Total Synthesis of Halicholactone and Neohalicholactone

Article abstract of DOI:10.1021/jo962312j

The total synthesis of the marine natural products neohalicholactone (1) and halicholactone (2), in enantiomerically pure form, are reported. Key steps in the synthesis of each compound include a cis-selective Wittig reaction, stereoselective cyclopropanation, nine-membered lactone formation using the Yamaguchi method and late-stage stereoselective Cr(II)/Ni(II) mediated coupling of vinyl iodides 39 and 50 with aldehyde 14. In the case of the neohalicholactone synthesis the two major components which were coupled in this convergent synthesis were each derived from the enantiomers of commercially available malic acid. The synthesis served to confirm the original assignment of absolute configuration which was made by Yamada and Clardy. We also demonstrated, through the preparation of diastereoisomers, that another reported compound closely related to neohalicholactone is likely to be the C-15 epimer 67.

Full text of DOI:10.1021/jo962312j

6638
J . Org. Chem. 1997, 62, 6638-6657
Tota l Syn th esis of Ha lich ola cton e a n d Neoh a lich ola cton e1
Douglas J . Critcher,† Stephen Connolly,‡ and Martin Wills*,§
School of Chemistry, University of Bath, Claverton Down, Bath, BA2 7AY, UK, Department of Chemistry,
University of Warwick, Coventry, CV4 7AL, UK, and Astra Charnwood, Bakewell Road, Loughborough,
Leicestershire, LE11 5RH, UK
Received December 10, 1996X
The total synthesis of the marine natural products neohalicholactone (1) and halicholactone (2), in
enantiomerically pure form, are reported. Key steps in the synthesis of each compound include a
cis-selective Wittig reaction, stereoselective cyclopropanation, nine-membered lactone formation
using the Yamaguchi method and late-stage stereoselective Cr(II)/Ni(II) mediated coupling of vinyl
iodides 39 and 50 with aldehyde 14. In the case of the neohalicholactone synthesis the two major
components which were coupled in this convergent synthesis were each derived from the
enantiomers of commercially available malic acid. The synthesis served to confirm the original
assignment of absolute configuration which was made by Yamada and Clardy. We also
demonstrated, through the preparation of diastereoisomers, that another reported compound closely
related to neohalicholactone is likely to be the C-15 epimer 67.
In tr od u ction
stereochemistry of 1 and 2, encouraged us to undertake
an enantioselective total synthesis.
Examination of the constituents of the marine sponge
Halichondria okadai, collected off the coast of J apan by
the Yamada group, led to the isolation of the fatty acid
metabolites neohalicholactone (1) and halicholactone
(2).2a The structural elucidation of these metabolites was
achieved by evaluation of spectral data, coupled with
chemical evidence. The absolute stereochemistry at car-
bon fifteen of halicholactone (2) was shown to be of R
configuration by chemical degradation of its diacetate.
This gave the R-diacetate of 1,2-heptanediol (3) as one
of the fragments, the absolute stereochemistry of which
was confirmed by comparison with authentic material.
In a more recent publication the relative stereochemistry
of 1 was established by an X-ray crystallographic study.2b
Since it is likely that a similar biosynthetic pathway
yields both 1 and 2 then, in combining the aforemen-
tioned stereochemical knowledge, it seems reasonable to
speculate that the absolute stereochemistry in both
molecules should be (8S,9R,11R,12R,15R) as shown
below.
Halicholactone and neohalicholactone are believed to
be formed from arachidonic acid (4) and eicosapentaenoic
acid (5), respectively, probably via lipoxygenation reac-
tions followed by intramolecular cyclizations. Although
the details of the mechanism have not yet been fully
determined, the structurally related marine natural
products of the constanolactone series 6, the related
lactone 7, and hydridalactone (8) provide some clues.
Corey’s general hypothesis for marine prostanoid bio-
synthesis involves an initial (8R) lipoxygenase oxidation
of arachidonic acid (4) followed by formation of an allene
Compound 2 has been shown to exhibit inhibitory
activity (IC50 ) 630 µm), against 5-lipoxygenase of guinea
pig polymorphonuclear leukocytes. These factors, as well
as the necessity to unambiguously confirm the absolute
† University of Bath.
‡ Astra Charnwood.
§ University of Warwick.
X Abstract published in Advance ACS Abstracts, August 15, 1997.
(1) Preliminary communications: (a) Critcher, D. J .; Connolly, S.;
Wills, M. J . Chem. Soc., Chem. Commun., 1995, 139. (b) Critcher, D.
J .; Connolly, S.; Wills, M. Tetrahedron Lett., 1995, 36, 3763.
(2) (a) Niwa, H.; Wakamatsu, K.; Yamada, K. Tetrahedron Lett.,
1989, 30, 4543. (b) Kigoshi, H.; Niwa, H.; Yamada, K.; Stout, T. J .;
Clardy, J . Tetrahedron Lett., 1991, 32, 2427.
oxide 9 and subsequent ring opening to give a cation 10
S0022-3263(96)02312-2 CCC: $14.00 © 1997 American Chemical Society

Total Synthesis of Halicholactone and Neohalicholactone
J . Org. Chem., Vol. 62, No. 19, 1997 6639
Sch em e 1
Sch em e 2. Retr osyn th etic An a lysis of
Neoh a lich ola cton e a n d Ha lich ola cton e
which cyclizes to give the prostanoid 11.3 Incubation of
arachidonic acid with an acetone powder from the coral
Plexaura homomalla gave, as well as the expected 11, a
small amount of an eicosanoid 12 containing a cyclopro-
pane, that was easily cyclized to the lactone 7 to aid
characterization.4
Recently, Gerwick has isolated a series of constano-
lactones (6), which are closely related to 7, from the red
marine alga Constantinea simplex off the Oregon coast.5a,b
Gerwick also assigned the absolute stereochemistry of
these compounds using a degradative method. During
the course of our studies, White reported an elegant syn-
thesis of 7 and constanolactones A (6a ) and B (6b) using
a route that mimics the proposed biosynthetic pathway
and that serves to confirm the indicated absolute stere-
ochemistry of these compounds.6 While the biosynthesis
of constanolactones fits within the extended Corey model,
Gerwick has proposed an alternative, but related, route
that depends critically on an initial 12-lipoxygenase
oxidation to give the 12S-epoxide and subsequently the
12-alcohol of the same configuration.5b The biosynthesis
of hybridalactone (8), from the marine alga Laurencia
hybrida, also depends critically on an initial S-selective
12-lipoxygenation reaction, in this case of eicosapen-
taenoic acid. The absolute stereochemistry and biosyn-
thetic route to this compound was determined by Corey.7
In the case of halicholactone and neohalicholactone,
therefore, it is likely that a similar pathway is operating
to that leading to constanolactones. A 12-lipoxygenation
pathway2a and a 15-lipoxygenation pathway (Scheme 1)8
have been proposed, the latter of which was suggested
in a paper by Gerwick during the course of our studies
and will be discussed in more detail later.8 The pathway
would, of course, require an initial R-selective lipoxygen-
ation process, which would not be in accord with the
pathway described above for the related materials,
although it would not be without precedent for marine
natural products.9 A total synthesis to confirm the
absolute configuration of these molecules was therefore
undertaken.
Syn th etic Ap p r oa ch
Our retrosynthetic approach is shown in Scheme 2.
Disconnection of the C12-C13 bond will give two frag-
ments 13 and 14. The “left hand” fragment 13 could be
made either from (R)-(+)-1-octyn-3-ol (15) (in the case of
halicholactone) or from alcohol 16 (in the case of neohali-
cholactone) which could be prepared from R-malic acid.
The synthesis of cyclopropanes, and in particular, their
enantioselective synthesis, has been an area of intensive
research in recent years. Disconnection of 14 to the
acyclic unsaturated synthons 17 may be appropriate. The
transformation of 17 to 14 may be achieved using
dimethylsulfoxonium methylide.10 This method neces-
sitates that 17 has either R,â-unsaturated ester or amide
functionality.11 The presence of R-alkoxy functionality
at C(8) would be expected to promote preferential reagent
attack to one face of the electron-poor olefin, thus
(3) (a) Corey, E. J .; Matsuda, S. P. T. Tetrahedron Lett., 1987, 28,
4247. (b) Corey, E. J .; d′Alarcao, M.; Matsuda, S. P. T.; Lansbury J r,
P. T.; Yamada, Y. J . Am. Chem. Soc., 1987, 109, 289.
(4) (a) Brash, A. R. J . Am. Chem. Soc., 1989, 111, 1891. (b)
Baertscki, S. W.; Brash, A. R.; Harris, T. M. J . Am. Chem. Soc., 1989,
111, 5003.
(5) (a) Nagle, D. G.; Gerwick, W. H. Tetrahedron Lett., 1990, 31,
2995. (b) Nagle, D. G.; Gerwick, W. H. J . Org. Chem., 1994, 59, 7227.
(c) For a recent review see Gerwick, W. H. Lipids, 1996, 31, 1215.
(6) (a) White, J . D.; J ensen, M. S. J . Am. Chem. Soc., 1993, 115,
2970. (b) White, J . D.; J ensen, M. S. J . Am. Chem. Soc., 1995, 117,
6224. (c) Related approach: Nagasawa, T.; Onoguchi, Y.; Matsumoto,
T.; Suzuki, K. Synlett, 1995, 1023. (d) .White, J . D.; J ensen, M. S.
Synlett, 1996, 31.
(9) Schneider, W. P.; Hamilton, R. D.; Rhuland, L. E. J . Am. Chem.
Soc., 1972, 94, 2122, and references therein.
(10) Corey, E. J .; Chaykovsky, M. J . Am. Chem. Soc., 1965, 87, 1353.
(11) (a) Kaiser, C.; Trost, B. M.; Beeson, J .; Weinstock, J . J . Org.
Chem., 1965, 30, 3972. (b) Rodriques, K. E. Tetrahedron Lett., 1991,
32, 1275.
(7) (a) Corey, E. J .; De, B.; Ponder, J . W.; Berg, J . M. Tetrahedron
Lett., 1984, 25, 1015. (b) Corey, E. J .; De, B. J . Am. Chem. Soc., 1984,
106, 2735.
(8) Proteau, P. J .; Rossi, J . V.; Gerwick, W. H. J . Nat. Prod., 1994,
57, 1717.

6640 J . Org. Chem., Vol. 62, No. 19, 1997
Critcher et al.
providing all the chemical information required for a
diastereoselective reaction. Synthon 17 would be pre-
pared from the precursor 18 that we anticipated would
in turn be available from S-malic acid.
hydroxy-γ-butyrolactone have been published.17,18 Sig-
nificantly, it is possible to differentiate the two carboxylic
acid groups in malic acid using a combination of borane-
dimethyl sulfide complex in the presence of a catalytic
amount of sodium borohydride.18 Subsequent lactoniza-
tion of the reduction product to (3S)-hydroxy-γ-butyro-
lactone has been accomplished using standard acidic
conditions in 90% yield. It was this three-step high-
yielding approach that we wished to exploit. Our syn-
thesis of the “right hand” portion is shown in Scheme 3.
(S)-Malic acid was dissolved in a 3% anhydrous hy-
drochloric acid/methanol solution to give dimethyl (S)-
malate in yields in excess of 80% after distillation.19 In
our hands, the selective reduction of the diester became
extremely exothermic upon addition of a catalytic amount
of sodium borohydride. We took the precaution of only
performing this reaction on a moderate scale (between 5
and 10 g). If the mixture was cooled to 0 °C prior to the
addition of NaBH4 and subsequently allowed to warm
slowly to room temperature, the reaction was controllable
and selective. It should be noted that while we were able
to isolate methyl (S)-3,4-dihydroxybutanoate in high yield
(>90%), methyl (S)-2,4-dihydroxybutanoate was also
present. Cyclization of methyl (S)-3,4-dihydroxybu-
tanoate to the hydroxylactone was achieved using an
aqueous 18 N H2SO4/THF solution in 75% yield.20
We investigated the (p-methoxyphenyl)methyl (which
will be abbreviated to the more commonly employed
PMB, or p-methoxybenzyl) protection of 3-hydroxy-γ-
butyrolactone using 2,2,2-trichloroacetimidate method-
ology.21-23 PMB-2,2,2-trichloroacetimidate is more reac-
tive than its benzyl analogue and extremely sensitive to
acids. It is best prepared immediately before use, and
the benzylation reaction requires only 0.3 mol % of TfOH;
addition of 10 mol % of TfOH causes instant decomposi-
tion. The use of 10 mol % of a weak acid, 10-camphor-
sulfonic acid (CSA), has also been reported as a suitable
alternative to 0.3 mol % TfOH.22 We found that the
PMB-protection of 3-hydroxy-γ-butyrolactone furnished
23 in yields in excess of 70% using either 0.3 mol % TfOH
or 5 mol % CSA.
Resu lts a n d Discu ssion
As discussed above, we anticipated that a key inter-
mediate in the synthesis of the right hand fragment of 1
and 2 would be a C8-protected derivative of methyl (S)-
8,9-(Z)-dihydroxynon-5-enoate (18). The synthesis of this
portion of a target molecule has been successfully achieved,
in numerous cases, via Wittig olefination using ylide 19
derived from (4-carboxybutyl)triphenylphosphonium
bromide.12-14
One example of the use of this strategy may be found
in the preparation of (8S,9R)-neodidemnilactone (20).13
This was efficiently completed by building on the
14E,12E,10Z side chain from the unprotected C10 hydroxy
group in 21. The acetonide protecting group was then
cleaved, and after basic hydrolysis of the ester group, the
resulting (8S,9R)-dihydroxy acid was preferentially cy-
clized to the ten-membered lactone product using the
method described by Yamaguchi. Significantly, the final
cyclization step had a choice of forming the ten-membered
lactone product or its nine-membered alternative; the
formation of the latter is, however, not mentioned in the
publication.
We anticipated that direct lactonization of a C9-hydroxy
protected derivative of (18) to a nine-membered lactone
would then be followed by a problematical C9-hydroxy
deprotection step. It is reasonable to expect such a
compound to be converted to its more stable ten-mem-
bered lactone isomer during C9-hydroxy deprotection via
a facile translactonization reaction.14 For this reason we
favored a late-stage lactonization strategy. We wished
to prepare (3S)-3-O-(((p-methoxyphenyl)methyl)oxy)-γ-
butyrolactol (22) and condense this lactol with ylide 19.
The synthesis of a derivative of 18 should then be possible
in just seven steps.
The reduction of lactone 23 was achieved using a slight
excess of diisobutylaluminum hydride, in toluene, at low
temperature (<-20 °C). The target lactol 22 was either
isolated in quantitative crude yield or could be chromato-
graphed in which case the isolated yield was slightly
lower (but still >80%). However, 22 was usually em-
ployed in the next Wittig olefination step without further
purification. The Wittig reaction was carried out using
a slight modification to the conditions first described by
Holmes14 using fresh sodium bis(trimethylsilyl)amide in
toluene. Before adding the deep red ylide (cooled to -78
°C) to the lactol at -78 °C, 22 was first deprotonated
using 1 equiv of sodium bis(trimethylsilyl)amide. This
presumably produces the acyclic aldehyde derivative,
which was kept at low temperature (-78 °C) in toluene.
Since the anion was quite insoluble in toluene at low
The antipodal trityl-protected analog of 22 has been
reported and has been employed in a Wittig olefination
during the synthesis of the C11-C20 segment of leuko-
triene B4.15 However, its preparation required nine steps
from L-ascorbic acid using the procedure first described
by Tanaka.16 Shorter routes to the precursor (3R)-
(16) Tanaka, A.; Yamashita, K. Synthesis, 1987, 570.
(17) Seebach, D.; Eberle, M. Synthesis, 1986, 37.
(18) Saito, S.; Hasegawa, T.; Inaba, M.; Nishida, R.; Fujii, T.; Nomizu
S.; Moriwake, T. Chem. Lett., 1984, 1389.
(12) Corey, E. J .; Weinshenker, N. M.; Schaaf, T. K.; Huber, W. J .
Am. Chem. Soc., 1969, 91, 5675.
(13) Niwa, H.; Inagaki, H.; Yamada, K. Tetrahedron Lett., 1991, 32,
5127.
(14) Congreve, M. S.; Holmes, A. B.; Hughes, A. B.; Looney, M G. J .
Am. Chem. Soc., 1993, 115, 5815.
(15) Wang, Y.; Li, J .; Wu, W.; Sun, X.; Wu, Y.; Xaioa, W.; Shi, L.;
Huang, Y. Prog. in Nat. Sci., 1991, 69.
(19) Mori, K.; Takigawa, T.; Matsuo, T. Tetrahedron, 1979, 35, 933.
(20) Luk, K-C.; Wei, C.-C. Synthesis, 1988, 226.
(21) (a) Iversen, T.; Bundle, D. R. J . Chem. Soc., Chem. Commun.,
1981, 1240. (b) Wessel, H.-P.; Iversen, T.; Bundle, D. R. J . Chem. Soc.,
Perkin Trans. 1, 1985, 2247.
(22) Nakajima, N.; Horita, K.; Abe, R.; Yonemitsu, O. Tetrahedron
Lett., 1988, 29, 4139.
(23) Takaku, H.; Ueda, S.; Ito, T. Tetrahedron Lett., 1983, 24, 5363.

Total Synthesis of Halicholactone and Neohalicholactone
Sch em e 3
J . Org. Chem., Vol. 62, No. 19, 1997 6641
temperatures, the volume of toluene was increased
relative to that used in the Holmes procedure. The
addition of the preformed ylide to the sodium alkoxide
of 22 was found to be more satisfactory than the inverse
of this addition. We found that 22 was prone to decom-
position if sufficiently low temperatures were not main-
tained. The crude (Z)-olefinic carboxylic acid 24 was then
converted to its methyl ester 18 (R ) PMB) using a HCl/
methanol solution at pH 4-5.
Using this approach, methyl (8S,5Z)-8-O-(PMB)-8,9-
dihydroxynon-5-enoate (18 (R ) PMB)) was consistently
isolated on a gram scale in yields greater than 60% for
the three-step transformation from lactone 23. Swern
oxidation24 gave the corresponding aldehyde 25 which
was subsequently condensed with tert-butyl dieth-
ylphosphonoacetate using conditions described by Masam-
une and Roush,25 to furnish (E)-γ-alkoxy-R,â-unsaturated-
tert-butyl ester (26) in 75% yield.
Ester 26 was treated with 2 equiv of DMSY (prepared
from trimethylsulfoxonium iodide and NaH) in dimethyl
sulfoxide, and the solution was heated at 90 °C for 20
h.26-29 An inseparable 5:2 mixture of two diastereomeric
cyclopropanes 27 were isolated in good yield (74%). No
products corresponding to addition of DMSY to the
exposed methyl ester were observed. We found that more
elevated temperatures were required in our reaction than
those described by Magnus.29
Deprotection of the (p-methoxybenzyl)methyl group in
27 using DDQ30 gave the two (8S)-hydroxycyclopropane
diastereoisomers 28 and 29 in quantitative yield. For-
tunately, these two compounds were separable by chro-
matography, and the major isomer 28 was isolated in
51% yield for the two-step transformation from tert-butyl
ester 26. From the coupling constants of the terminal
cyclopropane protons (H10a and H10b) in the 1H-NMR
spectrum of the major isomer 28, it was clear that a
trans-cyclopropane ring was in place. Excellent agree-
ment with coupling constants of the cyclopropane protons
H
7a and H7b of trans-cyclopropane, and lactone-containing
compound 7 were observed.6a,b
The cyclopropane protons gave unresolvable splitting
patterns in the minor diastereomer 29. However, sub-
sequent manipulations of 29 gave derivatives whose
coupling constants for H10a and H10b were in very good
agreement with those described for 28. An X-ray crystal
structure of the derived acid1a confirmed the trans-
substitution pattern and had the correct relative stereo-
chemistry (8S,9R,11R) required to complete the synthesis
of the right-hand fragment of neohalicholactone (1) and
halicholactone (2). The most likely explanation for the
exclusive formation of the trans-isomers 27 is direct
kinetic control in the cyclopropanation reaction.31,32
Assuming that the addition of DMSY to 26 is irrevers-
ible, we offered the following stereochemical arguments.
The reaction of γ-alkoxy-R,â-unsaturated esters with a
(24) Omura, K.; Swern, D. Tetrahedron, 1978, 34, 1651.
(25) Blanchette, M. A.; Choy, W.; Davies, J . T.; Essenfeld, A. P.;
Masamune, S.; Roush W. R.; Sakai, T. Tetrahedron Lett., 1984, 25,
2183.
(29) Magnus, P.; Schultz, J .; Gallagher, T. J . Am. Chem. Soc., 1985,
107, 4984.
(26) Corey, E. J .; Chaykovsky, M. J . Am. Chem. Soc., 1962, 84, 867.
(27) Corey, E. J .; Chaykovsky, M. J . Am. Chem. Soc., 1964, 86, 1640.
(28) Gololobov, Y. G.; Nesemeyanov, A. N.; Lysenko, V. P.; Boldeskul,
I. E. Tetrahedron, 1987, 43, 2609.
(30) Oikawa, Y.; Yoshioka, T.; Yonemitsu, O. Tetrahedron Lett.,
1982, 23, 885.
(31) Trost, B. M.; Mignani, S. M. Tetrahedron Lett., 1986, 27, 4137.
(32) Dechoux, L.; Doris, E. Tetrahedron Lett., 1994, 35, 2017.

6642 J . Org. Chem., Vol. 62, No. 19, 1997
Critcher et al.
F igu r e 2.
molecular orbital theory has been used to understand
qualitatively these preferences. When the s*C-A orbital
is aligned anti to the forming bond, its overlap with the
HOMO of the transition state (which consists of the
nucleophile HOMO and the alkene LUMO) is maximized,
and the transition state is stabilized. An electropositive
group prefers the outside position since the interaction
of an occupied sC-D orbital with the transition state
HOMO is destabilising. Mod el C summarizes this
information (Figure 2).
F igu r e 1.
variety of nucleophiles has been studied.31,33-35 For either
(E)- or (Z)-γ-alkoxy unsaturated esters, the reactive
conformation shown in m od el A (Figure 1) explains the
stereochemical outcome of reactions with reagents such
as 1-acetoxy-2-((trimethylsilyl)methyl)-2-propene, cyclo-
pentadiene, isopropylidenediphenylsulfurane, and amines.
The approach of the reagent takes place from the least
hindered side of the starting unsaturated ester adopting
the conformation shown in m od el A, whatever the
stereochemistry of its CdC bond.31 In this model, the
large alkyl substituent R’ is perpendicular to the conju-
gated double bond, and the hydrogen atom eclipses the
conjugated double bond. However, there is also a set of
reagents which will react with (Z)-γ-alkoxy unsaturated
esters to give products whose â-stereochemistry can also
be rationalized using m od el A, while the same nucleo-
philes will give the opposite â-stereochemistry if the (E)-
double bond isomer is used. The conformation shown in
m od el B is thought to result from a favorable interaction
between the p-orbitals on the double bond and an
unshared pair of electrons of the γ-oxygen. Such an
interaction is only active if an (E)-double bond is present.
Reagents that are consistent with these experimental
findings are osmium tetraoxide, organocopper-boron
trifluoride complexes, and isopropylidenetriphenyphos-
phorane. Therefore, the 5:2 mixture of cyclopropane
isomers was the result of (8S)-(E)-γ-alkoxy-R,â-unsatur-
ated-tert-butyl ester 26 preferentially reacting via the
conformation shown in m od el B, since only this model
predicts the generation of a 9R,11R cyclopropane product.
Houk36 has studied stereoselective additions to double
bonds using ab initio quantum mechanics to predict
transition state structures (see Figure 2). By researching
both nucleophilic and electrophilic additions to chiral
allylic ethers the following generalizations have been
made. For nucleophilic attack on π bonds, an electrone-
gative allylic ether group (A ) electron acceptor) would
prefer to adopt an anti position so that the withdrawal
of electrons from the π-system can be maximized. The
most electropositive allylic substituent (B ) electron
donor) prefers the outside position in order to minimize
the donation of electrons to the already electron rich
π-system of the transition state, see m od el C. Frontier
Application of the nucleophile addition m od el C to our
DMSY addition to R,â-unsaturated ester 26 predicts the
stereochemical preference that we observed experi-
mentally: re-face attack. However, in m od el C the large
sterically demanding alkyl chain is positioned in the
outside region where it presents a steric obstacle to the
approaching nucleophile. It is possible that this factor
accounts for the modest stereoselectivity we observed.
The alkyl chain would be less obstructive in the inside
position. Nucleophilic attack would occur, once again,
anti to the electron-withdrawing group as shown in
m od el D, and si-face attack is now predicted.
The cyclization of hydroxy acids to furnish saturated
nine-membered lactones is known to be problematical.
The three standard macrolactonization method, the
2-thiopyridyl ester method of Corey,37 the utilization of
1-methylchloropyridinium iodidesthe method of Mu-
kaiyama,38 and the formation of a carboxylic 2,4,6-
trichlorobenzoic anhydride, as described by Yamaguchi,39
afford at best nine-membered lactones in 25%, 13% and
18% yields, respectively. However, in 1982, Still re-
ported40 that the presence of a (Z)-olefinic linkage pro-
vides “substantial enthalpic, as well as entropic benefit
to the desired cyclization”. The enthalpic benefit is due
to the elimination of important transannular repulsions
in the product.
We were able to synthesize the required nine-mem-
bered lactones in good yield by employing the lactoniza-
tion conditions first described by Yamaguchi.39 Initially,
enantiomerically pure hydroxy trans-cyclopropane com-
pounds 28 and 29 were smoothly converted to their nine-
membered lactones by basic hydrolysis to give interme-
diate ω-hydroxy acids 30 and 31 that were then lactonized
to provide 32 and 33 in 75% and 76% yield, respectively.
The only observed side-product was a dimeric species
which formed readily if high dilution and slow addition
conditions were not employed. However, this unwanted
eighteen-membered lactone was usually formed in less
(33) Kreif, A.; Dumont, W.; Pasau, P.; Lecomte, P. Tetrahedron,
1989, 45, 3039.
(34) Yamatoto, Y.; Nishii, S.; Ibuka, T. J . Chem. Soc., Chem.
Commun., 1987, 464.
(35) Stork, G.; Kahn, M. Tetrahedron Lett., 1983, 24, 3951.
(36) (a) Houk, K. N.; Moses, S. R.; Wu, W.-D.; Rondan, N. G.; J ager,
V.; Schole, R.; Froncoek, F. R. J . Am. Chem. Soc., 1984, 106, 3880. (b)
Houk, K. N.; Paddon-Row, M. N.; Rondan, N. G.; Wu, Y. D.; Brown, F.
K.; Spellmeyer, D. C.; Metz, J . T.; Li, Y.; Loncharich, R. J . Science,
1986, 231, 1108.
(37) (a) Corey, E. J .; Nicolaou, K. C. J . Am. Chem. Soc., 1974, 96,
5614. (b) Corey, E. J .; Brundle, D. J . Tetrahedron Lett., 1976, 3409.
(38) Mukaiyama, T.; Usui, M.; Saigo, K. Chem. Lett., 1976, 49.
(39) Inanaga, J .; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M.
Bull. Chem. Soc. J pn., 1979, 52, 1989.
(40) Still, W. C.; Galynker, I. J . Am. Chem. Soc., 1982, 104, 1774.

Total Synthesis of Halicholactone and Neohalicholactone
Sch em e 4
J . Org. Chem., Vol. 62, No. 19, 1997 6643
than 5% yield and could be separated from the nine-
membered lactones by chromatography, hydrolyzed to the
hydroxy acid (using lithium hydroxide) in excellent yield
(84%) and relactonized.
Conversion of 32 and 33 to their carboxylic acid
derivatives 34 and 35 was achieved using trifluoroacetic
acid in dichloromethane. The X-ray structure of 341a
revealed that this isomer contains the correct 8S,9R,11R
relative stereochemistry for the completion of the syn-
thesis of both neohalicholactone (1) and halicholactone
(2). It is interesting that 34 adopted a very similar
conformation to the corresponding region of neohali-
cholactone itself.1a
3-ol (40) in 94% and 77% yields, respectively, from 37
and 38. The hydrozirconation reaction was far more
problematical with the PMB-protected substrate 38. The
reaction needed to be performed at low temperature (0
°C) and required 2.2 equiv of bis(cylopentadienyl)zirco-
nium chloride hydride to achieve full conversion.
With 8S,9R,11R-carboxylic acid 34 in hand, its reduc-
tion to an aldehyde involved the initial formation of a
carbonic-carboxylic anhydride that was then reduced
using sodium borohydride (NaBH4).41 Carboxylic acid 34
dissolved in tetrahydrofuran was treated with ethyl
chloroformate in the presence of triethylamine to yield
the corresponding mixed anhydride. The deposited tri-
ethylamine hydrochloride was filtered and the purified
mixed anhydride reduced with sodium borohydride to
afford 36 in 77% yield. Finally 36 was oxidized to
aldehyde 14 using TPAP.42 Therefore, the synthesis of
the right hand fragment was completed in 16 steps from
(S)-malic acid in a satisfactory 7.5% overall yield. Using
the reaction sequence described, we were able to prepare
several hundreds of milligrams of enantiomerically pure
8S,9R,11R-aldehyde 14.
We synthesized suitably protected 1-iodo-1-octene-3-
ol derivatives 37 and 38 from (R)-1-octyn-3-ol (15) in two
steps. Using the tert-butyldiphenylsilyl chloride/imida-
zole/N,N-dimethylformamide protocol, 37 was isolated in
89% yield. The (p-methoxyphenyl)methyl derivative 38
was also synthesized with using p-methoxybenzyl chloride/
sodium hydride/catalytic tetra-n-butylammonium iodide/
DMF in 56% yield. The hydrozirconation-iodination
procedure first reported by Schwartz43 gave (3R,1E)-3-
((tert-butyldiphenylsilyl)oxy)-1-iodo-1-octen-3-ol (39) and
(3R,1E)-3-(((p-methoxyphenyl)methyl)oxy)-1-iodo-1-octen-
For the preparation of the corresponding fragment of
neohalicholactone (Scheme 4), we found that the best
results were obtained when we performed the Wittig
reaction of 19 with (R)-22 without the prior formation of
the sodium alkoxide. In this case, 2.1 equiv of the ylide
were prepared as before, and this, at -78 °C, was added
to lactol (R)-22, also at -78 °C, via cannula. The
resulting mixture was maintained at low temperature
for 5 min, and warmed to 0 °C, followed by stirring at
this temperature for 30 min. The reaction was quenched
using a saturated aqueous ammonium chloride solution.
After chromatography, the desired (Z)-olefin 41 was now
isolated in 76% yield. The only detected byproduct was
the nonpolar trimethylsilyl primary hydroxyl protected
derivative of 41, in 11% yield. The partial trimethylsilyl
protection of the desired compound was presumably a
result of the protonation of hexamethyldisilazide during
workup. As for the “right-hand” fragment, we avoided
the use of silyl protecting groups in this transformation
due to their known tendancy to migrate to less hindered
positions.
The trimethylsilyl protecting group is very labile to
hydrolysis,44 and we were, therefore, able to cleave it in
(43) (a) Hart, D. W.; Blackburn, T. F.; Schwartz, J . J . Am. Chem.
Soc., 1975, 97, 679. (b) Schwartz, J .; Labinger, J . A. Angew. Chem.,
Int. Ed. Engl., 1976, 15, 333. (c) Corey, E. J .; Trybulski, E. J .; Melvin,
L. S.; Nicolaou, K. C.; Secrist, J . A.; Lett, R.; Sheldrake, P. W.; Falck,
J . R.; Brunelle, D. J .; Haslanger, M. F.; Kim, S.; Yoo, S. J . Am. Chem.
Soc., 1978, 100, 4618. (d) Larock, R. C.; Kondo, F.; Narayanan, K.;
Sydnes, L. K.; Hsu, M.-F. H. Tetrahedron Lett., 1989, 30, 5737.
(41) Ishizumi, K.; Koga, K.; Yamada, S.-I. Chem. Pharm. Bull., 1968,
16, 492.
(42) (a) Griffith, W. P.; Ley, S. V.; Whitcome, G. P.; White, A. D. J .
Chem. Soc., Chem. Commun., 1987, 1625. (b) Griffith, W. P.; Ley, S.
V. Aldrichemica Acta, 1990, 23, 13. (c) Ley, S. V.; Norman, J .; Griffith,
W. P.; Marsden, S. P. Synthesis, 1994, 639.

6644 J . Org. Chem., Vol. 62, No. 19, 1997
Critcher et al.
quantitative yield under mildly acidic conditions (dilute
hydrochloric acid/tetrahydrofuran solution) which left the
p-methoxyphenyl)methyl group intact. The total yield
for the olefination reaction was at least 87%. The
oxidation of the key alcohol 41 to give (2R,4Z)-2-O-(((p-
methoxyphenyl)methyl)oxy)-4-heptenal (42) was achieved
quantitatively using the Swern oxidation conditions.24
Aldehyde 42 could also be isolated in yields in excess of
96% after chromatography. A closely related approach
to a homologue of 42 has been reported, the subsequent
elaboration of which provided an excellent precedent for
our proposed strategy.45
exchange to give an R-halo-metallic compound. Dibro-
moalkenes have been reacted with alkyllithiums and
then quenched at low temperature to produce mono
bromoalkenes.53 By allowing the R-halo-metallic com-
pound to warm, then R-elimination of lithium bromide
takes place and an alkylidene carbene 47 is produced.54
Such species are known to undergo very rapid 1,2-
With aldehyde 42 in hand, we examined the Takai
olefination reaction,46 in the formation of (3R)-(1E,5Z)-
3-O-(((p-methoxyphenyl)methyl)oxy)-1-iodo-1,5-octadi-
ene (42 (R ) PMB)) directly in one step. However, the
reaction of 42 with 6.0 equiv of chromium chloride and
2.0 equiv of iodoform in tetrahydrofuran for 16 h at room
temperature resulted in only a 11% yield of pure, but
unstable, trans-iodo alkene 43. White has reported the
Takai olefination of both (Z)-4-decenal and (2S,4Z)-2-O-
((tert-butyldiphenylsilyl)oxy)-4-decenal to give the trans-
iodo alkenes in 95% yield in each case (after just 3 h at
0 °C) and a more modest 54% yield (after 18 h at 0 °C),
respectively.6a,b Therefore, it appears that oxygen func-
tionality in the substrate severely retards the desired
reaction which in our case appeared to permit alternative
reaction pathways to become competitive.
We then considered an alternative approach to a
suitably protected trans-iodo alkene which would involve
the intermediacy of alkyne 44. The synthesis of alkynes
directly from aldehydes has been described, where the
anion of dimethyl (diazomethyl)phosphonate (Gilbert’s
reagent) is generated using potassium tert-butoxide at
low temperature (-78 °C) in tetrahydrofuran.47,48
hydrogen shift to produce alkynes.55,56a However, alky-
lidene carbenes are also known to undergo intramolec-
ular 1,5-insertion57 as well as intermolecular insertion
reactions.58 It is the former intramolecular process that
we believe results in the formation of the 3,4-dihydrofu-
ran derivative 46. Baird and colleagues have generated
alkylidene carbenes from 1,1-dibromo-2-methyl alkenes.54
Their results clearly provide evidence for the well known
1,2-migration step to give alkynes, as well as the in-
tramolecular insertion of a carbenoid into a C(5)-H bond
to yield 3,4-dihydrofuran ring systems.
Conventional PMB deprotection of 44 was smoothly
achieved using DDQ,30 to provide a crude sample of
propargylic alcohol intermediate which was then silylated
under standard conditions to give the tert-butyldiphen-
ylsilyl derivative 48 in high yield (93%).
However, in view of the fact that the 71% yield for the
transformation of PMB-protected dibromide 45 was not
fully reproducible, we felt that the earlier introduction
of the tert-butyldiphenylsilyl protecting group (before the
alkyne was produced) would improve the overall ef-
ficiency of the synthesis of the desired left-hand fragment.
Therefore dibromide 45 was now deprotected (DDQ) and
then reprotected using tert-butyldiphenylsilyl chloride to
give 49 in 92% overall yield (Scheme 4).
We now anticipated that the reaction of tert-butyl-
diphenylsilyl-protected dibromide 49 with 2 equiv of
n-butyllithium would afford the desired alkyne 48 cleanly
in high yield. However, 48 was isolated in a modest yield
(67%). A substantial amount of nonpolar materials was
However, we decided to synthesize the desired alkyne
44 via a two-step process involving an intermediate
dibromide 45. Using a modification of the Corey-Fuchs
procedure49 (2.0 equiv of carbon tetrabromide, 4.0 equiv
of triphenylphosphine, and 8.0 equiv of triethylamine in
dichloromethane at -78 °C), (3R,5Z)-3-O-((p-methoxy-
phenylmethyl)oxy)-1,1-dibromo-1,5-octadiene (62 was syn-
thesized in good yield (73%).50-52 The reaction of dibro-
mide 45 with 2 equiv of n-butyllithium in tetrahydrofu-
ran at low temperature gave a mixture of compounds
from which alkyne 44 was isolated in 71% yield. The
major impurity formed in this reaction (10% yield) was
assigned as the 2,5-disubstituted 3,4-dihydrofuran de-
rivative 46 based on the 1H-NMR and mass spectral data.
Intriguingly, 46 appeared to be diastereomerically pure
(53) (a) Grandjean, D.; Pale, P. Tetrahedron Lett., 1993, 34, 1155.
(b) Harada, T.; Katsuhira, T.; Oku, A. J . Org. Chem. 1992, 57, 5805.
(54) (a) Baird, M. S.; Baxter, A. G. W.; Hoorfar, A.; J efferies, I. J .
Chem. Soc. Perkin Trans. 1, 1991, 2575. (b) Hartzler, H. D. J . Am.
Chem. Soc., 1964, 86, 526.
(55) A methylidene carbene to acetylene rearrangement barrier of
only 0.9 kcal/mol has been calculated; see (a) Stang, P. J . Acc. Chem.
Res., 1982, 15, 348. (b) Stang, P. J . Chem. Rev., 1987, 78, 383. (c)
Likhotvorik, I. R.; Brown, D. W.; J ones, J r., M. J . Am. Chem. Soc.,
1994, 116, 6175. (d) Ochiai, M.; Ito, T.; Takaoka, Y.; Masaki, Y.;
Kunishima, M.; Tani, S.; Nagao, Y. J . Chem, Soc., Chem. Commun.,
1990, 118.
(56) (a) Gilbert, J . C.; Giamalva, D. H.; Weerasooriya, U. J . Org.
Chem., 1983, 48, 5251. (b) Gilbert, J . C.; Giamalva, D. H.; Baze, M.
E. J . Org. Chem., 1985, 50, 2557. (c) Gilbert, J . C.; Blackburn, B. K.
J . Org. Chem., 1986, 51, 3656. (d) Kitamura, T.;Stang, P. J . Tetra-
hedron Lett. 1988, 29, 1887. (e) Kim, S.; Cho, C. M. Tetrahedron Lett.,
1994, 35, 8405. (f) Schildknegt, K.; Bohnstedt, A. C.; Feldman, K. S.;
Sambandam, A. J . Am. Chem. Soc., 1995, 117, 7544. (g) Taber, D. F.;
Meagley, R. P. Tetrahedron Lett., 1994, 35, 7909. (h) Taber, D. J .;
Walter, R.; Meagley, R. P. J . Org. Chem., 1994, 59, 6014.
(57) For intramolecular trapping of an alkylidenecarbene-iodonium
ylide with triethylsilane, see; (a) Kitamura, T.; Stang, P. J . Tetrahedron
Lett., 1988, 29, 1887. For the formation of enol ethers and enamines
via the intramolecular reaction of a free alkylidene carbene with an
alcohol or amine, see (b) Gilbert, J . C.; Weerasooriya, U. J . Org. Chem.,
1983, 48, 448.
1
since we did not observe any doubling of peaks in its H-
NMR spectrum. There is substantial evidence, that in
common with our own experimental finding, to suggest
that the mechanism involves initial metal-halogen
(44) Kocienski, P. J . Protecting Groups; Thieme: New York, 1994.
(45) Leblanc, Y.; Fitzsimmons, B. J .; Adams, J .; Perez, F.; Rokach,
J . J . Org. Chem., 1986, 51, 789.
(46) Takai, K.; Nitta, K.; Utimoto, K. J . Am. Chem. Soc., 1986, 108,
7408.
(47) (a) Gilbert, J . C.; Weerasooriya, U. J . Org. Chem., 1979, 44,
4997. (b) Gilbert, J . C.; Weerasooriya, U. J . Org. Chem., 1982, 47,
1837. (c) Delpech, B.; Lett, R. Tetrahedron Lett., 1989, 30, 1521.
(48) Ragan, J . A.; Nakatsuka, M.; Smith, D. B.; Uehling, D. E.;
Schrieber, S. L. J . Org. Chem., 1989, 54, 4267.
(49) Corey, E. J .; Fuchs, P. L. Tetrahedron Lett., 1972, 3769.
(50) Marshall, J . A.; Andersen, M. W. J . Org. Chem., 1993, 58, 3912.
(51) McIntosh, M. C.; Weinreb, S. M. J . Org. Chem., 1993, 58, 4823.
(52) Grandjean, D.; Pale, P.; Chuche, J . Tetrahedron Lett., 1994, 35,
3529.
(58) (a) Taber, D. F.; Ruckle, R. E., J r. Tetrahedron Lett., 1985, 26,
3059. (b) Wee, A. G. H.; Liu, B; Zhang, L. J . Org. Chem, 1992, 57,
4404. (c) Taber,. D. F.; Ruckle, J r., R. E. J . Am. Chem. Soc., 1986,
108, 7686.

Total Synthesis of Halicholactone and Neohalicholactone
J . Org. Chem., Vol. 62, No. 19, 1997 6645
also isolated after chromatography. Therefore, we ex-
amined the possibility of affecting a complete change in
the mechanism of the dibromide to acetylene transforma-
tion. By adding at least 1 equiv of lithium diisopropyl-
amide to the dibromide 49, literature precedent sug-
gested that the corresponding bromo alkyne should form
via an anti-dehydrobromination reaction.59 In situ metal-
halogen exchange using an alkyllithium would then be
expected to produce a terminal acetylene.
however, we were not unable to isolate any of the desired
allylic alcohol. Oxygen functionality in substrates as well
as oxygen-containing solvents can interfere with the
desired asymmetric alkenyl transfer reaction. Bulky silyl
ethers such as tert-butyldiphenylsilyl or triisopropylsilyl
ethers have been shown to diminish the chelating ability
of the oxygen functionality in the presence of certain
organometallic reagents.63
In 1994, Wipf and Xu reported that allylic alcohols
could be prepared via a hydrozirconation reaction of an
alkyne followed by transmetalation with dimethyl- or
diethylzinc and subsequent addition of the (1-alkenyl-
)ethylzinc species to an aldehyde.62c They were able to
show that a modest enantiomeric excess could be achieved
in the coupling of hex-1-yne with benzaldehyde in the
presence of 8 mol % of the proline-derived ligand (S)-
DPMPM. The reaction was tolerant to tert-butyldiphen-
ylsilyl ethers and ester functionality. We were pleased
to find that application of this methodology, in the ab-
sence of directing ligand (S)-DPMPM, enabled us to react
the oxygen-containing model substrate 51 with hexanal
to produce racemic allylic alcohol 52 in 51% yield.
However, the reaction of 48 with 14 afforded none of
the desired allylic alcohol product under the reaction
conditions used to synthesize 52. The crude 1H-NMR
spectrum indicated that the hydrozirconation reaction
had proceeded smoothly since almost all of the alkyne
48 had been consumed, but it had simply been converted
to its alkene derivative 53. It appeared that aldehyde
14 reacted with excess diethylzinc to give the alcohol
product 54.
The addition of 1.5 equiv of lithium diisopropylamide
to 49 at -78 °C in tetrahydrofuran followed by 2.2 equiv
of n-butyllithium (1 equiv of n-butyllithium is required
to neutralize the proton of diisopropylamine) afforded the
desired alkyne 48 in 92% yield.52
Alkyne 48 was converted to (3R)-(1E,5Z)-3-O-((tert-
butyldiphenylsilyl)oxy)-1-iodo-1,5-octadiene (50) using
Schwartz’s reagent (Cp2ZrHCl) in tetrahydrofuran, and
the resultant vinyl metal species was then treated with
iodine.43a After chromatography, 50 was isolated in high
yield (92%), but this material was contaminated with
<8% of unreacted starting material. trans-Iodo alkene
50 was used directly without further purification due to
the sensitivity of iodo alkenes to sunlight and acids.
In summary, we were able to prepare tert-butyldiphen-
ylsilyl-protected trans-iodo alkene 50 in 12 steps (24%
yield) starting from (R)-malic acid.
The chromium(II) chloride/nickel(II) chloride method-
ology developed by Kishi60a and Takai60b has excellent
precedent for its efficient application to the coupling
reactions of trans-iodo alkenes with aldehydes. However,
we believed that if high diastereoselectivity was to be
observed in the coupling reaction, then the use of asym-
metric catalysis would be essential.61 The presence of
the adjacent trans 1,2-disubstituted cyclopropane moiety
in aldehyde 14 would prevent any inherent control in the
coupling reaction, since the bulky lactone-containing
substituent would be oriented away from the aldehyde
functionality.
Oppolzer has reported that (S)-diphenyl(1-methylpyr-
rolidin-2-yl)methanol ((S)-DPMPM) directed the (1-al-
kenyl)zinc/aldehyde addition with excellent yields and
enantiomeric excesses, and we were able to show that
we could successfully repeat this work. By closely
following the reported reaction conditions, we were able
to synthesize the allylic alcohol from the coupling of oct-
1-yne and heptaldehyde using 1 mol % of (S)-DPMPM
in 79% yield.62 We then introduced oxygen functionality
into our model substrate by preparing tert-butyldiphen-
ylsilyl-protected propargyl alcohol 51. Unfortunately,
(59) (a) Kang, S.-K.; Lee, D.-H.; Lee, J .-M. Synlett, 1990, 591. (b)
Takano, S.; Sugihara, T.; Ogasawara, K. Tetrahedron Lett., 1991, 32,
2797. (c) Negishi, E.-I.; Zhang Y.; Bagheri, V. Tetrahedron Lett., 1987,
28, 5793.
Organochromium-mediated transformations were now
investigated.60 It has been established that the stereo-
chemistry of a disubstituted trans-iodo olefin is retained
in the Cr(II)/Ni(II)-coupled product and that dimethyl
sulfoxide solvent is critical; a substantial amount of an
R,â-unsaturated aldehyde was isolated if DMF or mix-
tures of DMF and DMSO were used. However it was
noted that reactions were generally faster in DMF or
DMF/DMSO mixtures whereas the use of DMSO alone
resulted in fewer byproducts.
(60) (a) J in, H.; Uenishi, J .; Christ, W. S.; Kishi, Y. J . Am. Chem.
Soc., 1986, 108, 5644. (b) Takai, K.; Tagashira, M.; Kuroda, T.;
Utimoto, K.; Nozaki, H. J . Am. Chem. Soc., 1986, 108, 6048.
(61) (a) Cintas, P. Synthesis, 1992, 248. (b) Hodgson, D. M. J .
Organomet. Chem., 1994, 476, 1. (c) Kishi, Y. Pure Appl. Chem., 1992,
64, 343. During the course of our work Kishi published in detail on
his attempts to design an efficient chiral ligand for the Ni(II)/Cr(II)
coupling reaction. To date the best ligand is based on a 2,2′-dipyridyl
ring system, which has only produced modest levels of diastereoselec-
tion in chosen coupling reactions. (d) Chen, C.; Tagami, K.; Kishi, Y.
J . Org. Chem., 1995, 60, 5386.
(62) (a) Oppolzer, W.; Radinov, R. N. Tetrahedron Lett., 1991, 32,
5777. (b) Oppolzer, W.; Radinov, R. N. Helv. Chem. Acta, 1992, 75,
170. (c) Wipf, P.; Xu, W. Tetrahedron Lett., 1994, 35, 5197. (d) Soai,
K.; Takahashi, K. J . Chem. Soc., Perkin Trans. 1, 1994, 1257. (e)
Oppolzer, W.; Radinov, R. N. J . Am. Chem. Soc., 1993, 115, 1593. (f)
Oppolzer, W.; Radinov, R. N.; Brabander, J . N. Tetrahedron Lett., 1995,
36, 2607.
The synthesis of the related natural product 7 was
reported in 1993 by White.6a One of the key steps in the
synthesis is the Takai-Kishi coupling of trans-iodo
(63) Chen, X.; Hortelano, E. R.; Eliel, E. L.; Frye, S. V. J . Am. Chem.
Soc., 1992, 114, 1778.

6646 J . Org. Chem., Vol. 62, No. 19, 1997
Critcher et al.
Sch em e 5
Therefore, we chose to recombine the two mixtures of
compounds and subject the mixture to pyridinium chlo-
rochromate oxidation. In doing so we hoped to be able
to recover the pure aldehyde 14, while the alcohols would
be converted to a single compound, enone 61. In the
event, we were able to recover 87% of 14 and isolated
enone 61 in 54% yield from the oxidation reaction after
chromatography.
The reaction of enone 61 with 3.0 equiv of tetrabutyl-
ammonium fluoride in tetrahydrofuran at room temper-
ature gave the desired product 62 in 81% yield after just
1 h. In view of this, the desilylation of 60a was slow due
to a retardation effect caused by the presence of the
proximal C12 hydroxyl group. We reduced 62 by employ-
ing Luche conditions64 and isolated halicholactone (2) and
its C12 epimer 63 as a 1:1 mixture of diastereomers in a
modest 40% yield. The reaction confirmed our suspicions
that carbonyl reduction adjacent to the trans-cyclopro-
pane ring would not be a diastereoselective process, and
this was in keeping with other literature reports on
alkene 55 with aldehyde 56. A 1:1 mixture of allylic
alcohols 57 resulted from the coupling reaction. This
result confirmed our suspicions that the trans cyclopro-
pane ring prevents any inherent stereocontrol in the
coupling reaction.
We therefore expected that any diastereocontrol in the
coupling of trans-iodo alkenes 39 and 50 with aldehyde
14 would have to come from the enantiomerically pure
vinylic anion species. This is obviously disfavored due
to the remote location of the chiral centre in the trans-
iodo olefin. In order to determine to what extent the (3R)-
ether center of the trans-vinylic chromium species could
control the addition, we reacted trans-iodo alkenes 40 and
39 with a prochiral aldehyde, 2-methylpropanal (Scheme
5). We isolated the (p-methoxyphenyl)methyl-protected
allylic alcohol 58 as a 1:1 mixture of diastereomers,
whereas a 2.1:1 mixture of allylic alcohol diastereomers
59 was isolated in the latter case (major product relative
configuration not determined). The differently protected
allylic alcohols 58 and 59 were both isolated in quantita-
tive yield following the experimental conditions reported
by Kishi.60a
1
related systems.65 Importantly, the 400 MHz H-NMR
spectrum of the 1:1 mixture of diastereomers (2 and 63)
provided conclusive evidence that the isomerically pure
compound we had earlier isolated by direct desilylation
of 60a was halicholactone (2). The signals which be-
1
longed to 63 in the 400 MHz H-NMR of the 1:1 mixture
could be determined by subtraction of the signals that
we knew belonged to 2.
We have therefore completed a convergent 20-step
synthesis of halicholactone (2) in 3.1% overall yield
starting from commercially available materials.
The synthesis of neohalicholactone (1) (Scheme 6) now
became relatively straightforward since we could draw
on the knowledge we had earlier gained in the synthesis
of 2. The coupling of 1 equiv of aldehyde 14 with 2 equiv
of trans-iodo alkene 50 mediated by 6 equiv of chro-
mium(II) chloride (containing 0.5 wt % NiCl2) gave as
before a 2:1 mixture of diastereomers 64a and 64b which
were isolated in 61% yield. Due to solubility problems,
this reaction was carried out in a 1:1 mixture of DMSO
and DMF, and the full consumption of aldehyde 14 was
observed after just 3 h at room temperature.
We applied this methodology to the synthesis of hali-
cholactone (2) (Scheme 6). The reaction of 1.0 equiv of
aldehyde 14 with 1.7 equiv of trans-iodo alkene 39 in the
presence of 3.3 equiv of chromium(II) chloride (containing
0.5 wt % NiCl2) provided the desired allylic alcohol 60
as a 2:1 mixture of diastereomers in good yield (73%).
The two diastereomers 60a (major) and 60b (minor) were
partially separated by repeated chromatography so that
the major isomer 60a was isolated in 44% yield. With
60a in hand, we attempted the final deprotection step
which would give, if this diastereomer possessed the
correct relative stereochemistry at C12, enantiomerically
pure halicholactone (2). This reaction required surpris-
ingly vigorous conditions; the reaction of 60a with 2.5
equiv of TBAF in THF at reflux for 3 h gave the de-
silylated product in almost quantitative yield (99%). The
A small amount (22 mg) of the 2:1 mixture of diaster-
eomers 64a and 64b was desilylated with 3.0 equiv of
tetrabutylammonium fluoride in tetrahydrofuran at re-
flux for 2 h to give a mixture of neohalicholactone 1 and
the C12-epimer of neohalicholactone 65 in 90% yield.
1
From the 400 MHz H-NMR spectrum (in C6D6) of this
1
400 MHz H-NMR (in C6D6) and 125 MHz 13C-NMR (in
diastereomeric mixture, it was clear that the major
compound, a result of the mildly selective coupling
reaction, was most likely to be 1. The distinguishing
CDCl3) spectra for this compound exactly matched the
data reported for halicholactone (2). The optical rotation
of our product, [R]18 ) -91.7 (c 0.29 in chloroform),
1
D
features of the H-NMR spectrum were the signals that
corresponded well with the literature value, [R]23
-85.4 (c 1.16 in chloroform).
)
D
belonged to the C10 cyclopropane protons and the C12
1
proton. In the original 500 MHz H-NMR data of the
We were able to confirm unequivocally that we had
synthesized halicholactone (2) and not the C12-epimer
by performing the following transformations. The reac-
tion of 1.2 equiv of 39 with 2.13 equiv of tert-butyllith-
ium followed by aldehyde 14 provided the coupled
product 60 in disappointing yield, 28%, but with im-
proved diastereoselectivity, 60a :60b ratio of 3.2:1. It was
likely that the vinyllithium anion was quenched by
abstraction of an R-hydrogen from aldehyde 14 at a
competitive rate to which it was adding to the aldehyde
center. While we were able to separate the diastereomers
60a and 60b by flash chromatography, both of these
compounds were contaminated by unreacted aldehyde 14.
authentic sponge-derived 1, these signals were found at
0.27 (H10a, ddd, J ) 8.5, 5.0, 5.0), 0.45 (H10b, ddd, J ) 8.5
,5.0, 5.0), and 3.52 (H12, m), respectively. Our 400 MHz
1H-NMR spectrum clearly showed that it was the major
isomer 1 that was in excellent agreement with this
published data: 0.31 (H10a, ddd, J ) 8.6, 5.2, 5.2), 0.51
(H10b, ddd, J ) 8.9, 5.2, 5.2), and 3.58 (H12, dd, J ) 6.7,
3.4). The agreement for the minor isomer 65 was less
consistent with the published data for neohalicholactone;
(64) Luche, J .-L. J . Am. Chem. Soc., 1978, 100, 2226.
(65) Delanghe, P. H. M.; Lautens, M. Tetrahedron Lett., 1994, 35,
9513.

Total Synthesis of Halicholactone and Neohalicholactone
Sch em e 6
J . Org. Chem., Vol. 62, No. 19, 1997 6647
0.31 (H10a, m), 0.42 (H10b, ddd, J ) 8.6, 4.9, 4.9), and 3.50
(H12, dd, J ) 6.7, 3.4). Therefore, we returned to our
diastereomeric mixture of 64a and 64b and separated
them, and the major compound 64a was desilylated to
provide a pure sample of neohalicholactone (1) (74%).
From our pure sample of 1, we obtained an identical 500
MHz 1H-NMR (in d6-benzene) spectrum and 125 MHz
13C-NMR (in CDCl3) spectrum to those generously sup-
plied by Professor K. Yamada for the sponge-derived
Concomitant with the publication of our results,1b
White published a total synthesis of constanolactones A
and B (6a and 6b).6b A key step in this process was the
Cr(II)/Ni(II)-mediated reaction of a vinyl iodide bearing
a silyloxy group with aldehyde 56, which gave a selectiv-
ity (2:1) in the same relative direction as our result.
However, the final seemingly trivial task of removing the
silyl protecting group from the addition products could
not be achieved. It is interesting to note that this
problem was solved by performing the chromium-medi-
ated coupling reaction using the desilylated derivative
of the vinyl iodide. The problematical final desilylation
step reported by White is in keeping with our own
findings.
sample of 1. The optical rotation we obtained, [R]18
)
D
-54.6 (c 0.76 in chloroform), was also in excellent
agreement with the literature value, [R]16 ) -54.2 (c
D
0.73 in chloroform). Finally, in an attempt to convert
the otherwise wasted C12-epimer 64b into 64a we per-
formed the following reactions. The minor isomer 64b
(containing a small amount of 64a ) was oxidized to 66
using 5 mol % TPAP/1.5 equiv of NMO conditions in 91%
yield. However, the reduction of the carbonyl group in
66 under Luche conditions68 returned a 3:1 mixture of
64b:64a (58%). The undesired isomer, 64b, was unfor-
tunately the diastereomer which formed in excess.
In 1993, Gerwick and Proteau reported that the brown
alga, Laminaria sinclairii, produced a variety of interest-
ing oxylipins including several hydroxy-containing com-
pounds including methyl (15S)-hydroxy-(5Z,8Z,11Z,13E,-
17Z)-eicosapentaenoate and methyl (15S)-hydroxy-
(11Z,13E,17Z)-eicosatetraenoate.66 The isolation of these
compounds suggests that L. sinclairii possesses an active
lipoxygenase with positional specificity for C15 in C-20
substrates, and significantly that the C15 hydroxy center
that is formed is of S stereochemistry. In December 1994,
when we were approaching the completion of the syn-
thesis of neohalicholactone (1), Gerwick and co-workers
reported the isolation of “neohalicholactone” from L.
sinclairii.8 From spectroscopic analysis, Gerwick re-
ported that his isolate of “neohalicholactone” was identi-
cal in every way, overall structure and relative and
absolute stereochemistry, to neohalicholactone (1) iso-
lated from the marine sponge H. okadai. The maximal
1
difference in the 13C-NMR was 0.1 ppm and in the H-
NMR it was 0.05 ppm. Further, the optical rotation value
they had obtained for L. sinclairii derived material was
[R]27 ) -77 (c 0.14 in chloroform). The corresponding
D
value for the H. okadai derived material was [R]16
)
D
-54.2° (c 0.73 in chloroform). Therefore, Gerwick and
colleagues concluded that since the optical rotation value
of the L. sinclairii derived material was not substantially
different from the original value and of the same sign
(negative rotation) that they had isolated an identical
sample of neohalicholactone (1). Since the absolute
stereochemistry of 1 had not been reported at this time
(Yamada proved the absolute configuration of 2), Ger-
wick’s group went on to perform a conclusive degradative
study which established the absolute stereochemistry at
C15 of L. sinclairii derived “neohalicholactone”.
Therefore in summary, we have completed the first
total synthesis of an enantiomerically pure sample of 1.
The convergent synthesis was achieved in 30 steps and
0.65% overall yield starting from commercially available
materials, (R)- and (S)-malic acid. The total synthesis
of 1 and 2 unambiguously confirmed that the relative
stereochemistry of both these Halichondria okadai me-
tabolites is 8S,9R,11R,12R,15R, as had been assigned by
Yamada and Clardy2 at the outset of this project.
This result was in keeping with other C15 hydroxyl
metabolites from the same source. They concluded that
(66) Proteau, P. J .; Gerwick, W. H. Lipids, 1993, 28, 783.

6648 J . Org. Chem., Vol. 62, No. 19, 1997
Critcher et al.
Sch em e 7
in both the algae- and sponge-derived neohalicholactone
samples, the absolute stereochemistry was in fact
(8R,9S,11S,12S,15S) and therefore opposite to that orig-
inally proposed. However, as already described in section
2.3, the (8S,9R,11R,12R,15R) sample of neohalicholac-
tone (1) we had synthesized matched exactly every piece
of spectroscopic data published for the original sponge-
derived material, including most significantly the sign
and magnitude of the optical rotation value.1 Therefore,
a curious disparity existed between our findings and
those reported by Gerwick.8 Since we knew the absolute
stereochemistry of our starting materials and the relative
configurations in the advanced intermediate 34 (X-ray),
we considered the possibility that Gerwick had actually
isolated a C15 epimer of neohalicholactone which pos-
sessed very similar spectroscopic properties to those of
H. okadai derived neohalicholactone.
pound 68b (contaminated with 17% of 68a ) was desily-
lated using tetrabutylammonium fluoride and 67 (con-
taminated with 17% of C15,C12-epi-neohalicholactone (69))
was isolated in 84% yield. We obtained 400 MHz and
1
270 MHz H-NMR spectra of this 5:1 mixture of isomers
(67:69), as well as a 67.8 MHz 13C-NMR spectrum. From
our proton NMR comparisons with the data Gerwick had
compiled from a 300 MHz 1H-NMR spectrum of L.
sinclairii derived “neohalicholactone”, it became apparent
that the latter material could indeed be C15-epi-neo-
1
halicholactone (67). Remarkably the H-NMR and 13C-
NMR spectra of 67 and 1 were indistinguishable at the
It was our belief that the C13-C14 trans olefin would
form a natural break in the stereochemical information
of the molecule such that inversion of stereochemistry
1
at C15 would not significantly manifest itself in the H-
or 13C-NMR spectra. Further, Gerwick had obtained his
1H-NMR at 300 MHz and 13C-NMR at 75 MHz, while we,
in common with Yamada, performed our corresponding
analyses at 500 MHz and 125 MHz, respectively. We
believed that if Gerwick had isolated C15-epi-neohali-
cholactone (67) then it would only be in the higher field
1H-NMR spectrum that any differences between 67 and
1 would reveal themselves. We prepared a sample of 67
in order to test this. (3S)-(1E,5Z)-3-O-((tert-butyldiphen-
ylsilyl))oxy)-1-iodo-1,5-octadiene (50) was prepared as
previously described but starting from (S)-malic acid. The
coupling of 2.0 equiv of trans-iodo alkene (S)-50 and 1.0
equiv of aldehyde 14 was performed using 4.4 equiv of
chromium(II) chloride (containing 0.5 wt % NiCl2) in
DMSO at room temperature over 16 h. A 1:1.3 mixture
of isomers 68a and 68b resulted, and after careful
separation of these isomers during five chromatography
runs, the minor isomer 68b was isolated in 24% yield
which was unfortunately still contaminated with 17%
of 68a . However, the major isomer 68a was isolated
pure in 30% yield, while a further 16% yield of a 1:1
mixture of 68a :68b was isolated. This implies that the
overall yield for the coupling reaction was at least 70%
(Scheme 7).
quoted field strengths. Also, due to the 17% of 69 (of
(8S,9R,11R,12S,15S) absolute stereochemistry) being
present in the 1H-NMR spectra, it was clear that this
isomer did not closely match Yamada’s and our own data
for 1. In order to obtain an accurate optical rotation
value for our sample of 67, the small amount of 69
contaminant was removed by very careful column chro-
matography purification. Our purified sample of 67 had
an optical rotation value of -84.9° (c 0.43 in chloroform).
This value was of the same sign and similar magnitude
to that reported for L. sinclairii derived “neohalicholac-
tone”, -77.0˚ (c 0.14 in chloroform).8 Finally we obtained
a 500 MHz 1H-NMR spectrum of 67 in d6-benzene.
Comparison of this spectrum with our 500 MHz spectrum
of 1, and that provided by Yamada for the sponge derived
1, was quite illuminating. Every signal appeared to be
identical in every respect (shape, size, position, and
coupling constants) throughout the spectra of 67 and 1
except for the signals resulting from the C16 protons. The
splitting pattern for the C16 protons in 67 was signifi-
cantly more dispersed than that found in 1.
1
Therefore, we found that 500 MHz H-NMR was able
to distinguish between 67 and 1. However to date only
a 300 MHz 1H-NMR of L. sinclairii derived “neohali-
cholactone” exists. A 500 MHz 1H-NMR of this material
is subject to it being reisolated from more collections of
Encouragingly, the H-NMR and 13C-NMR of 68b, the
1
minor isomer of the coupling reaction, were in excellent
agreement with those we had previously obtained for 64a
which had then been converted to 1. The minor isomer
would be expected, in this case, to be of R configuration
at C-12 based on the known directing effect of 50.
Further, the data we obtained from 68a were quite
different from both 64a and 68b. Therefore we presumed
that 68b possessed the correct relative stereochemistry
for the synthesis of C15-epi-neohalicholactone (67). Com-
1
L. sinclairii in the future. Therefore until this final H-
NMR experiment is performed we are only able to
conclude that the L. sinclairii derived “neohalicholactone”
is probably a C15-epimer of the sponge derived material
1 that we originally synthesized. The biosynthetic
implications of this conclusion, that two different marine
organisms possess lipoxygenases with different enan-
tiospecifities yet identical preferences for enantioselective

Total Synthesis of Halicholactone and Neohalicholactone
J . Org. Chem., Vol. 62, No. 19, 1997 6649
MgSO4. The solvent was removed under reduced pressure,
and the residue was purified by column chromatography on
silica, eluting with EtOAc/hexane (1:2). The product 23 (5.87
g, 73%) was isolated as a colorless oil.
formation of the nine-membered lactone, which is pre-
sumably also an enzyme-controlled process, must remain
the subject of future debate.
(3S)-3-O-(((p-Meth oxyp h en yl)m eth yl)oxy)-γ-bu tyr ola c-
tol (22). To lactone 23 (2.00 g, 9.00 mmol) in dry toluene (60
mL) at -20 °C under a nitrogen atmosphere was added
DIBAL-H (1.5 M, 6.6 mL, 9.9 mmol) slowly over a 5-min period
(making sure that the internal temperature did not rise above
-20 °C). After a few minutes the reaction was complete (by
TLC analysis) and was quenched with H2O (6.6 mL), and the
solution was then warmed to rt. Solid NaHCO3 and EtOAc
(50 mL) were added sequentially, and the mixture was stirred
vigorously for approximately 0.5 h. After this time the mixture
was suction filtrated, and the filter cake was washed with
EtOAc. The organic washings were evaporated under reduced
pressure to give the crude lactol 22 (2.02 g, 100%), as a mixture
of epimers (3:2) and as a colorless crystalline solid. Rf 0.3 (2:
1, hexane/ethyl acetate). Anal. Calcd for C12H16O4: C, 64.3,
H, 7.14. Found: C, 64.1; H, 7.16. IR (Nujol) 3419 (br, OH),
Exp er im en ta l Section
Tetrahydrofuran and ether were freshly distilled from
sodium using benzophenone as an indicator. Toluene was
freshly distilled from sodium; dichloromethane from phospho-
rus pentoxide; acetonitrile, dimethyl sulfoxide, and hexane
from calcium hydride. All the aforementioned solvents were
distilled under an inert atmosphere of either argon or nitrogen.
Reagents were either used as received, from commercial
sources or purified by recognized methods.67 Petroleum ether
refers to that fraction which boils in the range 60-80 °C, and
it was distilled prior to use as was ethyl acetate.
All reactions were carried out in vacuum-flame-dried Schlenk
tubes under an argon atmosphere, unless otherwise stated.
Flash chromatography68 was performed using Matrex nor-
mal phase silica, pore size 60 Å. All reactions were monitored
by thin layer chromatography using aluminum sheets pre-
coated with 250 nm silica gel and were visualized by UV light
and then by potassium permanganate solution, phosphomo-
lybdic acid solution, ninhydrin solution, or anisaldehyde
solution.
1
1780 cm-1; H-NMR (360 MHz, CDCl3) δ 1.84-2.20 (2H, m),
2.64 (0.4H, d, J ) 2.8), 3.68 (0.6H, d, J ) 11.2) 3.80 (1.2H, s),
3.81 (1.8H, s), 3.80-4.55 (5H, m), 5.42 (0.6H, dd, J ) 5.0, 11.1),
5.68 (0.4H, dd, J ) 2.7, 5.2), 6.87-6.90 (2H, m), 7.24-7.27
(2H, m); MS (FAB) m/z 247 ([M +Na]+, 11%), 224 (M, 8%),
223 ([M - 1]+, 8%), 121 (100); HRMS Found: m/ z 224.1046
(calcd for C12H16O4 (M+), 224.1049).
(3S)-3-O-(((p -Meth oxyp h en yl)m eth yl)oxy)-γ-bu tyr ola c-
ton e (23). To (3S)-hydroxy-γ-butyrolactone (5.41 g, 53.0
mmol) in dry dichloromethane (120 mL) and cyclohexane (50
mL) at room temperature under a nitrogen atmosphere was
added PMB-trichloroacetimidate (13.2 mL, 63.6 mmol). After
5 min a catalytic amount of TfOH (14.0 µL, 0.16 mmol) was
injected into the reaction mixture, and within a few minutes
trichloroacetamide could be seen precipitating out of solution
as the reaction progressed. After 2 h there was still starting
material visible by thin layer chromatography, and therefore
a further 2.0 mL (9.6 mmol) of (p-methoxyphenyl)methyl
trichloroacetimidate was required to achieve full consumption
of starting material. The reaction was quenched after a total
reaction time of 3 h by adding solid NaHCO3 until neutral pH
was obtained. The mixture was then diluted with a solution
of CH2Cl2 (100 mL):cyclohexane (50 mL) followed by suction
filtration. The filter cake was washed with a 2:1 mixture of
CH2Cl2 and cyclohexane (150 mL), and the solvent was
removed from the organic washings under reduced pressure.
The residue was purified by column chromatography on silica
eluting with EtOAc/ hexane (1:2) to afford 23 (8.9 g, 75%) as
a colorless oil. Rf 0.5 (2:1, hexane/EtOAc); [R]22D -18.1 (c 0.18,
Meth yl (8S,5Z)-8-O-(((p-Meth oxyp h en yl)m eth yl)oxy)-
8,9-d ih yd r oxyn on -5-en oa te (18 (R ) P MB)). To a suspen-
sion of (4-carboxybutyl)triphenylphosphonium bromide (19)
(5.15 g, 11.62 mmol) in dry toluene (50 mL) at room temper-
ature under an argon atmosphere was added sodium bis(tri-
methylsilyl)amide (20.3 mL, 20.3 mmol, 1.0 M solution in THF)
slowly over 5 min at 0 °C. The solution was then stirred at rt
for 0.5 h, and a red color developed, at which time the mixture
was cooled to -78 °C. Then in a different reaction vessel, to
a solution of lactol 22 (2.17 g, 6.89 mmol) in dry toluene (100
mL) at -78 °C under an argon atmosphere was added sodium
bis(trimethylsilyl)amide (10.17 mL, 10.17 mmol, a new bottle
as a 1.0 M solution in tetrahydrofuran) quite rapidly (ca 0.5
min). The solution now containing the sodium alkoxide of 22
was stirred for 15 min at -78 °C. The phosphonium ylide was
then transferred via cannula into the sodium alkoxide solution
of 22 over a 20 min period while both mixtures were kept at
-78 °C. After the addition, the resultant red solution was left
at -78 °C for 5 min and allowed to warm to 0 °C over 45 min.
At this time the solution was a pale yellow color, and TLC
analysis indicated that full consumption of starting material
had occurred. The reaction mixture was quenched by adding
MeOH (20 mL) and 5% HCl (10 mL) at -60 °C. After warming
to room temperature, CH2Cl2 (150 mL) and more 5% HCl
(enough to make the solution pH 2-3) were added, and the
two layers were then separated. The aqueous solution was
further extracted with CH2Cl2, and the combined organic
layers were dried over anhydrous Na2SO4. After the solvent
had been evaporated at reduced pressure, the residue was
dissolved in MeOH (20 mL), and a MeOH/AcCl (30 mL: 0.5
mL) mixture was then added until the reaction mixture was
at pH 4 and slow conversion to the methyl ester could be
monitored by TLC. After 3.5 h the solvent was removed under
reduced pressure and the residue purified by column chroma-
tography on silica eluting with EtOAc:petroleum ether (1:1).
The product 18 (R ) PMB) (2.15g, 69%) was a colorless oil. Rf
0.5 (2:1, hexane/EtOAc). Anal. Calcd for C18H26O5: C, 67.1;
CHCl3); IR (neat) 1780 (s, CO) cm-1 1H-NMR (360 MHz,
;
CDCl3) δ 2.60 (1H, dd, J ) 2.7, 18.0), 2.68 (1H, dd, J ) 5.4,
18.0), 3.81 (3H, s), 4.34-4.40 (3H, m), 4.44 (1H, d, J AB ) 11.4),
4.48 (1H, d, J AB ) 11.4), 6.89 (2H, d, J ) 8.6), 7.25 (2H, d, J
) 8.6); 13C-NMR (90.6 MHz, CDCl3) δ 34.9 (t), 55.2 (q), 70.9
(t), 72.5 (t), 73.5 (d), 113.8 (d), 128.7 (s), 129.4 (d), 159.5 (s),
175.4 (s); MS (EI) m/z 222 (M+, 15%), 121 (100); HRMS m/z
222.0903 (calcd for C12H14O4 (M+) 222.0892). A side product
was identified as (2S)-2-O-(((p-methoxyphenyl)methyl)oxy)-γ-
butyrolactone; IR (neat) 1780 (s, CO) cm-1; 1H-NMR (360 MHz,
CDCl3) δ 2.17-2.31 (1H, m), 2.39-2.47 (1H), 3.81 (3H, s), 4.15
(1H, t, J ) 7.7), 4.20-4.24 (1H, m), 4.36-4.44 (1H, m), 4.67
(1H, d, J AB ) 11.4), 4.86 (1H, d, J AB ) 11.4), 6.89 (2H, d, J )
8.5), 7.31 (2H, d, J ) 8.5); MS (EI) m/z 222 (M+, 5%), 137 (82),
121 (100). An alternative procedure used to obtain 23 was as
follows: To (3S)-hydroxy-γ-butyrolactone (3.70 g, 36.3 mmol)
in dichloromethane (40 mL) under at argon atmosphere at
room temperature was added (p-methoxyphenyl)methyl trichlo-
roacetimidate (11.3 mL, 54.4 mmol) followed by solid cam-
phorsulfonic acid (422 mg, 1.82 mmol). The reaction was
stirred at room temperature for 16 h and then diluted with
petroleum ether (100 mL), and the precipitate was removed
by filtration. The filtrate was then washed saturated aqueous
NaHCO3 (40 mL) and the organic layer dried over anhydrous
H, 8.15%. Found: C, 67.0; H, 8.07. [R]22 +23.3 (c 0.24 in
D
CHCl3); IR (neat) 3440 (br, OH), 1740 (s, CO) cm-1; H-NMR
1
(360 MHz, CDCl3) δ 1.69 (2H, pentet, J ) 7.5) 2.05-2.11 (2H,
m) 2.29-2.35 (4H, m), 3.48-3.52 (2H, m), 3.60-3.65 (1H, m)
3.66 (3H, s, OCH3), 3.80 (3H, s), 4.47 (1H, d J AB ) 11.2), 4.60
(1H, d, J AB ) 11.2), 5.40-5.50 (2H, m), 6.88 (2H, d, J ) 8.5),
7.27 (2H, d, J ) 8.5); 13C-NMR (90.6 MHz, CDCl3) δ 24.7 (t),
26.6 (t), 28.7(t), 33.4 (t), 51.5 (q, OCH3), 55.2 (q, OCH3), 64.0
(t), 71.2 (t), 79.1 (d), 113.9 (d), 125.7 (d), 130.1 (d), 130.4 (s),
131.1 (d, olefinic CH), 159.3 (s), 174.0 (s, CdO); MS (FAB) m/z
407/409 ([M + Rb]+, 5%), 323 ([M + H]+, 6%), 250 (5), 241 (11),
185 (4), 154 (10), 137 (15), 121 (100), 93 (14).
(67) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals; Pergamon: London, 1985.
(68) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem., 1978, 43, 2923.

6650 J . Org. Chem., Vol. 62, No. 19, 1997
Critcher et al.
Oxid a tion of 34 (R ) P MB) To Give Ald eh yd e 25. A
solution of oxalyl chloride (0.65 mL, 7.45 mmol) in CH2Cl2 (20
mL) was cooled to -78 °C under an argon atmosphere, and a
solution of DMSO (1.06 mL, 14.9 mmol) in CH2Cl2 (15 mL)
was added over a 5-min period via a dropping funnel. Stirring
was continued at -78 °C for 10 min followed by addition of 18
(R ) PMB) (2.0 g, 6.21 mmol) over ca. 5 min. The reaction
mixture was stirred for 15 min, and triethylamine (4.3 mL,
31.1 mmol) was added with stirring at -78 °C. The cooling
bath was then removed, and water (20 mL) was added at room
temperature. Stirring was continued for 10 min, and the two
layers were then separated. The aqueous phase was extracted
with dichloromethane (3 × 30 mL), the organic layers were
combined and dried over anhydrous magnesium sulfate, and
the solvent was evaporated to give crude aldehyde 25 (1.99 g,
quantitative yield) as a pale yellow colored oil which was used
directly in the next step.
minor isomer δ 13.1 (t), 18.3 (d), 24.7 (t), 25.1 (d), 26.6 (t),
28.1 (q), 32.7 (t), 33.4 (t), 51.4 (q), 55.2 (q), 70.3 (t), 79.3 (d),
80.1 (s), 113.7 (d), 126.2 (d), 129.0 (d), 130.5 (d), 130.6 (s), 159.1
(s), 172.9 (s), 173.0 (s); MS (FAB) m/z 433 ([M + H]+, 2%),
t
377 ([M - Bu]+, 3%), 239 (6), 189 (5), 137 (6), 121 (100),
57 (8).
DDQ Dep r otection of 37 a n d Su bsequ en t Sep a r a tion
of Isom er s 28 a n d 29. To 27 (2.13 g, 4.93 mmol) in a CH2Cl2/
H2O mixture (30 mL:1.6 mL) was added solid DDQ (1.23 g,
5.42 mmol) at room temperature. After 1 h the reaction was
complete, and the solvent was evaporated. Complete separa-
tion of the two hydroxy isomers 28 and 29 required repeated
column chromatography eluting with hexane/EtOAc (4:1). The
products (combined yield of 1.54 g, 100%) were both oils where
the major isomer 28 was isolated pure (1.06 g, 69%). Rf
0.5 (5:1, hexane/ethyl acetate) and Rf
0.47 (5:1, hexamnaejo/r
minor
EtOAc); major isomer 28 [R]21 -114.1 (c 0.7 in CHCl3); IR
D
(E)-γ-Alk oxy-r,â-u n sa tu r a ted ter t-Bu tyl Ester (26). To
a stirred suspension of LiCl (316 mg, 7.45 mmol) in MeCN
(40 mL) at rt under an argon atmosphere was added tert-butyl
diethylphosphonoacetate (2.33 mL, 9.94 mmol), DBU (0.93 mL,
6.21 mmol), and finally a solution of aldehyde 25 (1.99 g, 6.21
mmol) in dry MeCN (40 mL). The reaction mixture was stirred
at rt for 5 h at which time it was quenched by adding H2O (30
mL). The aqueous layer was then extracted with EtOAc, and
the combined organic layers were dried over anhydrous
MgSO4. After the solvent had been evaporated at reduced
pressure, the residue was purified by column chromatography
on silica eluting with hexane-EtOAc (6:1). Product 26 (1.95g,
75% for the two steps from 18 (R ) PMB)) was a colorless oil.
(neat) 3500, 1740, 1720 cm-1 1H-NMR (360 MHz, CDCl3) δ
;
0.79 (1H, ddd, J ) 4.3, 6.2, 8.3), 1.08-1.13 (1H, m), 1.43 (9H,
s), 1.47-1.57 (2H, m), 1.70 (2H, pentet, J ) 7.4), 1.86 (1H,
bs), 2.10 (2H, q, J ) 7.2,), 2.30-2.34 (4H, m), 3.14 (1H, q, J )
6.6), 3.66 (3H, s), 5.45-5.56 (2H, m); 13C-NMR δ (90.6 MHz,
CDCl3) 12.2 (t), 18.9 (d), 24.7 (t), 26.6 (t), 27.6 (d), 28.1 (q),
33.3 (t), 34.9 (t), 51.5 (q), 73.2 (d), 80.3 (s), 125.8 (d), 131.8 (d),
173.0 (s,), 174.1 (s); MS (FAB) m/z 313 ([M + H]+, 47%), 257
(65), 239 (97), 221 (57), 189 (85), 57 (100); HRMS m/z 313.2012,
(calcd for C17H29O5 (M + H)+ 313.2015). minor isomer 29;
1
[R]21 +30.4 (c 0.09 in CHCl3); H-NMR (360 MHz, CDCl3) δ
D
0.89-0.94 (1H, m), 1.06-1.11 (1H, m), 1.44 (9H, s), 1.48-1.57
(2H, m), 1.70 (2H, pentet, J ) 7.3), 2.11 (2H, q, J ) 7.2), 2.31-
2.40 (4H, m), 3.28 (1H, m), 3.67 (3H, s), 5.45-5.58 (2H, m);
13C-NMR (90.6 MHz, CDCl3) δ 11.7 (t), 18.9 (d), 24.7 (t), 26.6
(t), 27.0 (d), 28.1 (q), 33.3 (t), 35.2 (t), 51.5 (q), 72.3 (d), 80.3
(s), 125.8 (d), 132.0 (d), 173.0 (s), 174.1 (s); MS (FAB) m/z 313
([M + H]+, 44%), 257 (47), 239 (100), 221 (54), 207 (16), 189
(75), 57 (97); HRMS m/z 313.2009 (calcd for C17H29O5 (M +
H)+ 313.2015).
Rf 0.6 (5:1, hexane/EtOAc); [R]21 -35.2 (c 0.05 in CHCl3); IR
D
(neat) 1740 (s, CO), 1720 (s, CO) cm-1
;
1H-NMR (360 MHz,
CDCl3) δ 1.50 (9H, s), 1.65 (2H, pentet, J ) 7.4), 2.01-2.07
(2H, m), 2.28 (2H, t, J ) 7.5), 2.32-2.41 (2H, m), 3.66 (3H, s),
3.81 (3H, s), 3.92 (1H, q, J ) 6.5), 4.32 (1H, d, J AB ) 11.5),
4.52 (1H, d, J AB ) 11.5), 5.40-5.47 (2H, m), 5.92 (1H, d, J )
15.7), 6.73 (1H, dd, J ) 6.6, 15.6), 6.87 (2H, d, J ) 8.5), 7.25
(2H, d, J ) 8.5; 13C-NMR (90.6 MHz, CDCl3) δ 24.7 (t), 26.8
(t), 28.1 (q), 32.9 (t), 33.5 (t), 51.5 (q), 55.3 (q), 70.7 (t), 77.6
(d), 80.5 (s), 113.8 (d), 124.1 (d), 125.3 (d), 129.3 (d), 130.2 (s),
131.2 (d), 146.6 (d), 159.2 (s), 165.5 (s), 174.0 (s); MS (FAB)
m/z 505/503 ([M + Rb]+, 5%), 419 ([M + H]+, 2%), 363 (3), 241
(6), 207 (4), 154 (4), 137 (7), 121 (100); HRMS m/z 419.2423
(calcd for C24H34O6 (M+ + 1) 419.2434).
Ester Hyd r olysis of 28 to Give a Hyd r oxy Acid In ter -
m ed ia te (20) Wh ich w a s Su bsequ en tly La cton ized to 32.
To (8S,9R,11R)-28 (250 mg, 0.80 mmol) in a THF/H2O/MeOH
(4 mL:1 mL:1 mL) solution was added solid LiOH (67.0 mg,
1.60 mmol) at rt. After being stirred for 16 h, the mixture
was quenched with 5% HCl (2 mL). The aqueous layer was
then extracted with EtOAc. The combined organic layers were
dried over anhydrous MgSO4, and then the solvent was
removed under reduced pressure. The crude oil 30 (0.24 g,
100%) was used directly in the next step; IR (neat) 3500-2800,
Cyclop r op a n a tion Rea ction To Give 27 a s a Mixtu r e
of Tr a n s-Dia ster eom er s. Dimethylsulfoxonium methylide
(2.2 equivs generated from trimethylsulfoxonium iodide, 757
mg, 3.44 mmol, and sodium hydride (138 mg, 5.73 mmol, 60%
dispersion in mineral oil) in dry DMSO (6 mL) at rt was
treated with tert-butyl ester 26 (720 mg, 1.72 mmol) in DMSO
(6 mL). The mixture was stirred for 0.5 h at rt and then 20 h
at 90 °C. The reaction was then quenched at rt by adding
saturated aqueous NH4Cl (10 mL). The product was extracted
with EtOAc, the combined organic extracts were dried with
anhydrous Na2SO4, and the solvent was evaporated at reduced
pressure. The crude residue was then purified by column
chromatography on silica eluting with EtOAc/hexane (1:6)
which gave 27 (550 mg, 74%) as a colorless oil. The product
was a 5:2 mixture of inseparable but exclusively trans-
diastereomers where the major diastereomer contained the
desired relative stereochemistry (8S,9R,11R). Rf 0.5 (6:1,
hexane/ethyl acetate). Anal. Calcd for C25H36O6: C, 69.4; H,
8.33%. Found: C, 68.9; H, 8.54. Reverse phase HPLC (75%
to 95% MeOH/25% to 5% H2O + 0.1% trifluoroacetic acid)
retention time 5.4 min, 95% pure; IR (neat) 1740 (s, CO), 1720
(s, CO) cm-1; 1H-NMR (360 MHz, CDCl3) major isomer, δ 0.62-
1.55 (4H, m), 1.45 (9H, s), 1.69 (2H, pentet, J ) 7.4), 2.05-
2.11 (2H, m), 2.30 (2H, t, J ) 7.4), 2.35-2.42 (2H, m), 2.91
(1H, q, J ) 6.5), 3.66 (3H, s), 3.80 (3H, s), 4.45-4.58 (2H, m),
5.40-5.60 (2H, m), 6.86-6.88 (2H, m), 7.23-7.26 (2H, m),
minor isomer (360 MHz, CDCl3) δ 1.43 (9H, s), 3.00 (1H, q, J
) 6.5) all other peaks were coincidental with the major iso-
mer; 13C-NMR (67.8 MHz, CDCl3) major isomer δ 10.8 (t),
20.2 (d), 24.7 (t), 25.4 (d), 26.6 (t), 28.1 (q), 32.8 (t), 33.4 (t),
51.4 (q), 55.2 (q), 70.6 (t), 79.2 (d), 80.1 (s), 113.7 (d), 126.3
(d), 129.0 (d), 130.5 (d), 130.6 (s), 159.1 (s), 172.9 (s), 173.9 (s),
1
1712 cm-1, H-NMR (360 MHz, CDCl3) δ 0.80 (1H, ddd, J )
4.4, 6.3, 8.3), 1.10-1.15 (1H, m), 1.44 (9H, s), 1.49-1.60 (2H,
m), 1.72 (2H, pentet, J ) 7.3), 2.10-2.17 (2H, m), 2.35-2.43
(4H, m), 3.15 (1H, q, J ) 6.5), 5.47-5.57 (2H, m), 5.50-6.00
(2H, bs). To a solution of crude 8S,9R,11R-hydroxy acid 30
(650 mg, 2.18 mmol) in THF (6 mL) under an argon atmo-
sphere was added Et3N (0.49 mL, 3.49 mmol) at rt with stirring
for 10 min. Then at rt, 2,4,6-trichlorobenzoyl chloride (0.37
g, 2.40 mmol) was added slowly with stirring for 2 h. After
dilution with toluene (60 mL) this solution was added (using
a syringe pump) to excess DMAP (4.00 g, 32.7 mmol) in toluene
(100 mL) at reflux over 6 h. After the addition had been
completed, the reaction mixture was allowed to cool to rt and
the solvent was then evaporated. The residue was purified
by column chromatography on silica eluting with petroleum
ether/EtOAc (10:1 to 5:1). The product 32 (460 mg, 75%) was
isolated as a colorless oil. Rf 0.6 (10:1, petroleum ether/ethyl
acetate). The only observed side product was the dimeric
species (16.6 mg, 2.7%). Rf 0.4 (10:1, petroleum ether/ethyl
acetate) . The dimer was formed in much larger quantities if
the addition was too rapid or the mixture was too concentrated.
(In the event of significant amounts of dimer being formed it
was hydrolyzed back to the starting hydroxy acid). Monomer
32: Anal. Calcd for C16H24O4: C, 68.6; H, 8.57%. Found: C,
68.5; H, 8.75. [R]21D -81.5 (c 0.01, CHCl3); IR (neat) 1740 cm-1
;
the assignment of all the proton and carbon signals was
1
achieved using phase-sensitive H-1H COSY, normal 1H-13C
1
correlation, and long-range H-13C correlation; 1H-NMR (500
MHz, CDCl3) δ 0.80-0.86 (1H, m, cyclopropane CHH (10)),

Total Synthesis of Halicholactone and Neohalicholactone
J . Org. Chem., Vol. 62, No. 19, 1997 6651
t
1.13-1.18 (1H, m, cyclopropane CHH (10′)), 1.45 (9H, s, Bu
addition, the reaction mixture was allowed to cool to rt, and
the solvent was then removed under reduced pressure. The
residue was purified by column chromatography on silica
eluting with petroleum ether/ethyl acetate (7:1). The product
33 (329 mg, 76%) was isolated as a colorless crystalline solid.
Rf 0.6 (7:1, petroleum ether/ethyl acetate). [R]20D -2.1 (c 1.55,
(14)), 1.60-1.68 (2H, m, 2 × cyclopropane CH (9+11)), 1.77
(1H, dddd, J ) 6.5, 11.9, 11.9, 11.9, CHH (3)), 2.04-2.09 (2H,
m, CHH and CHH (4) and (3′)), 2.13-2.18 (1H, m, -CHH- (7)),
2.22-2.26 (1H, m, -CHH-(2)), 2.29-2.36 (1H, m, -CHH-(2′)),
2.45-2.56 (2H, m, 2 × CHH (7′) and (4′)), 4.27 (1H, ddd, J )
1.5, 7.5, 10.9, CHOCO), 5.45-5.49 (2H, m, CHdCH); 13C-NMR
(125.8 MHz, CDCl3) 12.5 (t, (10)), 19.6 (d, (9 or 11), 24.6 (d, (9
or 11)), 25.3 (t, (4)), 26.4 (t, (3)), 28.1 (q, (14)), 33.6 (t, (2)),
33.7 (t, (7), 74.6 (d, (8)), 80.5 (s, (13)), 124.4 (d, (6)), 134.9 (d,
(5)), 172.7 (s, (12)), 174.0 (s, (1)); MS (GC/EIMS) m/z 224 (3),
207 (7), 110 (100), 82 (92), m/ z (FAB) 281 ([M + H]+, 32%),
1
CHCl3); H-NMR (400 MHz, CDCl3) δ 0.94 (1H, ddd, J ) 4.3,
6.4, 8.5), 1.12 (1H, ddd, J ) 4.6, 4.6, 8.9), 1.45 (9H, s), 1.53
(1H, ddd, J ) 4.6, 4.6, 8.5), 1.58-1.65 (1H, m), 1.77 (1H, dddd,
J ) 6.7, 11.9, 11.9, 11.9), 2.02-2.34 (5H, m), 2.44-2.54 (2H,
m), 4.38 (1H, ddd, J ) 1.2, 7.3, 10.7), 5.44-5.49 (2H, m); 13C-
NMR (67.8 MHz, CDCl3) δ 12.5 (t), 19.2 (d), 24.4 (d), 25.3 (t),
26.4 (t), 28.1 (q), 33.5 (t), 33.8 (t), 74.1 (d), 80.6 (s), 124.4 (d),
134.8 (d, olefinic CH), 173.0 (s), 179.0 (s); MS (FAB) m/z 281
([M + H]+, 33%), 225 (100), 207 (85), 189 (45).
225 (100), 207 ([M - BuO-]+, 83%), 189 (43); HRMS m/z
t
281.1749 (calcd for C16H25O4 (M + H)+ 281.1753). Dimeric
compound: IR (neat) 1723 (CdO) cm-1; H-NMR (270 MHz,
1
CDCl3) δ 0.77-0.88 (2H, m, 2 × cyclopropane CH), 1.12-1.18
(2H, m), 1.44 (18H, s), 1.55-1.73 (8H, m), 2.05-2.35 (10H, m),
2.50-2.59 (2H, m), 4.45-4.56 (2H, m), 5.30-5.50 (4H, m); 13C-
NMR (67.8 MHz, CDCl3) δ 12.4 (t), 19.6 (d), 24.4 (d), 25.2 (t),
26.3 (t), 28.1 (q), 32.5 (t), 33.5 (t), 74.6 (d), 80.4 (s), 125.6 (d),
131.6 (d), 172.6 (s), 173.1 (s); MS (FAB) m/z 583 ([M + Na]+,
16%), 561 ([M + H]+, 11%), 449 (35), 225 (100), 207 (82); HRMS
m/z 559.3285 (calcd for C32H47O8 (M - 1)+ 559.3271).
Hyd r olysis of th e Dim er ic Com p ou n d . To the dimeric
compound (56.0 mg, 0.10 mmol) in a 3:1:1 THF (1.0 mL), H2O
(0.3 mL), and MeOH (0.3 mL) mixture was added solid LiOH
(17.0 mg, 0.4 mmol) in one portion. The resultant homoge-
neous solution was stirred at rt for 16 h. The reaction was
quenched by adding 2 N HCl until pH 1-2 was obtained. The
aqueous phase was extracted with EtOAc. The solvent was
then removed from the combined organic extracts under
reduced pressure to give the crude hydroxy acid 30 (50.3 mg,
84%).
Acid ic Hyd r olysis of th e ter t-Bu tyl Ester in 33. To 33
(300 mg, 1.07 mmol) in CH2Cl2 (5 mL) at rt under an argon
atmosphere was added freshly distilled CF3SO2H (1.5 mL) in
CH2Cl2 (5 mL) dropwise. The reaction was complete after 2
h, and the solvent was then removed under reduced pressure.
To ensure all the CF3SO2H had been removed, toluene (15 mL)
was added to the crude residue, and the solvent was again
removed under reduced pressure. This process was then
repeated to give a crude residue which was purified by column
chromatography on silica eluting with ethyl acetate/petrol-
eum ether (1:1). The product 35 (191 mg, 80%) was isolated
as a colorless oil. [R]21 -17.0 (c 0.5, CHCl3). Anal. Calcd
D
for C12H16O4: C, 64.3; H, 7.14. Found: C, 64.3; H, 7.25. IR
(neat) 1736 cm-1; 1H-NMR (270 MHz, CDCl3) δ 1.07-1.14 (1H,
m), 1.28 (1H, ddd, J ) 4.8, 4.8, 8.8), 1.63 (1H, ddd, J ) 4.4,
4.4, 8.4), 1.69-1.87 (2H, m), 2.00-2.37 (5H, m) 2.44-2.56 (2H,
m), 4.40 (1H, ddd, J ) 1.5, 7.3, 11.0), 5.40-5.54 (2H, m) 9.93
(1H, bs); 13C-NMR (67.8 MHz, CDCl3) δ 13.6 (t), 18.0 (d), 25.2
(t), 25.7 (d), 26.3 (t), 33.6 (t), 33.7 (t), 73.8 (d), 124.1 (d), 134.9
(d), 173.9 (s), 179.6 (s); MS (CI-isobutane) m/z 225 ([M + H]+,
15%), 207 ([M - OH]+, 100%), 189 (60), 179 (5), 110 (32).
Selective Red u ction of th e Ca r boxylic Acid F u n ction -
a lity in th e (8S,9R,11R) Isom er 34 To 8S,9R,11R-carboxylic
acid 34 (67.3 mg, 0.30 mmol) in THF (1 mL) under an argon
atmosphere at rt was added Et3N (30.3 mg, 0.3 mmol), and
the mixture was stirred for 5 min. After cooling the mixture
to -5 °C, EtOCOCl (32.4 mg, 0.3 mmol) in THF (0.5 mL) was
added dropwise. Stirring was continued at -5 °C for 0.5 h,
and then the precipitated Et3NHCl was filtered and washed
with THF (2 mL). The mixed anhydride was then slowly
added to a solution of NaBH4 in H2O (0.3 mL) at 0 °C. The
resultant mixture was stirred at 0 °C for 0.5 h and subse-
quently at rt until effervescence ceased (ca. 0.5 h). Addition
of 2 N HCl (1 mL) followed, and the aqueous layer was
repeatedly extracted with EtOAc. The combined organic
extracts were dried over anhydrous MgSO4, and the solvent
was removed under reduced pressure. The residue was
purified by column chromatography on silica eluting with
EtOAc/petroleum ether (1:2). The product 36 (48.5 mg, 77%)
was isolated as a colorless oil. Rf 0.4 (2:1, hexane/ethyl
Acid Hyd r olysis of ter t-Bu tyl Ester 32. To the lactone
32 (160 mg, 0.57 mmol) in CH2Cl2 (6 mL) at rt under an argon
atmosphere was added CF3SO2H (1.5 mL) dropwise. The
reaction was complete after 2 h, and the solvent was then
removed under reduced pressure. To ensure all the CF3SO2H
had been removed, toluene (4 mL) was added to the crude
residue, and the solvent was again removed under reduced
pressure. This process was then repeated several times to give
a crude solid which was then crystallized from CH2Cl2 and
cyclohexane to give (8S,9R,11R)-34 (103 mg, 80%) as a
colorless crystalline solid. {Some less pure material (22.2 mg,
17%) was also obtained by removal of the solvent under
reduced pressure from the mother-liquor}. Anal. Calcd for
C12H16O4: C, 64.3; H, 7.14%. Found: C, 64.3; H, 7.44. [R]21
D
-238.0 (c 0.06, CHCl3); IR (neat) 3300-2400, 1740 cm-1; H-
1
NMR (270 MHz, CDCl3) δ 1.01 (1H, ddd, J ) 4.6, 6.6, 8.1),
1.33 (1H, ddd, J ) 4.8, 4.8, 9.0), 1.71-1.79 (3H, m), 2.00-
2.10 (2H, m), 2.11-2.21 (1H, m), 2.23-2.37 (2H, m), 2.41-
2.60 (2H, m), 4.31 (1H, ddd, J ) 1.5, 7.5, 11.0), 5.42-5.55 (2H,
m); 13C-NMR (90.6 MHz, CDCl3) δ 13.5 (t), 18.2 (d), 25.3 (t),
26.0 (d), 26.3 (t), 33.4 (t), 33.7 (t), 74.2 (d), 124.1 (d), 135.0 (d),
173.9 (s), 179.4 (s); MS (ESI-LOOP, ammonium acetate) m/z
447 ((2M - H])+, 99%), 283 ((M + OAc)+, 100%), 223 ((M -
H)+, 59%. An X-ray crystal structure of 34 was obtained which
confirmed the absolute stereochemistry. Suitable crystals for
X-ray analysis were grown from a solution of hexane/EtOAc
(10:1) over a two-week period.1
acetate); [R]20 -86.2 (c 1.75, CHCl3) IR (neat) 1734 cm-1; 1H-
D
NMR (270 MHz, CDCl3) δ 0.55-0.64 (2H, m), 0.94-1.04 (1H,
m), 1.14-1.25 (1H, m), 1.75 (1H, bs), 1.70-1.85 (1H, m), 2.03-
2.60 (7H, m), 3.45 (1H, dd, J ) 7.2, 11.4), 3.52 (1H, dd, J )
6.6, 11.4), 4.19 (1H, ddd, J ) 1.5, 8.4, 10.4), 5.41-5.53 (2H,
m); 13C-NMR (67.8 MHz, CDCl3) δ 8.3 (t), 19.8 (d), 20.5 (d),
25.3 (t), 26.5 (t), 33.7 (t), 33.8 (t), 66.0 (t), 76.4 (d), 124.7 (d),
134.6 (d), 174.1 (s); m/z (GC/EIMS) 282 ([M - H + TMS]+,
1%), 182 ([M - H2O]+, 5%), 110 (100), 82 (85), and m/z (FAB)
233 ([M + Na]+, 100%), 211 ([M + H]+, 71%), 193 ([M + H -
H20]+, 85%).
Th e F ollow in g Set of Exp er im en ta l d a ta w a s Com -
p iled fr om th e Min or 8S,9S,11S Cyclop r op a n e Dia ste-
r eom er 29. Ya m a gu ch i La cton iza tion To Give 33.
A
similar procedure to that already described in the preparation
of 32 was used. A crude sample of 8S,9S,11S hydroxy acid
was prepared by basic hydrolysis of 31 (2.0 equiv of LiOH,
THF:H2O:MeOH (4:1:1), room temperature, 16 h, quantitative
crude yield). Then to this crude sample of 8S,9S,11S hydroxy
acid (461 mg, 1.55 mmol) in THF (10 mL), under an argon
atmosphere, were added Et3N (0.34 mL, 2.48 mmol) and 2,4,6-
trichlorobenzoyl chloride (0.27 mL, 1.70 mmol) at rt and the
resulting mixture was stirred for 2.5 h. The precipitate was
then removed by filtration, and the filtrate was diluted with
toluene (50 mL). This dilute solution was added (using a
syringe pump) to excess 4-(dimethylamino)pyridine (2.83 g,
23.2 mmol) in toluene (100 mL) at reflux over 3 h. After the
Ald eh yd e 14. To (8S,9R,11R) alcohol 36 (60.0 mg, 0.29
mmol) in CH2Cl2 (3 mL, containing powdered molecular sieves)
under an argon atmosphere at rt was added NMO (50.2 mg,
0.43 mmol) followed by TPAP (5.0 mg, 0.014 mmol). The
mixture was stirred at rt for 20 min and then filtered directly
down a silica column eluting with petroleum ether/EtOAc (2:
1). The product 14 (59 mg, quantitative yield) was isolated
as a colorless oil. Rf 0.7 (10:3, petroleum ether/ethyl acetate);
IR (neat) 1730 (CO + CHO) cm-1; 1H-NMR (270 MHz, CDCl3)
δ 1.12 (1H, ddd, J ) 4.9, 6.4, 8.4), 1.36-1.43 (1H, m), 1.69-

6652 J . Org. Chem., Vol. 62, No. 19, 1997
Critcher et al.
1.85 (2H, m), 1.98-2.58 (8H, m), 4.35 (1H, ddd, J ) 1.7, 7.9,
11.0), 5.41-5.54 (2H, m), 9.24 (1H, d, J ) 4.6). Aldehyde 14
was used directly in the next step.
stirred at rt for 35 min. Then an equal volume of saturated
aqueous NH4Cl and Et2O were added. The two layers were
separated, and the aqueous phase was extracted with Et2O.
The combined organic extracts were then washed with aqueous
Na2S2O3 and dried over anhydrous MgSO4. Removal of the
solvent under reduced pressure gave a residue which was
purified by column chromatography on silica eluting with neat
petroleum ether. 1H-NMR analysis of the purified material
revealed that we had isolated a 1:1 mixture of starting
material 38 and product 40. Therefore this 1:1 mixture of
compounds were redissolved in CH2Cl2 (4 mL) and treated
sequentially with Cp2ZrHCl (123 mg, 0.42 mmol), with stirring
at rt for 45 min, followed by the addition of iodide (107 mg,
0.42 mmol). After 30 min the reaction was worked up as
previously described and the product 40 (181 mg, 77%) isolated
pure as a colorless oil. 1H-NMR (270 MHz, CDCl3) δ 0.87 (3H,
t, J ) 6.8), 1.21-1.48 (8H, m), 3.70 (1H, q, J ) 7.3), 3.80 (3H,
s), 4.28 (1H, d, J AB ) 11.4), 4.52 (1H, d, J AB ) 11.4), 6.26
(1H, d, J ) 14.5,), 6.46 (1H, dd, J ) 7.7, 14.5), 6.87 (2H, d, J
) 8.8, aryl), 7.52 (2H, d, J ) 8.8). This compound was used
directly after its preparation and was therefore not fully
characterized.
(R)-3-O-((ter t-Bu tyld ip h en ylsilyl)oxy)-1-octyn -3-ol (37).
To (R)-(+)-octyn-3-ol (0.58 mL, 3.96 mmol) in DMF (10 mL)
under an argon atmosphere at rt was added imidazole (0.67
g, 9.90 mmol). Once the imidazole had dissolved, TBSCl (1.24
mL, 4.75 mmol) was added slowly over a 5 min period. The
reaction mixture was then stirred at rt for 3 h and then
quenched with ice. The two layers were separated, and the
aqueous layer was extracted with Et2O. The combined organic
layers were dried over anhydrous MgSO4, and the solvent was
evaporated. The residue was purified by column chromatog-
raphy on silica eluting with EtOAc/petroleum ether (1:10). The
product 37 (1.29 g, 89%) was a colorless oil. Rf 0.6 (neat
petroleum ether). Anal. Calcd for C24H32OSi: C, 79.1; H, 8.80.
Found C, 78.9; H, 9.00. [R]20 +39.2 (c 0.66 in CHCl3); IR
D
(neat) 2361 cm-1; 1H-NMR (270 MHz, CDCl3) δ 0.85 (3H, t, J
) 6.6), 1.09 (9H, s), 1.17-1.70 (8H, m), 2.31 (1H, d, J ) 2.01),
4.32 (1H, dt, J ) 2.2, 6.2), 7.35-7.47 (6H, m), 7.68-7.77 (4H,
m); MS (FAB) m/z 365 ((M + H)+, 13%), 319 (15), 308 ((M + H
t
- Bu)+, 85%), 237 (41), 207 (100), 199 (92).
(R,E)-((ter t-Bu t yld ip h en ylsilyl)oxy)-1-iod o-oct en -3-ol
(39). To 37 (255 mg, 0.70 mmol) in degassed THF (5 mL) at
rt under an argon atmosphere was added bis(cylopentadi-
enyl)zirconium chloride hydride (266 mg, 0.91 mmol) rapidly.
Once a homogeneous solution had formed, we analyzed the
reaction by TLC which indicated that all the starting material
had been consumed. Therefore iodine (212 mg, 0.84 mmol)
was added, and the solution was stirred at rt for 30 min. Then
an equal volume of saturated aqueous NH4Cl and Et2O were
added, the two layers were separated, and the aqueous phase
was extracted with Et2O. The combined organic extracts were
dried over anhydrous MgSO4, and the solvent was removed
at reduced pressure. The residue was then purified by column
chromatography on silica, eluting with neat petroleum ether.
The product 39 (322 mg, 94%) was a colorless oil. Rf 0.8 (neat
petroleum ether). Anal. Calcd for C24H33OSiI: C, 58.5; H,
(2R,4Z)-2-O-((p -Meth oxyp h en yl)m eth yl)-4-h ep ten e-1,2-
d iol (41). Propyltriphenylphosphonium bromide (2.82 g, 7.32
mmol) in toluene (10 mL) was treated with NaHMDS (7.14
mL, 7.14 mmol, 1 M solution in THF) at 0 °C under an argon
atmosphere. After 10 min at 0 °C and then 30 min at rt, the
red-orange mixture was cooled to -78 °C, and 22 (800 mg,
3.57 mmol) dissolved in toluene (5 mL) was added slowly via
cannula. After stirring at -78 °C for 5 min and 0 °C for 30
min the reaction mixture was quenched with saturated aque-
ous NH4Cl. The two layers were separated, and the aqueous
phase was extracted with EtOAc. The organic layers were
dried over anhydrous MgSO4, and the solvent was evaporated.
The residue was purified by column chromatography on silica,
eluting with EtOAc/petroleum ether (1:4). The desired product
41 (676 mg, 76%) and trimethylsilyl-protected derivative (126
mg, 11%) were both isolated as colorless oils. Data for 41: Rf
0.5 (4:1, petroleum ether/EtOAc); [R]24D -23.3 (c 0.22, CHCl3);
6.70. Found C, 58.4; H, 6.9. [R]20 +79.8 (c 0.48 in CHCl3);
D
IR (neat) 1607 cm-1; 1H-NMR (270 MHz, CDCl3) δ 0.83 (3H, t,
J ) 6.6), 1.06 (9H, s), 1.03-1.47 (8H, m), 4.06 (1H, qm, J )
6.6), 5.90 (1H, dd, J ) 1.0, 14.5), 6.45 (1H, dd, J ) 6.8, 14.5),
7.34-7.46 (6H, m), 7.60-7.67 (4H, m); MS (FAB) m/z 492
([M]+, 0.5%), 491 ((M - H)+, 2.5%), 435 (26), 415 (16), 365 ((M
- 1)+, 8%), 309 (40), 199 ((Ph2SiOH)+, 100%).
IR (neat) 1612, cm-1; H-NMR (270 MHz, CDCl3) d 0.96 (3H,
1
t, J ) 7.5), 2.02-2.12 (2H, m), 2.20-2.45 (2H, m), 3.49-3.67
(3H, m), 3.80 (3H, s), 4.46 (1H, d, J AB ) 11.2), 4.62 (1H, d, J AB
), 11.2), 5.29-5.53 (2H, m), 6.88 (2H, d, J ) 8.6), 7.28 (2H, d,
J ) 8.6); 13C-NMR (67.8 MHz, CDCl3) d 14.1 (q), 20.6 (t), 28.6
(t), 55.3 (q), 64.2 (t), 71.2 (t), 79.3 (d), 113.9 (d), 123.7 (d), 129.4
(d), 130.4 (s), 134.2 (d), 159.3 (s); MS (CI, NH3) m/z 268 ((M +
NH4)+, 6%), 138 (22), 121 (100); HRMS: 268.1913 (calcd for
C15H26O3N (M + NH4)+ 268.1913). Data for silylated deriva-
tive: 1H-NMR (270 MHz, CDCl3) δ 0.11 (9H, s), 0.95 (3H, t, J
) 7.5), 2.04 (2H, pentet, J ) 7.3), 2.20-2.40 (2H, m), 3.40-
3.50 (1H, m), 3.55-3.63 (2H, m), 4.53 (1H, d, J AB ) 11.2), 4.58
(1H, d, J AB ) 11.2), 5.32-5.52 (2H, m), 6.86 (2H, d, J ) 8.8),
7.27 (2H, d, J ) 8.8). [Note that by stirring the silylated
derivative in a mixture of tetrahydrofuran and dilute mineral
acid it could be quantitatively converted to 41].
(3R)-3-O-((p -Met h oxyp h en yl)m et h yl)oxy)-1-oct yn -3-
ol (38). NaH (190 mg, 4.75 mmol, 60% dispersion in mineral
oil) was washed twice with petroleum ether in order to remove
the mineral oil. The remaining NaH was then dried at reduced
pressure prior to the reaction. To the solid NaH was added
DMF (8 mL) followed by the slow addition of (R)-(+)-1-octyn-
3-ol (0.58 mL, 3.96 mmol) at 0 °C under an argon atmosphere.
The reaction was left at 0 °C until the effervescence had ceased
(ca. 30 min), and then tetrabutylammonium iodide (73.0 mg,
0.20 mmol) and 4-methoxybenzyl chloride (0.59 mL, 4.36
mmol, freshly distilled) were added sequentially at rt. The
resultant mixture was left at rt for 2 h and then quenched
using saturated aqueous NH4Cl. The aqueous layer was
repeatedly extracted with Et2O, and the combined organic
layers were dried over MgSO4. Evaporation of the solvent
under reduced pressure gave a crude residue which was
purified by column chromatography on silica eluting with
petroleum ether/EtOAc (10:1). The product 38 (543 mg, 56%)
(2R,4Z)-2-O-(((p-Meth oxyp h en yl)m eth yl)oxy)-4-h ep te-
n a l (42). A solution of oxalyl chloride (62.8 mL, 0.72 mmol)
in CH2Cl2 (1.5 mL) was cooled to -78 °C under an argon
atmosphere, and a solution of DMSO (0.1 mL, 1.44 mmol) in
CH2Cl2 (2 mL) was added over a 5-min period via syringe.
Stirring was continued at -78 °C for 10 min followed by
addition of 41 (150 mg, 0.60 mmol) in CH2Cl2 (2 mL) over ca.
10 min. The reaction mixture was stirred for 20 min, and Et3N
(0.42 mL, 3.0 mmol) was added slowly with stirring at -78
°C. The cooling bath was removed, and at rt water (4 mL)
was added. Stirring was continued for 5 min, and then the
layers were separated. The aqueous phase was extracted with
CH2Cl2, and the combined organic layers were dried over
anhydrous MgSO4, and the solvent was evaporated to give
crude aldehyde 42 (188 mg, greater than quantitative yield)
as a pale yellow oil. Crude aldehyde 42 was either used
directly or was purified by column chromatography eluting
with petroleum ether/ethyl acetate (5:1). Aldehyde 42 (144
mg, 96%) was now isolated as a colorless oil. Rf 0.6 (5:1,
was obtained as a colorless oil. Rf 0.8 (10:1 petroleum ether/
1
ethyl acetate); [R]18 +102.4 (c 1.27 in CHCl3); H-NMR (270
D
MHz, CDCl3) δ 0.83 (3H, t, J ) 7.0), 1.18-1.30 (4H, m), 1.37-
1.42 (2H, m), 1.64-1.74 (2H, m), 2.41 (1H, d, J ) 2.0), 3.76
(3H, s,), 3.99 (1H, dt, J ) 2.0, 6.6), 4.39 (1H, d, J AB ) 11.4),
4.69 (1H, d, J AB ) 11.4, 6.83 (2H, d, J ) 8.8), 7.52 (2H, d, J )
8.8); MS (GC/EI) m/z 246 ((M)+, 8%), 121 ((CH3OC6H4CH2)+,
100%).
(R,E)-(((p -Met h oxyp h en yl)m et h yl)oxy)-1-iod ooct en -
3-ol (40). To 38 (161 mg, 0.58 mmol) in THF (5 mL) at 0 °C
under an argon atmosphere was added Cp2ZrHCl (0.24 mg,
0.81 mmol) rapidly. Once a homogeneous solution had formed,
iodine (177 mg, 0.70 mmol) was added, and the solution was
1
petroleum ether/ethyl acetate); IR (neat) 1727 (s), cm-1; H-
NMR (270 MHz, CDCl3) δ 0.95 (3H, t, J ) 7.5), 2.00-2.20 (2H,

Total Synthesis of Halicholactone and Neohalicholactone
J . Org. Chem., Vol. 62, No. 19, 1997 6653
m), 2.42-2.50 (2H, m), 3.75 (1H, dt, J ) 2.2, 6.6), 3.81 (3H, s),
4.53 (1H, d, J AB ) 11.5), 4.59 (1H, d, J AB ) 11.5), 5.31-5.41
(1H, m), 5.48-5.57 (1H, m), 6.88 (2H, d, J ) 8.6), 7.27 (2H, d,
J ) 8.6), 9.61 (1H, d, J ) 2.2). No further data was obtained
since the aldehyde appeared to be quite unstable and was
therefore used directly in subsequent steps.
(s); MS (FAB) m/z 267 ((M + Na)+, 1.5%), 244 ((M)+, 3%), 243
((M)+, 4%), 121 (100). HRMS m/z 243.1387 (calcd for C16H19O2
(M - H)+ 243.1385. Data for 2,5-disubstituted 3,4-dihydro-
furan derivative 46. Rf 0.5 (15:1, petroleum ether/EtOAc); 1H-
NMR (270 MHz, CDCl3) δ 0.93 (3H, t, J ) 7.5), 2.03 (2H,
pentet, J ) 7.5), 2.30-2.52 (2H, m), 3.79 (3H, s), 4.81-4.90
(1H, m), 5.35-5.55 (2H, m), 5.69-5.72 (1H, m), 5.83 (1H, ddd,
J ) 2.2, 2.2, 6.1), 5.93 (1H, ddd, J ) 1.3, 2.4, 6.1), 6.86 (2H, d,
J ) 8.8, 7.24 (2H, d, J ) 8.8); MS (EI) m/z 244 ([M]+, 3%), 175
([M - C5H9]+, 90%), 121 ([MeOC6H4CH2]+, 100%).
(3R)-(1E,5Z)-3-O-(((p -Met h oxyp h en yl)m et h yl)oxy)-1-
iod o-1,5-octa d ien e (43). To a stirred suspension of CrCl2
(357 mg, 2.90 mmol) in THF (4 mL) under an argon atmo-
sphere was added dropwise a solution of pure 42 (120 mg, 0.48
mmol) and CHI3 in THF (3 mL). The mixture was stirred for
1 h at 0 °C and then left at rt for 16 h. Water (4 mL) was
then added followed by EtOAc (10 mL). The two layers were
separated, and the aqueous layer was extracted with EtOAc
(3 × 10 mL). The combined organic extracts were dried over
anhydrous MgSO4, and the solvent was evaporated. The
residue was purified by column chromatography on silica
eluting with petroleum ether/EtOAc (2:1). The product 43 (20
mg, 11%) was isolated as a colorless oil. Rf 0.5 (10:1, petroleum
(3R,5Z)-3-O-((ter t-Bu tyldiph en ylsilyl)oxy)-1,1-dibr om o-
1,5-octa d ien e (49). To a solution of 45 (309 mg, 0.77 mmol)
in CH2Cl2 (6 mL) and water (0.3 mL) was added DDQ (191
mg, 0.84 mmol) in one portion. The reaction was stirred at rt
for 30 min, and then the volume of solvent was reduced by
evaporation under reduced pressure. The concentrated mix-
ture was placed directly onto a column containing a plug of
silica eluting with petroleum ether/EtOAc (5:1). The desired
hydroxyl intermediate (7.65 mmol) was isolated contaminated
with 4-methoxybenzaldehyde. The crude hydroxyl intermedi-
ate was then reprotected. To the aforementioned material
dissolved in DMF (4 mL) was added imidazole (125 mg, 1.84
mmol) followed by TBDMSCl (0.24 mL, 0.92 mmol) at rt under
an argon atmosphere. After stirring the mixture at rt for 16
h, water (5 mL) was added. The aqueous phase was extracted
with Et2O, and the combined organic layers were dried over
anhydrous MgSO4. The solvent was removed by evaporation
under reduced pressure and the residue purified by column
chromatography eluting with neat petroleum ether. The
product 49 (368 mg, 92%) was isolated as a colorless oil. Rf
0.7 (neat petroleum ether). Anal. Calcd for C24H30OBr2Si: C,
1
ether/EtOAc); H-NMR (270 MHz, CDCl3) δ 0.95 (3H, t, J )
7.5), 2.02 (2H, pentet, J ) 7.1), 2.23-2.40 (2H, m), 3.73 (1H,
q, J ) 7.0), 3.80 (3H, s), 4.31 (1H, d, J AB ) 11.4), 4.52 (1H, d,
J AB ) 11.4), 5.35-5.25 (1H, m), 5.52-5.41 (1H, m), 6.29 (1H,
d, J ) 14.7), 6.48 (1H, dd, J ) 7.5, 14.7), 6.88 (2H, d, J ) 8.8),
7.24 (2H, d, J ) 8.8).
(3R,5Z)-3-O-(((p -Met h oxyp h en yl)m et h yl)oxy)-1,1-d i-
br om o-1,5-octa d ien e (45). To a solution of PPh3 (2.31 g, 8.8
mmol) in CH2Cl2 (10 mL) under an argon atmosphere at rt
was added CBr4 (1.46 g, 4.4 mmol) followed by Et3N (2.47 mL,
17.6 mmol). After 15 min at rt a deep burgundy colored
solution had developed, and the mixture was cooled to -78
°C. Crude aldehyde 42 (546 mg, 2.20 mmol) dissolved in
CH2Cl2 (10 mL) was then added slowly via cannula. The
temperature was allowed to warm to 0 °C slowly over a 2 h
period and was then left at rt for 16 h. The burgundy colored
solution was poured into 100 mL of stirring petroleum ether,
and the precipitated solids were removed by suction filtration.
The filtrate was concentrated by evaporation of the solvent,
and the residue was purified by column chromatography on
silica eluting with petroleum ether/EtOAc (15:1 to 10:1). The
product 45 (650 mg, 73% from 41 was isolated as a colorless
55.2; H, 5.75. Found: C, 55.5; H, 5.86. [R]18 +10.5 (c 0.73
365
in CHCl3) IR (neat) 1620 cm-1; H-NMR (270 MHz, CDCl3) δ
1
0.90 (3H, t, J ) 7.5), 1.06 (9H, s), 1.92 (2H, pentet, J ) 7.3),
2.18-2.40 (2H, m), 4.32-4.39 (1H, m), 5.29-5.43 (2H, m), 6.40
(1H, d, J ) 8.2), 7.35-7.43 (6H, m), 7.63-7.69 (4H, m); 13C-
NMR (67.8 MHz, CDCl3) δ 14.1 (q), 19.2 (s), 20.6 (t), 26.9 (q),
34.5 (t), 73.9 (d), 89.0 (s), 122.7 (d), 127.6 (d, aryl), 129.7 (d),
133.5 (s), 133.6 (s), 134.7 (d), 135.9 (d), 141.0 (d); MS (FAB)
t
m/z 521 ([M - 1]+, 1%), 465 ([M - Bu]+, 13%), 359 (10), 239
(16), 199 (73), 135 (100).
oil. Rf 0.6 (15:1, petroleum ether/ethyl acetate); [R]22 +20.8
(3R,5Z)-3-O-((ter t-Bu t yld ip h en ylsilyl)oxy)-5-oct en -1-
yn e (48). To a solution of dibromide 49 (120 mg, 0.23 mmol)
in THF (1 mL) at -78 °C under an argon atmosphere was
added LDA (0.35 mmol), prepared by adding n-BuLi (0.14 mL,
0.35 mmol, 2.5M solution in hexanes) to diisopropylamine (48
mL, 0.35 mmol) in THF (2 mL) at 0 °C and then warming to
rt after 15 min, via cannula. The reaction was stirred at -78
°C for 30 min and then allowed to warm to 0 °C for 20 min.
After cooling to -78 °C, n-BuLi (0.20 mL, 0.51 mmol) was
added slowly, and the mixture was allowed to warm to rt. The
reaction was quenched with saturated aqueous NH4Cl (2 mL)
and EtOAc (5 mL) was added. The two layers were separated
and the aqueous layer was extracted with EtOAc. The
combined organic layers were dried over anhydrous MgSO4,
and the solvent was evaporated under reduced pressure. The
residue was purified by column chromatography on silica
eluting with neat petroleum ether. The product 48 (76 mg,
92%) was isolated as a colorless oil. Rf 0.5 (petroleum ether/
D
(c 0.98, CHCl3); IR (neat) 1614 cm-1; H-NMR (270 MHz, CDCl3)
δ 0.95 (3H, t, J ) 7.5), 2.00-2.10 (2H, m), 2.25-2.45 (2H, m),
3.79 (3H, s), 4.04-4.12 (1H, m), 4.35 (1H, d, J AB ) 11.4), 4.53
(1H, d, J AB ) 11.4), 5.29-5.55 (2H, m), 6.41 (1H, d, J ) 8.6),
6.87 (2H, d, J ) 8.6), 7.28 (2H, d, J ) 8.6); 13C-NMR (67.8
MHz, CDCl3) δ 14.1 (q), 20.7 (t), 32.1 (t), 55.2 (q), 70.6 (t), 78.7
(d), 91.1 (s), 113.7 (d), 122.8 (d), 129.4 (d), 130.1 (s), 134.6 (d),
139.7 (d), 159.2 (s); MS (FAB) m/z 406 ([M]+, 0.7%), 404 ([M]+,
1.5%), 402 ([M]+, 0.7%), 121 (100); HRMS m/z 401.9816 (calcd
for C16H20O2Br2 (M+) 401.9830; also m/z 400.9750 (calcd for
C16H19O2Br2 (M - 1)+) 400.9752).
(3R,5Z)-3-O-(((p-Meth oxyp h en yl)m eth yl)oxy)-5-octen -
1-yn e (44) (a n d Sid e P r od u ct 46). To a solution of dibro-
mide 45 (123 mg, 0.30 mmol) in THF (1 mL) at -78 °C under
an argon atmosphere was added n-BuLi (0.40 mL, 0.61 mmol,
1.5M in hexanes) dropwise. The reaction was stirred at -78
°C for 30 min and then allowed to warm to 0 °C for 5 min.
Then saturated aqueous NH4Cl (1.5 mL) and EtOAc (3 mL)
were added and the two layers separated. The aqueous layer
was extracted with EtOAc (2 × 5 mL), and the combined
organic layers were dried over anhydrous MgSO4. The solvent
was evaporated and the residue purified by column chroma-
tography on silica eluting with petroleum ether/EtOAc (10:1).
The reaction produced several products, and we isolated the
desired product 44 (53 mg, 71%) and the side product 46 (7.7
mg, 10%). Data for 44. Rf 0.6 (10:1, petroleum ether/ethyl
ethyl acetate); [R]20 +14.5 (c 1.47, CHCl3); IR (neat) 2361,
D
1651 cm-1; 1H-NMR (270 MHz, CDCl3) δ 0.88 (3H, t, J ) 7.5),
1.09 (9H, s), 1.83-2.00 (2H, m), 2.34 (1H, d, J ) 2.0), 2.39-
2.43 (2H, m), 4.32 (1H, dt, J ) 2.0, 6.4), 5.38-5.50 (2H, m),
7.35-7.44 (6H, m), 7.69-7.78 (4H, m); 13C-NMR (67.8 MHz,
CDCl3) δ 14.2 (q), 19.3 (s), 20.7 (t), 26.9 (q), 36.2 (t), 63.6 (d),
72.6 (d), 84.8 (s), 123.2 (d), 127.4 (d), 127.6 (d), 129.7 (d), 129.8
(d), 133.4 (s), 133.6 (s), 134.6 (d), 135.9 (d), 136.0 (d); MS (FAB)
m/z 363 ((M + H)+, 9%), 319 (15), 305 ((M - tBu)+, 34%), 227
(29), 207 (31), 199 (100); HRMS m/z 363.2155 (calcd for
C24H30OSi (M + H)+ 363.2144).
acetate); [R]187 +87.3 (c 0.63 in CHCl3); IR (neat) 3292 cm-1
;
D
1H-NMR (270 MHz, CDCl3) δ 0.95 (3H, t, J ) 7.5), 2.06 (2H,
pentet, J ) 7.5), 2.47 (1H, d, J ) 2.0), 2.47-2.53 (2H, m), 3.79
(3H, s), 4.05 (1H, dt, J ) 2.0, 6.8), 4.45 (1H, d, J AB ) 11.5),
4.53 (1H, d, J AB ) 11.5), 5.37-5.58 (2H, m), 6.87 (2H, d, J )
8.6), 7.28 (2H, d, J ) 8.6); 13C-NMR (67.8 MHz, CDCl3) δ 14.1
(q), 20.7 (t), 33.5 (t), 55.2 (q), 67.9 (d), 70.1 (t), 73.8 (d), 82.7
(s), 113.7 (d), 123.2 (d), 129.6 (d), 129.8 (s), 134.5 (d), 159.2
(3R)-(1E,5Z)-3-O-((ter t-Bu tyld ip h en ylsilyl)oxy)-1-iod o-
1,5-octa d ien e (50). To 48 (322 mg, 0.89 mmol) in THF (10
mL) at rt under an argon atmosphere was added Cp2ZrHCl
(312 mg, 1.07 mmol) rapidly. Once an homogeneous solution
had formed we analyzed the reaction by TLC, and all the
starting material (48) had been consumed. Therefore iodine

6654 J . Org. Chem., Vol. 62, No. 19, 1997
Critcher et al.
(271 mg, 1.07 mmol) was added, and the solution was stirred
at rt for 45 min. Then an equal volume of saturated aqueous
NH4Cl and Et2O were added. The two layers were separated,
and the aqueous phase was extracted with Et2O. The com-
bined organic extracts were dried over anhydrous MgSO4
followed by evaporation of the solvent under reduced pressure.
The residue was then purified by column chromatography on
silica eluting with neat petroleum ether. The product 50 (404
mg, 92%) was a colorless oil which still contained ca. 8%
starting material 48. Rf (neat petroleum ether) 0.8; 1H-NMR
(270 MHz, CDCl3) δ 0.86 (3H, t, J ) 7.5), 1.06 (9H, s), 1.83
(2H, pentet, J ) 7.3), 2.10-2.25 (2H, m), 4.06 (1H, q, J ) 6.6),
5.16-5.24 (1H, m), 5.31-5.42 (1H, m), 5.97 (1H, dd, J ) 1.0,
14.5), 6.48 (1H, dd, J ) 6.4, 14.5), 7.34-7.46 (6H, m), 7.61-
7.77 (4H, m). No further purification was attempted, and this
material was used directly.
organic layers were dried over anhydrous MgSO4. The solvent
evaporated at reduced pressure to give a crude residue which
was purified by column chromatography eluting with petro-
leum ether/EtOAc (3:1). The product 58 (86.4 mg, 100%) was
isolated as a 1:1 mixture of diastereomers. Rf 0.4 (10:3,
petroleum ether/EtOAc). Anal. Calcd for C20H32O3: C, 75.0;
H, 10.0. Found: C, 74.9; H, 10.1; IR (neat) 3443 (br, OH), 1612
cm-1 1H-NMR (270 MHz, CDCl3) δ 0.83-0.97 (9H, m), 1.26-
.
1.80 (9H, m), 3.73 (1H, q, J ) 6.0), 3.80 (3H, s), 3.85-3.95
(1H, m), 4.27 (0.5H, d, J AB ) 11.5), 4.30 (0.5H, d, J AB ) 11.5),
4.51 (0.5H, d, J AB ) 11.5), 4.52 (0.5H, d, J AB ) 11.5), 5.54 (0.5H,
dd, J ) 7.3, 15.8), 5.55 (0.5H, dd, J ) 7.3, 15.8), 5.65 (0.5H,
dd, J ) 6.1, 15.6), 5.67 (0.5H, dd, J ) 6.1, 15.8), 6.87 (2H, d,
J ) 8.8), 7.22-7.38 (2H, m); MS (GC:EI, CF3C[dNSi(CH3)]-
OSiMe3. CH3CN) m/z 349 ((M - H + Si(Me3)3 - (CH3)2CH)+,
2%), 319 (M - H + Si(Me3)3 - (CH3)2CH - 2 × CH3)+, 1%),
259 (3), 241 (2), 137 (8), 121 (100); MS (FAB) m/z 320 ((M)+,
0.2%), 303 ((M - OH)+, 0.4%), 289 (1), 242 (1.6), 137 (7), 121
(100).
En a n tioselective Ad d ition of Oct-1-yn e to Hep ta n a l
Usin g (S)-Dip h en yl(1-m eth ylp yr r olid in -2-yl)m eth a n ol a s
Ca ta lyst. To borane dimethyl-sulfide complex (100 mL, 1.00
mmol, 10 M solution) in degassed hexane (1 mL) was added
cyclohexene (203 mL, 2.00 mmol) dropwise at 0 °C with stirring
at this temperature for 1 h. Then oct-1-yne (147 mL, 1.00
mmol) was added at 0 °C with stirring for 30 min followed by
warming to rt for 30 min. At -78 °C Et2Zn (1.0 mL, 1.00
mmol, 1.0M solution in hexane) was added slowly, followed
by (S)-DPMPM (2.7 mg, 0.01 mmol), and the solution was
warmed to -20 °C for 10 min. Then at -20 °C, heptaldehyde
(139 µL, 1.00 mmol) in hexane (4 mL) was added slowly. After
5 min the solution was warmed to 0 °C and left at this
temperature for 2 h followed by stirring at rt overnight.
Saturated aqueous NH4Cl (4 mL) and EtOAc (7 mL) were
added, and the two layers were separated. The aqueous layer
was extracted with EtOAc, and the combined organic layers
were dried over anhydrous MgSO4. The solvent was evapo-
rated at reduced pressure and the residue purified by column
chromatography eluting with petroleum ether/EtOAc (5:1). The
product allylic alcohol (179 mg, 79%) was isolated as a colorless
(6R,4E)-6-O-(ter t-Bu t yld ip h en ylsilyl)-2-m et h yl-4-u n -
d ecen e-3,6-d iol (59) a s a 2.1:1 Mixtu r e of Dia ster eom er s
a t C-3. To a mixture of trans-iodo alkene 39 (450 mg, 0.92
mmol) and 2-methylpropanal (28 mL, 0.31 mmol) in a mixture
of DMSO and DMF (5 mL:3 mL) at rt under an argon
atmosphere was added CrCl2 (225 mg, 1.83 mmol), containing
a catalytic amount of NiCl2 (0.5 wt %, 1.0 mg, 7.71 mmol) in
one portion. The dark green solution was stirred at rt for 1 h
and then quenched by adding saturated aqueous NH4Cl (5 mL)
and CHCl3 (6 mL). The mixture was extracted with EtOAc,
and the combined organic layers were dried over anhydrous
MgSO4. The solvent was evaporated at reduced pressure to
give a crude residue which was purified by column chroma-
tography, eluting with petroleum ether/EtOAc (10:1). The
product 59 (136 mg, 100%) was isolated as a 2.1:1 mixture of
diastereomers. Rf 0.3 (10:1, petroleum ether/EtOAc); IR (neat)
3420 cm-1; major isomer, 1H-NMR (270 MHz, CDCl3) δ 0.73-
0.88 (9H, m), 1.06 (9H, s), 1.08-1.30 (6H, m), 1.38-1.64 (3H,
m), 3.66-3.73 (1H, m), 4.15-4.24 (1H, m), 5.37 (1H, ddd, J )
1.5, 6.5, 15.2), 5.54 (1H, ddd, J ) 1.5, 6.8, 15.2), 7.30-7.44
(6H, m), 7.64-7.76 (4H, m). minor isomer, 1H-NMR (270 MHz,
CDCl3) δ 0.73-0.88 (9H, m), 1.05 (9H, s), 1.08-1.30 (6H, m),
1.38-1.64 (3H, m), 3.52-3.58 (1H, m), 4.10-4.18 (1H, m), 5.37
(1H, ddd, J ) 1.0, 7.5, 15.8), 5.54 (1H, ddd, J ) 1.0, 8.3, 15.8),
7.30-7.44 (6H, m), 7.64-7.76 (4H, m); MS major isomer, (GC:
EI, CF3C[dNSi(CH3)3]OSiMe3‚CH3CN) m/z 495 ((M - H +
Si(Me3)3 - CH3)+, 1%), 467 ((M - H + Si(Me3)3 - (CH3)2CH)+,
oil. Rf 0.6 (10:1, petroleum ether/ethyl acetate); [R]18 +4.3 (c
D
0.35 in CHCl3); 1H-NMR (270 MHz, CDCl3) δ 0.90 (6H, m),
1.25-1.45 (16H, m, 1.45-1.60 (2H, m), 1.98-2.06 (2H, m), 4.03
(1H, q, J ) 6.8), 5.43 (1H, dd, J ) 7.2, 15.4), 5.62 (1H, td, J )
6.6, 15.4); m/z (CI) 225 ((M - 1)+, 30%), 209 ((M - OH)+,
100%), 141 (72)
Ra cem ic (2E)-1-O-((ter t-Bu tyld ip h en ylsilyl)oxy)-oct-2-
en e-1,4-d iol (52). To alkyne 51 (64 mg, 0.22 mmol) in CH2Cl2
(1 mL) at rt under an argon atmosphere was added Cp2ZrHCl
(67 mg, 0.23 mmol). Once a homogeneous solution had formed
the mixture was cooled to -78 °C, and Et2Zn (0.20 mL, 0.23
mmol, 1.1M in toluene) was added over 5 min. After warming
the solution to 0 °C and stirring for 10 min, hexanal (23 mg,
0.23 mmol) in CH2Cl2 (2.5 mL) was added dropwise via syringe.
The mixture was stirred at 0 °C for 1.5 h and then saturated
aqueous NH4Cl (3 mL) was added. The two layers were
separated, and the aqueous layer was extracted with Et2O.
The combined organic layers were washed with saturated
aqueous NaHCO3 and dried over anhydrous MgSO4. The
solvent was evaporated at reduced pressure, and the residue
was purified by column chromatography on silica gel eluting
with hexane/EtOAc (3:1). The product 52 (43 mg, 51%) was
isolated as a colorless oil. Rf 0.5 (4:1, petroleum ether/EtOAc);
1H-NMR (400 MHz, CDCl3) δ 0.82 (3H, t, J ) 6.7), 0.99 (9H,
s), 1.22-1.47 (8H, m), 1.35 (1H, d, J ) 4.0), 4.10-4.29 (1H,
m), 4.44-4.48 (2H, m), 5.61-5.70 (2H, m), 7.28-7.37 (6H, m),
7.59-7.79 (4H, m); MS (EI) m/z 379 ((M - OH)+, 35%), 339
t
11%), 453 (M - H + Si(Me3)3 - Bu+, 31%), 439 (4), 381 (7),
271 (100), 211 (18), 199 (95), minor isomer, (GC:EI CF3C-
[dNSi(CH3)3]OSiMe3‚CH3CN) m/z 467 ((M - H + Si(Me3)3 -
t
(CH3)2CH)+, 12%), 453 (M - H + Si(Me3)3 - Bu+, 35%), 439
(5), 381 (7), 271 (100), 211 (17), 199 (95), and 95% pure by gas
chromatography; MS (FAB) m/z 437 ((M - H)+, 1%), 421 ((M
- OH)+, 28%), 339 (6), 239 (10), 199 (56), 135 (100); HRMS
m/z 421.2918 (calcd for C28H41OSi (M - OH)+ 421.2927).
15-O-(ter t-Bu tyld ip h en ylsilyl)h a lich ola cton e (60a ) a n d
12-epi-15-O-(ter t-bu tyld ip h en ylsilyl)h a lich ola cton e (60b)
a s a 2.1 Mixtu r e of Isom er s a t C-12. To a mixture of trans-
iodo alkene 39 (155 mg, 0.32 mmol) and 8S,9R,11R-aldehyde
14 (39.4 mg, 0.19 mmol) in DMSO (3 mL) at rt under an argon
atmosphere was added CrCl2 (77.4 mg, 0.63 mmol) containing
a catalytic amount of NiCl2 (approximately 0.5 wt %, 0.5 mg,
3.87 mmol) in one portion. The dark green solution was stirred
at rt for 3 h and then quenched by adding saturated aqueous
NH4Cl (3 mL) and CHCl3 (5 mL). The mixture was then
extracted with EtOAc, and the combined organic layers were
dried over anhydrous MgSO4. After the solvent had been
evaporated under reduced pressure the crude residue was
purified by column chromatography on silica eluting with
petroleum ether/EtOAc (7:1 to 5:1) (73% crude yield). By
repeating the column chromatography process a total of three
times the two diastereomers were sufficiently separated such
that the major isomer 60a (47.0 mg, 44%) was isolated as a
colorless oil. The minor isomer 60b was still contaminated
with 60a (32.0 mg, 30%, 60b:60a 4:1). Data for 60a . Rf 0.3
t
((M - Bu)+, 10%), 267 (6), 239 (19), 221 (20), 207 (16), 199
(100).
(6R,4E)-6-O-((p -Met h oxyp h en yl)m et h yl)-2-m et h yl-4-
u n d ecen e-3,6-d iol (58) a s a 1:1 Mixtu r e of Dia ster eom er s
a t C-3. To a mixture of trans-iodo alkene 40 (181 mg, 0.45
mmol) and 2-methylpropanal (24.3 mL, 0.27 mmol) in DMSO
(2 mL) at rt under an argon atmosphere was added CrCl2 (110
mg, 0.89 mmol), containing a catalytic amount of NiCl2 (0.5
wt %, 0.5 mg, 3.86 mmol) in one portion. The dark green
solution was stirred at rt for 2 h and quenched by adding
saturated aqueous NH4Cl (3 mL) and CHCl3 (5 mL). The
mixture was then extracted with EtOAc, and the combined
(5:1, petroleum ether/ethyl acetate); [R]15 -43.7 (c 0.59 in
D
CHCl3); IR (neat) 3744, 3430 (OH), 2929, 2859, 1738 (s), 1720,

Total Synthesis of Halicholactone and Neohalicholactone
J . Org. Chem., Vol. 62, No. 19, 1997 6655
1650, 1496, 1455 cm-1; 1H-NMR (500 MHz, CDCl3) δ 0.48 (1H,
ddd, J ) 5.0, 5.0, 8.5), 0.58 (1H, ddd, J ) 5.0, 5.0, 8.5), 0.84
(3H, t, J ) 8.0), 0.88-0.98 (2H, m), 1.05 (9H, s), 1.12-1.32
(6H, m), 1.40-1.55 (2H, m), 1.72-1.80 (1H, m), 2.00-2.12 (3H,
m), 2.16-2.28 (2H, m), 2.38-2.52 (2H, m), 3.53-3.58 (1H, m),
4.15-4.25 (2H, m), 5.39 (1H, dd, J ) 5.0, 15.6), 5.43-5.48 (2H,
m), 5.64 (1H, dd, J ) 7.5, 15.6), 7.32-7.42 (6H, m), 7.65-7.69
(4H, m); 13C-NMR (125.8 MHz, CDCl3) δ 7.7, 14.0, 19.1, 19.3
(s), 22.5, 22.6, 24.3, 25.3, 26.5, 27.0 (q), 31.7, 33.6, 33.9, 37.9,
73.4, 73.8, 76.0, 124.8, 127.4, 127.5, 129.5, 129.6, 131.1, 134.0,
134.4, 134.6, 134.7, 135.9, 136.0, 174.0 (s, C1); MS (FAB) m/z
597 ((M + Na)+, 12%), 557 ((M - OH)+, 14%), 301 ((M - OH
1.55 (2H, m), 4.13 (1H, q, J ) 6.4), 4.93-5.02 (2H, m), 5.79
(1H, ddd, J ) 6.4, 10.4, 17.0), 7.31-7.44 (6H, m), 7.64-7.07
(4H, m).
Oxid a tion of Allylic Alcoh ols 60a a n d 60b in th e
P r esen ce of Ald eh yd e 14. Th e Recover y of 14. To a
mixture of 60a , 60b, and 14 (60a + 60b, 23.9 mg, 41.6 mmol;
and 14, 15.8 mg, 75.9 mmol) in CH2Cl2 (2 mL) containing
powdered 4 Å molecular sieves at rt was added solid PCC (28
mg, 128 mmol) in one portion. The dark green solution was
stirred at rt for 30 min and then the volume of solvent reduced
before the mixture was placed directly on a silica column,
eluting with petroleum ether/EtOAc (5:1). The product 61
(12.8 mg, 54%) and aldehyde 14 (13.7 mg, 87%) were isolated
separately. Data for 61. Rf 0.8 (5:1, petroleum ether/ethyl
t
- BuPh2SiOH)+, 199 (100); HRMS m/z 557.3425 (calcd for
C36H49O3Si [M - OH]+ 557.3451). Data for 60b. Rf 0.4 (5:1,
petroleum ether/EtOAc); 1H-NMR (270 MHz, CDCl3) δ 0.43-
0.48 (2H, m), 0.84 (3H, t, J ) 6.6), 0.84-0.95 (1H, m), 0.97-
1.06 (1H, m), 1.05 (9H, s), 1.12-1.60 (8H, m), 1.72-2.60 (8H,
m), 3.36-3.42 (1H, m), 4.05-4.15 (2H, m), 5.39 (1H, dd, J )
6.2, 15.6), 5.43-5.48 (2H, m), 5.51 (1H, dd, J ) 7.0, 15.6), 7.32-
7.43 (6H, m), 7.64-7.69 (4H, m).
1
acetate); H-NMR (270 MHz, CDCl3) δ 0.81 (3H, t, J ) 6.8),
0.93 (1H, ddd, J ) 4.2, 6.4, 8.4), 1.08 (9H, s,), 1.10-1.34 (7H,
m), 1.42-1.51 (3H, m), 1.71-1.86 (2H, m), 2.06-2.34 (5H, m),
2.47-2.50 (2H, m), 4.27 (1H, ddd, J ) 1.3, 8.0, 10.7), 4.35 (1H,
q, J ) 5.3), 5.46-5.50 (2H, m), 6.16 (1H, dd, J ) 1.3, 15.9),
6.73 (1H, dd, J ) 5.5, 15.8), 7.30-7.46 (6H, m), 7.56-7.68
(4H, m).
Ha lich ola cton e 2.1a To a solution of 60a (34.0 mg, 59.4
mmol) in THF (2 mL) was added TBAF (150 mL, 0.15 mmol,
1.0 M in THF), and the solution was then heated at reflux for
3 h. The volume of solvent was then reduced and the solution
directly chromatographed on silica eluting with EtOAc/
petroleum ether (10:3). The product 2 (19.6 mg, 99%) was
isolated as a colorless oil. Rf 0.4 (10:3, EtOAc/petroleum ether);
[R]15D -91.7 (c 0.29 in CHCl3); IR (neat) 3408 (br), 2930, 2858,
Desilyla tion of En on e 61. To a solution of 61 (12.8 mg,
22.4 mmol) in THF (0.5 mL) was added TBAF (69 µL, 69 mmol,
1.0 M solution in THF) at rt. After 1 h the solution was
concentrated and then placed directly onto a silica column
eluting with petroleum ether/EtOAc (3:1). The product 62 (6.2
mg, 81%) was isolated as a colorless oil. Rf 0.4 (3:1, petroleum
1
ether/EtOAc); H-NMR (270 MHz, CDCl3) δ 0.81 (3H, t, J )
1
1738 (s,), 1716, 1652 cm-1; H-NMR (400 MHz, C6D6) δ 0.34
6.8), 0.93 (1H, ddd, J ) 4.2, 6.3, 8.3), 1.15-1.80 (12H, m), 1.97-
2.30 (6H, m), 2.37-2.50 (2H, m), 4.20 (1H, ddd, J ) 1.3, 7.7,
10.7), 4.22-4.35 (1H, m), 5.38-5.42 (2H, m,), 6.36 (1H, dd, J
) 1.5, 15.8), 6.79 (1H, dd, J ) 5.0, 15.8).
(1H, ddd, J ) 5.2, 5.2, 8.9), 0.53 (1H, ddd, J ) 5.0, 5.0, 8.9),
0.93 (3H, t, J ) 7.0), 0.89-0.96 (1H, m), 1.08-1.15 (1H, m),
1.30-1.69 (12H, m), 1.76-1.84 (1H, m, CHH of lactone), 1.94
(1H, ddd, J ) 1.2, 6.1, 12.2), 2.12-2.19 (2H, m), 2.36-2.48
(2H, m), 3.60 (1H, dd, J ) 3.5, 7.0), 3.98-4.04 (1H, m), 4.35-
4.40 (1H, ddd, J ) 1.2, 8.0, 12.0), 5.38-5.48 (2H, m), 5.71-
5.81 (2H, m); also (500 MHz, CDCl3) δ 0.60 (1H, ddd, J ) 5.1,
5.1, 8.5), 0.71 (1H, ddd, J ) 5.2, 5.2, 8.8), 0.89 (3H, t, J ) 6.9),
1.00-1.06 (1H, m), 1.08-1.13 (1H, m), 1.30-1.45 (6H, m,),
1.47-1.65 (4H, m), 1.94 (1H, dddd, J ) 5.9, 12.5, 12.5, 12.5),
2.03-2.10 (2H, m), 2.13-2.16 (1H, m), 2.22-2.32 (2H, m),
2.43-2.52 (2H, m), 3.70 (1H, dd, J ) 4.2, 7.4), 4.11 (1H, q, J
) 6.2), 4.22 (1H, ddd, J ) 1.5, 8.3, 10.6), 5.45-5.50 (2H, m),
5.72-5.80 (2H, m); 13C-NMR (125.8 MHz, CDCl3) δ 8.2, 14.0,
19.5, 22.6, 23.4, 25.0, 25.3, 26.5, 31.7, 33.6, 33.8, 37.2, 72.2,
74.1, 76.1, 124.7, 131.7, 134.0, 134.6, 174.0; MS (FAB) m/z 359
((M + Na)+, 33%), 337 ((M + H)+, 3%) 319 ((M - OH)+, 100%),
301 ((M - (OH + H2O))+, 15%), 149 (28); HRMS 319.2276
(calcd for C20H31O3 (M - OH)+ 319.2273).
Lu ch e Red u ction of En on e 62. To enone 62 (6.2 mg, 18.6
mmol) and CeCl3‚(H2O)7 (7.0 mg, 18.6 mmol) in MeOH (1 mL)
at rt was added NaBH4 (0.7 mg, 18.6 mmol) in one portion.
After 5 min saturated aqueous NH4Cl (1 mL) was added, and
the solution was acidified with 2 N HCl. The aqueous layer
was extracted with Et2O, and the combined organic layers were
dried over anhydrous MgSO4. The solvent was evaporated at
reduced pressure and the residue purified by column chroma-
tography on silica eluting with EtOAc/petroleum ether (10:3).
Compounds 63 and 2 (2.5 mg, 40%) were isolated together as
a 1:1 diastereomeric mixture. (We also observed some unre-
acted starting material by thin layer chromatography which
accounts for the low yield in this reaction). The two products
(2 and 63) had identical Rf values. By subtracting the signals
we had observed in the pure spectrum of 2 from the spectrum
we now had for the 1:1 mixture of 2 and 63, the unique and
distinguishing signals belonging to 63 could be determined and
were as follows. 1H-NMR (400 MHz, C6D6) δ 0.30 (1H, m),
0.40 (1H, ddd, J ) 4.7, 4.7, 8.6) and 3.47 (1H, dd, J ) 2.0,
7.0). The corresponding values for 2 in the mixed spectrum:
δ 0.31 (1H, ddd, J ) 5.2, 5.2, 8.9), 0.49 (1H, ddd, J ) 5.0, 5.0,
8.9), 3.57 (1H, dd, J ) 3.5, 7.0, C(12)). The latter values are
in better agreement with the published data for H. okadai
derived halicholactone: δ 0.29 (1H, ddd, J ) 5.0, 5.0, 8.0), 0.47
(1H, ddd, J ) 5.0, 5.0, 8.0), 3.53 (1H, dd, J ) 4.0, 7.0).
15-O-(ter t-Bu tyld ip h en ylsilyl)n eoh a lich ola cton e (64a )
an d C12-epi-15-O-(ter t-Bu tyldipen ylsilyl)n eoh alich olacton e
(64b) a s A 2.1 Mixtu r e of Isom er s a t C-12. A mixture of
trans-iodo alkene 50 (274 mg, 0.56 mmol) and 8S,9R,11R-
aldehyde 14 (58.0 mg, 279 mmol) in DMSO (2.5 mL) and DMF
(2.5 mL) at rt under an argon atmosphere was added CrCl2
(205 mg, 1.67 mmol) containing a catalytic amount of NiCl2
(approximately 0.5 wt %, 1.0 mg, 7.74 mmol) in one portion.
The dark green solution was stirred at rt for 3 h and then
quenched by adding saturated aqueous NH4Cl (5 mL) and
CHCl3 (8 mL). The mixture was extracted with EtOAc and
the combined organic layers were dried over anhydrous
MgSO4. After the solvent had been evaporated under reduced
pressure, the crude residue was purified on silica eluting with
petroleum ether/EtOAc (5:1). The products 64a and 64b (97
mg, 61%) were isolated together as a 2:1 mixture. Then 75
mg (of the total 97 mg) of the product mixture was rechro-
matographed, eluting with petroleum ether/EtOAc (5:1) in an
Alter n a tive Ap p r oa ch to th e P r ep a r a tion of 15-O-(ter t-
Butyldiphenylsilyl)halicholactone (60a) and Its C12 Isomer
(60b). Addition of the Vinyllithium Derivative of 39 to
Aldehyde 14. To trans-iodo alkene 39 (90.2 mg, 183 mmol) in
THF (2 mL) at -78 °C under an argon atmosphere was added
tert-BuLi (204 mL, 326 mmol, 1.6 M in pentanes) dropwise.
The solution was stirred at -78 °C for 30 min, and then
aldehyde 14 (31.8 mg, 153 mmol) in THF (3 mL) was added
via cannula over a 10 min period. After stirring at -78 °C
for 30 min the solution was warmed to 0 °C for 5 min.
Saturated aqueous NH4Cl (3 mL) and EtOAc (5 mL) were then
added and the two layers separated. The aqueous layer was
extracted with EtOAc, and the combined organic layers were
dried over MgSO4. The solvent was evaporated at reduced
pressure and the residue purified by column chromatography
eluting with petroleum ether/EtOAc (5:1). The two isomers
60a and 60b were separated from each other but not from
unreacted aldehyde 14. The major isomer 60a and 14 (23.3
mg in total; 18.2 mg 60a and 5.1 mg 14) were isolated together,
while the minor isomer 60b and 14 (16.4 mg in total; 5.7 mg
of 60b and 10.7 mg of 14) were also isolated together. In
addition the alkene resulting from deiodination (38.0 mg, 57%)
was recovered. Therefore the addition was diastereoselective,
providing a 3:1 ratio of 60a :60b (23.9 mg, 28%) which were
contaminated with 14 (15.8 mg, 50%). Data for 60a , 60b, and
14 has already been given. Data for alkene. Rf 0.9 (10:1,
1
petroleum ether/ethyl acetate); H-NMR (270 MHz, CDCl3) δ
0.82 (3H, t, J ) 7.0), 1.07 (9H, s), 1.07-1.30 (6H, m), 1.36-

6656 J . Org. Chem., Vol. 62, No. 19, 1997
Critcher et al.
effort to separate the two diastereomers. The two diastereo-
mers were sufficiently separated such that the major isomer
64a (44.3 mg) was isolated pure as a colorless oil. The minor
isomer 64b (30.0 mg) was still contaminated with 64a . Data
Neoh a lich ola cton e (1) a n d C12-epi-Neoh a lich ola cton e
(65) fr om th e 2:1 Mixtu r e of 80a a n d 64b. To a solution of
64a and 64b (2:1 mixture, 21.9 mg, 39.5 mmol) in THF (2 mL)
was added TBAF (120 mL, 120 mmol, 1.0 M in THF), and the
solution was then heated at reflux for 2 h. The volume of
solvent was then reduced and the solution directly chromato-
graphed on silica eluting with petroleum ether/EtOAc (10:3).
The products 1 and 65 (11.9 mg, 90%) were isolated together
in the 2:1 ratio. Rf 0.4 (10:3, EtOAc/petroleum ether). By
subtracting the signals we had observed in the pure spectrum
of 1 from the spectrum we then acquired from a 2:1 mixture
of 1 and 65, the unique and distinguishing signals belonging
to 65 were as follows: 1H-NMR (400 MHz, C6D6) δ 0.31 (1H,
m), 0.42 (1H, ddd, J ) 4.9, 4.9, 8.6) and 3.50 (1H, dd, J ) 4.9,
7.0). The corresponding values for 1 in the mixed spectrum:
δ 0.31 (1H, ddd, J ) 5.2, 5.2, 8.6), 0.51 (1H, ddd, J ) 5.2, 5.2,
8.9), 3.57 (1H, dd, J ) 3.5, 7.0). The latter values are in better
agreement with the published data for H. okadai derived
neohalicholactone. δ 0.27 (1H, ddd, J ) 5.0, 5.0, 8.5), 0.45 (1H,
ddd, J ) 5.0, 5.0, 8.5), 3.52 (1H, m).
Oxid a tion of Un d esir ed Isom er 64b to En on e 66. To
the C12-isomer 64b (30.0 mg, 52.4 mmol, containing a small
amount of 64a ) and MMO (9.5 mg, 81.2 mmol) in CH2Cl2 (1
mL) containing powdered 4 Å molecular sieves was added
TPAP (1.0 mg, 2.7 mmol) at rt. After 30 min the mixture was
added directly to the top of a silica column eluting with
petroleum ether/EtOAc (5:1). The product 66 (27.3 mg, 91%)
was isolated as a colorless oil. Rf (5:1, petroleum ether 60:80/
ethyl acetate) 0.8; 1H-NMR (270 MHz, CDCl3) δ 0.83 (3H, t, J
) 7.3), 0.84-0.96 (2H, m), 1.09 (9H, s,), 1.25-1.33 (2H, m),
1.62-2.38 (10H, m), 2.40-2.60 (2H, m), 4.26 (1H, ddd, J )
1.3, 8.0, 11.3), 4.32-4.42 (1H, m,), 5.16-5.32 (1H, m), 5.35-
5.50 (3H, m), 6.20 (1H, dd, J ) 1.3, 15.9), 6.75 (1H, dd, J )
5.3, 15.9), 7.31-7.47 (6H, m), 7.60-7.70 (4H, m).
Lu ch e Red u ction of En on e 66. To enone 66 (27.3 mg,
47.9 mmol) and CrCl2(H2O)7 (17.8 mg, 47.9 mmol) in MeOH
(2 mL) at rt was added NaBH4 (1.8 mg, 47.9 mmol) in one
portion. After 5 min saturated aqueous NH4Cl (2 mL) was
added, and the solution was acidified with 2 N HCl. The
aqueous layer was extracted with EtOAc, and the combined
organic layers were dried over anhydrous MgSO4. The solvent
was evaporated under reduced pressure, and the residue
purified by column chromatography on silica eluting with
petroleum ether/EtOAc (10:3). Compounds 64a and 64b (15.7
mg, 58%) were isolated together as a 1:3 diastereomeric
mixture, where the major isomer (64b) possessed the undes-
ired 8S,9R,11R,12S,15R relative stereochemistry. We also
recovered some unreacted starting material 66 (3.3 mg, 12%).
Data for 64b. H-NMR (270 MHz, CDCl3) δ 0.44-0.49 (2H,
m), 0.88 (3H, t, J ) 7.5), 0.96-1.06 (2H, m), 1.06 (9H, s), 1.77-
1.90 (3H, m), 2.00-2.28 (7H, m), 2.40-2.48 (2H, m), 3.41 (1H,
t, J ) 6.6), 4.10-4.23 (2H, m), 5.22-5.45 (5H, m), 5.58 (1H,
ddd, J ) 0.9, 6.7, 15.8), 7.34-7.48 (6H, m), 7.64-7.70 (4H,
m); MS (FAB) m/z 595 ([M + Na]+, 7%), 555 ([M - OH]+, 30%),
457 (6), 299 (5), 253 (5), 239 (11), 199 (100); HRMS m/z
555.3282 (calcd for C36H47O3Si (M - OH)+ 555.3294). The
nonsuperimpossible signals belonging to 64a : 1H-NMR (270
MHz, CDCl3) δ 0.51-0.59 (2H, m), 3.54-3.59 (1H, m), 5.64
(1H, ddd, J ) 1.3, 6.4, 15.6).
C15-ep i-15-O-(ter t-Bu t yld ip h en ylsilyl)n eoh a lich ola c-
ton e (68b) a n d C12,C15-epi-15-O-(ter t-Bu tyld ip en ylsilyl)-
n eoh a lich ola cton e (68a ) a s a 1:1.3 Mixtu r e of Isom er s
a t C-12. To a mixture of trans-iodo alkene (S)-50 (169 mg,
0.35 mmol) and 8S,9R,11R-aldehyde 14 (36.0 mg, 173 mmol)
in DMSO (2.0 mL) at rt under an argon atmosphere was added
CrCl3 (93 mg, 0.76 mmol) containing a catalytic amount of
NiCl3 (approximately 0.5 wt %, 0.5 mg, 3.87 mmol) in one
portion. The dark green solution was stirred at rt for 16 h
and then quenched by adding saturated aqueous NH4Cl (3 mL)
and CHCl3 (4 mL). The mixture was extracted with EtOAc,
and the combined organic layers were dried over anhydrous
MgSO4. After the solvent had been evaporated under reduced
pressure, the crude residue was purified on silica eluting with
petroleum ether/EtOAc (5:1). The products 68a and 68b (69.3
mg, 70%) were isolated together as a 1.3:1 mixture. This
mixture was rechromatographed a further four times, eluting
for 64a . Rf 0.5 (5:1, petroleum ether/ethyl acetate); [R]25
D
1H-
-44.6 (c 0.50 in CHCl3); IR (neat) 1736 (s), 1650 cm-1
;
NMR (500 MHz, CDCl3) δ 0.51 (1H, ddd, J ) 5.1, 5.1, 8.5),
0.59 (1H, ddd, J ) 5.2, 5.2, 8.7), 0.84 (3H, t, J ) 7.0), 0.90-
0.98 (2H, m), 1.05 (9H, s,), 1.76 (1H, dddd, J ) 6.4, 12.1, 12.1,
12.1), 1.87 (2H, pentet, J ) 7.2), 2.01-2.12 (3H, m), 2.18-
2.32 (4H, m), 2.39-2.48 (2H, m), 3.54-3.59 (1H, m), 4.19-
4.24 (2H, m), 5.26-5.30 (1H, m), 5.36-5.41 (1H, m), 5.44-
5.48 (3H, m), 5.68 (1H, ddd, J ) 1.3, 6.4, 15.6), 7.32-7.42 (6H,
m), 7.65-7.69 (4H, m); 13C-NMR (125.8 MHz, CDCl3) δ 7.8,
14.2, 19.1, 19.3 (s), 20.6, 22.6, 25.3, 26.5, 27.0 (q), 33.6, 33.9,
35.9, 73.5 (×2), 76.0, 124.1, 124.8, 127.4, 127.5, 129.5, 129.6,
131.2, 133.5, 133.6, 134.2, 134.4, 134.6, 135.9, 136.0, 174.0 (s,
C1); MS (FAB) m/z 595 ((M + Na)+, 40%), 555 ((M - OH)+,
27%), 457 (5), 239 (10), 199 (100); HRMS m/z 555.3262 (calcd
for C36H47O3Si (M - OH)+ 555.3294). Data for 64b; Rf 0.55
(5:1, petroleum ether/ethyl acetate); The 1H-NMR spectrum
of 64a :64b (2:1) revealed that only two signals from 64b were
not coincident with the corresponding signals for 64a ; 1H-NMR
(270 MHz, CDCl3) δ 3.39-3.43 (1H, m) and 5.58 (1H, ddd, J
) 0.9, 6.7, 15.8) for 64b, versus δ 3.52-3.60 (1H, m) and 5.67
(1H, ddd, J ) 1.3, 6.7, 15.2) for 64a . A pure 1H-NMR spectrum
of 64b was not obtained at this time.
Neoh a lich ola cton e (1).1a To a solution of pure 64a (34.0
mg, 61.4 mmol) in THF (2 mL) was added TBAF (123 mL, 123
mmol, 1.0 M in THF), and the solution was then heated at
reflux for 3.5 h. After this time a small amount of starting
material 64a was still visible by TLC. However the reaction
was worked-up since we did not wish to decompose any product
1. The volume of solvent was reduced, and the solution was
directly chromatographed on silica eluting with EtOAc/
petroleum ether (10:3). The product 1 (15.2 mg, 74%) was
isolated as a colorless oil, which later solidified on leaving at
0 °C for several hours. A small amount of the product was
recrystallized from an EtOAc/hexane mixture (1:5) over 3 days.
Rf 0.4 (10:3, EtOAc/petroleum ether); [R]18 -54.6 (c 0.76 in
D
CHCl3); mp 72 °C (literature value1a 69-70 °C). By performing
a
1H-1H COSY experiment the assignment of all proton
signals was possible. 1H-NMR (500 MHz, C6D6) δ 0.31 (1H,
ddd, J ) 5.0, 5.0, 8.6, cyclopropane C(10)H), 0.50 (1H, ddd, J
) 5.1, 5.1, 8.7, cyclopropane C(10)H), 0.89-0.92 (1H, m,
cyclopropane C(11)H), 0.94 (3H, t, J ) 7.5, C(20)H3CH2), 1.02
(1H, d, J ) 4.4, C(12)OH), 1.06-1.09 (1H, m, cyclopropane
C(11)H), 1.22 (1H, d, J ) 4.2, C(15)OH), 1.55-1.64 (2H, m,
C(3)H2), 1.76-1.84 (1H, m, C(4)HH), 1.95 (1H, ddd, J ) 1.5,
7.2, 13.3, C(7)HH of lactone), 2.02 (2H, dpentet, J ) 1.3, 7.4,
C(19)H2), 2.11-2.15 (2H, m, C(2)CH2), 2.18-2.29 (2H, m,
C(16)H2), 2.37-2.43 (2H, m, C(4)HH + C(7)HH), 3.55-3.59
(1H, m, C(12)HOH), 3.99-4.04 (1H, m, C(15)HOH), 4.38 (1H,
ddd, J ) 1.5, 8.3 and 11.0, C(8)HOC(O)), 5.41-5.48 (3H, m,
lactone olefinics (C5 + C6) and cis-C(17)H), 5.52-5.55 (1H,
m, cis-C(18)H), 5.71-5.79 (2H, m, 2 × trans-C(13 + 14)H); also
(500 MHz, CDCl3) δ 0.60 (1H, ddd, J ) 5.1, 5.1, 8.4, cyclopro-
pane CH), 0.71 (1H, ddd, J ) 5.2, 5.2, 8.8, cyclopropane CH),
0.97 (3H, t, J ) 7.5, CH3CH2), 1.00-1.11 (2H, m, 2 ×
cyclopropane CH), 1.60-1.70 (2H, bs, 2 × OH), 1.94 (1H, dddd,
J ) 6.4, 11.9, 11.9, 11.9, CH of lactone), 2.03-2.10 (4H, m),
2.13-2.17 (1H, m, CHH of lactone), 2.24-2.33 (4H, m), 2.45-
2.52 (2H, m), 3.70 (1H, dd, J ) 3.8, 7.4, C(12)HOH), 4.16-
4.19 (1H, m, C(15)HOH), 4.23 (1H, ddd, J ) 1.4, 8.3, 10.6,
C(8)HOC(O)), 5.34-5.36 (1H, m, cis-CH), 5.45-5.50 (2H, m,
2 × cis-CH), 5.56-5.59 (1H, m, cis-CH), 5.75-5.80 (2H, m, 2
1
× trans-CH). By performing a H-13C COSY experiment most
of the carbon signals were assigned. 13C-NMR (125.8 MHz,
CDCl3) δ 8.2 (C10), 14.2 (C20), 19.5 (C9), 20.7 (C19), 23.4 (C11),
25.3 (C4), 26.4 (C3), 33.6 (C2), 33.8 (C7), 35.2 (C16), 71.5 (C15),
74.2 (C12), 76.1 (C8), 123.7 (C17), 124.7 (C6), 131.8 (C14 or
13), 133.2 (C13 or 14), 134.7 (C18 or C5), 135.3 (C5 or C18),
174.1 (C1); MS (FAB) m/z 357 ((M + Na)+, 40%), 317 ((M -
OH)+, 100%), 299 ((M - (OH + H2O))+, 13%), 149 (41), 119
(52), 105 (63); HRMS m/z 317.2096 (calcd for C20H29O3 (M -
OH)+ 317.2117).1a

Total Synthesis of Halicholactone and Neohalicholactone
J . Org. Chem., Vol. 62, No. 19, 1997 6657
with petroleum ether/EtOAc (5:1) in an effort to separate the
two diastereomers. The two diastereomers were sufficiently
separated such that the major isomer 68a (27.7 mg) was
isolated pure as a colorless oil. The minor isomer 68b (23.5
mg) was still contaminated with 17% of 68a . Also a 1:1
mixture of 68a :68b (15.7 mg) was isolated. Data for 68a . Rf
cyclopropane C(10)H), 0.47 (1H, ddd, J ) 5.1, 5.1, 8.8,
cyclopropane C(10)H), 0.82-0.89 (1H, m, cyclopropane C(11)H),
0.89 (3H, t, J ) 7.4, C(20)H3CH2), 1.01-1.07 (1H, m, cyclo-
propane C(11)H), 1.51-1.58 (2H, m, C(3)H2), 1.73-1.78 (1H,
m, C(4)HH), 1.90 (1H, ddd, J ) 1.4, 7.3, 13.5, C(7)HH of
lactone), 1.97 (2H, dpentet, J ) 1.2 and 7.4, C(19)H2), 2.06-
2.13 (2H, m, C(2)CH2), 2.18-2.32 (2H, m, C(16)H2-), 2.31-
2.40 (2H, m, C(4)HH + C(7)HH), 3.54 (1H, dd, J ) 2.6, 6.8,
C(12)HOH), 3.97-4.00 (1H, m, C(15)HOH), 4.34 (1H, ddd, J
) 1.5, 8.3, 12.2, C(8)HOC(O)), 5.35-5.44 (3H, m, cis-C(5)H,
cis-C(6)H, and cis-C(17)H), 5.49-5.54 (1H, m, cis-C(18)H),
5.68-5.71 (2H, m, 2 × trans-C(13 + 14)H); 13C-NMR (67.8
MHz, CDCl3) 8.2, 14.2, 19.5, 20.7, 23.3, 25.3, 26.5, 33.6, 33.8,
35.2, 71.6, 74.2, 76.1, 123.6, 124.7, 131.9, 133.3, 134.7, 135.4,
174.1 (C1), and 13C-NMR (125.8 MHz, C6D6) δ 7.7, 14.4, 19.5,
21.0, 23.8, 25.6, 26.6, 33.7, 34.1, 35.8, 71.7, 73.5, 76.0, 124.7,
125.1, 132.3, 133.4, 134.6, 134.7, 172.9 (C1); MS (FAB) m/z 357
((M + Na)+, 100%), 317 ((M - OH)+, 93%), 299 ((M - -(OH +
H2O))+, 12%), 259 (18), 165 (42), 149 (70), 123 (71), 109 (74),
105 (94); HRMS m/z 317.2129 (calcd for C20H29O3 (M - OH)+
317.2117). The peaks that distinguished 67 from 69 in the
1H-NMR spectrum at 270 MHz: 0.49-0.53 (1H, m), 0.59-0.65
(1H, m), 3.47 (1H, dd, J ) 3.7, 6.3). Our data for 67 was
consistent with that reported for L. sinclairii derived “neo-
halicholactone”.8
0.6 (5:1, petroleum ether/EtOAc); [R]19 -50.3 (c 0.94 in
D
CHCl3); IR (neat) 1737 (vs) cm-1; 1H-NMR (400 MHz, CDCl3)
δ 0.45-0.48 (2H, m), 0.87 (3H, t, J 7.3), 0.91-1.06 (2H, m),
1.06 (9H, s), 1.77 (1H, dddd, J ) 6.4, 12.5, 12.5, 12.5), 1.87
(2H, pentet, J ) 7.3), 2.07-2.35 (7H, m), 2.43-2.51 (2H, m),
3.44 (1H, t, J ) 6.4), 4.13-4.21 (2H, m), 5.26-5.48 (5H, m),
5.63 (1H, ddd, J ) 1.2, 6.4, 15.6), 7.33-7.44 (6H, m), 7.65-
7.69 (4H, m); 13C-NMR (100.4 MHz, CDCl3) δ 7.5, 14.2, 19.3,
20.3, 20.6, 23.2, 25.2, 26.5, 27.0 (q), 33.6, 33.8, 35.8, 73.5, 74.5,
76.3, 124.0, 124.7, 127.4, 129.5, 129.6, 131.2, 133.6, 134.2,
134.3, 134.6, 135.9, 136.0, 174.2; MS (FAB) m/z 595 ((M +
Na)+, 16%), ((M - 1)+, 3%), 555 ((M - OH)+, 57%), 503 (8),
457 (8), 317 (7), 299 (8), 239 (10), 199 (100); HRMS m/z
555.3313 (calcd for C36H47O3Si (M - OH)+ 555.3294). Data
for 68b (contaminated with 68b). 1H-NMR (500 MHz, CDCl3)
δ 0.52 (1H, ddd, J ) 5.2, 5.2 and 8.2), 0.60 (1H, ddd, J ) 5.0,
5.0, 8.5), 0.88 (3H, t, J ) 7.3), 0.90-0.98 (2H, m), 1.06 (9H, s,
tBu), 1.72-1.80 (1H, m), 1.87 (2H, pentet, J ) 7.3), 2.01-2.12
(3H, m), 2.18-2.35 (4H, m), 2.43-2.50 (2H, m), 3.45-3.41 (1H,
m), 4.10-4.21 (2H, m), 5.26-5.41 (2H, m), 5.43-5.48 (3H, m),
5.64 (1H, ddd, J ) 1.2, 6.4, 15.6), 7.33-7.44 (6H, m), 7.65-
7.69 (4H, m); 13C-NMR (125.8 MHz, CDCl3) δ 7.9, 14.2, 19.1,
19.3 (s), 20.6, 22.7, 25.2, 26.5, 27.0 (q), 33.5, 33.9, 35.9, 73.8,
74.1, 76.0, 124.1, 124.7, 127.4, 127.5, 129.5, 129.6, 131.4, 133.4,
133.8, 134.1, 134.3, 134.6, 135.9, 136.0, 174.0 (s, C1); MS (FAB)
m/z 595 ((M + Na)+, 29%), 555 ((M - OH)+, 89%), 503 (13),
497 (6), 457 (11), 317 (11), 299 (11), 239 (11), 199 (100). HRMS
m/z 555.3301 (calcd for C36H47O3Si (M - OH)+ 555.3294).
C15-epi-n eoh a lich ola cton e (67). To a 5:1 mixture of 68b:
68a (23.5 mg, 41.1 mmol) in THF (2 mL) was added TBAF
(103 mL, 103 mmol, 1.0M in THF), and the solution was then
heated at reflux for 5 h. After this time a small amount of
starting material was still visible by TLC. However the
reaction was worked-up after this time since we did not wish
to decompose any product. The volume of solvent was reduced,
and the solution was directly chromatographed on silica
eluting with EtOAc/petroleum ether (10:3). The product 67
(11.5 mg, 84%) was isolated as a colorless oil which still
contained 17% of the C12-isomer 69. The mixture of 67 (major)
and 69 were rechromatographed eluting with petroleum ether/
ethyl acetate (1:1). We now isolated some fractions that
contained pure 67 (4.3 mg) while many of the fractions were
still mixtures of 67 and 69 (6.0 mg). Data for pure 67. Rf 0.5
Ack n ow led gm en t. We thank the EPSRC and Astra
Charnwood for support of a CASE award (to D.J .C.),
Professor K. Yamada (Nagoya) for providing 1H- and
13C-NMR samples of 1 and 2 for comparison, and
Professor R. J . K. Taylor (York) and Dr A. B. Holmes
(Cambridge) for advice on certain aspects of the syn-
thesis. Professor W. Gerwick (Oregon) is thanked for
providing advice and copies of spectra. We also thank
Dr J . Ballantine of the EPSRC National Mass Spec-
troscopy Service (Swansea) for HRMS FAB analyses of
certain compounds. M.W. thanks Professor David A.
Evans (Harvard University) for accommodating him and
for the generous provision of computer and office facili-
ties during the preparation of this manuscript.
Su p p or tin g In for m a tion Ava ila ble: Details of experi-
mental work with O-silyl protected lactols and the preparation
of early intermediates, photocopies of 1H-NMR spectra of
compounds not supported by elemental analyses, and 500 MHz
1
and COSY H-NMR spectra of 1, 2, and 67 (49 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
(10:3, EtOAc/petroleum ether); [R]20 -84.9 (c 0.43 in CHCl3);
D
IR (neat) 1738 (vs, CO), cm-1. By performing a H-1H COSY
1
experiment the assignment of all proton signals was possible.
1H-NMR (500 MHz, C6D6) δ 0.27 (1H, ddd, J ) 4.9, 4.9, 8.6,
J O962312J