement.
Keywords: Radical; Cyclization; Mechanism.
*
Corresponding author. Tel.: +82 428692835; fax: +82 428698370;
e-mail: leehy@kaist.ac.kr
z Deceased, 31 July 2003.
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.08.031

7226
H.-Y. Lee et al. / Tetrahedron Letters 45 (2004) 7225–7229
Table 1. Tandem radical cyclization to tricyclo[4.3.2.01,5]undecanes5
W
i) Bu3SnH, AIBN (cat.)
R
Y
Y
ii) SiO2
X
X
W
R
Entry
1
Substrate
Producta a:bc
Yieldb (%)
89
O
O
8a:8b (2:1)
2
76
O
O
4:1
3
4
5
6
80
Ph
O
O
Ph
a only
60
O
O
Ph
Ph
Ph
Ph
3:1
61
O
O
64d
O
O
a The reaction conditions: the substrate was refluxed with Bu3SnH (1.5equiv) and AIBN (0.1equiv) in benzene (0.01M) for 2h and the cooled
reaction mixture was stirred with SiO2 overnight.
b Isolated yield after column chromatography.
c The ratio was determined by 1H NMR.
d The reaction concentration was 0.005M.
Lowering the concentration of the reaction to promote
the rearrangement did not show any sign of the forma-
tion of the rearranged product. The result in entry 6
clearly supported the tandem radical cyclization–rear-
rangement–cyclization sequence as shown in Scheme 1
to form the tricyclic compounds. Since the substrate
does not have the methyl group to stabilize the rear-
ranged intermediate, the rearrangement did not occur
and the reaction stopped after the first radical cycliza-
tion reaction and was quenched to form the diquinane
structure.
for the synthesis of tricyclo[4.3.2.01,5]undecanes. The
cyclization reaction proceeded smoothly to produce
the desired tricyclic compounds (Table 2).
Contrary to the tricyclo[4.3.2.01,5]undecanes where
tandem radical cyclization reaction has exclusive prefer-
ence toward one isomeric product of the ring skele-
ton, others could produce mixtures of skeletal isomers.
When the ring size of the substrate was six (n = 2), a
mixture of two skeletal isomers was obtained in
equal amount. When the ring size of the substrate was
larger than six, the skeletal isomer that was not found
in the tricyclo[4.3.2.01,5]undecane became the sole
isomeric product from the cyclization reaction. This
unexpected selectivity could be due to the increasing
steric interaction between the existing ring and the five-
membered ring during the final cyclization sequence
for the formation of the isomer like 8. This result is
also crucial for the applicability of the cyclization reac-
tion to the total synthesis of pleuromutiline, as pleuro-
mutiline has the same skeletal isomer as the
With the successful construction of tricyclo-
[4.3.2.01,5]undecanes, we decided to explore the possibil-
ity of generalization of the reaction to construct
tricyclo[4.3.n.01,5]alkanes since generalization of this
tandem cyclization reaction for other ring sizes would
include tricyclo[4.3.5.01,5]tetradecane system that could
serve as the key intermediate for the synthesis of pleu-
romutiline.7 Precursors for the cyclization were prepared
from cyclic enones in the same way as for the substrates

H.-Y. Lee et al. / Tetrahedron Letters 45 (2004) 7225–7229
Table 2. Tandem radical cyclization to tricyclo[4.3.n.01,5]alkanes
7227
Next we turned our attention to the formation of carbo-
cyclic compounds. The precursors for the carbocyclic
compounds were prepared through the same sequence
as the preparation of the precursor for the radical
cyclization in suberosenone synthesis. After direct but-
enylation or butynylation of 2-carboethoxy cyclopent-
anone, the ketone was protected as the corresponding
acetal and the remaining ester was converted into
the other required alkene or alkyne appendage for the
tandem cyclization reaction (entries 3 and 4). For the
substrates in entries 1 and 2, 2-carboethoxy cyclopenta-
none was alkylated twice as dianions, and the decarb-
oxylation produced the corresponding cyclopentanone
with proper appendages. The result of the radical cycliz-
aiton reaction was summarized in Table 3.
i) Bu SnH, AIBN (cat.)
3
products
()
n
ii) SiO
2
O
n
Producta,b
Yieldc (%)
2
64
O
O
O
2:3
3
4
80
40
9
Though W could be any functional group capable of sta-
bilizing the rearranged radical intermediate, the size of
W also played an important role for the tandem cycliza-
tion as OTBS group blocked the second cyclization to
form the tricyclic structure after the rearrangement
(entry 1). When the OTMS group replaced the OTBS
group, the tandem cyclization reaction proceeded
smoothly (entries 2–4). Contrary to the expectation,
the structure of the product in entry 4 was not the one
we expected from the tandem radical cyclization involv-
ing the rearrangement. Since only the terminal alkyne
can form the initial vinyl radical intermediate, the prod-
uct should have been 3 with the hydroxyl group on the
five-membered ring (Scheme 2).
O
10
a The reaction conditions: the substrate was refluxed with Bu3SnH
(1.5equiv) and AIBN (0.1equiv) in benzene (0.01M) for 2h and the
cooled reaction mixture was stirred with SiO2 overnight.
b Relative stereochemistry was assigned through NOESY experiment.
c Isolated yield after column chromatography.
tricyclo[4.3.5.01,5]tetradecane compound 10 shown in
Table 2.
Table 3. Tandem radical cyclization to tricyclo[4.3.2.01,5]undecanes
W
i) Bu3SnH, AIBN (cat.)
R
Y
Y
ii) SiO2
R
X
X
W
Entry
1
Substrate
OTBS
Producta a:bc
Yieldb (%)
64
TBSO
OTMS
2
3
4
59
OTMS
1:1
OTMS
55d
OH
OH
TMSO
TMSO
1:1
OTMS
TMS
OH
60d
OH
TMS
1:1
a The reaction conditions: the substrate was refluxed with Bu3SnH (1.5equiv) and AIBN (0.1equiv) in benzene (0.01M) for 2h and the cooled
reaction mixture was stirred with SiO2 overnight.
b Isolated yield after column chromatography.
c The ratio was determined by 1H NMR.
d The reaction mixture was treated with 1% HCl after the radical cyclization.

7228
H.-Y. Lee et al. / Tetrahedron Letters 45 (2004) 7225–7229
also explained why 2 was the exclusive product from 1
i) Bu3SnH,
AIBN(cat.)
TMSO
TMS
during the total synthesis of subserosenone. When 1 re-
acted with the tin radical, the tin radical added to either
of the two terminal alkynes to form two different inter-
mediates. While one rearranged to form more stable
radical intermediate before the second radical cycliza-
tion reaction, the other intermediate underwent the sec-
ond radical cyclization reaction before the
rearrangement. Then, both cyclization products eventu-
ally produced 2. It became evident that addition of a
substituent with the proper stereochemistry facilitated
the cyclization reaction of the tether9 and could alter
the reaction pathway.
OH
ii) HCl
TMS
OH
OTMS
Bu3SnH
SnBu3
4
OH
OH
OH
C
OH
3
Scheme 2. Expected cyclization route to 3 and 4.
The formation of 4 was initially rationalized by acid cat-
alyzed rearrangement of 3 to 4, as the tertiary alcohol of
3 could form an allylic cation easily and could undergo
similar rearrangement as the cyclopropyl methyl radical
rearrangement. However, a careful examination of the
structure of 4 revealed that the stereochemistry of the
alcohol on the tether might play an important role in
forming 4 as the stereochemistry of the alcohol in the
product was the more stable equatorial isomer. The
equatorially positioned alcohol might have promoted
the direct cyclization into the six-membered ring from
the intermediate similar to A. To probe this possibility,
we subjected 5a and 5b to the radical cyclization reac-
tion separately. Two stereoisomers of the alcohol would
provide the information on the effect of stereochemistry
of the alcohol on the cyclization reaction. When 5a and
5b were subjected to the radical cyclization reaction con-
dition, structurally different products were obtained
(Scheme 3).8
In conclusion, a reliable and general synthetic method to
construct tricyclo[4.3.n.01,5]alkane skeleton was devel-
oped. It was clearly demonstrated that the reaction
pathway could be different to form the tricyclic com-
pounds depending on the presence and the stereochem-
istry of substituents in the tether.
Acknowledgements
This work was supported by the grant from the Korea
Research Foundation (KRF-2002-070-C-00058).
References and notes
1. Stork, G.; Baine, N. H. J. Am. Chem. Soc. 1982, 104, 2321.
2. (a) Curran, D. P. Aldrichim. Acta 2000, 33, 104; (b) Giese,
B.; Kopping, B.; Goebel, T.; Dickhaut, J.; Thoma, G.;
Kulicke, J. J.; Trach, F. Org. React. 1996, 48, 301; (c)
Jasperse, C. P.; Curran, D. P.; Fevig, T. L. Chem. Rev.
1991, 91, 1237.
3. (a) Gomez, A. M.; Company, M. D.; Uriel, C.; Valverde,
S.; Lopez, J. C. Tetrahedron Lett. 2002, 43, 4997; (b) Stork,
G., Jr.; Mook, R., Jr. Tetrahedron Lett. 1986, 27, 4525.
4. (a) Stork, G.; West, F.; Lee, H. Y.; Isaacs, R. C. A.;
Manabe, S. J. Am. Chem. Soc. 1996, 118, 10660; (b) Clive,
D. L. J.; Magnuson, S. R.; Manning, H. W.; Mayhew, D.
L. J. Org. Chem. 1996, 61, 2095; (c) Toyota, M.; Yokota,
M.; Ihara, M. Org. Lett. 1999, 1, 1627.
In the reaction of 5a, the anticipated product 6 was ob-
tained following the radical cyclization–rearrangement–
cyclization sequence. On the other hand, 5b produced
the unanticipated product 7 that was clearly the product
of the tandem radical cyclization reaction without the
rearrangement. While the hydroxyl group of 5a was lo-
cated at the axial position and showed 1,3-diaxial inter-
action with the other hydroxyl group that could not
provide stabilization of the transition state for the direct
cyclization, the hydroxyl group of 5b was positioned at
the equatorial position during the cyclization reaction
to stabilize the six-membered ring transition state for
the direct cyclization right after the first radical cycliza-
tion before the rearrangement could proceed. This result
5. Lee, H. Y.; Kim, B. G. Org. Lett. 2000, 2, 1951.
6. Stork, G.; Mook, R., Jr. J. Am. Chem. Soc. 1987, 109, 2829.
7. Bacque, E.; Pautrat, F.; Zard, S. Z. Org. Lett. 2003, 5, 325,
and references cited therein.
8. All the cyclization products gave satisfactory spectroscopic
data for the structures. Spectral data of selected com-
1
pounds; 4 H NMR (400MHz, CDCl3): d 5.09 (t, 1H, J =
2.6Hz), 5.04 (t, 1H, J = 2.2Hz), 4.84 (m, 2H), 3.98 (m, 1H),
2.73 (d, 1H, J = 5.8Hz), 2.68 (m, 1H), 2.54 (m, 1H), 2.49 (d,
1H, J = 6.3Hz), 2.16 (m, 1H), 1.92 (ddd, 1H, J = 15.3, 10.3,
5.2Hz), 1.78–1.69 (m, 3H), 1.55 (b s, 1H), 1.50(b s, 1H),
1.37 (ddd, 1H, J = 14.6, 10.3, 4.3Hz), 1.15 (m, 1H). 13C
NMR (100MHz, CDCl3): d 153.9, 148.3, 111.0, 108.1, 86.9,
TMSO
a
+
60%
OH
OH
OH
OH
OTMS
5a
6a 1 : 2
6b
72.9, 55.4, 50.2, 37.3, 28.7, 28.6, 25.5, 24.7. IR (neat, cmÀ1
)
3411, 2944, 1651, 1086, 1041, 999, 892. HRMS: calcd for
C17H24O3: 206.1307, found: 206.1307. 6a 1H NMR
(400MHz, CDCl3): d 4.78 (t, 1H, J = 2.3Hz), 4.70(t, 1H,
J = 2.3Hz), 3.92 (m, 1H), 2.66 (d, 1H, J = 6.3Hz), 2.52 (m,
1H), 2.36 (s, 1H), 2.34 (m, 1H), 2.22–2.15 (m, 2H), 2.01 (m,
1H), 1.85–1.76 (m, 2H), 1.66 (m, 1H), 1.56–1.45 (m, 2H),
1.28 (m, 1H), 1.05 (d, 3H, J = 7.7Hz). 13C NMR (100MHz,
TMSO
a
OH
55%
OH
OTMS
5b
7
Scheme 3. Reagents: (a) Bu3SnH, AIBN; SiO2.

H.-Y. Lee et al. / Tetrahedron Letters 45 (2004) 7225–7229
7229
CDCl3): d 151.0, 110.4, 90.2, 77.8, 54.2, 52.0, 42.6, 41.1,
30.5, 27.8, 26.6, 25.7, 17.6. 6b 1H NMR (400MHz, CDCl3):
d 4.54 (t, 1H, J = 2.3Hz), 4.50(t, 1H, J = 2.3Hz), 3.88 (d,
1H, J = 3.8Hz), 2.79 (m, 1H), 2.66 (t, 1H, J = 5.8Hz), 2.39–
2.31 (m, 2H), 2.24 (dd, 1H), 2.14 (m, 1H), 2.01 (dd, 1H,
J = 14.2, 8.4Hz), 1.68 (b, 1H), 1.58 (m, 1H), 1.47–1.38 (m,
3H), 1.26 (dd, 1H, J = 11.6, 3.0Hz), 1.03 (d, J = 7.2Hz).
13C NMR (100MHz, CDCl3): d 153.7, 104.5, 89.8, 76.8,
58.9, 45.9, 44.2, 42.9, 41.3, 40.6, 27.8, 27.7, 18.7. 7 1H NMR
(400MHz, CDCl3): d 5.03 (t, 1H, J = 2.6Hz), 4.94 (t, 1H,
J = 2.3Hz), 3.97 (m, 1H), 2.63 (m, 1H), 2.50(m, 1H), 2.36
(m, 1H), 1.87 (dd, 1H, J = 4.0, 2.1Hz), 1.78–1.65 (m, 4H),
1.62 (m, 1H), 1.41 (m, 1H), 1.32 (m, 2H), 0.99 (m, 1H), 0.87
(d, 3H, J = 6.9Hz), 0.85 (m, 1H). 13C NMR (100MHz,
CDCl3): d 156.0, 106.8, 87.7, 72.3, 54.8, 45.0, 36.7, 28.7,
28.6, 28.5, 24.9, 19.7, 18.8. 8a 1H NMR (CDCl3, 400MHz):
d 4.56 (t, 1H, J = 2.04Hz), 4.50 (t, 1H, J = 2.04Hz), 4.20 (t,
1H, J = 9.0Hz), 3.44 (t, 1H, J = 9.0Hz), 2.70 (t, 1H,
J = 7.2Hz), 2.54 (m, 1H), 2.36 (m, 1H),2.22 (dd, 1H,
J = 14.1, 7.2Hz), 1.97 (m, 2H), 1.81 (m, 2H), 1.18 (dd, 1H,
J = 14.1, 2.0Hz), 1.11 (m, 1H), 1.12 (s, 3H), 0.95 (d, 3H,
J = 7.2Hz). 13C NMR (CDCl3, 100MHz): d 152.60, 106.80,
91.77, 73.21, 48.41, 45.16, 45.04, 43.68, 37.94, 31.25, 27.72,
25.74, 17.18, IR (neat, cmÀ1) 2953, 2870, 2285, 1646, 1446,
1375, 1235, 1153, 1112, 1058, 1028, 949, 883. HRMS: calcd
for C13H20O: 192.1514, found: 195.1525. 8b 1H NMR
(CDCl3, 400MHz): d 4.59 (t, 1H, J = 2.08Hz), 4.54 (t, 1H,
J = 2.08Hz), 4.12 (t, 1H, J = 8.6Hz), 3.50(t, 1H,
J = 9.9Hz), 2.52 (m, 1H), 2.49 (d, 1H, J = 6.0Hz), 2.25
(m, 2H), 2.10(m, 1H), 1.87 (m, 1H), 1.74–1.66 (m, 3H), 1.35
(m, 1H), 0.88 (d, 3H, J = 6.8Hz), 0.69 (s, 3H). 13C NMR
(CDCl3, 100MHz): d 150.74, 107.82, 90.54, 73.60, 51.33,
50.86, 36.92, 33.40, 31.33, 26.94, 25.49, 12.42, 11.88. IR
(neat, cmÀ1) 2959, 2874, 1775, 1720, 1684, 1644, 1471, 1380,
1339, 1194, 1151, 1084, 1062, 1009, 993, 907, 883 HRMS:
1
calcd for C13H20O: 192.1514, found: 195.1525. 9 H NMR
(CDCl3, 400MHz): d 4.58 (m,2H), 3.90(t, 1H, J = 8.7Hz),
3.41 (dd, 1H, J = 9.8, 8.3Hz), 2.57 (m, 2H), 2.39 (d, 1H,
J = 9.8Hz), 2.19 (m, 2H), 1.92 (m, 2H), 1.75 (m, 4H), 1.66
(m, 2H), 1.45 (m, 1H), 0.98 (s, 3H), 0.75 (d, 3H, J = 6.7Hz).
13C NMR (CDCl3, 100MHz): d 151.50, 108.25, 85.29,
71.01, 48.00, 45.33, 40.39, 37.78, 34.55, 30.92, 30.24, 26.56,
22.79, 15.64, 10.33 IR (neat, cmÀ1) 2924, 2867, 1647, 1468,
1381, 1253, 1111, 1043, 1030, 1018, 958, 932, 885 HRMS:
calcd for C15H24O: 220.1827, found: 220.1832. 10 1H NMR
(CDCl3, 400MHz): d 4.82 (t, 1H, J = 2.3Hz), 4.71 (t, 1H,
J = 2.3Hz), 3.82 (t, 1H, J = 8.7Hz), 3.30(dd, 1H, J = 9.8,
8.2Hz), 2.66–2.55 (m, 2H), 2.43 (m, 1H), 2.29 (dd, 1H),
2.00–1.80 (m, 5H), 1.74–1.56 (m, 5H), 1.40 (m, 2H), 0.88 (s,
3H), 0.75 (d, 3H, J = 6.7Hz). 13C NMR (CDCl3, 100MHz):
d 148.96, 111.78, 84.33, 68.80, 47.00, 45.09, 41.87, 38.36,
34.13, 33.17, 33.06, 29.31, 26.50, 26.09, 15.82, 10.88. IR
(neat, cmÀ1) 2922, 2859, 1641, 1465, 1452, 1434, 1384, 1264,
1235, 1196, 1128, 1088, 1035, 952, 925, 889. HRMS: calcd
for C16H26O: 234.1984, found: 234.1986.
9. Beckwith, A. L. J.; Moad, G. Chem. Commun. 1974, 472.