Synthesis of (-)-astrogorgiadiol

Article abstract of DOI:10.1021/jo001519g

Reaction of Rh₂(S)-PTPA₄ with the (R)-citronellol-derived α-diazo-β-ketoester 1 led to the formation of cyclic β-ketoester 2 in 95% yield and 48% diastereomeric excess. The purity of 2 was increased to >99% de after one crystallization. To demonstrate its utility in steroid total synthesis, the β-ketoester 2 was carried on to secosteroid (-)-astrogorgiadiol (3), a naturally occuring vitamin D analogue with antiproliferative properties.

Full text of DOI:10.1021/jo001519g

944
J . Org. Chem. 2001, 66, 944-953
Syn th esis of (-)-Astr ogor gia d iol
Douglass F. Taber* and Scott C. Malcolm
Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716
taberdf@udel.edu
Received October 24, 2000
Reaction of Rh2(S)-PTPA4 with the (R)-citronellol-derived R-diazo-â-ketoester 1 led to the formation
of cyclic â-ketoester 2 in 95% yield and 48% diastereomeric excess. The purity of 2 was increased
to >99% de after one crystallization. To demonstrate its utility in steroid total synthesis, the
â-ketoester 2 was carried on to secosteroid (-)-astrogorgiadiol (3), a naturally occuring vitamin D
analogue with antiproliferative properties.
In tr od u ction
malignant cells.4 Unfortunately, it is unlikely that 1,25
could be used clinically as an anticancer agent, since
large doses also lead to dangerously high levels of calcium
in the blood, a condition known as toxic hypercalcemia.
Consequently, much research has gone into separating
the cell differentiation properties of vitamin D analogues
from their effect on calcium homeostasis.4 Most of the
analogues prepared so far (∼80%)4 have modifications on
the upper side chain. A-ring manipulation is rare because
of difficulties in synthesis, and aromatic A-ring analogues
are almost unheard of. Thus, aromatic A-ring secosteroids
constitute a largely unexplored class of vitamin D3
analogues.5
The development of general methods for the prepara-
tion of polycarbocyclic natural products with control of
both relative and absolute configuration has been a
longstanding challenge in organic synthesis. The problem
of controlling the configuration of a stereogenic center
on a pendant side chain relative to the ring to which it
is attached has been particularly troublesome. We
report that Rh-mediated cyclization of enantiomerically
pure 1 can proceed with appreciable selectivity for
insertion into HR, to give the nicely crystalline â-ketoester
2. The utility of 2 as a chiron for the preparation of
physiologically active natural products is illustrated by
the conversion of 2 to the antiproliferative marine seco-
steroid (-)-astrogorgiadiol (3).1,2
Astrogorgiadiol (3) was isolated1 from a J apanese
marine sponge of the genus Astrogorgia after extracts
showed significant biological activity. A purified sample
of 3 was subsequently found to inhibit cell division of
fertilized starfish eggs at 50 µg/mL. The structural
similarity of 3 to vitamin D suggested the possibility of
a similar mode of action. Further investigation of this
activity, however, depended on this substance being made
available by total synthesis. We envisioned that 3 could
be prepared from the symchiral â-ketoester 2.
P r ep a r a tion of th e D-Rin g Ch ir on . The starting
point for our synthesis was the inexpensive (R)-citronellal
5 (Scheme 1). Citronellol 6, prepared by lithium alumi-
num hydride reduction of 5, was converted to its ben-
zenesulfonate, and this was homologated with the dian-
ion of methyl acetoacetate.6 The resulting â-ketoester 8
so obtained had all the required carbons for the D-ring
Astr ogor gia d iol. Vitamin D and its metabolites (for
example, 1R,25-dihydroxy vitamin D3 (4)) control a
variety of processes in the body related to bone calcifica-
tion and the regulation of serum calcium levels.3 The
observation of vitamin D receptors in roughly 60% of all
cancer cell lines has led to the finding that high concen-
trations of 1,25 will suppress or differentiate many
(1) For the isolation, structure and biological activity of
(-)-astrogorgiadiol, see: Fusetani, N.; Nagata, H.; Hirota, H.; Tsuyuki,
T. Tetrahedron Lett. 1989, 30, 7079.
(4) Vitamin D; Feldman, D., Glorieux, F. H., Pike, J . W., Eds.;
Academic Press: New York, 1997.
(5) Zhu, G.-D.; Okamura, W. H. Chem. Rev. 1995, 95, 1877 and
references cited within.
(6) Huckin, S. N.; Weiler, L. J . Am. Chem. Soc. 1974, 96, 1082-
1087.
(2) While this work was in progress,
a
partial synthesis of
(-)-astrogorgiadiol from vitamin D was reported: Della Sala, G.; Izzo,
I.; De Riccardis, F.; Sodano, G. Tetrahedron Lett. 1998, 39, 4741.
(3) DeLuca, H. FASEB J ournal 1988, 2, 224.
10.1021/jo001519g CCC: $20.00 © 2001 American Chemical Society
Published on Web 01/12/2001

Synthesis of (-)-Astrogorgiadiol
J . Org. Chem., Vol. 66, No. 3, 2001 945
Sch em e 1
Sch em e 2
Davies.11 Through the combination of five catalysts and
six substrates (Scheme 2), we briefly explored the opti-
mization of asymmetric induction in the rhodium-medi-
ated cyclization reaction (Table 1). The pentyl side chain
substrates 9a and 9b (entries 1 and 2) were chosen as
controls. With these prochiral substrates it was possible
to gauge the effect solely of the catalyst. Similarly, the
achiral rhodium acetate catalyst was chosen to allow the
determination of the effect of the existing stereogenic
center on the dimethylhexenyl side chain.
In entries 3-6, the substrates were enantiomerically
pure. In these cases, we expected, with the enantiomeri-
cally pure catalysts, to have a matched and mismatched
pair of reacting molecules. Thus, it was important to
determine the diastereomeric excess for each pair. In
practice, it was easier to secure the enantiomer of the
substrates (entries 5 and 6) rather than the enantiomer
of each catalyst.
Hashimoto10 had noted significant improvement in the
product ratios when employing the sterically demanding
dimethylpentyl ester. It was assumed that the bulk of
the ester helped to further discriminate the two modes
of cyclization, but this sort of adjustment had not been
explored with the other catalysts. Thus, it was important
to test those substrates in each case (entries 2, 4, and 6).
The cyclizations were carried out as summarized in
Table 1. The pentyl products (10) were analyzed as
before.12 The dimethylhexenyl products (2) were analyzed
by gas chromatography on a chiral column. Entries 3 and
4 are of particular interest since they represent the
substrate that leads to the correct absolute configuration
of the natural steroids. For these entries, the combination
of the (S)-BiTISP catalyst and the methyl ester (R-1a )
gave the highest ratio of RR to SR (58% de), but (R-1a )
with (R)-PTPA gave a better overall yield of 2.
and for the upper side chain. Ketone 8 reacted smoothly
with mesyl azide7 to provide 1, the substrate for rhodium-
mediated cyclization.
The cyclization of the R-diazo-â-ketoester 1 with rhod-
ium octanoate resulted in the formation of a moderate
excess of (R,R)-2, the desired diastereomer (14% de),
presumably as a result of the effect of the stereogenic
center in close proximity to the insertion site.8 This
mixture could not be separated by silica gel chromatog-
raphy, but the desired (R,R)-2 could be brought to
diastereomeric purity by recrystallization.
Im p r oved Ca ta lysis of th e C-H In ser tion . If we
could improve the ratio of diastereomers from the cy-
clization of the diazo compound (1 f 2), we would
increase the final yield of the D-ring â-ketoester. To this
end, we screened several enantiomerically pure catalysts,
the MEPY catalyst of Doyle,9 the PTPA catalyst of
Hashimoto,10 and the DOSP and BiTISP catalysts of
(7) Taber, D. F.; Ruckle, R. E., J r.; Hennessy, M. J . J . Org. Chem.
1986, 51, 4077.
(8) For the first report of such an influence of a side chain
stereocenter on the diastereoselectivity of Rh-mediated C-H insertion,
see: Daniewski, A. R.; Warchol, T. Polish J . Chem. 1992, 66, 1985.
(9) For the MEPY catalyst, see: (a) Doyle, M. P.; Dyatkin, A. B.;
Roos, G. H. P.; Canas, F.; Pierson, D. A.; Vanbasten, A.; Muller, P.;
Polleux, P. J . Am. Chem. Soc. 1994, 116, 4507-4508. (b) Doyle, M. P.;
McKervey, M. A.; Ye, T. In Modern Catalytic Methods for Organic
Synthesis with Diazo Compounds; J ohn Wiley & Sons: New York,
1997.
The methyl ester cyclization was scaled up. The initial
3:1 mixture of diastereomeric products was recrystallized
(10) For the PTPA catalyst, see: Hashimoto, S.; Watanabe, N.; Sato,
T.; Shiro, M.; Ikegami, S. Tetrahedron Lett. 1993, 34, 5109-5112.

946 J . Org. Chem., Vol. 66, No. 3, 2001
Ta ble 1. R:S Ra tios a n d Isola ted Yield s for Rh -Med ia ted Cycliza tion s
Taber and Malcolm
acetate
(R)-PTPA
(5S)-MEPY
(S)-DOSP
(S)-BiTISP
1
2
3
9a
9b
R-1a
50:50 (98%)a
50:50 (99%)a
57:43 (92%)a
60:40 (99%)a
61:39 (98%)a
74:26 (98%)a
74:26 (62%)e
78:22 (89%)a
31:69 (92%)a
23:77 (88%)a
51:49 (82%)d
50:50 (75%)d
57:43 (66%)d
51:49 (95%)a
48:52 (99%)a
64:36 (94%)a
79:21 (38%)b
4
5
6
R-1b
S-1a
S-1b
61:39 (99%)a
45:55 (92%)a
38:62 (91%)a
57:43 (48%)d
42:58 (64%)d
38:62 (44%)d
73:27 (95%)a
46:54 (95%)a
35:65 (96%)a
67:33 (12%)c
75:25 (32%)b
40:60 (10%)c
2 h; 20 °C. 2 h, 60 °C. c 4 h, 60 °C. 16 h, 60 °C. 96 h, -10 °C.
a
b
d
e
Sch em e 3
and was used without further purification. The reduction
reaction was performed as we have previously de-
scribed13a,14 except the hydrogen pressure was reduced
from 50 to 10 psig. This permitted easy analysis of the
progress of the reaction by TLC. It was important to
carefully optimize the relative amounts of water and HCl
in the methanol solvent, to minimize the formation of
byproducts 12 and 13. The reduction was halted after
approximately 50% of the starting material had been
consumed.
After hydrogenation, the alcohol 11 was easily removed
by chromatography, allowing the production of gram-
scale batches of diastereomerically enriched 2 (>10:1).
The combination of these two procedures, the use of the
appropriate enantiomer of the Hashimoto catalyst10 to
give an initial preference for pro-R or pro-S C-H inser-
tion, followed by selective Ru-BINAP hydrogenation13,14
to polish the enantiomeric purity of the cyclized â-keto
ester, will be a powerful method for the preparation of
cyclopentane products of high enantiomeric purity.
Syn th esis of th e A-Rin g Syn th on . Although direct
Robinson annulation of R-methyl cyclohexanones usually
proceeds with substantial regioselectivity, the regiose-
lectivity has not always been satisfactory with R-methyl
cyclopentanones, so alternative strategies have been
developed.15 Nonetheless, the direct Robinson annulation
appeared so straightforward that we felt compelled to
investigate it. To this end, we prepared (Scheme 4) the
A-ring enone 20.
Our approach was based on the report of the specific
dibromination of m-methylanisole.16 The crystalline di-
bromide 15 was coupled with allylmagnesium chloride,17
and the resultant alkene was converted to the primary
alcohol by hydroboration. The methyl group was then
installed by Ni-catalyzed Grignard coupling, following the
procedure of Kumada.18 Oxidation of the alcohol 18 with
the Dess-Martin reagent afforded the aldehyde 19.
Addition of vinylmagnesium bromide to the crude alde-
hyde followed by reoxidation then led to the enone 20.
We found that the enone 20 so prepared was contami-
nated with variable amounts (∼5-10%) of the enone 22
lacking the aromatic methyl substituent. This impurity
presumably arose from reduction of the aromatic bromide
17 by an adventitious nickel hydride species. Since this
impurity could not be removed by chromatography, the
enone 20 was purified by preparation of the crystalline
phenyl ether 21.
once to give > 99% pure (R,R)-2 in an overall 40% yield
for the cyclization.
For all save one run in Table 1, the influence of the
side chain stereogenic center was dominant, not the
influence of the catalyst. The single exception was
uncovered using the combination of the (S)-BiTISP
catalyst11 and the methyl esters (entry 3 and entry 5 for
BiTISP). With BiTISP, the catalyst was dominant, de-
livering an excess of pro-R insertion even with (S)-1.
Kin etic Resolu tion by Ru -BINAP Hyd r ogen a -
tion . It was apparent that even the best of the chiral Rh
catalysts would not directly deliver products of high
enantiomeric purity. We therefore considered strategies
for improving the enantiomeric purity of the cyclopentane
products. In particular, kinetic resolution of the mixture
seemed possible, by the application of our modification
(Scheme 3) of the asymmetric ruthenium-BINAP hydro-
genation of â-ketoesters.13
Complexation of the chiral catalyst to the chiral
ketoester can result in either a matched or a mismatched
set. The matched set should be lower in energy and
therefore lead to a faster rate of reduction. The critical
question was, would the ratio of rates be sufficiently high
to allow the practical preparation of enantiomerically
pure products?
We briefly explored this question using the 1.4:1 ratio
of products from Rh2(octanoate)4 cyclization of 1. The
ruthenium catalyst was prepared by the known method13b
(14) Taber, D. F.; Wang, Y. J . Am. Chem. Soc. 1997, 119, 22.
(15) J ankowski, P.; Marczak, S.; Wicha, J . Tetrahedron 1998, 54,
12071 and references cited within.
(16) Hoye, T. R.; Mi, L. J . Org. Chem. 1997, 62, 8586.
(17) Taber, D. F.; Green, J . H.; Geremia, J . M. J . Org. Chem. 1997,
62, 9342.
(18) Tamao, K.; Sumitani, K.; Kiso, Y.; Zembayashi, M.; Fujioka,
A.; Kodama, S.-I.; Nakajima, I.; Minato, A.; Kumada, M. Bull. Soc.
Chem. J pn. 1976, 49, 1958.
(11) For the DOSP and BiTISP catalysts, see: Davies, H. M. L. Eur.
J . Org. Chem. 1999, 2459, 9.
(12) Taber, D. F.; Malcolm, S. C. J . Org. Chem. 1998, 63, 3717.
(13) (a) Taber, D. F.; Silverberg, C. J . Tetrahedron Lett. 1991, 32,
4227. (b) For the original report of the preparation of the Ru-BINAP
catalyst, see: Ikariya, T.; Ishii, Y.; Kawano, H.; Arai, T.; Saburi, M.;
Yoshikawa, S.; Akutagawa, S. J . Chem. Soc., Chem. Commun. 1985,
922.

Synthesis of (-)-Astrogorgiadiol
J . Org. Chem., Vol. 66, No. 3, 2001 947
Sch em e 4
Sch em e 5
of the trans ring fusion. As expected, reduction (Pd-C,
1 atm H2) of the side chain alkene proceeded quickly.
Reduction of the tetrasubstituted alkene proceeded more
slowly (90% conversion after 2 days), to give a mixture
of 28-31. It was convenient to equilibrate (KOH/MeOH)
this mixture prior to isolation. Under these conditions,
29 (13C δ 215.5) was converted to 28 (13C δ 213.0), while
30 (13C δ 214.9) and 31 (13C δ 215.1) were received as a
∼1:1 mixture. The ratio of 28 to 30 + 31 was 75:25.
Although the side chain alkene of 24 was not needed
for (-)-astrogorgiadiol 3, other related natural products
do have more highly substituted side chains. We there-
fore surveyed several of the methods for the selective
reduction of conjugated alkenes that have been devel-
oped. Where applicable, mixtures of 1,2 and 1,4 reduction
were oxidized and the epimeric product ketones were
equilibrated prior to quantitative analysis by 13C NMR.
As before, 33 (Scheme 6) was converted under these
conditions to 32 (13C δ 213.0; 1H δ 0.98, 3H, s), while 34
Robin son An n u la tion . Methylation of 2 (Scheme 5)
followed by decarbomethoxylation gave the R-methyl
ketone 25. The crude product so prepared was found to
be contaminated with ∼10% of a 1:1 mixture of the
isomeric cyanohydrins (24). Exposure of the crude prod-
uct from the decarbomethoxylation reaction to methanolic
KOH reversed the cyanohydrin formation, and the meth-
yl ketone 2519 was isolated in 75% yield from 23.
With 21 and 25 in hand, we were prepared to explore
the direct Robinson annulation. We were pleased to
observe that NaOMe-mediated condensation of 25 with
the phenyl ether 21 in refluxing methanol in fact
proceeded smoothly, to give the desired enone 26 as the
preponderant product. The major product 25 (1H NMR δ
0.97, 3H, d) contained ∼10% of the regioisomer 27 (1H
NMR δ 0.86, 3H, d). This minor regioisomer was conve-
niently separated at a later stage of the synthesis.
Red u ction of th e En on e. We next addressed the
problem of reducing the CD enone 26 to the trans-
hydrindane (Scheme 6). Literature results15 suggested
that catalytic hydrogenation would give a preponderance
1
1
(13C δ 214.9; H δ 1.11, 3H, s) and 35 (13C δ 215.1; H δ
0.88, 3H, s) were received as a ∼1:1 mixture.
The results from the screening of several different
reduction procedures15 are summarized in Table 2. It was
apparent that the most promising of these methods was
CuH (entry 5),23 so this procedure was optimized. The
CD enone 26 was added to a 10-fold excess of 2:1 DIBAL/
(19) For previous preparations of variations on the symchiral D-ring
ketone, see: (a) Desmaele, D.; Ficini, J .; Guingant, A.; Khan, P.
Tetrahedron Lett. 1983, 24, 3079. (b) Hatakeyama, S.; Hirotoshi, N.;
Takano, S. Tetrahedron Lett. 1984, 25, 3617. (c) Pan, L.-R.; Tokoy-
orama, T. Tetrahedron Lett. 1992, 33, 1469. (d) He, M.; Tanimori, S.;
Ohira, S.; Nakayama, M. Tetrahedron 1997, 53, 13307.
(20) Keinan, E.; Greenspoon, N. J . Am. Chem. Soc. 1986, 108, 7314.
(21) Enholm, E. J .; Kinter, K. S. J . Org. Chem. 1995, 60, 4850.
(22) Ojima, I.; Kogure, T. Organometallics 1982, 1, 1390-1399.

948 J . Org. Chem., Vol. 66, No. 3, 2001
Taber and Malcolm
Sch em e 6
Sch em e 7
with L-Selectride gave the expected axial alcohol 36,
which could at this stage be isolated free from the
contaminating regioisomer derived from 27. Deprotec-
tion of the aryl methyl ether proceeded24 uneventfully
to provide a compound that was identical to natural
1
(-)-astrogorgiadiol by H and 13C NMR and [R]D (obs )
-7.4°, c 0.095, CHCl3; lit.1 ) -16.4°, c 0.058, CHCl3; lit.2
1
) -4.6°, c 0.2, CHCl3). Comparison H NMR spectra were
provided by N. Fusetani and G. Sodano.
Ta ble 2. Red u ction of En on e 24
yield of yield of
Con clu sion
entry
reagent
conv
32
34 + 35
20
1
2
3
4
5
(Ph3P)4Pd,ZnCl2, Ph2SiH2
Bu3SnH, AIBN21
24%
68%
100%
80%
8%
25%
We expect that the two enantiomers of the crystalline
â-ketoester 2, readily prepared in gram quantities from
commericially available (R)- and (S)-citronellol, will be
valuable chirons for the construction of terpene-derived
natural products. The enantioselective Ru-BINAP hy-
drogenation of cyclopentanone carboxylates described
here should also open a general route to cyclopentane
derivatives of high enantiomeric purity.
31%
23%
13%
Li/liq NH3
(Ph3P)3RhCl,Et3SiH22
CuCN/DIBAL/BuLi23
25%
53%
49%
n-BuCu reagent at -50 °C, and the reaction temperature
was then raised to -20 °C over 10 min. The starting
material was completely consumed to give mixtures of
1,2, 1,4, and complete reduction. For simplicity, the crude
product was reoxidized and equilibrated before analysis.
The cis-fused ketones 34 and 35 could not be observed
in the mixture of products. The reoxidation step was
initially found to be problematic as a large amount of a
diene was isolated at the expense of the starting enone.
This side reaction could be minimized by slow addition
of the Dess-Martin reagent to the crude reduction
mixture.
Exp er im en ta l Section
1H NMR and 13C NMR spectra were obtained as solutions
in deuteriochloroform (CDCl3). 13C multiplicities were deter-
mined with the aid of a J VERT pulse sequence, differentiating
the signals for methyl and methine carbons as “d” from
methylene and quaternary carbons as “u”. The infrared (IR)
spectra were determined as neat oils. Mass spectra (MS) were
(23) For CuH reduction of trisubstituted CD enones, see: (a) Tsuda,
T.; Hayashi, T.; Satomi, H.; Kawamoto, T.; Saegusa, T. J . Org. Chem.
1986, 51, 537. (b) Tsuda, T.; Kawamoto, T.; Kumamoto, Y.; Saegusa,
T. Synth. Commun. 1986, 16, 639. (c) Daniewski, A. R.; Kiegiel, J .
Synth. Commun. 1988, 18, 115. (d) Loughlin, W. A.; Haynes, R. K. J .
Org. Chem. 1995, 60, 807.
Syn th esis of (-)-Astr ogor gia d iol. Hydrogenation
(Scheme 7) of the ketone 32 obtained from the CuH
reduction proceeded rapidly to produce material that was
identical to the major product (28) obtained from catalytic
hydrogenation of the enone 26. Reduction of this ketone
(24) Gevorgyan, V.; Liu, J .-X.; Rubin, M.; Benson, S.; Yamamoto,
Y. Tetrahedron Lett. 1999, 40, 8919.

Synthesis of (-)-Astrogorgiadiol
J . Org. Chem., Vol. 66, No. 3, 2001 949
obtained at an ionizing potential of 15 eV. Substances for
which C,H analyses are not reported were purified as specified
and gave spectroscopic data consistent with being >95% the
assigned structure. Optical rotations were determined as
solutions in dichloromethane unless otherwise noted. Rf values
indicated refer to thin-layer chromatography (TLC) on 2.5 ×
10 cm, 250 µm analytical plates coated with silica gel GF,
unless otherwise noted, and developed in the solvent system
indicated. All glassware was flame dried under a dry nitrogen
stream before use. Tetrahydrofuran (THF), diethyl ether, and
1,2-dimethoxyethane (DME) were distilled from sodium/ben-
zophenone ketyl under dry nitrogen. Dichloromethane (CH2-
Cl2) and toluene were distilled from calcium hydride under
dry nitrogen. All reaction mixtures were stirred magnetically,
unless otherwise noted.
yellow oil. The residue was distilled (bulb-to-bulb, 140-160
°C, 0.8 mmHg) to yield crude 8 (31.3 g, 64% of theoretical based
on citronellol) as a yellow oil, TLC Rf (10% MTBE/petroleum
1
ether) ) 0.37. H NMR (CDCl3, δ): 5.08 (m, 1H), 3.74 (s, 3H),
3.45 (s, 2H), 2.52 (t, J ) 7.34 Hz, 2H), 1.96 (m, 2H), 1.68 (s,
3H), 1.60 (s, 3H), 1.6-1.2 (m, 6H), 1.13 (m, 1H), 0.87 (d, J )
6.85 Hz, 3H). 13C NMR (CDCl3, δ): (u) 202.6, 167.5, 130.9, 48.8,
43.2, 36.8, 36.1, 25.3, 20.8; (d) 124.7, 52.1, 32.1, 25.6, 19.2, 17.5.
IR (cm-1): 2925, 1748, 1716.
Triethylamine (42 mL, 301 mmol) and mesyl azide (18 g,
149 mmol) were added to a solution of the acetoacetate 8 in
CH3CN (130 mL). Although the reaction is usually complete
in 2 h under these conditions, in this case, the mixture was
allowed to stir overnight. One-half of the solvent was removed
using a rotovap (50 °C) and the remaining liquid was parti-
tioned between petroleum ether and, sequentially, 3 M aque-
ous NaOH and brine. The solvent was removed in vacuo,
leaving 36.8 g of a brown oil. The residue was chromato-
graphed and the purest fraction provided 1 (26.4 g, 49% from
citronellol) as a pale yellow oil, TLC Rf (10% MTBE/petroleum
(R)-Citr on ellol (6). Powdered LiAlH4 (12.5 g, 329 mmol)
was introduced into a 5-L round-bottom flask along with 2 L
of THF (dried over 3 Å molecular sieve). The slurry was stirred
mechanically under N2, and the temperature was lowered to
0 °C. (R)-Citronellal (5, Takasago) (100 g, 648 mmol) was added
neat over about 15 min, and the residue was rinsed into the
flask with an additional 500 mL of THF. Notwithstanding the
fact that TLC indicated complete consumption of citronellal,
the mixture was warmed to a gentle reflux and subsequently
cooled to 0 °C. The reaction was worked up by sequential
dropwise addition of the following to the vigorously stirring
mixture: 12.5 mL of water, 12.5 mL of 3 M NaOH, and a
solution of 37.5 mL of water in 37.5 mL of THF. After filtration
through Celite and thorough washing of the solids with MTBE,
the solvent was evaporated to yield 116 g of crude citronellol.
This oil was distilled (bulb-to-bulb, 120-140 °C, 0.8 mmHg)
to give 6 (93.4 g, 92%) as a clear oil, TLC Rf (10% MTBE/
petroleum ether) ) 0.09. 1H NMR (CDCl3, δ): 5.10 (m, 1H),
3.66 (m, 2H), 2.1-1.8 (m, 3H), 1.68 (s, 3H), 1.7-1.5 (m, 2H),
1.60 (s, 3H), 1.4-1.3 (m, 2H), 1.17 (m, 1H), 0.90 (d, J ) 6.66
Hz, 3H). 13C NMR (CDCl3, δ): (u) 131.2, 61.0, 39.8, 37.1, 25.4;
(d) 124.6, 29.1, 25.6, 19.4, 17.6. IR (cm-1): 3331 (b), 2927, 1454,
1377, 1058. MS (m/z, %): 55 (64), 69 (100), 82 (44), 95 (33),
109 (14), 123 (18), 138 (6), 156 (6). HRMS calcd for C10H20O:
1
ether) ) 0.47. H NMR (CDCl3, δ): 5.09 (m, 1H), 3.84 (s, 3H),
2.83 (t, J ) 7.85 Hz, 2H), 1.96 (m, 2H), 1.68 (s, 3H), 1.60 (s,
3H), 1.7-1.5 (m, 2H), 1.5-1.3 (m, 3H), 1.15 (m, 2H), 0.88 (d,
J ) 6.49 Hz, 3H). 13C NMR (CDCl3, δ): (u) 193.0, 161.8, 131.1,
40.5, 36.9, 36.4, 25.5, 21.9; (d) 124.9, 52.1, 32.2, 25.7, 19.4, 17.6.
IR (cm-1): 2914, 2133, 1725, 1659, 1309. This substnace was
not stable to analysis by mass spectrometry.
Meth yl 5-Dim eth ylh exen yl-2-oxocyclop en ta n eca r box-
yla te (2) via Rh 2Oct4 Ca ta lysis. The CH2Cl2 used in the
following reaction was distilled from CaH2 and passed through
a column (20 × 150 mm) of anhydrous K2CO3 prior to use.
The R-diazo-â-ketoester 1 (25.5 g, 101 mmol) was dissolved in
1 L of CH2Cl2. A solution of rhodium octanoate (400 mg, 0.56
mol %) in 5 mL of dry CH2Cl2 was added in one portion. The
solution was allowed to stir for 12 h before being evaporated.
The green residue was chromatographed to provide cyclopen-
tanone 2 (17.0 g, 74%) as a light-green oil. Bulb-to-bulb
distillation (170 °C at 0.4 mmHg) yielded a clear oil which
formed moist crystals on standing. These did not show a sharp
melting point. TLC Rf (10% MTBE/petroleum ether) ) 0.28.
1H NMR (CDCl3, δ): 5.06 (m, 1H), 3.75 (s, 3H), 2.97 (d, J )
11.6 Hz, 0.43H), 2.95 (d, J ) 11.3 Hz, 0.57H), 2.55 (m, 1H),
2.5-2.3 (m, 2H), 2.3-1.9 (m, 3H), 1.69 (s, 3H), 1.61 (s, 3H),
1.6-1.3 (m, 3H), 1.17 (m, 1h), 0.93 (d, J ) 6.83 Hz, 1.3H), 0.91
(d, J ) 6.85 Hz, 1.7H). 13C NMR (CDCl3, δ): (Major) (u) 212.4,
170.7, 131.7, 38.7, 33.8, 25.3, 25.1; (d) 124.2, 59.6, 52.4, 46.7,
36.8, 25.7, 17.6, 16.9. 13C NMR (CDCl3, δ): (Minor) (u) 212.2,
170.3, 131.6, 38.5, 33.8, 25.3, 25.1; (d) 124.1, 59.7, 52.4, 46.3,
35.7, 25.7, 15.5, 16.9. IR (cm-1): 3447 (br), 2966, 1757, 1726.
Integration of the peaks (59.6 versus 59.7) and (46.7 versus
46.3) indicated a 1.4:1.0 mixture of components (14% de). The
result was nicely corroborated by GC-FID. Samples were
analyzed on a Hewlett-Packard HP6890 gas chromatograph
using a Chiraldex γ-cyclodextrin trifluoroacetyl capillary GC
column (30 m × 0.25 mm). Signals were obtained from a flame-
ionization detector. Upon injection of the sample (5 µL of a 1
mg/mL solution in ethyl acetate), the oven temperature was
maintained at 50 °C for 20 min. Then the temperature was
increased at 1 °C/min for 100 min. For the methyl ester, the
peaks at t ) 95.5 (minor) and t ) 96.6 (major) integrated at
43:57 (14% de). The dimethylpentyl ester was detected at t )
97.3 (minor) and t ) 98.9 (major).
156.1514, found 156.1514. [R]17 ) +3.74 (c 1.02, EtOH).
D
(S)-Meth yl 2-Diazo-7,11-dim eth yl-3-oxo-10-dodecen oate
(1). A solution of citronellol 6 (30.0 g, 192 mmol) in 400 mL of
CH2Cl2 was allowed to stir at 0 °C under N2 in a 1-L round-
bottom flask. Triethylamine (60 mL, 430 mmol) was added in
one portion, followed by DMAP (2.0 g, 1.64 mmol). Finally,
benzenesulfonyl chloride (30 mL, 240 mmol) was added
dropwise via syringe. After the mixture was stirred for 2 h at
0 °C, it was partitioned between petroleum ether and, sequen-
tially, 3 M aqueous HCl, water, and brine. The organic layer
was dried (Na2SO4) and the solvent was removed in vacuo at
30 °C. Removal of residual solvent under high vacuum yielded
citronellyl benzenesulfonate (62.4 g, 110% of theoretical) as a
clear yellow oil, TLC Rf (10% MTBE/petroleum ether) ) 0.41.
1H NMR (CDCl3, δ): 7.92 (m, 2H), 7.66 (m, 1H), 7.56 (m, 2H),
5.03 (m, 1H), 4.09 (m, 2H), 1.91 (m, 2H), 1.68 (m, 1H), 1.67 (s,
3H), 1.57 (s, 3H), 1.52 (m, 1H), 1.44 (m, 1H), 1.24 (m, 1H),
1.11 (m, 1H), 0.81 (d, J ) 6.49 Hz, 3H). 13C NMR (CDCl3, δ):
(u) 136.1, 131.4, 69.2, 36.6, 35.5, 25.1; (d) 133.6, 129.1, 127.7,
124.2, 28.7, 25.6, 18.9, 17.5. IR (cm-1): 2914, 1449, 1360, 1188,
944.
Sodium hydride (30 g, 60% in mineral oil, 750 mmol) was
suspended in 400 mL of THF in a 1-L three-neck round-bottom
flask fitted with a mechanical stirrer. The slurry was cooled
in an ice-water bath and methyl acetoacetate (46 mL, 426
mmol) was added rapidly dropwise. After the addition was
complete, the mixture was stirred for 10 min, then n-BuLi (165
mL, 2.33 M in hexanes, 384 mmol) was added rapidly drop-
wise. The mixture spontaneously warmed to reflux and then
cooled over the next 10 min. Finally, a solution of the crude
citronellyl benzenesulfonate (192 mmol theoretical) in 100 mL
of THF was added via cannula, and the mixture was stirred
at room temperature for 1 h. The mixture was quenched
cautiously by pouring into saturated aqueous NH4Cl before
being partitioned between MTBE and, sequentially, water and
brine. The solvent was removed in vacuo, yielding 67 g of a
Recrystallization was effected by adding 1 g of the crude 2
to 10 mL of ethanol and 10 mL of water and heating to 50 °C.
More ethanol was added to the oily mixture as it was
maintained at this temperature until all of the oil was
dissolved (up to an additional 10 mL). Seeding can be helpful
but, usually, cooling to room temperature produced signifi-
cantly diastereomerically enriched material (>99% de) as
1
white flakes, mp ) 58-59 °C. H NMR (CDCl3, δ): 5.07 (m,
1H), 3.75 (s, 3H), 2.95 (d, J ) 11.26 Hz, 1H), 2.55 (m, 1H),
2.5-2.3 (m, 2H), 2.20 (m, 1H), 2.06 (m, 1H), 1.94 (m, 1H), 1.69
(s, 3H), 1.61 (s, 3H), 1.6-1.4 (m, 3H), 1.16 (m, 1h), 0.91 (d, J
) 6.85 Hz, 3H). 13C NMR (CDCl3, δ): (u) 212.4, 170.7, 131.7,
38.7, 33.8, 25.3, 25.1; (d) 124.2, 59.6, 52.4, 46.7, 36.8, 25.7, 17.6,

950 J . Org. Chem., Vol. 66, No. 3, 2001
Taber and Malcolm
16.9. IR (cm-1): 2969, 1744, 1724, 1287, 1118. MS (m/z, %):
55(26), 69 (36), 81 (10), 109 (60), 141 (100), 168 (5), 220 (13),
252 (2). HRMS calcd for C15H24O3: 252.1725; found: 252.1731.
on 2 not recovered) as a green oil, TLC Rf (50% MTBE/
petroleum ether) ) 0.51.
An analogous experiment that was not vigorously stirred
yielded moderately enriched 2 (3.06 g, 78% de, 47% recovery
of this diasteromer). Recrystallization was effected by adding
this recovered 2 to 24 mL of ethanol and 12 mL of water and
heating to 30 °C. The temperature was reduced to 20 °C and
the solution was seeded with a few 97% de crystals. The
solution was cooled to -5 °C over 15 min, which produced large
crystals. The solution was further cooled to -20 °C and filtered.
The crystals were rinsed with a few milliliters of cold 50%
aqueous EtOH and dried in vacuo to give significantly dia-
stereomerically enriched 2 (1.46 g, 99% de, 29% overall
recovery of this diastereomer) as white flakes, mp 58-59 °C.
1H NMR (CDCl3, δ): 5.07 (m, 1H), 3.75 (s, 3H), 2.95 (d, J )
11.26 Hz, 1H), 2.55 (m, 1H), 2.5-2.3 (m, 2H), 2.20 (m, 1H),
2.06 (m, 1H), 1.94 (m, 1H), 1.69 (s, 3H), 1.61 (s, 3H), 1.6-1.4
(m, 3H), 1.16 (m, 1h), 0.91 (d, J ) 6.85 Hz, 3H). 13C NMR
(CDCl3, δ): (u) 212.4, 170.7, 131.7, 38.7, 33.8, 25.3, 25.1; (d)
124.2, 59.6, 52.4, 46.7, 36.8, 25.7, 17.6, 16.9. IR (cm-1): 2969,
1744, 1724, 1287, 1118. MS (m/z, %): 55(26), 69 (36), 81 (10),
109 (60), 141 (100), 168 (5), 220 (13), 252 (2). HRMS calcd for
[R]17 ) -51.45 (c 1.14, EtOH).
D
Rh 2(S)-P TP A4. (R)-(+)-Phenylalanine (500 mg, 3.03 mmol)
and phthalic anhydride (500 mg, 3.38 mmol) were combined
in a test tube. Three times the solid was heated until it melted
and then was allowed to cool. The solid was recrystallized from
50% ethanol/water (5 mL) to yield (R)-N-phthaloylphenylala-
nine (0.65 g, 73%) as a white solid, mp 179-180 °C. The (R)-
N-phthaloylphenylalanine (600 mg, 2.03 mmol) and rhodium
trifluoroacetate (164 mg, 0.25 mmol) were combined in a
round-bottomed flask. Three times the solid was dissolved in
dichloroethane (20 mL) and heated to dryness under a stream
of nitrogen. The green oil was chromatographed. The major
green fraction was collected and recrystallized from 10%
acetone/heptane (5 mL) to give small green needles (114 mg,
33% based on rhodium added). TLC Rf (1:1:3 acetone/CH2Cl2/-
petroleum ether) ) 0.28.
The mother liquor was combined with the other column
fractions and evaporated. More rhodium trifluoroacetate was
added (100 mg, 0.15 mmol). Two times the sample was
dissolved in dichloroethane (20 mL) and heated to dryness
under a stream of nitrogen. The green oil was dissolved in hot
2:1 methanol/CH2Cl2 (30 mL). Overnight a large mass of green
crystals separated. These were collected and dissolved in CH2-
Cl2 (10 mL). Methanol was added (30 mL) and the solution
was allowed to sit overnight in an open container. Large green
crystals separated and were collected. The crystals (presum-
ably the methanol adduct of the catalyst) were dissolved in
ethyl acetate to displace the methanol. Then, all of the solvent
was removed by heating in vacuo to yield Rh2(S)-PTPA4 (150
mg, 27% additional yield based on total rhodium added) as a
mint green powder.
C
15H24O3: 252.1725; found: 252.1731. [R]17D ) -51.45 (c 1.14,
EtOH).
4-Meth yl-3-(4-h yd r oxybu tyl)a n isole (16). A three-neck
round-bottom flask was charged with magnesium (4.2 g, 102.0
mmol), a few crystals of iodine, and 15 mL of THF. The mixture
was heated until the iodine color was discharged, then cooled
to 0 °C. A solution of allyl chloride (2.8 mL, 34.35 mmol) in 20
mL of THF was added slowly dropwise over about 1.5 h. After
the addition was complete, the gray solution was warmed to
reflux briefly and then cooled back down to 0 °C. This Grignard
solution was added rapidly dropwise via cannula to a solution
of dibromide 15 (4.75 g, 16.96 mmol) in 35 mL of THF that
was stirring at 0 °C. After 30 min, the solution was quenched
cautiously with saturated aqueous NH4Cl and partitioned
between MTBE and, sequentially, water and brine. The
combined organic extract was dried (Na2SO4) and evaporated
to give crude 16 (3.96 g, 97% of expected) as a pale-yellow oil,
TLC Rf (10% MTBE/petroleum Ether) ) 0.69. The NMR
spectrum of 16 indicated that it was 95% pure. 1H NMR
(CDCl3, δ): 7.33 (d, J ) 8.53 Hz, 1H), 6.69 (d, J ) 3.07 Hz,
1H), 6.55 (dd, J ) 8.53, 3.07 Hz, 1H), 5.80 (m, 2H), 5.00 (d, J
) 17.1 Hz, 1H), 4.93 (d, J ) 10.2 Hz, 2H), 3.70 (s, 3H), 2.70
(m, 2H), 2.29 (m, 2H). 13C NMR (CDCl3, δ): (u) 158.8, 142.0,
115.2, 114.9, 35.8, 33.8; (d) 137.6, 133.2, 116.1, 112.4, 55.4. IR
(cm-1): 2934, 1571, 1472, 1241, 1017. MS (m/z, %): 63 (21),
77 (44), 91 (38), 105 (15), 120 (32), 147 (10), 161 (100), 171
(16), 199 (84), 201 (82), 240 (41), 242 (40). HRMS calcd for
Meth yl 5-Dim eth ylh exen yl-2-oxocyclop en ta n eca r box-
yla te (2) via Rh 2(S)-P TP A4 Ca ta lysis. The R-diazo-â-ke-
toester 1 (2.00 g, 7.13 mmol) was dissolved in 100 mL of CH2Cl2
(dry, see above) and a solution of Rh2(S)-PTPA4 (47.8 mg, 0.48
mol %) in 5 mL of CH2Cl2 was added rapidly dropwise. After
2 h at room temperature, the solvent was evaporated and the
green residue was chromatographed to yield 2 (1.73 g, 96%)
as a clear oil. The mixture thus obtained was a 74:26 mixture
of (R,R)-2 and (S,R)-2 (48% de). The product was distilled and
crystallized as above to yield (R,R)-2 (722 mg, 40%) that was
1
diastereomerically pure by H and 13C NMR.
(S)-Ru BINAP Hyd r ogen a tion . A 5 mL reactivial was
charged with [RuCl2(COD)]n (39 mg, 0.139 mmol), (S)-BINAP
(100 mg, 0.151 mmol), triethylamine (200 mg, 1.98 mmol), and
toluene (4 mL). The vial was sealed and heated in a 140 °C oil
bath for 3 h. The solution was allowed to cool slightly and
transferred to a round-bottom flask (rinsing with hot THF)
under a N2 stream. The volatile material was removed in vacuo
to yield a brown solid. This material was dissolved in THF
(10 mL) to provide a 13.9 mM (S)-RuBINAP solution.
C
11H13BrO: 240.0149, found 240.0136.
Approximately 5% of the biaryl could be separated as a
1
white solid, TLC Rf (10% MTBE/petroleum ether) ) 0.53. H
NMR (CDCl3, δ): 7.44 (d, J ) 8.53 Hz, 2H), 6.72 (d, J ) 3.07
Hz, 2H), 6.65 (dd, J ) 8.53, 3.07 Hz, 2H), 3.78 (s, 6H), 3.96 (s,
4H).
Cyclopentanone 2 (5.07 g, 20.1 mmol, 14% de) was dissolved
in MeOH (100 mL) in a modified Parr bottle. Water (0.5 mL,
28 mmol) and a solution of HCl in MeOH (5.0 mL of a stock
solution made from 1 mL of concentrated aqueous HCl and
99 mL of MeOH, 0.60 mmol) were added and the solution was
purged with N2 for 5 min. The 13.9 mM (S)-RuBINAP solution
(10 mL, 0.69 mol % based on 2) prepared above was added.
Four times the flask was evacuated using a water aspirator
and then refilled with H2 (10 psig) causing the green solution
to turn brown. The flask was heated in a 60 °C oil bath with
vigorous stirring under an H2 atmosphere until the starting
material was about half consumed (2 h, TLC). The mixture
was transferred to a round-bottom flask and the solvent was
removed in vacuo. The residue was diluted in CH2Cl2 (100 mL).
Flash silica gel (10 g) and 3 M aqueous HCl (1 mL) were added
and the mixture was allowed to stir for 30 min. The solvent
was then evaporated and the resultant solid was chromato-
graphed to yield the recovered cyclopentanone 2 (2.10 g, 84%
de, 66% recovery of this diastereomer) as a clear oil. Also
isolated was the reduced compound 11 (2.64 g, 90% yield based
A round-bottom flask was charged with a 2 M solution of
BH3‚DMS in THF (15 mL, 30 mmol) and cooled to 0 °C. A
solution of cyclohexene (6.0 mL, 59.22 mmol) in 15 mL of THF
was added dropwise. The cooling bath was removed and the
mixture was allowed to stir at ambient temperature for 1 h.
A solution of crude alkene 16 (15.42 mmol theoretical) in 15
mL of THF was added in one portion and the resultant mixture
was allowed to stir for 2 h. At this time, the mixture was
quenched by cautious addition of ethanol (17 mL), 3 M aqueous
NaOH (17 mL), and 30% H2O2 (51 mL) and allowed to stir
overnight. The mixture was partitioned between saturated
aqueous NH4Cl and MTBE, dried over Na2SO4, and evapo-
rated. Bulb-to-bulb distillation (60 °C at 0.5 mmHg) of the
residue afforded 17 (5.6 g. 140% of expected) as a yellow oil.
The material is contaminated with residual cyclohexanol but
was used directly in the next step. An analytically pure sample
was obtained by chromatography, TLC Rf (50% MTBE/
petroleum ether) ) 0.27. 1H NMR (CDCl3, δ): 7.38 (d, J ) 8.53
Hz, 1H), 6.76, (d, J ) 3.07 Hz, 1H), 6.61 (dd, J ) 8.53, 3.07
Hz, 1H), 3.76 (s, 3H), 3.70 (m, 2H), 2.72 (m, 2H), 1.68 (m, 3H).

Synthesis of (-)-Astrogorgiadiol
J . Org. Chem., Vol. 66, No. 3, 2001 951
13C NMR (CDCl3, δ): (u) 158.7, 142.4, 114.7, 62.4, 35.9, 32.1,
25.9; (d) 133.1, 115.9, 112.9, 55.3. IR (cm-1): 3347 (b), 1595,
1572, 1472, 1241. MS (m/z, %): 77 (24), 91 (25), 105 (11), 121
(47), 146 (9), 161 (100), 178 (8), 199 (28), 201 (26), 212 (35),
214 (30), 258 (36), 260 (31). HRMS calcd for C11H15BrO2:
258.0255, found 258.0255.
The crude vinyl carbinol was dissolved in 30 mL of CH2Cl2
and Dess-Martin periodinane (3.90 g, 9.20 mmol) was added
as a suspension in 20 mL of CH2Cl2. The reaction warmed to
reflux briefly on its own accord. After 30 min, the mixture was
evaporated onto silica gel and chromatographed to afford enone
20 (1.72 g, 86% from the vinyl carbinol) as a clear oil. The
three step yield (from alcohol 18) was 59% based on starting
material not recovered. For 20, TLC Rf (50% MTBE/petroleum
The crude 17 (15.42 mmol theoretical) was diluted with 140
mL of dry ether and cooled to 0 °C. NiCl2(dppp) (180 mg, 0.314
mmol) was added to produce a red suspension. A 3 M solution
of MeMgBr in ether (11.3 mL, 33.9 mmol) was added dropwise
to produce, initially, a milky white precipitate. After about two-
thirds of the Grignard had been added, the red particles
became sticky and adhered to the walls of the flask. As the
solution was warmed to reflux, the red particles were dissolved
and the liquid turned yellow. The reaction was monitored over
the next 84 h as additional amounts of catalyst and Grignard
were added to complete the reaction. The mixture was cooled
and quenched with saturated aqueous NH4Cl solution, ex-
tracted with ether, and washed with brine. The organic layer
was dried over Na2SO4 and the solvent removed in vacuo. Bulb-
to-bulb distillation (170-180 °C at 0.5 mmHg) afforded 18
(2.27 g, 76% yield from dibromide) as a clear oil, TLC Rf (50%
1
ether) ) 0.65. H NMR (CDCl3, δ): 7.04 (d, J ) 8.19 Hz, 1H),
6.69 (d, J ) 2.73 Hz, 1H), 6.65 (dd, J ) 8.19, 2.73 Hz, 1H),
6.35 (dd, J ) 17.4, 10.2 Hz, 1H), 6.19 (dd, J ) 17.8, 1.02 Hz,
1H), 5.80 (dd, J ) 10.2, 1.02 Hz, 1H), 3.76 (s, 3H), 2.64 (t, J )
7.17 Hz, 2H), 2.59, (m, 2H), 2.23, (s, 3H), 1.90 (m, 2H). 13C
NMR (CDCl3, δ): (u) 200.4, 157.7, 140.9, 128.0, 127.8, 38.1,
32.6, 23.9; (d) 136.4, 130.8, 114.6, 110.9, 55.1, 18.2. IR (cm-1):
2945, 1681, 1612, 1500, 1253. MS (m/z, %): 91 (12), 135 (13),
136 (10), 148 (100), 218 (25). HRMS calcd for C14H18O2:
218.1307, found 218.1303.
The three-step yield for the same procedure using freshly
prepared vinylmagnesium bromide (0.5M, 1.5 equiv) was 63%.
There was no recovered alcohol 18.
The enone 20 prepared by the procedure outlined above was
contaminated with about 5% of the desmethyl enone 22. These
could be separated by formation of the crystalline phenoxy
ketone 21.
1
MTBE/petroleum ether) ) 0.34. H NMR (CDCl3, δ): 7.03 (d,
J ) 8.19 Hz, 1H), 6.71 (d, J ) 2.73 Hz, 1H), 6.65 (dd, J )
8.19, 2.73 Hz, 1H), 3.77 (s, 3H), 3.67 (d, J ) 6.94 Hz, 2H),
2.59 (m, 2H), 2.22 (s, 3H), 1.64 (m, 2H). 13C NMR (CDCl3, δ):
(u) 157.7, 141.2, 127.9, 62.2, 33.8, 29.6; (d) 130.8, 114.6, 110.7,
55.1, 18.2. IR (cm-1): 3361 (br), 2936, 1609, 1499, 1252, 1036.
MS (m/z, %): 65 (11), 77 (16), 79 (11), 91 (28), 105 (10), 121
(35), 123 (27), 135 (100), 136 (43), 148 (20), 149 (11), 161 (9),
176 (8), 194 (53). HRMS calcd for C12H18O2: 194.1307, found
194.1301.
P h en oxy Keton e 21. NaH (25 mg, 1.1 mmol) was added
in small portions to a solution of phenol (1 mL) in THF (1 mL).
After 30 min, enone 20 (100 mg, 0.458 mmol) was added as a
solution in 1 mL of THF. The mixture was allowed to stir at
ambient temperature for 3 h before being partitioned between
saturated aqueous NH4Cl and MTBE. The organic layer was
dried over Na2SO4 and the solvent was removed in vacuo. The
residue was chromatographed and recrystallized (10% EtOAc/
petroleum ether) to yield 21 (109 mg, 76%) as a white solid,
mp 44.0-45.0 °C. TLC Rf (20% MTBE/petroleum ether) ) 0.41.
1H NMR (CDCl3, δ): 7.3 (m, 2H), 7.04 (d, J ) 8.19 Hz, 1H),
6.94 (t, J ) 7.17 Hz, 1H), 6.88 (d, J ) 8.53 Hz, 2H), 6.69 (d, J
) 2.73 Hz, 1H), 6.66 (dd, J ) 8.19, 2.73 Hz, 1H), 4.22 (t, J )
6.14 Hz, 2H), 3.76 (s, 3H), 2.86 (t, J ) 6.14 Hz, 2H), 2.6 (m,
4H), 2.23 (s, 3H), 1.9 (m, 2H). 13C NMR (CDCl3, δ): (u) 208.3,
158.5, 157.7, 140.9, 128.0, 62.8, 42.8, 42.1, 32.6, 23.6, 18.3; (d)
130.9, 129.4, 120.9, 114.7, 114.4, 110.9, 55.2, 18.3. IR (cm-1):
2939, 1714, 1600, 1496, 1246. HRMS calcd for C12H24O3:
312.1726, found 312.1720.
The material was contaminated with about 5% of the
corresponding desmethyl compound, which could not be ef-
1
ficiently separated by column chromatography. H NMR (δ):
7.18 (t, 7.85 Hz, 1H), 6.7-6.8 (m, 2 H), 3.77 (s, 3H). 13C NMR
(δ): (u) 159.4, 143.9; (d) 129.1, 120.7, 114.1, 110.8, 55.0.
En on e 20. The Dess-Martin periodinane (6.80 g, 16.03
mmol) was suspended in 40 mL of dry CH2Cl2 and a solution
of arylbutanol 18 (3.10 g, 15.96 mmol) in 40 mL of CH2Cl2 was
added to it. The white precipitate was dissolved immediately
but reformed slowly. After 30 min at ambient temperature,
the reaction was partitioned between ether and, sequentially,
1:1 10% aqueous Na2S2O3/saturated aqueous NaHCO3, water,
and brine. The aqueous washes were back extracted with ether
and washed with water and brine. The combined organic
extracts were dried over Na2SO4, filtered through silica, and
evaporated to give crude 19 (3.64 g, 119% of expected) as a
pale-yellow oil, TLC Rf (50% MTBE/petroleum ether) ) 0.60.
A sample was further purified by chromatography for analysis.
1H NMR (CDCl3, δ): 9.77 (t, J ) 1.71 Hz, 1H), 7.05 (d, J )
8.19 Hz, 1H), 6.68 (m, 1H), 6.66 (d, J ) 2.73 Hz, 1H), 3.77 (s,
3H), 2.60 (m, 2H), 2.50 (dt, J ) 7.17, 1.71 Hz, 2H), 2.23 (s,
3H), 1.91 (m, 2H). 13C NMR (CDCl3, δ): (u) 157.1, 140.6, 127.9,
43.3, 32.6, 22.3; (d) 202.3, 131.0, 114.7, 111.0, 55.1, 18.3. IR
(cm-1): 2945, 1723, 1609, 1504, 1252. MS (m/z, %): 65 (17),
77 (23), 91 (41), 105 (13), 121 (42), 135 (100), 135 (50), 148
(45), 164 (26), 192 (39). HRMS calcd for C12H16O2: 192.1150,
found 192.1158.
Meth yla ted D-Rin g Ch ir on (23). To a solution of cyclo-
pentanone 2 (2.00 g, 7.93 mmol) in 40 mL of acetone was added
methyl iodide (1.0 mL, 16.06 mmol) and K2CO3 (4.0 g, 28.94
mmol). The mixture was heated to reflux for 5 h before being
cooled and partitioned between saturated aqueous NH4Cl and
EtOAc. The organic layer was dried over Na2SO4, filtered
through silica, and evaporated. The residue was chromato-
graphed to yield the C-methylated compound 23 (1.80 g, 85%)
as a clear oil, TLC Rf (10% MTBE/petroleum ether) ) 0.40.
1H NMR (CDCl3, δ): 5.08 (m, 1H), 3.67 (s, 3H), 2.57 (m, 1H),
2.24 (m, 1H), 2.15 (m, 1H), 2.04 (m, 1H), 2.0-1.7 (m, 3H), 1.69
(s, 3H), 1.61 (s, 3H), 1.49 (m, 2H), 1.40 (2, 3H), 1.15 (m, 1H),
1.01 (d, J ) 6.83 Hz). 13C NMR (CDCl3, δ): (u) 216.9, 171.4,
131.4, 59.2, 37.5, 34.1, 25.1, 24.5; (d) 124.3, 55.0, 51.8, 35.2,
25.6, 21.3, 18.2, 17.6. IR (cm-1): 2967, 1753, 1732, 1222, 1172.
MS (m/z, %): 67 (35), 82 (26), 97 (41), 109 (100), 123 (38), 137
(31), 160 (28), 188 (56), 219 (10), 233 (42), 248 (89), 266 (2).
HRMS calcd for C16H26O3: 266.1882, found 266.1882.
Also isolated was the O-methylated compound (0.22 g, 10%
yield) as a clear oil, TLC Rf (10% MTBE/petroleum ether) )
The crude aldehyde 19 (15.95 mmol theoretical) was dis-
solved in 25 mL of THF and cooled to 0 °C. A 1 M solution of
commercial vinylmagnesium bromide (25 mL, 25.00 mmol) in
THF was added dropwise. After 1 h at 0 °C, the mixture was
partitioned between saturated aqueous NH4Cl and MTBE,
dried over Na2SO4 and evaporated. Silica gel chromatography
afforded the vinyl carbinol (2.02 g, 57% yield from alcohol 18)
and alcohol 18 (0.48 g, 15% recovered). For the vinyl carbinol,
TLC Rf (50% MTBE/petroleum ether) ) 0.56. 1H NMR (CDCl3,
δ): 7.04 (d, J ) 8.19 Hz, 1H), 6.70 (d, J ) 2.73 Hz, 1H), 6.65
(dd, J ) 8.19, 2.73 Hz, 1H), 5.87 (ddd, J ) 17.07, 10.24, 6.49,
1H), 5.23 (dt, J ) 17.07, 1.37 Hz, 1H), 5.11 (dt, J ) 10.24,
1.37 Hz, 1H), 4.14 (m, 1H), 3.78 (s, 3H), 2.7-2.5 (m, 2H), 2.23
(s, 3H), 1.7-1.5 (m, 4H). 13C NMR (CDCl3, δ): (u) 157.7, 141.7,
127.9, 114.8, 36.8, 33.3, 25.9; (d) 141.1, 130.8, 114.7, 110.7,
73.2, 55.2, 18.4. IR (cm-1): 3415 (br), 2934, 1609, 1499, 1251.
1
0.07. H NMR (CDCl3, δ): 5.04 (m, 1H), 3.83 (s, 3H), 3.70 (s,
3H), 2.97 (m, 1H), 2.58 (m, 2H), 2.02 (m, 1H), 2.0-1.7 (m, 3H),
1.66 (s, 3H), 1.58 (s, 3H), 1.7-1.6 (m, 1H), 1.32 (m, 2H), 0.89
(d, J ) 6.83 Hz, 3H). 13C NMR (CDCl3, δ): (u) 169.4, 165.6,
130.9, 106.3, 30.5, 30.3, 26.1, 20.6; (d) 124.9, 57.5, 50.6, 47.7,
34.6, 25.6, 17.7, 17.5. IR (cm-1): 2951, 1690, 1627, 1376, 1233,
1060. MS (m/z, %): 95 (12), 155 (100), 182 (10), 195 (13), 266
(16). HRMS calcd for C16H26O3: 266.1882, found 266.1901.
The O-alkylated material (0.22 g, 0.827 mmol), as a solution
in 30 mL of THF, could be recycled by brief exposure to
concentrated aqueous HCl (1 mL). After 15 min, the mixture

952 J . Org. Chem., Vol. 66, No. 3, 2001
Taber and Malcolm
was filtered through silica, evaporated, and chromatographed
to give 2 (0.11 g, 53%) as clear oil. The overall yield of 23 then
became 90% based on starting material not recovered.
Dim eth ylh exen yl-m eth ylcyclop en ta n on e (25). To a so-
lution of methylated compound 23 (1.80 g, 6.76 mmol) in 67
mL of HMPA was added NaCN (1.32 g, 26.93 mmol). The
mixture was warmed to 75-80 °C for 5 h before being
partitioned between MTBE and, sequentially, saturated aque-
ous NaHCO3 and water. The organic layer was evaporated.
When chromatographed at this stage, the mixture was found
to contain about 10% of the diastereomeric cyanohydrins 24
in roughly equal amounts, TLC Rf (10% MTBE/petroleum
ether) ) 0.16. 1H NMR (CDCl3, δ): 5.08 (m, 1H), 2.35 (br s,
1H), 2.2-1.7 (m, 6H), 1.69 (s, 3H), 1.60 (s, 3H), 1.6-1.5 (m,
2H), 1.3-1.2 (m, 2H), 1.10 (d, J ) 6.83 Hz, 3H), 0.93 (m, 1H),
0.80 (d, J ) 6.83 Hz, 3H). IR (cm-1): 3438 (br), 2239 (w).
TLC Rf (10% MTBE/petroleum ether) ) 0.14. 1H NMR
(CDCl3, δ): 5.08 (m, 1H), 2.66 (br s, 1H), 2.23 (m, 1H), 2.1-
1.8 (m, 4H), 1.8-1.4 (m, 4H), 1.69 (s, 3H)1.61 (s, 3H), 1.4-1.1
(m, 2H), 1.16 (d, J ) 6.83 Hz, 3H), 0.83 (d, J ) 6.83 Hz, 3H).
IR (cm-1): 3432 (br), 2239 (w).
10 min and then the mixture was allowed to stir at -20 °C
for 1 h. The reaction was quenched with a solution of 1:1
saturated aqueous NH4Cl and 3 M aqueous HCl (20 mL) at
-20 °C and allowed to warm to ambient temperature over 30
min. The mixture was filtered and extracted with MTBE. The
organic layer was evaporated and then partitioned between
MTBE and, sequentially, 1 M aqueous HCl, water, and brine.
The organic layer was dried (Na2SO4), filtered through silica
gel, and evaporated to yield the crude reduction product (105
mg) as a pale-yellow oil.
A 0.1 M solution of the Dess-Martin periodinane in CH2-
Cl2 (1 mL, 0.1 mmol) was added to a solution of the crude oil
from reduction in CH2Cl2. After 30 min, and again after 1 h,
additional portions (1 mL each) of periodinane were added.
After 2 h, the reaction was complete (TLC). A solution of 1:1
10% aqueous Na2S2O4/1 M aqueous NaOH (20 mL) was added
and the mixture was allowed to stir for 30 min. The mixture
was then partitioned between MTBE and, sequentially, water
and brine. The organic layer was dried (Na2SO4), filtered
through silica, and evaporated to yield the crude oxidation
product (104 mg) as a yellow oil.
The crude residue from decarbomethoxylation was instead
diluted with 20 mL of 10% KOH in MeOH. After 5 min at
ambient temperature, the mixture was partitioned between
petroleum ether and, sequentially, water and brine. The
organic layer was dried over Na2SO4 and evaporated. The
residue was chromatographed to give ketone 25 (1.06 g, 75%)
as a clear oil, TLC Rf (10% MTBE/petroleum ether) ) 0.46.
1H NMR (CDCl3, δ): 5.10 (m, 1H), 2.30 (dd, J ) 18.43, 8.53
Hz, 1H), 2.10 (m, 2H), 2.0-1.8 (m, 3H), 1.69 (s, 3H), 1.62 (s,
3H), 1.7-1.6 (m, 2H), 1.6-1.4 (m, 2H), 1.15 (m, 1H), 1.08 (d,
J ) 6.83 Hz, 3H), 1.00 (d, J ) 6.49 Hz, 3H). 13C NMR (CDCl3,
δ): (u) 221.6, 131.5, 37.2, 32.3, 25.8, 23.1; (d) 124.4, 50.1, 46.8,
34.0, 25.6, 17.6, 13.8. IR (cm-1): 2963, 1742, 1456, 1378, 1159.
MS (m/z, %): 55 (91), 69 (87), 82 (34), 97 (100), 110 (17), 120
(21), 137 (32), 138 (27), 152 (5), 208 (71). HRMS calcd for
The crude oil from oxidation was dissolved in a 1:1:4 solution
of 1% aqueous KOH/MeOH/THF (20 mL) and stirred at
ambient temperature for 1 h. The mixture was partitioned
between MTBE and saturated aqueous NH4Cl. The organic
layer was evaporated and then partitioned between MTBE
and, sequentially, water and brine. The organic layer was
evaporated and chromatographed to yield the recovered enone
26 (50.7 mg, 51%) and the trans-hydrindanone 32 (26.1 mg,
52% based on 26 not recovered) as a clear oil, TLC Rf (10%
1
MTBE/petroleum ether) ) 0.42. H NMR (CDCl3, δ): 7.03 (d,
J ) 8.53 Hz, 1H), 6.72 (d, J ) 2.73 Hz, 1H), 6.64 (dd, J )
8.53, 2.73 Hz, 1H), 5.08 (m, 1H), 3.77 (s, 3H), 2.70 (m, 1H),
2.6-2.2 (m, 4H), 2.27 (s, 3H), 2.17 (m, 1H), 2.00 (m, 2H), 1.9-
1.5 (m, 6H), 1.69 (s, 3H), 1.61 (s, 3H), 1.5-1.0 (m, 6H), 0.98
(s, 3H), 0.95 (d, J ) 6.49 Hz, 3H). 13C NMR (CDCl3, δ): (u)
213.0, 157.7, 142.3, 131.2, 128.1, 42.8, 38.5, 38.3, 35.7, 31.2,
29.0, 27.7, 25.1, 24.6; (d) 130.8, 124.9, 114.5, 110.8, 55.22,
55.21, 55.0, 50.4, 35.4, 25.7, 18.4, 18.3, 17.6, 11.5. IR (cm-1):
2952, 2869, 1708, 1500, 1251.
C
14H24O: 208.1827, found 208.1833. Anal. Calcd for C14H24O:
C, 80.71; H, 11.61. Found: C, 80.83; H, 11.93. [R]17D ) -48.51
(c 1.14, EtOH).
CD En on e 26 fr om th e A-Rin g P h en yl Eth er 21. A
solution of 5 M NaOMe in MeOH (0.14 mL, 0.70 mmol) was
added to a solution of ketone 25 (79.6 mg, 0.382 mmol) in 2
mL of MeOH at 0 °C. After 10 min, a solution of A-ring phenyl
ether 21 (109 mg, 0.349 mmol) in 10 mL of MeOH was added
dropwise. The solution was allowed to stir at ambient tem-
perature for 3 h and then warmed to reflux for 26 h. The cooled
solution was then partitioned between saturated aqueous NH4-
Cl and ether. The organic layer was dried over Na2SO4 and
evaporated. The residue was chomatographed to yield recov-
ered 25 (23.1 mg, 29%), and after a base wash (3 M aqueous
NaOH, to remove phenol), the CD enone 26 (82.7 mg, 78%
based on 25 not recovered) as a clear oil, TLC Rf (20% MTBE/
petroleum ether) ) 0.45. 1H NMR (CDCl3, δ): 7.01 (d, J ) 8.19
Hz, 1H), 6.7-6.6 (m, 2H), 5.08 (m, 1H), 3.76 (s, 3H), 2.7-2.5
(m, 3H), 2.5-2.3 (m, 3H), 2.28 (s, 3H), 2.2-2.1 (m, 3H), 2.02
(m, 1H), 2.0-1.8 (m, 3H), 1.69 (s, 3H), 1.61 (s, 3H), 1.6-1.2
(m, 4H), 1.11 (m, 1H), 1.02 (s, 3H), 0.97 (d, J ) 6.83 Hz, 3H).
13C NMR (CDCl3, δ): (u) 198.3, 174.3, 157.5, 141.4, 131.3,
130.2, 128.0, 45.1, 36.6, 35.6, 33.6, 32.1, 27.4, 27.0, 26.6, 24.5;
(d) 130.6, 124.6, 115.4, 110.8, 55.8, 55.1, 33.7, 25.7, 18.7, 18.3,
17.6, 16.4. IR (cm-1): 2961, 1661, 1503, 1455, 1251.
The bicyclic enone 26 was contaminated with ∼10% of its
regioisomer 27 that could not be separated by chromatography.
This regioisomer was identified by its diagnostic resonance in
the 1H NMR (δ 0.86, d).
Cop p er Hyd r id e Red u ction . Copper cyanide (220 mg,
2.46 mmol) was suspended in THF (10 mL) and chilled to -20
°C. A 2.25 M solution of n-BuLi in hexanes (1.1 mL, 2.2 mmol)
was added dropwise. The brown solution was stirred at -20
°C for 30 min, and then the temperature was lowered to -50
°C. A 1 M solution of DIBAL in hexanes (4.9 mL, 4.9 mmol)
was added slowly dropwise. The dark brown solution was
allowed to stir at -50 °C for 1 h before the CD enone 26 (99.0
mg, 0.242 mmol) was added as a solution in 1:1 THF/HMPA
(3.6 mL). The temperature was raised to -20 °C over the next
Also isolated was a diene (11.0 mg, 11%), presumably formed
from elimination of the 1,2 reduction product, as a pale-yellow
oil, TLC Rf (10% MTBE/petroleum ether) ) 0.73. 1H NMR
(CDCl3, δ): 7.04 (d, J ) 8.53 Hz, 1H), 6.70 (d, J ) 2.73 Hz,
1H), 6.65 (dd, J ) 8.53, 2.73 Hz, 1 H), 5.51 (m, 2H), 5.12 (m,
1H), 3.78 (s, 3H), 2.7 (m, 2H), 2.5-1.8 (m, 9H), 2.25 (s, 3H),
1.69 (s, 3H), 1.62 (s, 3H), 1.6 (m, 1H), 1.44 (m, 2H), 1.1 (m,
1H), 0.98 (d, J ) 6.14, 3H), 0.88 (s, 3H). 13C NMR (CDCl3, δ):
(u) 157.8, 148.7, 142.1, 133.0, 131.1, 127.9, 45.8, 36.4, 36.1,
35.6, 33.6, 33.1, 24.6, 23.9; (d) 130.7, 125.1, 124.4, 119.1, 114.7,
110.8, 56.9, 55.2, 33.8, 25.7, 18.8, 18.4, 17.6, 15.5. IR (cm-1):
2926, 1609, 1503, 1250, 1046.
Astr ogor gia d iol Meth yl Eth er (36). A sample of 5% Pd/C
(20 mg) was added to a solution of the trans-hydrindanone 32
(38.3 mg, 0.0933 mmol) in ethanol (10 mL). Three times in
succession, the flask was evacuated and refilled with H2. After
2 h at ambient temperature and pressure, the mixture was
filtered and evaporated. Column chromatography afforded the
saturated trans-hydrindanone (32.4 mg, 84%) as a clear oil,
TLC Rf (10% MTBE/petroleum ether) ) 0.42. 1H NMR (CDCl3,
δ): 7.03 (d, J ) 8.53 Hz, 1H), 6.72 (d, J ) 2.73 Hz, 1H), 6.64
(dd, J ) 8.53, 2.73 Hz, 1H), 3.77 (s, 3H), 2.7 (m, 1H), 2.6-2.2
(m, 4H), 2.27 (s, 3H), 2.17 (m, 1H), 1.97 (m, 1H), 1.9-0.8 (m,
16H), 0.98 (s, 3H), 0.93 (d, J ) 6.49 Hz, 3H), 0.871 (d, J )
6.49, 3H), 0.867 (d, J ) 6.49 Hz, 3H). 13C NMR (CDCl3, δ): (u)
213.0, 157.7, 142.3, 128.1, 42.8, 39.4, 38.5, 38.3, 35.8, 31.2, 29.0,
27.7, 25.1, 23.7; (d) 130.8, 114.5, 110.8, 55.22, 55.20, 55.0, 50.4,
35.6, 28.0, 22.8, 22.5, 18.5, 18.3, 11.5.
A 1.0 M solution of L-Selectride in THF (160 µL, 0.16 mmol)
was added dropwise to a solution of the trans-hydrindanone
(32.4 mg, 0.0785 mmol) in THF (2 mL) at -78 °C. After 2 h,
the starting material was 90% consumed (TLC). Additional
L-Selectride (80 µL, 0.080 mmol) was added. After an ad-
ditional 1 h at -78 °C, the mixture was quenched with acetone
(400 µL) and allowed to warm to ambient temperature. The

Synthesis of (-)-Astrogorgiadiol
J . Org. Chem., Vol. 66, No. 3, 2001 953
1
solution was partitioned between CH2Cl2 and saturated aque-
ous NH4Cl. The organic layer was dried (Na2SO4) and evapo-
rated. The residue was chromatographed to yield astrogorgia-
diol methyl ether 36 (17.8 mg, 47% from 32) as a clear oil,
TLC Rf (20% MTBE/petroleum ether) ) 0.47. 1H NMR (CDCl3,
δ): 7.04 (d, J ) 8.19 Hz, 1H), 6.72 (d, J ) 2.73 Hz, 1H), 6.65
(dd, J ) 8.19, 2.73 Hz, 1H), 4.04 (bs, 1H), 3.77 (s, 3H), 2.7 (m,
1H), 2.4 (m, 1H), 2.24 (s, 3H), 1.9-0.9 (m, 22H), 0.92 (d, J )
6.49, 3H), 0.863 (d, J ) 6.49 Hz, 3H), 0.858 (d, J ) 6.49 Hz,
3H). 13C NMR (CDCl3, δ): (u) 157.8, 142.5, 127.9, 42.9, 39.5,
36.1, 34.1, 31.1, 30.3, 30.1, 27.7, 24.4, 23.7; (d) 130.8, 114.5,
110.7, 67.2, 56.2, 55.2, 47.7, 40.9, 35.7, 28.0, 22.8, 22.5, 18.7,
18.4, 11.0. IR (cm-1): 3453 (br), 2932, 2867, 1499, 1251. HRMS
calcd for C28H46O2: 414.3500, found 414.3487.
CHCl3). H NMR (C6D6, δ): 6.93 (d, J ) 8.19 Hz, 1H), 6.60 (d,
J ) 2.73 Hz, 1H), 6.45 (dd, J ) 8.19, 2.73 Hz, 1H), 4.09 (bs,
1H), 3.8 (m, 1H), 2.7 (m, 1H), 2.3 (m, 1H), 2.22 (s, 3H), 1.9-
0.8 (m, 22H), 0.98 (d, J ) 6.49 Hz, 3H), 0.925 (d, J ) 6.49 Hz,
3H), 0.923 (d, J ) 6.49 Hz, 3H), 0.59 (s, 3H). 13C NMR (CDCl3,
δ): (u) 155.2, 143.3, 43.4, 40.3, 37.0, 34.8, 31.6, 31.3, 31.0, 28.5,
25.1, 24.7; (d) 131.7, 116.4, 113.2, 67.2, 56.8, 48.1, 41.6, 36.5,
28.8, 23.4, 23.1, 19.3, 11.6, 0.4. IR (cm-1): 3387 (br), 2932,
2869, 1463. The 1H NMR spectrum of this substance was
superimposable on those provided.1,2
Ack n ow led gm en t. We thank M. P. Doyle and H.
M. L. Davies for providing samples of their enantio-
merically pure rhodium catalysts. We thank the Takasa-
go Corporation for a generous gift of (R)-citronellal and
N. Fusetani and G. Sodano for sharing spectra with us.
This work is dedicated to Professor Gilbert Stork on the
occasion of his 80th birthday.
Astr ogor gia d iol (3). The aryl methyl ether 36 (7.7 mg, 18.6
mmol) was dissolved in CH2Cl2 (1 mL). Triethylsilane (60 µL,
376 mmol) and a 21 mM solution of tris(pentafluorophenyl)
boron (100 µL, 1.95 mmol) in CH2Cl2 were added. After 1 h,
the reaction was quenched with triethylamine (200 µL). The
solution was filtered through silica and evaporated. The
residual oil was diluted with 1 M tetrabutylammonium
fluoride (1 mL) in THF. After 24 h, the solvent was removed
in vacuo and the residual oil was partitioned between water
and ether. The organic layer was evaporated and chromato-
graphed to yield 3 (6.3 mg, 85% yield) as a clear oil, TLC Rf
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra for all new compounds. This material is available free
of charge via the Internet at http://pubs.acs.org.
(30% MTBE/petroleum ether) ) 0.23. [R]18 ) -7.4 (c 0.095,
J O001519G
D