Free-radical-5-exo-trig cyclization of chiral 3-silylhepta-1,6-dienes: Concise approach to the A-B-C ring core of hexacyclinic acid

Article abstract of DOI:10.1021/jo051062g

Free-radical-mediated 5-exo-trig cyclization of hepta-1,6-dienes 6a-c incorporating an allylsilane moiety was shown to provide at low temperature exclusively the trans-cis cyclopentanes 7a-c in good yield. In contrast, the same reaction with the alkyl analogue 9a led to the formation of all four possible stereoisomers. Interestingly, extension of this sequence of radical processes to alkoxy analogues 12-14 provided the complementary cis-cis stereoisomers with modest to excellent stereoselectivity. It is noteworthy that under such conditions allylsilanes were found to be much more reactive than their alkyl and alkoxy analogues. Beckwith-Houk-type models were proposed that rationalize the stereochemical course of these 5-exo-trig cyclizations. Finally, an illustration of the value of this methodology was proposed with the synthesis of the A-B-C tricyclic unit of polyketide hexacyclinic acid 3a.

Full text of DOI:10.1021/jo051062g

Free-Radical-5-exo-Trig Cyclization of Chiral
3-Silylhepta-1,6-dienes: Concise Approach to the A-B-C Ring
Core of Hexacyclinic Acid†
Philippe James,‡ Franc¸ois-Xavier Felpin,‡ Yannick Landais,*,‡ and Kurt Schenk§
Universite´ Bordeaux-I, Laboratoire de Chimie Organique et Organome´tallique,
351, Cours de la Libe´ration, 33405 Talence Cedex, France and Universite´ de Lausanne,
Institut de Cristallographie, BPS Dorigny, CH-1015 Lausanne, Suisse
y.landais@lcoo.u-bordeaux1.fr
Received May 27, 2005
Free-radical-mediated 5-exo-trig cyclization of hepta-1,6-dienes 6a-c incorporating an allylsilane
moiety was shown to provide at low temperature exclusively the trans-cis cyclopentanes 7a-c in
good yield. In contrast, the same reaction with the alkyl analogue 9a led to the formation of all
four possible stereoisomers. Interestingly, extension of this sequence of radical processes to alkoxy
analogues 12-14 provided the complementary cis-cis stereoisomers with modest to excellent
stereoselectivity. It is noteworthy that under such conditions allylsilanes were found to be much
more reactive than their alkyl and alkoxy analogues. Beckwith-Houk-type models were proposed
that rationalize the stereochemical course of these 5-exo-trig cyclizations. Finally, an illustration
of the value of this methodology was proposed with the synthesis of the A-B-C tricyclic unit of
polyketide hexacyclinic acid 3a.
SCHEME 1. Natural Products Containing
Cyclopentanes
Introduction
Five-membered ring systems are common structural
units in natural products (i.e., 1-3, Scheme 1) and are
ubiquitous in Nature.1 Important efforts have thus been
devoted to construct a polysubstituted five-membered-
ring core in a diastereo- and enantiocontrolled manner.2
The cyclopentane skeleton may be embedded in a com-
plex polycyclic system as in 3a,b or may represent the
† Taken as a part of the Ph.D. thesis of P.J., University Bordeaux-I
(Oct 2004).
‡ Universite´ Bordeaux-I.
§ Universite´ de Lausanne.
(1) (a) 1 is produced in vivo in Humans, see: Taber, D. F.; Xu, M.;
Hartnett, J. C. J. J. Am. Chem. Soc. 2002, 124, 13121-13126 and
references therein. (b) 2 was isolated from marine diatom, see: Wang,
R.; Shimizu, Y. J. Chem. Soc., Chem. Commun. 1990, 413-413. (c) 3a
was isolated from Streptomyces cellulosae subsp, see: Ho¨fs, R.; Walker,
M.; Zeeck, A. Angew. Chem., Int. Ed. 2000, 39, 3258-3261. (d) 3b was
isolated from Streptomyces sp. No9885, see: Sato, B.; Muramatsu, H.;
Miyauchi, M.; Hori, Y.; Takase, S.; Hino, M.; Hashimoto, S.; Terano,
H. J. Antibiot. 2000, 53, 123-130.
(2) (a) Herndon, J. W. Tetrahedron 2000, 56, 1257-1280. (b) Hartley,
R. C.; Caldwell, S. T. J. Chem. Soc., Perkin Trans 1 2000, 477-501.
(c) Hassner, A.; Ghera, E.; Yechezkel, T.; Kleiman, V.; Balasubrama-
nian, T.; Ostercamp, D. Pure Appl. Chem. 2000, 72, 1671-1683. (d)
Thebtaranonth, C.; Thebtaranonth, Y. Tetrahedron 1990, 46, 1385-
1489.
key structural element as in 1 or 2. 5-exo-Trig cyclizations
are known to be efficient processes under anionic3 and
10.1021/jo051062g CCC: $30.25 © 2005 American Chemical Society
Published on Web 09/02/2005
J. Org. Chem. 2005, 70, 7985-7995
7985

James et al.
SCHEME 2. 5-exo-Trig Cyclization of Chiral
in systems such as 5 would allow, during step 5 f 4
(Scheme 2), control of the relative configuration between
stereogenic centers at C2 and C3 (1,2-stereocontrol) but
also between C2 and C6 (1,5-stereocontrol), taking ad-
vantage of both the steric and electronic effects of the
silicon group. Our study was effectively based on the
assumption that the steric hindrance of SiR3 would lead
to a decrease in the number of reactive conformations at
the cyclization transition state9a,d and that the partial
positive charge developing at C2 upon cyclization could
be stabilized by the â-silicon group,10,11 thus reinforcing
the inherent steric effect of this group. Precedent in the
laboratory on radical functionalization of chiral allylsi-
lanes supported such a hypothesis.12
We thus provide here a full account of our studies on
free-radical-mediated 5-exo-trig cyclization of dienes of
type 5 having an allylsilane moiety. The study was also
extended to models having various substituents at the
allylic position (X ) alkyl, OR) with the aim of presenting
a consistent picture of the effects controlling the stereo-
chemistry of this 5-exo-trig process. Finally, an illustra-
tion of the utility of such an approach for the stereose-
lective construction of cyclopentanoids is provided with
a concise synthesis of a precursor of the ABC ring core
of hexacyclinic acid 3a, the total synthesis of which has
not yet been disclosed.
1,6-Heptadienes
free-radical conditions4 and thus constitute methods of
choice to elaborate five-membered ring skeletons in a
limited number of steps with generally high selectivity
and yield. Among the cyclization processes, radical-
mediated cyclizations4 have attracted a great deal of
attention and hold a special place. The kinetics of these
transformations have been studied in depth, and rate
constants are available for most systems possessing a
wide range of atoms within the ring (C, O, N, Si,...).5
We recently started a study on the elaboration of
polysubstituted five-membered ring systems 4 relying on
the radical cyclization of 1,6-dienes 5 containing an
allylsilane moiety (Scheme 2).6 Our strategy was based
on utilization of a radical cascade involving addition of a
tosyl group, 5-exo-trig cyclization followed, by â-fragmen-
tation. This approach is well known to lead, with gener-
ally high efficiency and under mild conditions, to the
corresponding five-membered ring (i.e., 4) having two
new contiguous stereogenic centers.7 Several examples
of this transformation have been reported which empha-
size the problem of stereocontrol arising in this process.8
While an excellent level of 1,2-stereocontrol is generally
granted with systems having an allylic substituent (at
C3), generally poor to modest 1,5-stereocontrol is ob-
served in those cases. Studies in our group and others9
have convincingly demonstrated that in various processes
involving chiral organosilicon compounds with a silicon
group on a stereogenic center a high level of stereocontrol
could be attained. The presence of a bulky silicon group
within a chain restricts the number of reactive conforma-
tions at the transition state, leading to a high level of
transfer of chiral information. It was thus anticipated
that the presence of a silicon group in the allylic position
Results and Discussion
5-exo-Trig Cyclizations of Allylsilane Precursors.
Preliminary investigations were carried out using two
distinct 1,6-diene models 6a6 and 6b,c6 having, respec-
tively, one and two stereogenic centers. We started our
studies by performing the radical 5-exo-trig cyclizations
under thermal conditions using AIBN as an initiator and
a catalytic amount of p-TolSO2SePh. Reaction of 6a led
to the desired cyclopentanes 7a and 8a in reasonable
yield with modest stereocontrol (Entry 1, Table 1). A
combination of 1D and 2D NMR studies unambiguously
established that the major isomer 7a possessed the trans-
cis relative configuration, while the minor isomer 8a
possessed the trans-trans configuration. 1,2-Stereocontrol
was thus excellent, in good agreement with previous
reports from the literature,8 but 1,5-stereocontrol re-
mained modest. To our delight, much better results were
obtained by performing the reaction at room temperature
(3) Clayden, J. Organolithiums: Selectivity for Synthesis; Baldwin,
J. E., Williams, R. M., Eds.; Tetrahedron Organic Chemistry Series;
Pergamon: Amsterdam, 2002; Vol. 23, Chapter 7, pp 274-335.
(4) (a) Beckwith, A. L. J. Tetrahedron 1981, 37, 3073-3100. (b)
Majumdar, K. C.; Mukhopadhyay, P. P.; Basu, P. K. Heterocycles 2004,
63, 1903-1958 and references therein.
(9) (a) Fleming, I.; Dunogue`s, R.; Smithers, R. Org. React. 1989, 37,
57-575. (b) Fleming, I.; Barbero, A.; Walter, D. Chem. Rev. 1997, 97,
2063-2192. (c) Masse, C. E.; Panek, J. S. Chem. Rev. 1995, 95, 1293-
1316. (d) Fleming, I. J. Chem. Soc., Perkin Trans 1 1992, 3363-3369.
(e) Chabaud, L.; James, P.; Landais, Y. Eur. J. Org. Chem. 2004, 3173-
3199.
(10) For reviews on the â-silicon effect, see: (a) Panek, J. S. Silicon
Stabilization. In Comprehensive Organic Chemistry; Trost, B., Fleming,
I., Eds.; 1989; Vol. 1, 579-627. (b) Lambert, J. B. Tetrahedron 1990,
46, 2677-2689. (c) Jarvie, A. W. P. Organomet. Chem. Rev., Sect. A
1970, 6, 153-207. (d) White, J. M. Aust. J. Chem. 1995, 48, 1227-
1251.
(11) For the stabilization of â-carboradicals by a silicon group, see:
(a) Kawamura, T.; Kochi, J. K. J. Am. Chem. Soc. 1972, 94, 648-650.
(b) Griller, D.; Ingold, K. U. J. Am. Chem. Soc. 1974, 96, 6715-6720.
(c) Jackson, R. A.; Ingold, K. U.; Griller, D.; Nazran, A. S. J. Am. Chem.
Soc. 1985, 107, 208-211. (d) Auner, N.; Walsh, R.; Westrup, J. J. Chem.
Soc., Chem. Commun. 1986, 207-207.
(12) (a) Chabaud, L.; Landais, Y.; Renaud, P. Org. Lett. 2002, 4,
4257-4260. (b) Porter, N. A.; Wu, J. H.; Zhang, G.; Reed, A. D.
Tetrahedron Lett. 2000, 41, 5773-5777. (c) Chabaud, L.; Landais, Y.;
Renaud, P. Org. Lett. 2005, 7, 2587-2590.
(5) (a) Beckwith, A. L. J.; Glover, S. A. Aust. J. Chem. 1987, 40,
157-173. (b) Wilt, J. Tetrahedron 1985, 41, 3979-4000. (c) Wilt, J.
W.; Lusztyk, J.; Peeran, M.; Ingold, K. U. J. Am. Chem. Soc. 1988,
110, 281-287.
(6) James, P.; Landais, Y. Org. Lett. 2004, 6, 325-328.
(7) (a) Bertrand, M.-P.; De Riggi, I.; Lesueur, C.; Gastaldi, S.;
Nouguier, R.; Jaime, C.; Virgili, A. J. Org. Chem. 1995, 60, 6040-
6045. (b) Wang, C.; Russell, G. A. J. Org. Chem. 1999, 64, 2346-2352.
(c) Harvey, I. W.; Philips, E. D.; Whitham, G. H. Tetrahedron 1997,
53, 6493-6508. (d) Harvey, I. W.; Whitham, G. H. J. Chem. Soc., Perkin
Trans 1 1993, 185-190. (e) Harvey, I. W.; Whitham, G. H. J. Chem.
Soc., Perkin Trans 1 1993, 191-196. (f) Smith, T. A. K.; Whitham, G.
H. J. Chem. Soc., Perkin Trans 1 1989, 319-325. (g) Chuang, C.-P.
Tetrahedron 1991, 47, 5425-5436.
(8) (a) Bertrand, M.-P.; Gastaldi, S.; Nouguier, R. Tetrahedron Lett.
1996, 37, 1229-1232. (b) Bertrand, M.-P.; Gastaldi, S.; Nouguier, R.
Tetrahedron 1998, 54, 12829-12840. (c) Nouguier, R.; Gastaldi, S.;
Stien, D.; Bertrand, M.-P.; Renaud, P. Tetrahedron Lett. 1999, 40,
3371-3374. (d) Denes, F.; Chemla, F.; Normant, J. F. Angew. Chem.,
Int. Ed. Engl. 2003, 47, 4043-4046.
7986 J. Org. Chem., Vol. 70, No. 20, 2005

Cyclization of Chiral 3-Silylhepta-1,6-dienes
TABLE 1. Sulfonyl Radical-Mediated 5-exo-Trig
Cyclization of Dienes 6a-c
SCHEME 3. Sulfonyl Radical-Mediated 5-exo-Trig
Cyclization of Dienes 9a,b
entry diene solvent cat. T (°C) time (h)
d.r.a
yield (%)b
and large alkyl groups in allylic positions and allow
comparisons with the silicon group.13 Oxygenated sub-
stituents (OH, OR, and OCOR groups) were also studied
as it was anticipated that an alkoxy group (or an ester)
would introduce opposite electronic effects to that of a
silicon group.14 It was thus envisioned that this series of
experiments would provide some insight into the effects
(steric or/and electronic) governing the stereochemistry
of these 5-exo-trig cyclizations.
Precursor 9a bearing an allylic methyl substituent was
thus prepared in six steps from (R)-(-)-citronellene.6 9b
possessing a t-Bu group in allylic position was also
prepared in four steps from 4-tert-butylhex-5-enal (itself
prepared in five steps from 4-tert-butylcyclohexanone).15
Precursors 9a,b were then treated as above for allylsi-
lanes 6a-c using 0.25 equiv of p-TolSO2SePh in benzene
or CHCl3 at various temperatures (Scheme 3). 9a thus
afforded the expected cyclopentane 10 but as an insepa-
rable mixture of four diastereomers whatever the tem-
perature. When the reaction was carried out at -78 °C,
only trace amounts of cyclopentanes were observed after
8 h, showing the lack of reactivity of alkyl precursors as
compared to their silylated analogues.16 Similarly, pre-
cursor 9b exhibited no reactivity at room temperature,
and cyclization product 11 was only formed after 7 h at
80 °C with a modest conversion (57%) but as a single
diastereomer having trans-cis stereochemistry, as shown
by X-ray crystallographic analysis (Supporting Informa-
tion). Finally, under similar conditions the ramified
isomer of 9b also led to cyclopentane 11 with only 15%
conversion.
1
2
6a C6H6
6a C6H6
6a CCl4
6a CHCl3 0.15 -50
6a CHCl3 0.25 -50
6b C6H6
0.15
0.15
80c
16
10
4
86:14
96:4
69
83
85
84
85
85
72
79
56
61
3
0.15 -15
2.5
98:2
4
3
>99:<1
>99:<1
81:19
95:5
95:5
>99:<1
>99:<1
5
0.5
8
6
0.15
16
7
6b CHCl3 0.15 -50
6b CHCl3 0.25 -50
3
8
0.5
3.5
0.5
9
6b CH2Cl2 0.5
-78
10
6c CHCl3 0.25 -50
a Estimated from 1H NMR and GC analysis of the crude reaction
mixture. b Isolated yields after purification through chromatog-
raphy. c AIBN was used as an initiator.
under irradiation (sun lamp), where cyclopentane 7a was
obtained in 92% de in an improved 83% yield (entry 2).
Lowering the temperature to -15 °C and then -50 °C
finally led to complete stereocontrol with 7a obtained as
a unique diastereomer (GC) (entries 3 and 4). Increasing
the amount of catalyst from 0.15 to 0.25 equiv also
allowed the reaction to be carried out in only 0.5 h at
-50 °C with the same yield (entry 5). It is worth noting
that the stereochemistry of the starting sulfone 6a had
no effect on the stereocontrol as the reaction carried out
on a 10:1 E/Z mixture led to a 98:2 7a/8a ratio at -50 °C
(with 0.25 equiv of p-TolSO2SePh in CHCl3, 0.5 h). These
conditions were then applied to allylsilane 6b, which
afforded the corresponding cyclopentanes 7b/8b with a
high level of stereocontrol. At -50 °C (entries 7 and 8)
the stereocontrol was slightly lower than that observed
for analogue 6a. Complete stereocontrol, albeit with
moderate yield, was however observed at -78 °C using
0.5 equiv of catalyst (entry 9). The relative configuration
of 7b was finally secured through X-ray crystallographic
structure determination, thus supporting the configura-
tion of 7a established through NMR studies in advance.
Similarly, cyclization of diene 6c under optimized condi-
tions led to cyclopentane 7c as a unique stereoisomer
(entry 10).
Precursor 12 having an allylic alcohol moiety was
prepared from hexa-1,5-diene.6,17 Tuning the electronic
(13) An A value of 4.7-4.9 has been estimated for a t-Bu group as
compared to 2.5 for a silicon group, see: Eliel, E. L.; Wilen, S. H.;
Mander, L. N. Stereochemistry of Organic Compounds; John Wiley &
Sons: New York, 1994; Chapter 11, pp 696-697.
(14) (a) Fleming, I.; Sarkar, A. K.; Doyle, M. J.; Raithby, P. R. J.
Chem. Soc., Perkin Trans 1 1989, 2023-2030. (b) Tripathy, R.; Franck,
R. W.; Onan, K. D. J. Am. Chem. Soc. 1988, 110, 3257-3262. (c)
Landais, Y.; Planchenault, D.; Weber, V. Tetrahedron Lett. 1995, 36,
2987-2990. (d) Andrey, O.; Landais, Y. Tetrahedron Lett. 1993, 34,
8435-8438.
(15) Inoue, H.; Murata, S.; Suzuki, T. Liebigs Ann. Chem. 1994,
901-909.
(16) Cyclization of the achiral analogue of 9a, lacking the allylic Me
group, provided a 3:1 cis/trans ratio at best, in good agreement with
earlier reports.7
5-exo-Trig Cyclizations of Dienes Bearing various
Substituents in the Allylic Position (C3). Intrigued
by these results, we then examined the 5-exo-trig cycliza-
tion of a series of analogues of 6a-c having various
substituents in the allylic position. Alkyl substituents
such as methyl and tert-butyl groups were selected to
study the conformational preferences induced by small
J. Org. Chem, Vol. 70, No. 20, 2005 7987

James et al.
SCHEME 4. Preparation of Oxygenated
their configurations were determined using NOESY
experiments. Confirmation of the stereochemistry of the
major isomer 15a was obtained through X-ray structure
determination of the nitrobenzoate 18. The minor isomer
15b was shown to possess the trans-cis relative config-
uration (NOESY), similar to that of silylated analogues
7a-c (vide supra). The study was next extended to
silylated analogue 13, which at 5-10 °C with 0.45 equiv
of p-TolSO2SePh in CH2Cl2 led to cyclopentane 16 in 51%
isolated yield as a unique stereoisomer. The reactivity
of 13 was found to be much lower than that of the parent
alcohol, with a 31:52:17 13/16/15a ratio (1H NMR)
observed after 8 h of irradiation. Desilylation of 16 using
TBAF led to alcohol 15a having cis-cis stereochemistry,
indicating that cyclopentane 13 had the same cis-cis
relative configuration.18 Reaction of trifluoroacetate 14
afforded cyclopentane 17 as the only observable isomer,
albeit in modest yield due to low conversion. Similar to
the silylated precursor 13, trifluoroacetate 14 was poorly
reactive at -50 °C, producing only trace amounts of
cyclopentane 17 after several hours. At -25 °C and
irradiation for 7 h a 64:33 14/17 ratio was estimated by
1H NMR. Finally, the best levels of conversion with a 48:
52 14/17 ratio were obtained at 5-10 °C with 0.45 equiv
of p-TolSO2SePh in CH2Cl2. Under these conditions a
clean reaction occurred as shown by the 1H NMR
spectrum of the crude reaction mixture, where only the
starting material and the desired cyclopentane 17 were
obtained in quantitative yield. Attempts to separate
cyclopentane 17 from the starting material were plagued
by its sensitivity to silica gel. Nevertheless, diastereo-
isomerically pure 17 could be isolated, albeit in a 17%
yield, after careful chromatography onto silica gel. Re-
moval of the trifluoroacetate group under basic conditions
led to alcohol 15a with a cis-cis relative configuration,
demonstrating that sulfonyl-mediated 5-exo-trig cycliza-
tion of alcohol 12 and its trifluoroacetate analogue 14
proceeded with the same stereochemistry and complete
1,2- and 1,5-stereocontrol for the latter. The results
obtained in the alkoxy series are appealing since the
stereochemistry observed in this case is complementary
to that of the silicon series. Considering that a C-Si bond
can be oxidized into the corresponding C-OH bond with
retention of configuration,19 we thus have complementary
access to stereochemically pure trans-trans and trans-
cis trisubstituted cyclopentanols.
Precursors 12-14
SCHEME 5. Sulfonyl Radical-Mediated 5-exo-Trig
Cyclization of Dienes 12-14
DiscussionsTransition-State Models
and steric properties of the allylic alkoxy group was also
made possible through protection of the alcohol function
of 12 with two different protective groups as in silyl ether
13 and trifluoroacetate 14 (Scheme 4).
Dienes 12-14 were then treated as above with
p-TolSO2SePh in CHCl3 or CH2Cl2 to provide the desired
cyclopentanes with modest to excellent levels of stereo-
control (Scheme 5). Alcohol 12 was thus shown to provide
a 73:27 mixture of two stereoisomers 15a/15b, with the
major one having the cis-cis relative configuration. Both
stereoisomers were separated by chromatography, and
The concomitant 1,2- and 1,5-stereocontrol observed
during 5-exo-trig cyclizations of allylsilanes 6a-c into
cyclopentanes 7a-c was rationalized, assuming a chair-
like transition state A, with the bulky silicon group in a
pseudoequatorial position (Figure 1). Such a conformation
at the transition state is closely related to that proposed
for simpler systems by Beckwith and Schiesser20a-c and
later by Houk et al.20d The minor isomer would be formed
through a boatlike transition state B with a silicon group
again in a pseudoequatorial position. Such a boat con-
(18) Roland, A.; Durand, T.; Egron, D.; Vidal, J.-P.; Rossi, J.-C. J.
Chem. Soc., Perkin Trans 1 2000, 245-251.
(19) (a) Fleming, I. Chemtracts, Org. Chem. 1996, 9, 1-64. (b)
Tamao, K. Adv. Silicon Chem. 1996, 3, 1-62. (c) Jones, G. R.; Landais,
Y. Tetrahedron 1996, 52, 7599-7662.
(17) (a) Baylon, C.; Heck, M.-P.; Mioskowski, C. J. Org. Chem. 1999,
64, 3354-3360. (b) Alcaraz, L.; Harnett, J. J.; Mioskowski, C.; Martel,
J. P.; Le Gall, T.; Shin, D.-S.; Falck, J. R. Tetrahedron Lett. 1994, 35,
30, 5449-5452.
7988 J. Org. Chem., Vol. 70, No. 20, 2005

Cyclization of Chiral 3-Silylhepta-1,6-dienes
FIGURE 1. Transition-state models for 5-exo-trig cyclizations of precursors 6a-c.
formation does not suffer here from the flagpole interac-
tions observed in cyclohexane and can thus be invoked.20d
Such a conformation would be more favorable than a
chairlike conformation in which the silicon group would
be forced into a pseudoaxial arrangement, generating
important 1,3-diaxial interactions. TS-B however expe-
riences more important torsional strain, explaining its
higher energy as compared to TS-A. Interestingly, the
Newman representation of TS-A and TS-B shows that
the C-Si bond is nearly aligned with the incipient C-C
bond and thus stabilizes the developing positive charge
at C2. Such an effect, reminiscent of the â-silicon effect,10
may explain in part the unusual reactivity of allylsilanes
in this radical cascade process (i.e., stabilization of the
transition state). The matched polarity between electron-
rich allylsilanes21 and electrophilic tosyl radical may also
contribute to the high observable rate, as compared with
the alkyl and alkoxy analogues. It has effectively been
with the respective A value of these groups.13 Cyclopen-
tane 11 is thus likely to be formed through a chairlike
transition state such as A, efficiently locked by the t-Bu
group (Figure 1). It is known in radical hex-5-enyl
cyclizations that conformationally more rigid systems
lead to higher stereocontrol.23 The bulky silicon and t-Bu
groups on the heptadienyl chain probably limit the
number of conformations available for cyclization and
efficiently lock these conformations. In contrast, substitu-
tion of the silicon group with a smaller methyl substitu-
ent (i.e., as in precursor 9a) leads to a conformationally
more flexible system, where nonchairlike conformations
(i.e., TS-B) are also highly populated, leading to a
mixture of 2,6-cis and 2,6-trans diastereomeric products.
Additional electronic effects may also be invoked and
cannot be ruled out. In his seminal papers20a,b,24 Beckwith
rationalized formation of the major 2,6-cis stereoisomer
(1,5-stereocontrol), considering that these 5-exo-trig radi-
cal cyclizations likely proceed through a dipolar transition
state, involving an electrostatic interaction between the
“negatively charged (δ-)” olefinic moiety and the posi-
tively charged radical center C2 (δ+) (Figure 1).24b,c
Stabilization of this partial positive charge by the SiR3
group, through hyperconjugation, might reinforce this
electrostatic interaction in our case. Therefore, both steric
and electronic effects would cooperate to favor the forma-
tion of the trans-cis compounds.25
‚
shown that electrophilic radical species such as MeSO2
or CCl3‚ add faster to allylsilanes than to the correspond-
ing olefins lacking a silicon group.22 The origin of the high
1,2- and 1,5-stereocontrol observed during cyclization of
allylsilanes (and allylic t-Bu precursor 9b) is more
difficult to rationalize. Complete 1,2- and 1,5-stereocon-
trol observed during cyclization of t-Bu precursor 9b at
80 °C tends to support the steric origin of the stereose-
lectivity in these systems. It is noteworthy that 11 was
obtained as a unique stereoisomer at 80 °C, while at the
same temperature silylated analogues 7a/8a were formed
in a 86:14 ratio, indicating that the t-Bu group is a more
efficient “chiral inducer” that a PhMe2Si group, in line
The contrasting formation of the major cis-cis stereo-
isomer 15a in the alkoxy series may arise through
transition state C (Figure 2) in which the alkoxy group
at C3 prefers the pseudoaxial position for electronic
reasons.26 As shown in the Newman representation, a
(20) (a) Beckwith, A. L. J.; Schiesser, C. H. Tetrahedron Lett. 1985,
26, 373-376. (b) Beckwith, A. L. J.; Schiesser, C. H. Tetrahedron 1985,
41, 3925-3941. (c) Beckwith, A. L. J.; Zimmermann, J. J. Org. Chem.
1991, 56, 5791-5796. (d) Spellmeyer, D. C.; Houk, K. N. J. Org. Chem.
1987, 52, 959-974.
(21) Ground-state interactions between the σC-Si bond and the olefin
raise the HOMO of the allylsilane, making it more reactive towards
electrophiles, see: (a) Kahn, S. D.; Pau, C. F.; Chamberlin, A. R.; Hehre,
W. J. J. Am. Chem. Soc. 1987, 109, 650-663. (b) Kahn, S. D.; Pau, C.
F.; Hehre, W. J. J. Am. Chem. Soc. 1986, 108, 7396-7398.
(22) (a) Sakurai, H.; Hosomi, A.; Kumada, M. J. Org. Chem. 1969,
34, 1764-1768. (b) Gozdz, A. S.; Maslak, P. J. Org. Chem. 1991, 56,
2179-2189. (c) Chatgilialoglu, C.; Ferreri, C.; Ballestri, M.; Curran,
D. P. Tetrahedron Lett. 1996, 37, 6387-6390. (d) Jarvie, A. W. P.;
Rowley, R. J. J. Chem. Soc., B 1971, 2439-2442.
(23) RajanBabu, T. V. Acc. Chem. Res. 1991, 24, 139-145 and
references therein.
(24) (a) Beckwith, A. L. J. Tetrahedron 1981, 37, 3073-3100. (b)
Beckwith, A. L. J.; Blair, I.; Phillipou, G. J. Am. Chem. Soc. 1974, 96,
1613-1614. (c) Shono, T.; Nishigushi, I.; Ohmizu, H.; Mitani, M. J.
Am. Chem. Soc. 1978, 100, 545-550.
(25) Tripp, J. C.; Schiesser, C. H.; Curran, D. P. J. Am. Chem. Soc.
2005, 127, 5518-5527.
(26) (a) Guindon, Y.; Yoakim, C.; Gorys, V.; Ogilvie, W. W.; Delorme,
D.; Renaud, J.; Robinson, G.; Lavalle´e, J.-F.; Slassi, G.; Jung, J.;
Rancourt, J. J. Org. Chem. 1994, 59, 1166-1178. (b) Durkin, K.; Liotta,
D.; Rancourt, J.; Lavalle´e, J.-F.; Boisvert, L.; Guindon, Y. J. Am. Chem.
Soc. 1992, 114, 4912-4914.
J. Org. Chem, Vol. 70, No. 20, 2005 7989

James et al.
effect and the hydrogen-bonding interaction discussed
above could not operate.
Approach toward the A-B-C Ring Core of
Hexacyclinic Acid. Having demonstrated that high 1,2-
and 1,5-stereocontrol could be attained during radical
5-exo-trig cyclizations of the hepta-1,5-dienyl system
having an allylsilane moiety, we then attempted to
illustrate the utility of such a methodology in the
preparation of the A-B-C ring core of hexacyclinic acid
3a, a polyketide recently isolated from Streptomyces
cellulosae subsp. griserubiginosus (strain S1013).1c This
polycyclic system is structurally very close and probably
biosynthetically related to (-)-FR-182877 3b, isolated
from Streptomyces sp. No. 9885,1d which, however, ex-
hibits much higher cytotoxicity than 3a. While (-)-FR-
182877 3b has recently been synthesized using biomi-
metic strategies, involving as key steps aldol and
transannular Diels-Alder processes,27 hexacyclinic acid
3a has not yet succumbed to total synthesis. However,
recent studies have shown that the A-B-C ring of 3a
may also be assembled through a Diels-Alder approach,28a
while construction of the D-E-F ring was performed
through a transannular iodocyclization of a nine-mem-
bered ring â-ketoester.28b,c Our approach to A-B-C ring
precursor I follows a different pathway in which the
tricyclic skeleton would be constructed in two steps, i.e.,
a 5-exo-trig radical cyclization as described above and a
Pauson-Khand reaction29 (Scheme 6). The alcohol func-
tion at C8 (following hexacyclinic acid numbering1c) could
then be generated at a suitable time through Tamao-
Kumada-Fleming oxidation19 of the C-Si bond and the
methyl group at C7 generated from the corresponding
hydroxymethyl moiety. It was envisioned that the tricy-
clic skeleton of I could be easily assembled through
Pauson-Khand cyclization of a suitable enyne II, itself
obtained through alkylation of the sulfone III.30 Although
many examples of such intramolecular Pauson-Khand
processes have been reported before,29 it was however not
possible at this point to make certain the control of the
stereochemistry at C4. Sulfone III would be obtained
through the sulfonyl addition-5-exo-trig-â-elimination
cascade studied above with the concomitant control of the
relative configuration at C5, C8, and C9. Cyclization of
precursor IV would in turn be formed using an aldol
process between a â-silyl ester such as VI and aldehyde
V. Previous studies indicate that the relative configura-
FIGURE 2. Transition-state models for 5-exo-trig cyclizations
of precursors 12-14.
pseudoaxial OR group would be orthogonal to the electron-
deficient forming C-C bond and to the developing partial
positive charge at C2. This charge would thus be better
stabilized by the nearly aligned C3-H bond as a σC-H
bond is a better electron donor than a σC-O bond. The
occurrence of such an effect is further supported by the
complete stereocontrol, albeit with modest conversion,
observed with precursor 14 in which the strong electron-
withdrawing CF3CO2 group at C3 should reinforce this
electronic effect. The incomplete conversion during cy-
clization of diene 14 may be ascribed to the low reactivity
of the reacting olefin which is strongly deactivated by the
trifluoroacetate group, indicating again the importance
of polar effects in addition of electrophilic sulfonyl groups
onto olefins.22b Similar selectivities were obtained with
precursor 13 having a sterically hindered TBDMS group.
These results were at first glance relatively surprising
as it was expected that a bulky TBDMS group would
behave very much like a PhMe2Si group to lead predomi-
nantly to the trans-cis isomer.18 It is however plausible
that the silicon group on the oxygen is pointing away and
would thus be quite remote, thus having little influence
on the stereocontrol. An alternative model D may also
be proposed, invoking in the case of precursor 12,
hydrogen bonding between the alcohol function and the
closest sulfonyl group (Figure 2). Such an interaction
would thus maintain the OH group in a pseudoaxial
position. Although such an interaction seems to be
disproved by the better stereocontrol observed with
precursors 13 and 14, in which hydrogen bonding is
precluded, interactions between the sulfonyl group and
the trifluoroacetate carbonyl group (R ) COCF3) in 14
or the TBDMS group (R ) SiMe2t-Bu) in 13 cannot be
ruled out. The electronic effect discussed above and the
p-TolSO2 T R group interaction may also reinforce one
another and contribute to the high level of 1,2- and 1,5-
stereocontrol observed with models 13 and 14. Finally,
the minor isomer 15b, only observed upon cyclization of
the alcohol precursor 12, is likely to be formed through
a transition state close to TS-A (Figure 1) with the OH
group in a pseudoequatorial position. Such a conforma-
tion would not be favored as the OH group electronic
(27) (a) Vosburg, D. A.; Vanderwal, C. D.; Sorensen, E. J. J. Am.
Chem. Soc. 2002, 124, 4552-4553. (b) Vanderwal, C. D.; Vosburg, D.
A.; Weiler, S.; Sorensen, E. J. J. Am. Chem. Soc. 2003, 125, 5393-
5407. (c) Evans, D. A.; Starr, J. T. Angew. Chem., Int. Ed. Engl. 2002,
41, 10, 1787-1790.
(28) (a) Stellfeld, T.; Bhatt, U.; Kalesse, M. Org. Lett. 2004, 6, 3889-
3892. (b) Clarke, P. A.; Grist, M.; Ebden, M.; Wilson, C. Chem.
Commun. 2003, 1560-1561. (c) Clarke, P. A.; Grist, M.; Ebden, M.
Tetrahedron Lett. 2004, 45, 927-929. (d) Clarke, P. A.; Cridland, A.
P. Org. Lett. 2005, 7, 4221-4224.
(29) (a) Schore, N. E. In Comprehensive Organic Synthesis; Trost,
B. M., Fleming, I., Paquette, L. A., Eds.; Pergamon: Oxford, 1991; Vol.
5, pp 1037-1064. (b) Breczinski, P. M.; Stumpf, A.; Hope, H.; Krafft,
M. E.; Casalnuovo, J. A.; Schore, N. E. Tetrahedron 1999, 55, 6797-
6812. (c) Jeong, N.; Chung, Y. K.; Lee, S. H.; Yoo, S. E. Synlett 1991,
204-206. (d) Hoye, T. R.; Suriano, J. A. J. Org. Chem. 1993, 58, 1659-
1660. (e) Chung, Y. K.; Lee, B. Y.; Jeong, N.; Huddecek, M.; Pauson,
P. L. Organometallics 1993, 12, 220-223. (f) Sugihara, T.; Yamada,
M.; Ban, H.; Yamaguchi, M.; Kaneko, C. Angew. Chem., Int. Ed. Engl.
1997, 36, 2801-2804. (g) Sugihara, T.; Yamada, M.; Yamaguchi, M.;
Nisshizawa, M. Synlett 1999, 771-773.
(30) Simpkins, N. S. Sulfones in Organic Synthesis; Tetrahedron
Organic Chemistry Series; Pergamon Press: New York, 1993; Vol. 10.
7990 J. Org. Chem., Vol. 70, No. 20, 2005

Cyclization of Chiral 3-Silylhepta-1,6-dienes
SCHEME 6. Retrosynthetic Analysis of Hexacyclinic Acid A-B-C Tricyclic Skeleton
SCHEME 7. Pauson-Khand Cyclization of Enyne
19
tion between C6, C7, and C8 stereogenic centers should
be controlled efficiently.31 The anti stereochemistry be-
tween C7 and C8 should thus be secured through a face-
selective approach of the aldehyde on the â-silyl ester
enolate, anti relative to the silicon group,9a,d through a
transition-state model governed by A1,3 strain.32,33 In the
meantime, the relative configuration between C6 and C7
would depend on the enolate geometry, with the (E)- and
(Z)-enolates reported to produce predominantly the anti
and syn isomers, respectively.
Following these considerations we started the synthe-
sis of I by first investigating the intramolecular Pauson-
Khand reaction on a model compound. Cyclopentane 7a,
described above, was thus alkylated using a TMS-
protected propargyl bromide34 to produce 19 in good yield
as a 2:1 mixture (Scheme 7). This was then treated with
Co2(CO)8 to form the alkyne-cobalt complex, which after
removal of the solvents was treated with trimethylamine-
N-oxide (TMANO‚2H2O) in THF to afford the desired
tricyclic structure 20 as a 2:1 mixture of stereoisomers,
possessing the requisite C4-C5-trans configuration.
NOESY experiments on the major isomer of 20 showed
unambiguously that the stereochemistry at C4 had been
totally controlled during the enyne cyclization process,29b
thus demonstrating that this methodology was suitable
for elaboration and control of the stereochemistry of the
contiguous C4, C5, C8, and C9 stereocenters of the
A-B-C ring core of 3a.
through a three-step procedure, including a reduction of
the triple bond of commercial butyn-1,4-diol using LiAlH4
(70% yield) followed by acetylation (80%) and oxidation
of the remaining alcohol function using Dess-Martin
periodinane.35 The aldol reaction between 22 and the
enolate of 21, generated using LDA in THF, led to the
aldol product as a 93:7 mixture of anti-syn 23b and the
desired anti-anti 23a stereoisomers, respectively, in good
yield (Scheme 8). The stereoselectivity observed in our
case was in good agreement with previous reports from
the literature,32a and although both stereoisomers could
be isolated pure by chromatography, the amount of aldol
23a formed under these conditions was clearly not
suitable for our purpose. It was thus envisioned to reverse
the ratio of stereoisomers by generating the (Z)- instead
of (E)-enolate, known to produce predominantly the anti-
syn isomer. This problem could be solved by generating
the stereochemically defined enolate using LDA in a
mixture of THF and HMPA.36 When the reaction was
carried out using LDA and 23% of HMPA in THF, a
partial reversal of stereocontrol was observed with 23a
now formed predominantly in an unoptimized 40% yield.
These encouraging results then prompted us to extend
the strategy to the genuine system. We thus focused our
attention on the synthesis of ester IV (Scheme 6) and
control of the stereochemistry of the C6, C7, and C8
stereocenters through an aldol reaction between aldehyde
22 and ester 21. Aldehyde 22 was easily prepared
(31) Fleming, I.; Kilburn, J. D. J. Chem. Soc., Chem. Commun. 1986,
1198-1201.
(32) (a) Fleming, I.; Hill, J. H. M.; Parker, D.; Waterson, D. J. Chem.
Soc., Chem. Commun. 1985, 6, 318-320. (b) Panek, J. S.; Beresis, R.;
Xu F.; Yang, M. J. Org. Chem. 1991, 56, 7341-7344. (c) Beresis, R.
T.; Solomon, J. S.; Yang, M. G.; Jain, N. F.; Panek, J. S. Org. Synth.
1997, 75, 78-82. (d) Ward, R. A.; Procter G. Tetrahedron 1995, 51,
12821-12836.
(33) Hoffmann, R. W. Chem. Rev. 1989, 89, 1841-1860.
(34) Wu, R.; Schumm, J. S.; Pearson, D. L.; Tour, J. M. J. Org. Chem.
1996, 16, 6906-6921.
(35) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277-
7287.
J. Org. Chem, Vol. 70, No. 20, 2005 7991

James et al.
SCHEME 8. Aldol Reaction between Ester 21 and
Aldehyde 22
FIGURE 3. Transition-state model for the 5-exo-trig cycliza-
tion of precursor 24a.
SCHEME 10. Sulfonyl Radical-Mediated
5-exo-Trig Cyclization of Dienes 24a,b
SCHEME 9. Preparation of Allyl Sulfone 24a,b
Anti-anti acetate 23a was then converted into the cor-
responding sulfone 24a using a palladium-catalyzed sul-
fonylation.37 Similarly, anti-syn 23b gave sulfone 24b
(Scheme 9). Both sulfones were then submitted to the
above radical sulfonyl-mediated cascade to afford the cor-
responding cyclopentanes 25 and 26a,b in excellent yield
(Scheme 10). Interestingly, 24a led to 2538 with complete
stereocontrol, while its isomer 24b gave an 82:18 mixture
of two diastereomers 26a,b which could not be separated
through chromatography. The structure of stereoisomers
26a,b was however assumed by analogy with the stere-
ochemistry observed for the related substrates 7a-c
described in the preliminary studies (Table 1).
This difference of selectivity may be explained consid-
ering that with stereoisomer 24a the 5-exo-trig cyclization
likely proceeds through a chairlike transition state such
as A′ (Figure 3) similar to TS-A (Figure 1) in which all
substituents are in a pseudoequatorial arrangement.20,23
Such is not the case with stereoisomer 24b, where the
OH at C6 would be pseudoaxial in a similar chair con-
formation. It must be added that, as pointed out earlier
by RajanBabu,23 addition of alkyl radical onto olefins is
a nucleophilic process. Therefore, one may expect activa-
tion of the olefin π* bond by the C6 allylic σ*C-O bond
when it is properly aligned, as is the case with isomer
24a in TS-A′.
Vinylcyclopentane 25, having the correct relative con-
figuration, was then converted in two steps into the
MOM-protected diol 27 before being alkylated with the
required propargylic bromide (Scheme 11). The conditions
described above for the alkylation of model compound 7a
unfortunately led to no reaction. The same conditions
applied to the less sterically congested TBDMS-protected
alcohol 7c also led to complete recovery of the starting
material. After extensive experiments we finally found
that using Schlosser’s superbase,39 deprotonation oc-
curred to provide the sulfonyl carbanion which was
alkylated to provide enyne 28 as a 4:1 ratio of two
separable epimers at C10, albeit with modest conversion
(35% with a 65% corrected yield). Enyne 28 was then
treated under Pauson-Khand conditions as described
above, providing the desired tricyclic compound 29 in 74%
yield after filtration through a short pad of alumina.
Extensive NOESY experiments on the major isomer of
29 (with stereochemistry as shown) established the trans
relative configuration between C4 and C5, demonstrating
that as before for 20 complete stereocontrol was observed
during enyne cyclization. To conclude, using a straight-
forward approach relying on a radical 5-exo-trig cycliza-
tion/alkylation/intramolecular Pauson-Khand reaction
(36) Ireland, R. E.; Mueller, R. H.; Willard, A. K. J. Am. Chem. Soc.
1976, 98, 2868-2877.
(37) (a) Inomata, K.; Yamamoto, T.; Kotake, H. Chem. Lett. 1981,
1357-1360. (b) Felpin, F.-X.; Landais, Y. J. Org. Chem. 2005, 70,
6441-6446.
(38) It is also noteworthy that a low yield (26%) of 25 was first
obtained upon chromatography over silica gel, demonstrating its
sensitivity toward acid. This problem was finally overcome by adding
5% Et3N to the eluent, leading after chromatography to pure 25 (77%
yield). The acid sensitivity of 25 was attributed to the presence of the
γ-silylcarbinol moiety, which is susceptible to react under these
conditions. Examination of molecular models reveals that in 25 the
Si-C8-C7-C6-O moiety adopts an ideal W conformation in which
an interaction between the σC8-Si and a developing vacant p orbital at
C6 (percaudal interaction) may be favored. Such an interaction (γ-
silicon effect) was first proposed by Shiner et al. to rationalize the
reaction rate acceleration observed during solvolysis of 1,3-cis- as
compared to 1,3-trans-silylcylohexyl brosylates in CF3CH2OH, see.
Shiner, V. J., Jr.; Ensinger, M. W.; Rutowske, R. D. J. Am. Chem. Soc.
1987, 109, 804-809.
(39) Schlosser, M. Superbases as Powerful Tools in Organic Syn-
theses. In Modern Synthetic Methods; Scheffold, R., Ed.; VCH: Wein-
heim, 1992; pp 227-271.
7992 J. Org. Chem., Vol. 70, No. 20, 2005

Cyclization of Chiral 3-Silylhepta-1,6-dienes
was refluxed under a nitrogen atmosphere until TLC revealed
completion of the reaction (more AIBN was added if necessary).
The solvent was then evaporated, and the residue was purified
by chromatography on silica gel.
SCHEME 11. Preparation of the A-B-C Ring
Core of Hexacyclinic Acid
General Procedure for 5-exo-Trig Cyclization of Dienes
through Irradiation. A mixture containing the diene (1
mmol) and p-TolSO2SePh (0.15-0.50 mmol) in degassed
solvent (0.013 M) under a nitrogen atmosphere was irradiated
with a sun lamp (300 W) placed 10-15 cm from the flask until
TLC revealed completion of the reaction. The solvent was then
evaporated, and the residue was purified by chromatography
on silica gel.
Dimethylphenyl-[2-(toluene-4-sulfonylmethyl)-3-vinyl-
cyclopentyl]silane (7a). Following the general procedure
described above (irradiation), diene 6a (43 mg, 0.11 mmol) and
p-TolSO2SePh (8.4 mg, 0.03 mmol) in degassed CHCl3 (8.5 mL)
were irradiated for 30 min at - 50 °C. Purification by
chromatography (petroleum ether/EtOAc 9/1) afforded 7a as
a pale yellow oil (34 mg, 84%). IR (neat) ν ) 3068, 2951, 2864,
1597, 1426, 1315, 1301, 1249, 1149, 1113, 812, 772, 735, 701
1
cm-1. H NMR (CDCl3, 400 MHz) δ ) 7.61 (d, J ) 8.3 Hz, 2
H), 7.44 (m, 2 H), 7.35 (m, 3 H), 7.31 (d, J ) 8.3 Hz, 2 H), 5.61
(ddd, J ) 17.1, 10.4, 8.9 Hz, 1 H), 5.07 (dd, J ) 10.4, 2.1 Hz,
1 H), 5.01 (ddd, J ) 17.1, 2.1, 0.8 Hz, 1 H), 3.00 (dd, ABX, JAB
) 14.2 Hz, JAX ) 9.7 Hz, ∆ν/J ) 9, 1 H), 2.82 (m, 1 H), 2.79
(dd, ABX, JAX ) 9.7 Hz, JBX ) 2.7 Hz, ∆ν/J ) 9, 1 H), 2.44 (s,
3 H), 2.38 (m, 1 H), 1.89 (m, 1 H), 1.64-1.53 (m, 3 H), 1.15
(m, 1 H), 0.26 (s, 6 H). 13C NMR (CDCl3, 75 MHz) δ ) 144.7,
138.3, 137.9, 137.6, 134.3, 129.5, 128.6, 128.4, 128.3, 128.2,
117.5, 57.3, 47.3, 39.9, 31.9, 28.2, 25.3, 22.0, -3.4, -4.9. MS
(FAB+) m/z ) 421 (M + Na+, 20), 399 (M + H+, 4), 398 (M, 2),
383 (17), 321 (100). HRMS (FAB+) calcd for C23H30O2SSi +
Na+: 421.1633. Found: 421.1644. Anal. Calcd for C23H30O2-
Ssi: C, 69.30; H, 7.59. Found: C, 69.16; H, 7.62.
we have been able to build up and control the relative
configuration of the six stereogenic centers of the A-B-C
ring core of hexacyclinic acid 3a in only seven steps
starting from the readily available â-silylester 21.32b-d
Dimethylphenyl-[2-(toluene-4-sulfonylmethyl)-3-vinyl-
cyclopentyl]silane (8a). Following the general procedure
(thermal conditions), diene 6a (400 mg, 1.00 mmol), p-TolSO2-
SePh (46.7 mg, 0.15 mmol), and AIBN (14.2 mg, 0.10 mmol)
in degassed benzene (75 mL) were refluxed for 10 h. The
solvent was then evaporated, and the residue (86:14 mixture
of the two diastereoisomers 7a/8a) was purified by chroma-
tography (petroleum ether/EtOAc 9/1) to afford a pale yellow
oil (35 mg, 9% for pure 8a; 7a and 8a were isolated in 69%
overall yield). 8a: IR (neat) ν ) 3068, 2951, 1637, 1597, 1427,
Conclusion
In summary, we reported here on the remarkable effect
of a silicon group on the 1,2- and 1,5-stereocontrol of
radical 5-exo-trig cyclizations of hepta-1,6-dienes incor-
porating an allylsilane moiety. Such cyclizations at low
temperature produced trisubstituted cyclopentanes hav-
ing trans-cis stereochemistry as single diastereomers in
excellent yield. This appeared to be specific to heptadienyl
systems possessing a large group in the allylic position
(i.e., a R3Si or t-Bu group) since the same process carried
out on a related precursor having a methyl group led to
a very low level of stereocontrol and a mixture of the four
possible stereoisomers. In contrast, the reaction per-
formed on precursors having an alkoxy group in the
allylic position gave complementary selectivity with the
cis-cis isomer formed predominantly. Introduction of a
strongly electron-withdrawing trifluoroacetate group led
to complete stereocontrol under these conditions. Beck-
with-Houk-type models were found to be appropriate to
rationalize the stereochemistry observed. The stereose-
lectivity is assumed to be essentially steric in origin, as
indicated by the complete stereocontrol observed during
cyclization at 80 °C of precursor 9b, having a large allylic
t-Bu group. The value of this methodology was finally
illustrated with a concise approach to the densely func-
tionalized A-B-C ring core of hexacyclinic acid.
1316, 1301, 1250, 1147, 1112, 1087, 1018, 997, 734, 701 cm-1
.
1H NMR (CDCl3, 300 MHz) δ ) 7.60 (d, J ) 7.9 Hz, 2 H), 7.40-
7.27 (m, 7 H), 5.54 (ddd, J ) 17.3, 10.2, 7.9 Hz, 1 H), 4.96 (br
d, J ) 16.9 Hz, 1 H), 4.86 (br d, J ) 10.2 Hz, 1 H), 2.92 (dd, J
) 14.1, 8.8 Hz, 1 H), 2.83 (dd, J ) 14.1, 3.2 Hz, 1 H), 2.77 (m,
1 H), 2.45 (s, 3 H), 2.05 (m, 1 H), 1.80-1.60 (m, 2 H), 1.57-
1.37 (m, 2 H), 1.08 (m, 1 H), 0.23 (s, 6 H). 13C NMR (CDCl3, 63
MHz) δ ) 144.2, 141.4, 137.7, 137.1, 133.9, 129.6, 129.0, 128.2,
128.0, 127.8, 113.6, 61.3, 49.1, 41.1, 33.3, 32.6, 27.2, 21.6, -4.1,
-5.1. MS (FAB+) m/z ) 421 (M + Na+, 22), 383 (19), 321 (100),
211 (38). HRMS (FAB+) calcd for C23H30O2SSi + Na+: 421.1633.
Found: 421.1630.
Methyl 6-Acetoxy-3-hydroxy-2-(1-(dimethyl(phenyl)-
silyl)allyl)hex-4-enoate (23a). A solution of LDA (8.32 mmol)
in dry THF (22.7 mL) under a nitrogen atmosphere was cooled
at -78 °C, and HMPA (6.8 mL) was added. To this vigorously
stirred solution was added dropwise ester 21 (2.07 g, 8.32
mmol) in THF (1 mL). After 30 min at -78 °C, aldehyde 22
(1.12 g, 8.74 mmol) in THF (1 mL) was added dropwise to the
deep orange solution. The mixture was stirred 1 h at this
temperature and then quenched with glacial acetic acid (1 mL).
The mixture was then allowed to warm to room temperature,
and an aqueous solution of NH4Cl was added. The organic
layer was separated, and the aqueous layer was extracted with
ether. The combined organic layers were then washed with
aqueous NaHCO3 and saturated NaCl, dried over MgSO4, and
filtered, and the solvents were concentrated under vacuum.
Purification of the crude material (23a/23b 65:35) by chroma-
Experimental Section
General Procedure for 5-exo-Trig Cyclization of Dienes
under Thermal Conditions. A mixture containing the diene
(1 mmol), p-TolSO2SePh (0.15 mmol), and AIBN (0.10 mmol,
added by portions every 1.5 h) in degassed benzene (0.013 M)
J. Org. Chem, Vol. 70, No. 20, 2005 7993

James et al.
tography gave a pale yellow oil [23a (661 mg, 21%), 23b (602
mg, 19%), 40% overall yield]. 23a: IR (neat) ν ) 3507, 3070,
2952, 2902, 1739, 1712, 1626, 1428, 1380, 1361, 1247 1113,
1027, 969, 702, 656 cm-1. 1H NMR (CDCl3, 300 MHz) δ ) 7.47
(m, 2 H), 7.33 (m, 3 H), 5.80-5.60 (m, 2 H), 5.57 (dt, J ) 16.8,
10.1 Hz, 1 H), 5.07 (br dd, J ) 7.3 Hz, 1 H), 5.01 (br dd, J )
13.4 Hz, 1 H), 4.49 (d, J ) 5.5 Hz, 2 H), 4.32 (m, 1 H), 3.25 (s,
3 H), 3.05 (d, J ) 9.5 Hz, 1 H), 2.57 (m, 1 H), 2.53 (m, 1 H),
2.02 (s, 3 H), 0.32 (s, 3 H), 0.23 (s, 3 H). 13C NMR (CDCl3, 62.5
MHz) δ ) 173.9, 170.4, 136.3, 135.6, 134.9, 134.1, 129.0, 127.4,
124.5, 116.1, 69.7, 63.9, 51.0, 49.5, 33.2, 20.7, -3.4, -5.2. MS
(FAB+) m/z ) 399 (M + Na+, 100), 247 (12). HRMS (FAB+)
calcd for C20H28O5Si + Na+: 399.1603. Found: 399.1605.
Methyl 6-Acetoxy-3-hydroxy-2-(1-(dimethyl(phenyl)-
silyl)allyl)hex-4-enoate (23b). A solution of LDA (8.6 mmol)
in dry THF (10 mL) under a nitrogen atmosphere was cooled
at -78 °C, and ester 21 (1.94 g, 7.8 mmol) in THF (4 mL) was
added dropwise. After 45 min at -78 °C, aldehyde 22 (1.0 g,
7.8 mmol) in THF (2 mL) was added dropwise, and the mixture
was stirred 30 min at this temperature and then quenched
with glacial acetic acid (1 mL). The mixture was then allowed
to warm to room temperature, and an aqueous solution of NH4-
Cl was added. The organic layer was separated, and the
aqueous layer was extracted with ether. The combined organic
layers were then washed with aqueous NaHCO3, saturated
NaCl, dried over MgSO4, and filtered, and the solvents were
concentrated under vacuum. Purification of the crude material
(23b/23a 93:7) by chromatography (petroleum ether/ethyl
acetate 8/2 then 7/3) gave 23b a pale yellow oil (2.02 g, 69%
yield). IR (neat) ν ) 3496, 3070, 1736, 1624, 1248, 1196, 1145,
1113, 1084, 1026, 971, 903, 837, 817, 702, 657 cm-1. 1H NMR
(CDCl3, 300 MHz) δ ) 7.46 (m, 2 H), 7.33 (m, 3H), 5.81 (dd, J
) 15.5, 6.4 Hz, 1 H), 5.71 (dt, J ) 15.5, 5.7 Hz, 1 H), 5.55 (dt,
J ) 17.0, 10.6 Hz, 1 H), 4.99 (dd, J ) 10.2, 1.9 Hz, 1 H), 4.86
(dd, J ) 17.3, 1.5 Hz, 1 H), 4.52 (d, J ) 5.7 Hz, 2 H), 4.30 (q,
J ) 5.7 Hz, 1 H), 3.36 (s, 3 H), 2.86 (dd, J ) 11.3, 5.7 Hz, 1 H),
2.37 (br d, J ) 4.5 Hz, 1 H), 2.23 (t, J ) 10.9 Hz, 1 H), 2.03 (s,
3 H), 0.30 (s, 3 H), 0.26 (s, 3 H). 13C NMR (CDCl3, 62.5 MHz)
δ ) 172.8, 170.5, 136.6, 136.2, 134.2, 132.2, 129.0, 127.5, 126.5,
115.3, 72.4, 64.1, 51.2, 50.9, 34.2, 20.8, -3.7, -4.6. MS (FAB+)
m/z ) 399 (M + Na+, 100), 359 (6), 327 (7), 299 (11), 288 (16),
247 (33), 221 (17), 207 (22). Anal. Calcd for C20H28O5Si: C,
63.80; H, 7.50. Found: C, 63.99; H, 7.75.
silane (27). Cyclopentane 25 (0.91 g, 1.93 mmol) was dissolved
in dry CH2Cl2 (25 mL) under a nitrogen atmosphere, and the
solution was cooled to -78 °C. A 1 M solution of DIBAL-H in
CH2Cl2 (12.5 mL, 12.5 mmol) was then added dropwise, and
the mixture was stirred for 3 h at this temperature and then
quenched with MeOH. The reaction mixture was then allowed
to warm to room temperature, and an aqueous solution of
NaHCO3 was added. The mixture was filtered through Celite,
and the gelatinous precipitate was washed extensively with
CH2Cl2. The organic layer was washed with brine, dried over
MgSO4, filtered, and concentrated under vacuum. The crude
product was finally dissolved in dry CH2Cl2 (75 mL) under a
nitrogen atmosphere, and diisopropylethylamine (4.4 mL, 25
mmol) was added. The flask was cooled to 0 °C, and MOMCl
(1.4 mL, 15 mmol) was added dropwise. The solution was
stirred at room temperature overnight and then poured into
a NaHCO3 solution. The layers were separated, and the
aqueous solution was extracted with dichloromethane. The
combined organic layers were washed with NH4Cl and brine,
dried over MgSO4, and filtered, and the solvent was evaporated
under vacuum. Purification by chromatography (petroleum
ether/EtOAc/NEt3 85/10/5) afforded 27 as a pale yellow oil (615
mg, 60%, 2 steps). IR (neat) ν ) 2924, 1314, 1300, 125, 1148,
1110, 1088, 814, 702 cm-1 1H NMR (CDCl3, 300 MHz) δ )
.
7.65 (d, J ) 8.3 Hz, 2 H), 7.50 (m, 2 H), 7.38 (m, 3 H), 7.31 (d,
J ) 8.3 Hz, 2 H), 5.58 (dt, J ) 16.9, 9.4 Hz, 1 H), 5.12 (m, 2
H), 4.61 (d, J ) 3.0 Hz, 2 H), 4.54 (s, 2 H), 3.93 (br t, J ) 3.8
Hz, 1 H), 3.38 (m, 1 H), 3.33 (s, 3 H), 3.32 (s, 3 H), 3.15 (dd, J
) 9.4, 7.2 Hz, 1 H), 3.08 (dd, J ) 13.9, 7.5 Hz, 1 H), 2.93 (dd,
J ) 14.3, 4.9 Hz, 1 H), 2.80 (m, 2 H), 2.72 (m, 1 H), 2.45 (s, 3
H), 2.14 (m, 1 H), 1.22 (t, J ) 8.3 Hz, 1 H), 0.38 (s, 6 H). 13C
NMR (CDCl3, 62.5 MHz) δ ) 144.2, 137.4, 137.1, 134.7, 133.8,
129.5, 129.1, 127.8, 127.8, 119.1, 96.2, 95.0, 83.8, 68.9, 56.3,
55.1, 55.0, 53.4, 48.0, 37.4, 30.1, 21.4, -4.2, -5.1. MS (FAB+)
m/z ) 555 (M + Na+, 52), 501 (9), 471 (9), 439 (3), 425 (6), 379
(3), 305 (7), 275 (5), 213 (33), 179 (5), 149 (54), 135 (100). HRMS
(ES+) calcd for C28H40O6SSi + Na+: 555.2213. Found: 555.2213.
Enyne (28). To a stirred solution of silylcyclopentane 27
(420 mg, 0.81 mmol) in dry THF (3 mL) under an argon
atmosphere was added dropwise a solution of sublimed t-BuOK
(112 mg, 1.00 mmol) in THF (9 mL) at -78 °C. A 2.5 M solution
of n-BuLi in hexanes was then added dropwise at this
temperature, and the red-brown solution was stirred for 1 h
at -78 °C. 3-Bromo-1-trimethylsilylpropyne34 was added drop-
wise below -78 °C, the solution was stirred for 1 h before being
quenched with MeOH (1 mL), and the mixture was allowed
to warm to 0 °C. An aqueous solution of NH4Cl was added,
and the layers were separated. The aqueous layer was
extracted with ether; the combined organic layers were washed
with brine, dried over MgSO4, and filtered, and the solvents
were removed under vacuum. Purification by chromatography
(petroleum ether/EtOAc/NEt3 85/10/5) afforded a pale yellow
oil (104 mg of the major diastereomer was isolated; 180 mg
(35%) of a combined fraction of both diastereomers, and 131
mg of starting 27 was recovered, 65% corrected yield). 28: IR
(neat) ν ) 2954, 2177, 1596, 1427, 1312, 1250, 1146, 1109,
1086, 761, 702, 662 cm-1. 1H NMR (CDCl3, 300 MHz) Major δ
) 7.69 (d, J ) 8.3 Hz, 2 H), 7.61 (m, 2 H), 7.38 (m, 3 H), 7.33
(d, J ) 8.3 Hz, 2 H), 5.93 (dt, J ) 17.0, 10.1 Hz, 1 H), 5.12 (dd,
J ) 10.2, 2.0 Hz, 1 H), 5.05 (dd, J ) 17.1, 1.2 Hz, 1 H), 4.56 (s,
2 H), 4.45 (s, 2 H), 4.09 (dd, J ) 8.9, 7.0 Hz, 1 H), 3.37 (dd, J
) 10.0, 3.5 Hz, 1 H), 3.33-3.20 (m, 2 H), 3.26 (s, 6 H), 3.04 (br
dd, 1 H), 2.99 (m, 1 H), 2.63 (dd, J ) 18.8, 4.5 Hz, 1 H), 2.46
(m, 1 H), 2.44 (s, 3 H), 2.04 (m, 1 H), 1.67 (dd, J ) 8.3, 4.5 Hz,
1 H), 0.50 (s, 3 H), 0.48 (s, 3 H), 0.00 (s, 9 H). 13C NMR (CDCl3,
62.5 MHz) Major δ ) 144.4, 137.6, 136.6, 136.0, 134.2, 129.8,
129.1, 128.6, 127.7, 119.0, 103.0, 96.3, 95.8, 87.2, 82.9, 68.0,
64.0, 55.3, 55.2, 55.1, 48.3, 39.8, 24.1, 21.6, 17.0, -0.3, -3.1,
-4.9. Minor δ ) 144.4, 137.5, 136.6, 136.1, 134.1, 129.6, 129.3,
128.7, 128.0, 117.5, 103.3, 96.2, 95.1, 87.2, 86.1, 69.7, 64.0, 55.3,
55.2, 55.1, 48.4, 40.7, 27.4, 21.6, 18.0, -0.2, -4.8, -4.9. MS
(FAB+) m/z ) 665 (M + Na+, 7), 425 (6), 363 (4), 305 (5), 259
General Procedure for the Preparation of Allyl Sul-
fones. To a solution of sodium p-tolylsulfinate tetrahydrate
(1.1 mmol) in methanol was successively added a solution of
allyl acetate or allyl bromide (1 mmol) in THF and a solution
of Pd(PPh3)4 (0.05 mmol) in THF at room temperature under
a nitrogen atmosphere. The reaction mixture was stirred at
room temperature until disappearance of the starting material
(TLC) and then treated with KCN (0.2 mmol), and the solvents
were evaporated under vacuum. The residue was then purified
by chromatography on silica gel.
Methyl 3-Hydroxy-2-(1-(dimethyl(phenyl)silyl)allyl)-6-
tosylhex-4-enoate (24a). 24a was prepared according to the
general procedure above. Usual workup and chromatography
(petroleum ether/EtOAc 6/4) yielded 24a (E/Z: 95:5) as a
yellow viscous oil (785 mg, 94%). IR (neat) ν ) 3500, 3071,
2926, 1732, 1625, 1596, 1317, 1302, 1248, 1138, 1113, 1086,
1
702, 667 cm-1. H NMR (CDCl3, 300 MHz) δ ) 7.69 (m, J )
8.3 Hz, 2 H), 7.47 (m, 2 H), 7.34 (m, 3 H), 7.28 (d, J ) 7.9 Hz,
2 H), 5.58 (m, 3 H), 5.53 (m, 2 H), 4.27 (br d, J ) 9.4 Hz, 1 H),
3.71 (d, J ) 6.4 Hz, 2 H), 3.29 (s, 3 H), 3.24 (d, J ) 9.8 Hz, 1
H), 2.53 (m, 2 H), 2.38 (s, 3 H), 0.34 (s, 3 H), 0.24 (s, 3 H). 13
C
NMR (CDCl3, 62.5 MHz) δ ) 173.7, 144.2, 140.9, 136.1, 135.5,
135.4, 133.8, 129.4, 129.4, 128.9, 128.0, 127.9, 127.3, 116.7,
115.9, 69.8, 59.0, 51.1, 49.1, 33.2, 21.2, -3.5, -5.3. MS (FAB+)
m/z ) 495 (M + Na+, 100), 455 (5), 395 (3), 247 (6), 213 (14).
HRMS (FAB+) calcd for C25H32O5SSi + Na+: 495.1637.
Found: 495.1631.
(3-(Methoxymethoxy)-2-((methoxymethoxy)methyl)-5-
(tosylmethyl)-4-vinylcyclopentyl)dimethyl(phenyl)-
7994 J. Org. Chem., Vol. 70, No. 20, 2005

Cyclization of Chiral 3-Silylhepta-1,6-dienes
(4), 213 (23), 165 (13), 151 (17), 139 (27), 137 (15), 135 (100),
133 (49). HRMS (ES+) calcd for C34H50O6SSi2 + Na+: 665.2750.
Found: 665.2764.
128.4, 127.7, 96.9, 96.6, 85.9, 67.8, 63.8, 55.9, 55.3, 52.9, 49.4,
42.8, 42.7, 39.6, 29.9, 24.6, 21.6, -0.9, -3.2, -4.6. HRMS (ES+)
calcd for C35H50O7SSi2 + Na+: 693.2718. Found: 693.2714.
Tricyclic Ketone (29). Co2(CO)8 (41.5 mg, 0.121 mmol) was
added to enyne 28 (65 mg, 0.101 mmol) in dry ether (3 mL) at
room temperature. The reaction mixture was stirred for 1 h
and then concentrated under vacuum. The crude mixture was
then dissolved in THF (3 mL), and TMANO‚2H2O (73.8 mg,
0.666 mmol) was added by portions. The reaction mixture was
stirred for 3 h at room temperature, filtered, and concentrated
under vacuum. Purification through a short pad of neutral
alumina (CH2Cl2) yielded 29 as a yellow oil (50 mg, 74%). IR
(neat) ν ) 2950, 1687, 1549, 1302, 1280, 1248, 1146, 1110, 840,
703 cm-1. 1H NMR (CDCl3, 300 MHz) δ ) 7.70 (d, J ) 7.9 Hz,
2 H), 7.42-7.26 (m, 7 H), 4.67 (s, 2 H,), 4.43 (q, J ) 6.4 Hz, 2
H), 4.11 (m, 1 H), 3.40-3.27 (m, 3 H), 3.38 (s, 3 H), 3.30 (s, 3
H), 3.07-2.95 (m, 2 H), 2.75-2.42 (m, 3 H), 2.50 (s, 3 H), 2.13
(m, 1 H), 2.05 (m, 1 H), 1.96 (m, 1 H), 1.79 (m, 1 H), 0.37 (s, 3
H), 0.35 (s, 3 H), 0.03 (s, 9 H). 13C NMR (CDCl3, 62.5 MHz) δ
) 212.1, 184.5, 144.8, 139.1, 138.5, 137.7, 134.0, 129.9, 129.1,
Acknowledgment. This work was supported by the
Ministe`re de la Recherche et de la Technologie through
a Ph.D. grant to P.J. Y.L. gratefully acknowledges the
Institut Universitaire de France and the CNRS for
financial support. Finally, we thank Prof. M.-P. Ber-
trand (University of Marseille) and Prof. M. Malacria
(University Paris-VI) for stimulating discussions.
Supporting Information Available: Experimental pro-
cedures and structural data for new compounds; X-ray crystal-
lographic data for compounds 7b, 11, and 18 (cif files). This
material is available free of charge via the Internet at
http://pubs.acs.org.
JO051062G
J. Org. Chem, Vol. 70, No. 20, 2005 7995