On the synthesis of cepacin A

Article abstract of DOI:10.1016/j.tet.2007.03.144

Efforts directed toward a total synthesis of cepacin A is presented in full detail. The C-7, C-8, and C-9 stereogenic centers in the target molecule were derived from d-arabinose. The configuration of the allene axis was controlled at the bromoallenation step by the C-10 configuration of the precursor. An unexpected yet very interesting phenomenon was observed with the bromoallenation, where the α-isomer of the propargylic alcohol 31 was entirely resistant to the conditions that worked so well for its β-counterpart. The problem was eventually solved by careful tuning of the size of the neighboring groups based on the clue obtained from conformational analysis. The diyne moiety was incorporated into the molecular framework through a coupling of the TMS protected diyne with a proper bromoallene under the Sonogashira conditions with EtOAc as the solvent. Use of other solvents at this step led to complete failure.

Full text of DOI:10.1016/j.tet.2007.03.144

Tetrahedron 63 (2007) 4887–4906
On the synthesis of cepacin A
*
Chao-Jun Tang and Yikang Wu
State Key Laboratory of Bio-organic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry,
Chinese Academy of Sciences, 354 Fenglin Road, Shanghai 200032, China
Received 29 January 2007; revised 12 March 2007; accepted 23 March 2007
Available online 25 March 2007
Abstract—Efforts directed toward a total synthesis of cepacin A is presented in full detail. The C-7, C-8, and C-9 stereogenic centers in the
target molecule were derived from D-arabinose. The configuration of the allene axis was controlled at the bromoallenation step by the C-10
configuration of the precursor. An unexpected yet very interesting phenomenon was observed with the bromoallenation, where the a-isomer of
the propargylic alcohol 31 was entirely resistant to the conditions that worked so well for its b-counterpart. The problem was eventually solved
by careful tuning of the size of the neighboring groups based on the clue obtained from conformational analysis. The diyne moiety was in-
corporated into the molecular framework through a coupling of the TMS protected diyne with a proper bromoallene under the Sonogashira
conditions with EtOAc as the solvent. Use of other solvents at this step led to complete failure.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
2. Results and discussions
Cepacin A (1), a potent antibacterial substance produced by
Pseudomonas cepacia SC 11,783, was isolated by Parker and
co-workers in the 1980s.1 By comparison of the UV and IR
data with those of nemotin2 (2) along with 1H NMR analysis
and chemical derivatization/degradation, the structure of
cepacin A was proposed to be 1. The absolute allene config-
uration was assigned on the basis of the optical rotation
according to the rules of Lowe and Brewster.3
Our general strategy is shown in Scheme 1. Three of the four
stereogenic centers (C-7, C-8, and C-9) were derived from
D-arabinose, while the remainder (C-10) was intended to
be created by a substrate-controlled asymmetric addition
of acetylene to aldehyde. Installation of the diyne fragment
was arranged at a late stage through a coupling reaction with
a proper bromoallene. The stereochemistry of the allene axis
was controlled by the configuration of the leaving group
at the propargylic position (C-10). The lactone part was in-
troduced through a Wittig reaction in connection with con-
struction of the trans C]C double bond.
1
4
H
12
O
9
O
C
7
1
4
H
1:1 mixture
Sonogashira
16
OH
O
O
12
O
1 (Cepacin A)
Coupling
9
O
C
H
7
16
OH
O
C
1
2 (Nemotin)
O
1
4
H
12
O
As cepacin A represents a novel type of antibacterials with
an interesting structure containing a tightly packed challeng-
ing array of diyne–allene–epoxide–allylic alcohol–lactone
functionalities, it makes a very attractive target for synthetic
studies. Here in this article we wish to detail4 our efforts to-
ward synthesis of cepacin A in a hope to confirm the natural
structure and find an entry to the chiral diyne–allene system.
9
O
Wittig
Br
C
7
OH
O
O
RO
OR''
OH
10
HO
HO
D-Arabinose
9
7
CHO
TsO
OH
OR'
Scheme 1.
As in essentially all carbohydrate chiron-based syntheses,
a major task in our endeavor is to differentiate the hydroxyl
Keywords: Antibiotics; Diynes; Allenes; Epoxides; Lactones.
* Corresponding author. E-mail: yikangwu@mail.sioc.ac.cn
0040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.03.144

4888
C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
O
groups in the starting material. An approach (Scheme 2) in
the literature5 seemingly suitable for our purpose was then
attempted.
O
O
PPh3
a
b
Cl
CO2Me
11
O
CO2Me
d
O
9
10
OBn
OH
OH
OH
O
O
O
OBn
OH
O
O
O
a
b
Br
c
O
HO
HO
HO
HO
CO2Me
CO2H
13
12
D-arabinose
3
4
c
Scheme 4. (a) (i) MeOH/reflux; (ii) SOCl2, 50% over two steps; (b)
Ph3P]CH2/ꢀ78 ꢁC to rt; (c) Br2/MeOH/rt to reflux, 34%; (d) (i) Ph3P/
PhH/reflux; (ii) Na2CO3/H2O, 59% over two steps.
OH
OTBS
O
O
OBn
OTBS
O
O
O
d
+
O
OTBS
O
either only recovered starting materials or a complicated
product mixture.
OH
(?)
O
O
6
7
5
O
O
OH
OR
O
O
Scheme 2. (a) PhCH2OH/AcCl (cat), 84%; (b) Me2CO/CuSO4 (7.0 equiv)/
concd H2SO4 (cat), 81%; (c) TBSCl (2.0 equiv)/imidazole/DMF/rt, 91%;
(d) see the text.
CO2Me
HO
O
PPh3
H
O
See
text
+
O
CO2Me
OR
Following a more recent literature procedure,6 the hemi-
acetal OH in D-arabinose was replaced with a benzyl group.
The resultant 3 was treated with acetone in the presence of
CuSO4/H2SO4 to yield the corresponding acetonide 4. Pro-
tection of the hydroxyl group with TBSCl afforded 55
smoothly. The final cleavage of the benzyl mixed acetal,
however, did not proceed so well as expected. Catalytic hy-
drogenolysis over 10% Pd/C completely failed despite re-
peated tries. Li/naphthalene7 system gave only a low yield
(25%) of the desired 6 along with some unexpected products
(seemed to be an a/b mixture of the TBS migrated product 7).
10 R = TBS
10' R = H
9
6 R = TBS; 6' R = H
Scheme 5.
As the above failures were most likely caused by the great
tendency of the aldehyde functionality to form a hemiacetal,
we decided to modify our strategy of manipulating the car-
bohydrate chiron to avoid any possibility of cyclization to
form a cyclic hemiacetal.
To this end, D-arabinose was treated sequentially with EtSH/
6 N HCl and acetone/H2SO4 to yield 11 using a literature13
procedure. The thioacetal protecting group was then re-
moved with HgCl2/HgO14 (red) to free the aldehyde group.
Further treatment with 9 afforded a,b-unsaturated ester 12
in 91% yield.
Through a more careful search of literature we then found
that Kiso and Hasegawa8 had developed a convenient proce-
dure for making 8. By using conventional silylation condi-
tions we easily obtained the desired 6 (Scheme 3).
OH
OH
OH
O
O
OH
OH
O
O
O
The ketone carbonyl group in 12 was reduced under the
Luche15 conditions (CeCl3/NaBH4), giving alcohol 13 as
a mixture of the C-4 (cepacin numbering) epimers. As the
epimers in this case were inseparable on silica gel, the mix-
ture was used as such in the next step to give lactone 14. The
terminal acetonide was then selectively hydrolyzed with
CeCl3$7H2O/(CO2H)2/MeCN16 (Scheme 6) to yield diol
15, paving the way for further elaborations.
a
b
OTBS
O
O
HO
HO
8
6
D-arabinose
Scheme 3. (a) p-TsOH (cat)/Me2C(OMe)2/DMF/rt, 69%; (b) TBSCl/imid-
azole/DMF/rt, 60%.
The Wittig reagent 9 needed for construction of the lactone
moiety was then attempted using combinations of some
known reactions (Scheme 4). We first tried to use the acid
halide 119 to react with Ph3P]CH2 as reported by Ronald
and Wheeler10 without success. Then we turned to the second
route, making 9 through reaction of 13 with PPh3.11 Although
preparation of 13 was somewhat tedious because of concur-
rent formation of over-brominated side products, enough
amounts of 9 indeed could be obtained this way.
O
EtS
O
SEt
O
O
OH
OH
O
O
O
CO2Me
b
a
O
11
HO
O
12
HO
O
f
D-arabinose
c
O
O
HO
O
H
H
O
O
CO2Me
O
O
e
HO
HO
d
With both 6 and 9 in our hands, we proceeded to examine the
Wittig reaction shown in Scheme 5. This was expected to be
facile in the beginning as similar reactions12 of 6 with
Ph3P]CHCO2Me were quite successful. However, to our
surprise neither 6 nor its desilyl analogue 60 led to the antic-
ipated product(s). We tried many sets of reaction conditions
(in CH2Cl2, DMF or toluene, at ambient or elevated temper-
atures, with or without added benzoic acid12) and always got
O
O
O
O
O
13
O
14
O
15
Scheme 6. (a) (i) EtSH/6 N HCl, 78%; (ii) Me2CO/concd H2SO4, 92%; (b)
(i) HgCl2/HgO (red)/MeCN/H2O, 80%; (ii) 9/toluene/reflux, 91%; (c)
NaBH4/CeCl3$7H2O/MeOH, 80%; (d) p-TsOH/Me2CO/reflux, 70%; (e)
CeCl3$7H2O/(CO2H)2/MeCN, 40%; (f) p-TsOH/MeOH/reflux, 64%.

C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
4889
Murugensan and Pandurangan17 reported that quinolinium
fluorochromate (QFC) could oxidize TBS ether directly
into the corresponding aldehyde. Such a procedure would
offer an expeditious access to aldehyde 17, the precursor
we needed for introducing an acetylene moiety. Hence we
prepared 16 and attempted the oxidation with QFC (Scheme
7). Unfortunately, the anticipated oxidation did not18 occur at
all, which forced us to take a round-about strategy to achieve
desired transformations—to mask the primary hydroxyl
group before protecting the secondary hydroxyl group as a
TBS ether and then to re-free the primary hydroxyl group
for oxidation.
The modified route emerged from the known 11. Interest-
ingly, under the same16 conditions as utilized in the hydroly-
sis of 14 to 15 (40% yield), the intermediate diol in this case
could be obtained in 92% yield. However, when running the
reaction on preparative scales, the solid reagents tended to
clog and thus hampered efficient stirring. The workup was
also tedious. For these reasons, the 80% AcOH/60 ꢁC19
conditions were preferred in large-scale runs.
The primary and secondary hydroxyl groups were converted
to a Piv (pivaloyl) ester and a TBS ether, respectively. The
carbonyl group was released using an Hg(II)-mediated hy-
drolysis and the intermediate aldehyde was treated with 9
to afford a,b-unsaturated ester 22. The ketone carbonyl
group was reduced with NaBH4, leading to a 1:1 mixture
of the epimers. The isomers (23a and 23b) in this case could
be separated on silica gel under carefully chosen conditions.
The subsequent steps were therefore performed using a
single epimer as the starting material (cf. Section 4).
O
O
O
H
O
H
H
b
O
H
O
a
TBSO
HO
HO
O
O
TBSO
O
O
TBSO
O
O
O
16
17
15
O
O
Lactonization was achieved by exposure of either 23a or
23b to PPTS20 (pyridinium p-toluenesulfonate) in reflux-
ing benzene. The Piv protecting group was cleaved
with DIBAL-H with concurrent reduction of the lactone
to a lactol. Subsequent oxidation of 25a or 25b with IBX21
(o-iodoxybenzoic acid) gave the corresponding aldehyde
(26a or 26b) in 78% yield. PCC22 (pyridium chlorochro-
mate) or Dess–Martin periodinane23 could also fulfill the
same task, but the yields were significantly lower.
H
O
c
H
O
PivO
d
PivO
O
HO
O
TBSO
O
O
19
18
Scheme 7. (a) TBSCl/imidazole/DMAP/DMF, 54%; (b) QFC (cf. the text);
(c) PivCl/Et3N/CH2Cl2, 58%; (d) TBSCl/imidazole/DMF, 52%.
These steps were indeed executable. However, the corre-
sponding yields were not high enough to ensure facile accu-
mulation of the intermediates required for going through the
many remaining steps along the synthetic sequence. There-
fore, instead of proceeding further, we explored an alternative
route shown in Scheme 8, where the two previously low-
yielding steps were placed at an early stage of the synthesis.
Addition of TMS protected acetylenide to the aldehyde
group in 26b was not so smooth as we expected. Despite ex-
cess amounts of the added acetylenide, the reaction did not
go to completion. And the addition appeared to be totally
non-selective, which meant complete loss of control over
the allene configuration in a later stage. We reasoned that
the sluggish reaction might result from too much steric
crowding around the aldehyde group. Presence of the lactone
functionality could also contribute. Therefore, the synthetic
plan was modified one more time, with introduction of the
acetylene moiety shifted to an earlier stage preceding incor-
poration of the lactone part.
SEt
O
SEt
O
EtS
EtS
PivO
SEt
O
EtS
PivO
HO
O
b
a
TBSO
O
O
O
O
21
20
11
c
CO2Me
d
HO
PivO
As shown in Scheme 9, sequential TES protection and reduc-
tive removal of the Piv protecting group gave alcohol 29,
which on treatment with SO3$Py24 followed by addition
of TMS/acetylenide resulted in a 1.4:1 mixture of 31a
(b-OH, the more polar component) and 31b (a-OH, the
less polar component). The two epimers were separable on
silica gel. However, because the configurations of these
compounds were unknown at that moment, we selected
31a by chance to perform the remaining steps.
O
O
O
e
PivO
CO2Me
PivO
O
TBSO
O
TBSO
O
TBSO
22
O
O
23a, 23b
O
f
24a
24b
(Separated)
O
O
HO
O
O
O
HO
g
h
H
O
TMS
TBSO
At this point, we once considered the alternative of convert-
ing the propargylic alcohol 31 into bromoallene before in-
corporation of the lactone moiety at the other end of the
array of the arabinose-derived stereogenic centers. However,
as we could not rule out the possibility of undesired sulfur
alkylation (leading to formation of sulfonium salts), to be
on the safe side we decided to build the lactone moiety first.
O
TBSO
O
O
TBSO
O
HO
O
O
25a
25b
26a
27b (from 26b)
26b
Scheme 8. (a) (i) CeCl3$7H2O/(CO2H)2/MeCN, 82%, or 80% AcOH/60 ꢁC/
2 h, 75%; (ii) PivCl/Et3N/CH2Cl2, 77%; (b) TBSCl/imidazole/DMF, 95%;
(c) (i) HgCl2/HgO (yellow)/MeCN/H2O; (ii) 9/toluene/reflux, 94% from
21; (d) NaBH4/MeOH, 71%; (e) PPTS/PhH/reflux, 100%; (f) DIBAL-H/
CH2Cl2/ꢀ78 ꢁC, 95%; (g) IBX/DMSO, 78%, or PCC/NaOAc/CH2Cl2,
50%, or Dess–Martin oxidation, 41%; (h) TMSC^CLi/THF/ꢀ78 ꢁC, 40%.
Thus, the hydroxyl group in 31a was masked as an acetate to
pave the way for the following transformations. The

4890
C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
EtS
O
SEt
O
EtS
SEt
O
EtS
HO
EtS
TESO
SEt
O
SEt
O
PivO
HO
SEt
O
EtS
O
PivO
HO
TMS
7
a
b
a
10
9
TESO
TESO
O
c
HO
O
HO
O
31b
28
29
20
36
H
O
EtS
TESO
SEt
EtS
TESO
H
SEt
TMS
(Separated)
CH(SEt)2
10
9
d
NOE
b
7
O
O
H
37
O
O
HO
O
e
O
O
O
30
31a (β-OH)/31b (α-OH) = 1.4:1
Scheme 10. (a) K2CO3 (1.1 equiv)/MeOH–THF (2:1), 94%; (b) p-TsOH
(cat)/Me2C(OMe)2, 95%.
O
MeO2C
EtS
TESO
SEt
O
TMS
f
TMS
TESO
AcO
O
propargylic position, the acetyl group must be removed first.
Unfortunately, such a simple operation turned out to be ex-
tremely difficult in the presence of the silyl protecting
groups and the lactone functionality. To get around this prob-
lem we decided to put a tosyl group there in the first place.
AcO
O
33
O
32 (From 31a)
O
g
OH
4
O
MeO2C
TESO
4
h
TMS
TMS
TESO
7
7
9
O
By then we also found out that by using the I2 conditions, the
thioacetal in 31a could be hydrolyzed satisfactorily (Scheme
11). The intermediate aldehyde was readily converted into
38 by reaction with 9. Unlike in the previous cases, the
C]C double bond formed in this reaction was no longer
trans only. Instead, a 4:1 mixture of trans/cis isomers was
formed. The isomers were separated and only the trans one
was utilized in the subsequent steps.
O
9
AcO
O
34
35
AcO
O
Scheme 9. (a) TESCl/imidazole/DMF/rt, 90%; (b) DIBAL-H/CH2Cl2/
ꢀ78 ꢁC, 91%; (c) SO3$Py/i-Pr2NEt/DMSO–CH2Cl2 (1:1)/0 ꢁC to rt, 82%;
(d) TMSC^CLi/THF/ꢀ78 ꢁC, 81%; (e) Ac2O/Et3N/CH2Cl2/rt, 82%; (f)
(i) I2/NaHCO3/Me2CO–H2O (5:1)/0 ꢁC, 79%; (ii) 9/toluene/reflux, 83%;
(g) BH3/(S)-2-methyl-CBS-oxazaborolidine/0 ꢁC, 61%; (h) PPTS/PhH/
40 ꢁC, 59%.
O
MeO2C
TESO
EtS
TESO
SEt
O
subsequent hydrolysis of the thioacetal was not as smooth
as observed with other substrates. Many conventional
conditions such as HgCl2/HgO, NCS/AgNO3,25 NBS,26
PhI(AcO)2,27 and PhI(TFA)228 all failed to give the desired
TMS
a
TMS
38
O
HO
O
31a
HO
O
29
b
aldehyde. Finally, the deprotection was realized using I2
in the presence of powdered NaHCO3. Once the deprotection
problem was solved, the chain extension from the aldehyde
group was readily achieved by treatment with 9.
O
MeO2C
TESO
O
MeO2C
c
TMS
TESO
TsO
O
O
Reduction of the ketone group in this case was performed
under the CBS30 conditions. The configuration of the major
isomer at the C-4 was not experimentally determined yet, but
by comparison with similar substrates in the literature it is
expected to be as drawn in the scheme.31 The alcohol 34
was then readily transformed into lactone 35 using the con-
ventional PPTS conditions.
TsO
O
39
40
O
d
O
MeO2C
MeO2C
O
Br
e
Br
HO
C
TESO
C
H
H
O
O
H
O
H
O
42
41
The C-10 (cepacin numbering) configuration of 32 drawn in
Scheme 9 was actually established only after completion of
most of the syntheses disclosed here, through derivatization
of 31b using the sequence shown in Scheme 10. As distinct
NOE was observed between the H-9 and H-10 of 37, these
two Hs must be cis to each other. Therefore, the configura-
tion of 31b must be as drawn (the a-isomer). Consequently,
31a should have the configuration at the C-10 position as
drawn for 32 in Scheme 9.
Scheme 11. (a) (i) I2/NaHCO3/Me2CO–H2O (5:1)/0 ꢁC; (ii) 9/toluene/70–
80 ꢁC, 73% from 31a; (b) p-TsCl/NEt3/DMAP/CH2Cl2/0 ꢁC to rt, 74%;
(c) K2CO3/MeOH–THF (2:1)/0 ꢁC, 92%; (d) LiBr/CuBr$SMe2/THF/reflux,
61% (or 84% based on the consumed 40); (e) Bu4NF/THF, 97%.
The hydroxyl group was then tosylated. The product was
treated with LiBr/CuBr$SMe2 in a hope to obtain the corre-
sponding bromoallene. However, the bromide attack did not
occur at the triple bond. Instead, a simple tosylate to bromide
substitution at the propargylic position was observed. We
reasoned that the TMS protecting group might obstruct the
entry of the bromide. To circumvent this factor, the TMS
protecting group was then removed with K2CO3 in MeOH/
THF giving 40 for further transformation. Now starting
from the unprotected acetylene the desired bromoallene 41
Successful arrival at 35 seemingly left only a few steps to be
done before completing the whole synthesis—construction
of a bromoallene terminal, introduction of a diyne moiety,
and elaboration of the diol at C-8/C-9 into an epoxide. As
we planned to employ Mann’s32 method to construct the bro-
moallene moiety, which required a good leaving group at the

C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
4891
R
was indeed formed smoothly, which on treatment with
TBAF gave alcohol 42 in 97% yield.
MeO2C
MeO2C
O
O
Previous
Br
conditions
TsO
C
TsO
C
H
H
Tosylation of 42 under conventional conditions followed by
the CBS reduction and lactonization afforded 43. The aceto-
nide protecting group was then removed with HS(CH2)3SH/
BF3$OEt2 in CH2Cl2 to free the two hydroxyl groups. Fur-
ther treatment with K2CO3 in 50:1 Et2O/H2O provided the
epoxide 46 in 77% yield (Scheme 12).
O
O
43
H
O
49
H
O
50
MeO2C
O
R
MeO2C
Previous
conditions
O
Br
TESO
C
TESO
C
H
H
O
MeO2C
O
MeO2C
O
O
H
O
51
52
H
O
Br
41
Br
TsO
C
HO
C
H
H
O
O
Scheme 13.
43
H
b
O
H
O
42
reaction discovered by Andrus36 came to our view, where re-
placingTHF (or toluene or DMF) withEtOAc turned a failure
to success. Encouraged by that work, we re-examined the
coupling of 41 with first TMSC^CH and then TMSC^C–
C^CH under otherwise the same conditions. To our satisfac-
tion, this time we also obtained the coupling products
(Scheme 14). However, such a change in the solvent still
did not lead to any discernible improvements in the coupling
of epoxide 46.
O
O
O
O
c
Br
TsO
C
Br
TsO
H
H
C
O
OH
H
O
44
45
H
OH
TMS
d
R
MeO2C
O
MeO2C
O
1
1
Br
O
O
Br
H
a or b
TESO
C
H
TESO
C
H
51
O
e
H
O
4
7
4
O
C
C
O
H
H
52
H
O
H
O
7
H
41
H
(major isomer)
O
O
47 R = H
OR
OH
H
H
46
48 R = TES
Scheme 14. (a) TMSC^CH/Pd(Ph3P)2Cl2/CuI/i-Pr2NEt/EtOAc/ꢀ20 ꢁC to
rt, 77% for 51 (R¼TMS–); (b) TMSC^C–C^CH/Pd(Ph3P)2Cl2/CuI/
i-Pr2NEt/EtOAc/ꢀ20 ꢁC to rt, 59% for 52 (R¼TMSC^C–).
Scheme 12. (a) p-TsCl/NEt3/DMAP/CH2Cl2/0 ꢁC to rt, 72%; (b) (i)
BH3$SMe2/(S)-2-methyl-CBS-oxazaborolidine/0 ꢁC, 77% (93% based on
the consumed ketone); (ii) PPTS/PhH/40 ꢁC, 85%; (c) HS(CH2)3SH/
BF3$OEt2/CH2Cl2, 69%; (d) K2CO3/Et2O–H2O (50:1), 77%; (e) see text.
The 1H NMR spectrum of 51 and 52 suggested that in neither
case the coupling reaction proceeded with complete allene
configuration retention, although one of the isomers did pre-
dominate. The optical rotation was determined to be +50 and
+45 for 51 and 52, respectively. As the sign of the optical
rotation of diyne–allene containing compounds was domi-
nated1 by the contribution of allene moiety, the configuration
of the major isomer of 52 (and 51 a priori) is expected to be
opposite to that of cepacin A (as drawn in Scheme 14). This is
in line with the prediction of the bromoallene configuration
derived from 31a. Considering that 52 would not lead to nat-
ural cepacin A and the amount of 52 we had was not enough
to go forward any further, we discontinued the work along
this line.
Up to this point, we were seemingly only one step away from
the target structure (but with an allene of wrong configuration
as revealed by a later structural establishment of 31b)—to
install a diyne fragment onto the allene. However, this turned
out to be an impossible task. Direct treatment of 46 with
TMSC^C–C^CH33 (prepared from TMSC^C–C^
CTMS34) under the Sonogashira35 conditions (Pd(Ph3P)4 or
Pd(Ph3P)2Cl2/CuI, with different base (Et2NH, Et3N or
i-Pr2NEt) and/or solvent (THF or toluene, DMF)). Masking
the C-7 hydroxyl group as a TES ether did not lead to any
discernible improvements.
All the negative results with the epoxy-containing substrates
seemingly suggested that a labile epoxy functionality in the
coupling reaction might be the culprit. However, the coupling
reaction of 41 and 43 also completely failed under the same
conditions (Scheme 13). Although 43 does have a TsO–
group at the propargylic position and is therefore similar to
the epoxides 46 in a sense, the TESO– group in 41 is defi-
nitely not a good leaving group. Since this compound also
failed to deliver the coupling product, the failures could not
be attributed simply to the presence of a good leaving group.
In parallel to the work described above derived from 31a, we
also utilized 31b to perform the subsequent steps (after es-
tablishing the configuration of 31b experimentally as shown
in Scheme 10) in a hope to arrive at a bromoallene of correct
configuration. To make full use of 31a we still had, we tried
to oxidize it into the corresponding ketone 31c and then re-
duce the carbonyl group into alcohol again as shown in
Scheme 15.
Because 31a contained sulfur ether linkage partial struc-
tures, oxidation was more difficult than otherwise. We first
attempted to use mild MnO2 as the oxidant. However, the
When we almost completely lost the direction of further
exploration a dramatic solvent effect in a similar coupling

4892
C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
EtS
TESO
EtS
TESO
SEt
O
SEt
O
TMS
EtS
TESO
TESO
HO
CHO
O
SEt
O
TMS
TMS
TMS
a
a
O
b
O
O
b
HO
O
HO
O
31b
31c
31a
O
MeO2C
TESO
EtS
SEt
O
O
MeO2C
TESO
TMS
TESO
TMS
O
c
TMS
HO
O
O
O
31b
TsO
O
39'
HO
O
O
38'
Scheme 15. (a) MnO2/CH2Cl2/reflux, 50% or Dess–Martin oxidation, 72%;
(b) DIBAL-H/CH2Cl2/ꢀ78 ꢁC, 80%.
d
O
MeO2C
MeO2C
TESO
H
yield was only 50% and a large excess of the oxidant was re-
quired. These unfavorable factors made us to seek other al-
ternatives. The SO3$Py protocol, which was quite effective
on e.g. 29, did not work here. Finally, we gratifyingly found
that Dess–Martin oxidation could give 70% yield of the de-
sired ketone 31c.
TESO
C
e
Br
O
O
H
TsO
O
40'
41'
Scheme 16. (a) I2/NaHCO3/Me2CO–H2O (5:1)/0 ꢁC/20 min; (b) 9/toluene/
70 to 80 ꢁC, 73% over two steps, cis/trans¼1:4; (c) p-TsCl/DMAP/Et3N–
CH2Cl2/0 ꢁC to rt, 74%; (d) K2CO3/MeOH–THF (2:1)/0 ꢁC, 92%; (e)
LiBr/CuBr$SMe2/THF/reflux.
Reduction of 31c was best achieved with DIBAL-H, which
delivered a 10:1 mixture of 31b/31a. The results with L-Se-
lectride or K-Selectride were much less satisfactory, al-
though they were even bulkier than DIBAL-H. As we had
already developed a means to separate 31a and 31b, with
the help of the oxidation–reduction sequence we could ob-
tain enough amounts of 31b to perform the following steps.
the results of the bromoallenation—the bromide attack in the
case of 400 appears to be significantly more hindered than
that with 40 (Fig. 1).
The conformational analysis revealed that to overcome the
doubly hindered situation encountered with 400 a smaller
protecting group must be used in place of the TES. Acetyl
group is substantially smaller than TES and was therefore
chosen. However, by then we already ran out of both 390
and 400. Further testing could start only from 380.
As shown in Scheme 16, 31b was subjected to the same se-
ries of transformations as in converting 31a to 40, giving 400
smoothly as expected. However, in contrast to the smooth
transformation of 40 to 41, the a-epimer 400 was surprisingly
resistant to the same LiBr/CuBr$SMe2/THF reflux condi-
tions. We also tried to use some other leaving groups such
as 2,4,6-triisopropyl-benzenesulfonate (TPS) or triflate to
replace the tosyl group, but the results were all the same.
No traces of 410 could be detected.
Removal of the silyl groups in 380 with n-Bu4NF resulted in
diol 53 without any complications. The subsequent task was
to activate the hydroxyl group at the propargylic position and
mask the other hydroxyl group as an acetate. As the differ-
ence in steric crowding between the two hydroxyl groups is
not so large, a bulky sulfonating reagent is expected to be
beneficial. Indeed, treatment of 53 with TPSCl (2,4,6-triiso-
propyl-benzenesulfonyl chloride, a reagent much bulkier
than tosyl chloride) led nicely to the desired regio-selective
At first sight this outcome seems ridiculous, because the
structural difference between 40 and 400 is almost negligible.
However, a detailed conformational analysis of the two iso-
mers provided us with a possible clue to the drastic change in
C
B
(s)
A
(m)
(l)
OTs
H
TESO
(s)
(m)
(m)
(s)
H
(m)
OTES
H
OTES
OTES
(s)
R
H
OTs
O
TsO
40
H
(l)
R
(l)
(s)
(m)
H
R
OTs
(l)
(m)
R
(l)
(l)
(s)
R =
CO2Me
H
H
D
E
(l)
O
(m)
F
(s)
H
O
(m)
OTES
OTs
(s)
(s)
(m)
(m)
OTES
H
H
H
(s)
OTES
H
TESO
OTs
OTs
(l)
TsO
(s)
R
(l)
R
H
R
(l)
(m)
(m)
(s)
R
(l)
(l)
40'
Figure 1. The Newman presentations of the conformers of 40 (A–C) and 400 (D–F). The relative size of the substituents on each of the two central carbons of the
Newman projections is shown by (s), (m), and (l) (standing for small, medium, and large, respectively). The conformation with (l)–(l) interaction is of the highest
internal energy and thus is least stable. Consequently, the most stable conformation for 40 is B, while that for 400 should be F. Note that in conformer B the attack
of Brꢀ at the alkyne is mainly hindered by the R, whereas in F it is hindered by both the OTES and the R.

C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
4893
(m)
sulfonylation at the propargylic position as expected. The
C
B
(s)
H
A
(l)
OTPS
remaining free hydroxyl group at the C-9 was then readily
acylated with Ac2O. The resultant 54 underwent the bromo-
allenation smoothly, giving the long anticipated 55 in 86%
yield (Scheme 17).
(m)
(s)
(s)
H
(m)
(s)
OAc
H
H
OAc
OAc
(m)
(s)
OTPS
(l)
TPSO
(l)
H
H
R
(l)
(s)
(m)
R
(l)
R
(l)
(m)
Br−
O
MeO2C
TESO
O
MeO2C
HO
O
SEt
O
a
TMS
57 R =
H
CO2Me
54 R =
SEt
H
H
O
H
O
HO
O
O
O
HO
53
O
38'
O
SO2
b
Figure 2. Newman projections of the conformers of 54 and 57. The relative
size of the substituents on each of the two central carbons of the Newman
projections is shown by (s), (m), and (l) (standing for small, medium, and
large, respectively).
=
TPS
O
MeO2C
AcO
MeO2C
O
H
6
c
AcO
C
As removal of the acetyl group could be rather difficult at
later stages once the labile diyne moiety was installed, we
preferred to execute this deprotection task before the cou-
pling reaction during the elaboration of the lactone moiety.
In the event, the ketone carbonyl group in 55 was reduced
under the CBS conditions. As (R)-2-methyl-CBS-oxazaboro-
lidine was used here as the chiral additive, the configuration
of 59 at the C-4 is expected to be opposite to that in 44
(Scheme 19).
7
Br
9
TPSO
54
O
O
O
H
O
55
Scheme 17. (a) n-Bu4NF/THF/0 ꢁC, 85%; (b) TPSCl/DMAP/CH2Cl2/
40 ꢁC, then Ac2O/DMAP, 64% from 53; (c) CuBr$SMe2/LiBr/THF/reflux,
86%.
It is interesting to note that another substrate 57, which was
derived from 36 through the route in Scheme 18 and very
similar to 54 as far as the reaction centers are concerned,
completely failed to give the corresponding bromoallene un-
der the otherwise identical conditions. As these two com-
pounds (54 and 57) differ only at the substituents on the
C-6, at first sight this seems to be very difficult to understand.
However, if one looks into the conformation details, some
substantial difference still can be spotted.
MeO2C
AcO
O
MeO2C
AcO
OH
H
H
6
6
a
Br
Br
7
7
C
C
9
9
O
O
O
H
O
H
O
55
59
b
O
O
O
EtS
HO
EtS
HO
SEt
O
SEt
O
H
H
a
c
TBSO
C
HO
Br
Br
C
O
O
HO
O
TPSO
56
O
36
H
O
60
H
O
61
b
O
d
O
O
H
EtS
EtS
SEt
SEt
O
O
c
AcO
AcO
6
4
H
7
Br
H
C
TBSO
C
9
O
HO
7
TMS
TPSO
57
O
H
O
C
O
TMS
2
O
2
9
58
H
O
H
O
62
63
Scheme 18. (a) TPSCl/DMAP/CH2Cl2/50 ꢁC; (b) Ac2O/DMAP/CH2Cl2,
94% from 36; (c) CuBr$SMe2/LiBr/THF/reflux.
Scheme 19. (a) (R)-2-Methyl-CBS-oxazaborolidine/BH3$SMe2/THF/0 ꢁC/
10 min, 83%; (b) 0.2 M NaOMe/0 ꢁC/25 min, then PPTS (cat)/toluene/
60 ꢁC, 75%; (c) TBSOTf/2,6-lutidine/CH2Cl2/0 ꢁC, 80%; (d) TMSC^C–
C^CH/Pd(Ph3P)2Cl2/CuI/i-Pr2NEt/EtOAc/ꢀ20 ꢁC to rt, 77%.
As shown in Figure 2, the most stable conformer for both
compounds is expected to be C, where the interaction be-
tween the two largest substituents is avoided. The relevant
structural difference between 54 and 57 is highlighted with
boxes in the two lower partial structures. For the former,
the substituent on the C-6 is essentially linear, stretching
out through a carbon–carbon double bond. In the case of
the latter, however, the substituents are two –SEt groups,
which spread out and thus are obviously much bulkier than
their counterpart in 54. Hence, the larger steric crowding
created by the branched dithioacetal in 57 appeared to be
the culprit.
The resultant 59 was further converted into 60 by sequential
treatment with NaOMe and PPTS. The hydroxyl group was
then protected as a TBS ether before concatenation with
TMSC^C–C^CH under the same conditions as in the syn-
thesis of 52. Although the coupling product was still an in-
separable mixture of the allene isomers as in the previous
cases, one of them predominated with the optical rotation
of the same direction as of cepacin A indicating a different
allene configuration from that in 52.

4894
C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
Up to this point it seemed that we were only a few steps away
from the target structure. We only needed to convert the TBS
protected hydroxyl group into a good leaving group and re-
move the acetonide protecting group to free the diol before
eventually making the epoxy ring at a final step. However,
things turned out much more complicated than we expected.
All efforts37 to remove the TBS group failed. This made it
impossible to proceed to make a good leaving group at
C-9. To get around this problem, we next attempted the
coupling reaction with alcohol 60.
Varian Mercury 300 or a Bruker Avance 300 instrument (op-
erating at 300 MHz for proton), Bruker Avance 400
1
1
(400 MHz for H), or Varian Inova-600 (600 MHz for H).
The FTIR spectra were scanned with a Perkin Elmer 983 or
a Nicolet Avatar 360 FT-IR spectrometer. The EIMS,
EIHRMS, ESIMS, and ESIHRMS were recorded on an
HP5989A, a Finnigan MAT 8430, a PE Mariner API-TOF
(or an Agilent Technologies LC/MSD SL), and a Bruker
APEXIII 7.0 Tesla FT-MS spectrometer, respectively. Ele-
mental analyses were performed on an Elementar VarioEL
III instrument. Optical rotations were measured on a
Perkin–Elmer 341 or an Agilent Technologies P-1030 polari-
meter. Dry THF, Et2O, toluene, and petroleum ether (PE) and
n-hexane were distilled over Na/Ph2CO under argon. Dry
CH2Cl2, EtOAc, Et3N, Et2NH, i-Pr2NEt, and HNTMS2
With a free hydroxyl group at the C-9, the coupling could
also occur, though the yield was somewhat lower (Scheme
20). However, the subsequent activation of the hydroxyl
group ran into problem again. No matter what sulfonation re-
agent (MsCl, p-TsCl or p-Ts2O) was used, in the absence of
DMAP no reaction took place. Addition of DMAP did lead
to a rather fast reaction. The starting 64 disappeared very
quickly. The product, however, was not the expected 65
but the triyne–ene 66. It seemed that 65 was highly unstable
and extremely labile to the undesired elimination.
˚
were distilled over CaH2 under argon and kept over 4 A mo-
lecular sieves. Dry DMSO, DMF, and pyridine were stirred
with CaH2 under N2 for 24 h and distilled under reduced
pressure. Dry MeOH was refluxed/distilled over Mg turnings
under argon. Dry acetone was distilled over P2O5 under
argon.
O
4.1.1. Synthesis of ketone-ester 12. HgO (red, 800 mg,
3.69 mmol) and HgCl2 (800 mg, 2.95 mmol) were added to
a solution of 11 (500 mg, 1.49 mmol) in MeCN/H2O
(10:1 v/v, 11 mL) stirred at ambient temperature. One hour
later the mixture was filtered through Celite (washing with
CH2Cl2). The filtrate and washings were washed with aq
20% KI and satd NaHCO3 in turn and dried over anhydrous
MgSO4. Removal of the solvent left a colorless oil (the inter-
mediate aldehyde, 308 mg, 1.34 mmol, 90% yield). Part of
this aldehyde (93 mg, 0.40 mmol) and the Wittig reagent 9
(190 mg, 0.50 mmol) were dissolved in toluene (10 mL)
and heated to reflux with stirring overnight. The solvent
was removed on a rotary evaporator. The residue was
chromatographed on silica gel (3:7 Et2O/PE) to give 12 as a
yellowish oil (109 mg, 0.32 mmol, 80% from 11): FTIR
O
O
O
H
H
HO
HO
a
Br
C
TMS
C
O
2
O
H
O
H
O
64
60
c
TMS
b
O
O
O
O
L = Ts or Ms
H
LO
C
TMS
O
2
H
O
O
O
66
65
1
(film) 3000, 1740, 1700, 1370, 1220, 1160, 960 cmꢀ1; H
NMR (300 MHz, CDCl3) d 6.91 (dd, J¼15.9, 4.4 Hz, 1H),
6.47 (dd, J¼15.9, 1.3 Hz, 1H), 4.56 (ddd, J¼6.1, 4.4,
1.7 Hz, 1H), 4.18–4.08 (m, 2H), 4.00–3.92 (m, 1H), 3.73–
3.64 (m, 4H), 2.92 (t, J¼6.9 Hz, 2H), 2.65 (t, J¼6.6 Hz,
2H), 1.43 (s, 3H), 1.42 (s, 3H), 1.41 (s, 3H), 1.35 (s, 3H);
EIMS m/z (%) 342 (M+, 2.3), 327 (16.2), 43 (100), 101
(89.6), 55 (69.6), 115 (56.7), 169 (39.4), 59 (36.5), 94 (33.5),
41 (25.9); EIHRMS calcd for C16H23O7 ([MꢀCH3]+)
327.1444, found 327.1420.
Scheme 20. (a) TMSC^C–C^CH/Pd(Ph3P)2Cl2/CuI/i-Pr2NEt/EtOAc/
ꢀ20 ꢁC to rt, 54%; (b) MsCl/Py, or MsCl/Et3N, or p-TsCl/Py or p-TsCl/
Et3N, or p-Ts2O/Py; (c) p-TsCl/Py/DMAP (cat) or p-Ts2O/Et3N/DMAP
(cat), 75%.
3. Conclusions
The unassuming size of cepacin A belies its complexity as
a target for chemical synthesis. Many unexpected yet inter-
esting problems turned up along our way to arrive at the final
target structure. We managed to solve some of them, includ-
ing the bromoallenation through tuning the size of neighbor-
ing groups and the coupling with diyne by changing the
solvent. The eventual construction of the epoxy ring in the
presence of the labile diyne–allene motif is still to be
done. The control over the allene configuration is also to
be improved.
4.1.2. Synthesis of lactone 14. NaBH4 (127 mg, 3.34 mmol)
was added in portions to a solution of 12 (541 mg,
1.58 mmol) and CeCl3$7H2O (1.280 g, 3.44 mmol) in anhy-
drous MeOH (7.5 mL) stirred at 0 ꢁC. After completion of
the addition, the mixture was stirred for another 5 min before
the reaction was quenched by addition of H2O. The mixture
was then diluted with EtOAc, washed with aq satd NaHCO3,
and dried over anhydrous MgSO4. The residue left by re-
moval of the solvent was chromatographed on silica gel
(3:1 EtOAc/PE) to afford 13 as a colorless oil (434 mg,
1.26 mmol, 80% yield): FTIR (film) 3485, 2936, 1776,
4. Experimental
4.1. General
1
1734, 1371, 1167, 1065, 976, 919, 847 cmꢀ1; H NMR
(300 MHz, CDCl3)
d 5.98–5.88 (m, 2H), 4.38 (t,
The 1H NMR and 13C NMR spectra were recorded in CDCl3
at ambient temperature using the following instruments:
J¼7.7 Hz, 1H), 4.28–4.18 (m, 1H), 4.16–4.08 (m, 2H),
3.98–3.90 (m, 1H), 3.73–3.64 (m, 4H), 2.47 (td, J¼7.1,

C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
4895
1
2.4 Hz, 2H), 2.00–2.09 (m, 2H), 1.41 (s, 3H), 1.35 (s, 3H);
EIMS m/z (%) 329 (M+ꢀCH3, 5.1), 43 (100), 101 (70.6),
59 (48.4), 115 (47.4), 85 (28.2), 167 (25.9), 41 (25.9).
54% yield): FTIR (film) 2930, 1785, 1255, 836 cmꢀ1; H
NMR (300 MHz, CDCl3) d 5.92–5.86 (m, 2H), 4.96 (br,
1H), 4.55(d, J¼7.4 Hz, 1H), 3.97–3.88 (m, 2H), 3.60–3.50
(m, 2H), 2.55 (td, J¼6.6, 3.6 Hz, 2H), 2.48–2.37 (m, 1H),
2.08–1.95 (m, 1H), 1.43 (s, 3H), 1.39 (s, 3H), 0.90 (s, 9H),
0.89 (s, 9H), 0.10 (s, 3H), 0.09 (s, 3H), 0.06 (s, 6H); ESIMS
m/z 523.3 ([M+Na]+). Anal. Calcd for C25H48O6Si2: C,
59.96; H, 9.66. Found C, 59.82; H, 9.64.
A solution of 13 (943 mg, 2.74 mmol) and p-TsOH (60 mg,
0.32 mmol) in dry acetone (40 mL) was heated to reflux with
stirring under argon for 10 h. After cooling to ambient tem-
perature, the acid in the mixture was neutralized with pow-
dered Na2CO3 and traces of H2O. The solids were filtered
off. The filtrate was concentrated on a rotary evaporator
and the residue was chromatographed on silica gel (4:1
CH2Cl2/EtOAc) to afford 14 as a yellowish oil (600 mg,
1.92 mmol, 70%): FTIR (film) 3413, 2947, 2894, 1723,
4.1.6. Protection of the primary hydroxyl group in 15
with PivCl (18). Pivaloyl chloride (70 mL, 0.45 mmol)
was added to a solution of 15 (118 mg, 0.43 mmol) and
dry Et3N (0.2 mL) in dry CH2Cl2 (4 mL) stirred in an ice-
water bath. The mixture was then stirred at ambient temper-
ature overnight before being diluted with EtOAc, washed in
turn with 0.5 N HCl and satd aq NaHCO3, and dried over an-
hydrous MgSO4. Rotary evaporation and column chromato-
graphy on silica gel (1:12 EtOAc/PE) gave 16 as a colorless
oil (85 mg, 0.25 mmol, 58% yield): FTIR (film) 3483 (br),
1670, 1192, 1048, 964, 835 cmꢀ1 1H NMR (300 MHz,
;
CDCl3) d 5.90–5.75 (m, 2H), 5.10–4.90 (m, 1H), 4.45–
4.38 (m, 1H), 4.18–4.05 (m, 2H), 3.68 (q, J¼6.0 Hz, 1H),
3.70–3.60 (m, 1H), 2.60–2.50 (m, 2H), 2.49–2.34 (m, 1H),
2.10–1.95 (m, 1H), 1.42 (s, 3H), 1.40 (s, 3H), 1.39 (s, 3H),
1.34 (s, 3H); EIMS m/z (%) 297 (M+ꢀCH3, 2.5), 43 (100),
99 (45.4), 101 (39.1), 55 (33.5), 87 (31.4), 59 (27.8), 41
(20.0), 113 (13.9); EIHRMS calcd for C15H21O6
([M+ꢀCH3]) 297.1338, found 297.1352.
2983, 1777, 1729, 1165 cmꢀ1
;
1H NMR (300 MHz,
CDCl3) d 5.92 (t, J¼3.5 Hz, 2H), 5.04–4.95 (m, 1H), 4.51
(dd, J¼7.7, 4.9 Hz, 1H), 4.26 (dd, J¼11.8, 3.3 Hz, 1H),
4.15 (dd, J¼11.8, 6.8 Hz, 1H), 4.04–3.94 (m, 1H), 3.68–
3.59 (m, 1H), 2.56 (td, J¼6.6, 1.7 Hz, 2H), 2.50–2.40 (m,
1H), 2.10–1.95 (m, 1H), 1.43 (s, 3H), 1.27 (s, 3H), 1.24 (s,
9H); ESIMS m/z 379.1 ([M+Na]+); ESIHRMS calcd for
C18H28O7Na: 379.1727, found 379.1721.
4.1.3. Removal of terminal acetonide in 13 (15). A solution
of 13 (37 mg, 0.11 mmol) and p-TsOH (5 mg, 0.03 mmol) in
anhydrous MeOH (5 mL) was stirred at ambient temperature
overnight. The reaction mixture was diluted with EtOAc,
washed with satd aq NaHCO3, and dried over anhydrous
MgSO4. Rotary evaporation and column chromatography
on silica gel (9:1 EtOAc/MeOH) gave 15 as a yellowish
oil (19 mg, 0.07 mmol, 64% yield): FTIR (film) 3427 (br),
4.1.7. TBS protection of 18 (19). To a solution of 18 (80 mg,
0.23 mmol) in dry DMF (0.4 mL) stirred at ambient temper-
ature under argon were added in turn imidazole (70 mg,
0.51 mmol), TBSCl (75 mg, 0.50 mmol), and a catalytic
amount of DMAP. The mixture was stirred at ambient tem-
perature overnight before being diluted with EtOAc, washed
with 0.5 N HCl and satd aq NaHCO3, and dried over anhy-
drous MgSO4. Rotary evaporation and column chromato-
graphy on silica gel (1:3 EtOAc/PE) gave 19 as a colorless
2987, 1770, 1182, 977 cmꢀ1; H NMR (300 MHz, CDCl3)
1
d 5.95–5.88 (m, 2H), 5.05–4.95 (m, 1H), 4.48 (dd, J¼8.0,
5.0 Hz, 1H), 3.90–3.81 (m, 1H), 3.80–3.68 (m, 3H), 2.57
(dt, J¼3.5, 1.3 Hz, 2H), 2.50–2.40 (m, 1H), 2.08–1.96 (m,
1H), 1.44 (s, 3H), 1.43 (s, 3H); EIMS m/z (%) 257
(M+ꢀCH3, 4.9), 43 (100), 59 (56.7), 55 (48.6), 41 (38.9),
85 (29.7), 137 (29.2), 79 (26.5), 81 (25.9); EIHRMS calcd
for C12H17O7 ([M+ꢀCH3]) 257.1025, found 257.1039.
1
oil (852 mg, 0.11 mmol, 52% yield): H NMR (300 MHz,
CDCl3) d 5.98–5.82 (m, 2H), 5.06–4.92 (m, 1H), 4.52–
4.42 (m, 1H), 4.10–3.90 (m, 2H), 3.88–3.70 (m, 1H),
2.10–1.95 (m, 1H), 1.44 (s, 1.5H), 1.43 (s, 1.5H), 1.29 (s,
1.5H), 1.26 (s, 1.5H), 1.23 (s, 4.5H), 1.22 (s, 4.5H), 0.92
(s, 4.5H), 0.90 (s, 4.5H), 0.13 (s, 1.5H), 0.11 (s, 1.5H),
0.06 (s, 1.5H), 0.05 (s, 1.5H). (IR and MS were the same
as that for 24a/24b, the enantiopure samples.)
4.1.4. Removal of terminal acetonide in 14 (15). Alterna-
tively, 15 could also be obtained from 14. A mixture of 14
(560 mg, 1.79 mmol), CeCl3$7H2O (1.38 g, 3.70 mmol),
and oxalic acid (17 mg, 0.19 mmol) in MeCN (9 mL) was
stirred at ambient temperature until TLC showed disappear-
ance of 14. The acid was neutralized with powdered Na2CO3
and the solvent was removed on a rotary evaporator. The resi-
due was dissolved in EtOAc, washed with satd aq NaHCO3,
and dried over anhydrous MgSO4. Rotary evaporation and
column chromatography on silica gel (100:6 EtOAc/
MeOH) gave 15 as a colorless oil (193 mg, 0.71 mmol,
40%).
4.1.8. Conversion of 11 into 20. A solution of 11 (1.121 g,
3.34 mmol), CeCl3$7H2O (2.437 g, 6.54 mmol), and oxalic
acid (22 mg, 0.24 mmol) in MeCN (20 mL) was stirred at
ambient temperature until TLC showed complete disappear-
ance of 11. With cooling (ice-water bath), the excess acid
was neutralized with Na2CO3. The solids were filtered off.
The filtrate was diluted with EtOAc, washed with satd aq
NaHCO3, and dried over anhydrous MgSO4. Rotary evapo-
ration and column chromatography on silica gel (2:3 EtOAc/
PE) gave the intermediate diol as a colorless oil (811 mg,
2.74 mmol, 82% yield).
4.1.5. Protection of 15 with TBSCl (16). A solution of 15
(10 mg, 0.037 mmol), imidazole (20 mg, 0.15 mmol),
DMAP (4.5 mg, 0.037 mmol), and TBSCl (50 mg,
0.33 mmol) in anhydrous DMF (0.2 mL) was stirred at am-
bient temperature under argon until TLC showed complete
disappearance of the starting 15. The reaction mixture was
diluted with EtOAc, washed in turn with 0.5 N HCl and
satd aq NaHCO3, and dried over anhydrous MgSO4. Rotary
evaporation and column chromatography on silica gel (1:2
EtOAc/PE) gave 16 as a colorless oil (10 mg, 0.02 mmol,
Alternatively, the hydrolysis could also be performed as fol-
lows. A solution of 11 (10.505 g, 31.30 mmol) in 80%
HOAc (80 mL) was stirred at 60 ꢁC for 2 h. After cooling
to ambient temperature, the HOAc was removed on a rotary
evaporator. The residue was diluted with EtOAc, washed in

4896
C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
turn with satd aq NaHCO3 and brine, and dried over anhy-
drous MgSO4. Rotary evaporation and column chromato-
graphy on silica gel (2:3 EtOAc/PE) gave the intermediate
diol as a colorless oil (6.948 g, 23.47 mmol, 75% yield):
(15 mL) and treated with Wittig reagent 9 (470 mg,
1.21 mmol) at reflux temperature until TLC showed disap-
pearance of the starting aldehyde (ca. 10 h). Toluene was re-
moved by rotary evaporation. The residue was diluted with
Et2O and cooled in a refrigerator overnight. The supernatant
ethereal solution was poured out and concentrated on a rotary
evaporator to leave a residue, which was chromatographed
on silica gel (1:9 EtOAc/PE) to give 22 as a yellowish oil
(533 mg, 1.07 mmol, 94% from 21): [a]2D5 +6.62 (c 1.80,
CHCl3). FTIR (film) 2919, 1686, 1650, 1298, 1141, 1021,
1
FTIR (film) 3403 (br), 2929, 1454, 1240, 883 cmꢀ1; H
NMR (300 MHz, CDCl3) d 4.32 (dd, J¼6.8, 4.2 Hz, 1H),
4.12 (t, J¼6.9 Hz, 1H), 4.02 (d, J¼4.1 Hz, 1H), 3.90–3.70
(m, 3H), 2.80–2.60 (m, 5H), 2.01 (t, J¼6.5 Hz, 1H), 1.48
(s, 3H), 1.41 (s, 3H), 1.15 (t, J¼7.4 Hz, 6H).
Pivaloyl chloride (1.7 mL, 13.70 mmol) was added dropwise
to a solution of the above mentioned diol (3.880 g,
12.83 mmol) and dry Et3N (6 mL) in CH2Cl2 (25 mL) stirred
in an ice-water bath. After completion of the addition, the
bath was removed and the mixture was stirred at ambient
temperature (overnight) until TLC showed completion of
the reaction. The mixture was diluted with EtOAc, washed
with 0.5 N HCl and satd aq NaHCO3, and dried over anhy-
drous MgSO4. Rotary evaporation and column chromato-
graphy on silica gel (1:12 EtOAc/PE) gave the intermediate
diol as a yellowish oil (3.605 g, 9.90 mmol, 77% yield):
[a]2D5 +61.1 (c 1.0, CHCl3). FTIR (film) 3485 (br), 2973,
969 cmꢀ1 1H NMR (300 MHz, CDCl3) d 6.87 (dd,
;
J¼15.7, 4.4 Hz, 1H), 6.45 (dd, J¼15.9, 1.4 Hz, 1H), 4.61
(dt, J¼7.4, 1.3 Hz, 1H), 4.06 (dd, J¼11.8, 5.0 Hz, 1H),
4.08–4.00 (m, 1H), 3.86 (dd, J¼7.42, 5.5 Hz, 1H), 3.68 (s,
3H), 2.91 (t, J¼6.6 Hz, 2H), 2.64 (t, J¼6.9 Hz, 2H), 1.45
(s, 3H), 1.39 (s, 3H), 1.21 (s, 9H), 0.91 (s, 9H), 0.14 (s,
3H), 0.13 (s, 3H); 13C NMR (75 MHz, CDCl3) d 197.5,
178.2, 173.1, 143.7, 129.1, 110.0, 80.5, 77.0, 65.2, 51.8,
35.3, 27.7, 27.2, 27.0, 26.5, 25.7, 18.0, ꢀ4.47, ꢀ4.53;
EIMS m/z (%) 485 (M+ꢀCH3, 1.7), 443 (2.0), 57 (100),
159 (39.7), 73 (26.4), 115 (22.8), 41 (18.5), 283 (18.1), 94
(18.0), 259 (17.0); EIHRMS calcd for C24H41O8Si
([MꢀCH3]+) 485.2571; found 485.2531.
1
1731, 1481, 1241, 882 cmꢀ1; H NMR (300 MHz, CDCl3)
d 4.42 (dd, J¼11.8, 2.4 Hz, 1H), 4.36 (dd, J¼3.3 Hz, 1H),
4.19 (dd, J¼11.8, 6.1 Hz, 1H), 4.10 (dd, J¼8.2, 6.9 Hz,
1H), 4.03 (d, J¼3.6 Hz, 1H), 3.91–3.86 (m, 1H), 1.46 (s,
3H), 1.39 (s, 3H), 1.28 (t, J¼7.4 Hz, 3H), 1.27 (t,
J¼7.4 Hz, 3H), 1.24 (s, 9H); EIMS m/z (%) 319 (M+ꢀEt,
1.2), 303 (4.0), 262 (29.8), 261 (39.1), 217 (22.3), 187
(43.7), 159 (28.4), 143 (36.9), 135 (100), 85 (36.1), 57
(81.5). Anal. Calcd for C17H32O5S2: C, 53.65; H, 8.48. Found
C, 53.38; H, 8.35.
4.1.11. Reduction of 22 with NaBH4 (23a and 23b).
NaBH4 (118 mg, 3.11 mmol) was added in portions to a so-
lution of 22 (705 mg, 1.41 mmol) in MeOH (4 mL) stirred in
an ice-water bath. After completion of the addition, the mix-
ture was stirred for 5 min before addition of water to quench
the reaction. The solvent was removed by rotary evaporation
and the residue was diluted with EtOAc, washed in turn with
0.5 N HCl and satd aq NaHCO3 before being dried over an-
hydrous MgSO4. Rotary evaporation and column chromato-
graphy on silica gel (3:7 EtOAc/PE) gave a mixture of 23a
and 23b as a colorless oil (501 mg, 1.0 mmol, 71% yield),
which could be further separated into pure 23a and 23b by
a second chromatography on silica gel eluting with 8:1
CH2Cl2/EtOAc.
4.1.9. TBS protection of 20 (21). A solution of 20 (83.605 g,
9.90 mmol), imidazole (2.841 g, 20.71 mmol), and TBSCl
(2.203 g, 14.69 mmol) in dry DMF (4 mL) was stirred at am-
bient temperature under argon overnight. The mixture was
extracted with petroleum ether (10 mL) thrice. The com-
bined petroleum ether layers were diluted with Et2O, washed
in turn with 0.5 N HCl and satd aq NaHCO3 before being
dried over anhydrous MgSO4. Rotary evaporation and col-
umn chromatography on silica gel (1:24 EtOAc/PE) gave
21 as a colorless oil (4.660 g, 9.53 mmol, 95% yield):
[a]2D6 +52.02 (c 0.50, CHCl3). FTIR (film) 2920, 2686,
Data for 23a (the more polar component): [a]2D5 +20.2 (c
1.05, CHCl3). FTIR (film) 3431, 2917, 1686, 1427, 1144,
1
1020, 989, 853 cmꢀ1; H NMR (300 MHz, CDCl3) d 5.90
(dd, J¼15.4, 5.2 Hz, 1H), 5.73 (dd, J¼15.4, 6.6 Hz, 1H),
4.48 (t, J¼7.5 Hz, 1H), 4.24–4.16 (m, 1H), 4.12–4.04 (m,
2H), 3.94 (d, J¼7.9 Hz, 1H), 3.80 (dd, J¼7.7, 3.5 Hz, 1H),
3.67 (s, 3H), 2.46 (t, J¼7.4 Hz, 2H), 2.08–2.00 (m, 1H),
1.98–1.78 (m, 1H), 1.41 (s, 3H), 1.39 (s, 3H), 1.20 (s, 9H),
0.89 (s, 9H), 0.11 (s, 6H); EIMS m/z (%) 57 (100), 73
(33.3), 259 (31.9), 159 (29.4), 85 (22.5), 43 (15.4), 55
(14.7), 75 (14.2); ESIMS m/z 525.3 ([M+Na]+); ESIHRMS
calcd for C25H46O8SiNa ([M+Na]+), 525.2877; found
525.2854.
1
1445, 1424, 1250, 1139, 1069, 991, 872 cmꢀ1; H NMR
(300 MHz, CDCl3) d 4.32–4.25 (m, 3H), 4.10 (dd, J¼11.8,
3.6 Hz, 1H), 3.98–3.90 (m, 2H), 2.86–2.67 (m, 4H,), 1.46
(s, 3H), 1.38 (s, 3H), 1.31–1.20 (m, 15H), 0.85 (s, 9H),
0.15(s, 3H), 0.14 (s, 3H); EIMS m/z (%) 419
(M+ꢀSC2H5ꢀCH3, 3.9), 375 (13.8), 335 (3.6), 57 (100),
73 (45.6), 199 (40.8), 159 (32.5), 75 (29.5); ESIMS 517.30
([M+Na]+); HRMS calcd for C23H46O5S2SiNa (M+ꢀCH3)
517.2448; found, 517.2464.
Data for 23b (the less polar component): [a]2D5 +5.9 (c 2.25,
CHCl3). FTIR (film) 3431, 2919, 1686, 1427, 1404, 1020,
4.1.10. Conversion of 21 to 22. A mixture of 21 (565 mg,
1.14 mmol), yellow HgO (653 mg, 3.0 mmol), HgCl2
(626 mg, 2.3 mmol) in MeCN (10 mL) and H2O (1 mL)
was stirred at ambient temperature for 1 h. The solids were
filtered off through Celite (washing with 150 mL of
CH2Cl2). The filtrate and washings were combined, washed
with 20% aq KI and satd aq NaHCO3, and dried over anhy-
drous MgSO4. Rotary evaporation left the intermediate alde-
hyde as a colorless oil, which was dissolved in dry toluene
1
988, 961 cmꢀ1; H NMR (300 MHz, CDCl3) d 5.83 (dd,
J¼15.3, 6.3 Hz, 1H), 5.63 (dd, J¼15.4, 7.1 Hz, 1H), 4.45
(t, J¼7.7 Hz, 1H), 4.24–4.10 (m, 3H), 3.83 (d, J¼6.6 Hz,
1H), 3.80 (dd, J¼8.3, 3.0 Hz, 1H), 3.68 (s, 3H), 2.45 (t,
J¼7.4 Hz, 2H), 1.98–1.80 (m, 2H), 1.42 (s, 3H), 1.40 (s,
3H), 1.21 (s, 9H), 0.89 (s, 9H), 0.13 (s, 6H); 13C NMR
(75 MHz, CDCl3) d 178.4, 174.2, 137.1, 129.3, 108.9,
80.5, 76.4, 71.2, 69.0, 64.8, 51.7, 38.8, 31.4, 30.0, 27.2,

C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
4897
27.0, 26.9, 25.8, 18.0, ꢀ4.4, ꢀ4.8; ESIMS m/z 525.3
([M+Na]+); ESIHRMS calcd for C25H46O8SiNa 525.2863
([M+Na]+); found 525.2854.
([M+Na]+); ESIHRMS calcd for C19H36O6SiNa
([M+Na]+) 411.2173; found 411.2179.
Data for 25b (derived from 24b): FTIR (film) 3423 (br),
2953, 1379, 1055 cmꢀ1 1H NMR (300 MHz, CDCl3)
;
4.1.12. Lactonization of 23 (24a and 24b). A solution of
23a or 23b (63 mg, 0.13 mmol) and PPTS (8 mg) in benzene
(10 mL) was heated to reflux for 2 h. After cooling to ambi-
ent temperature, a small amount of powdered Na2CO3 and
traces of water were added. The mixture was stirred for
a while before the solvent was removed on a rotary evapora-
tor. The residue was diluted with EtOAc, washed with satd
aq NaHCO3, and dried over anhydrous MgSO4. Removal
of the solvent left 24a or 24b as a yellowish oil (100% yield),
which was rather pure as shown by 1H NMR.
d 5.96–5.70 (m, 2H), 5.56 (br, 1H), 5.47 (br, 1H), 4.65 (q,
J¼6.6 Hz, 1H), 4.50–4.30 (m, 2H), 3.90–3.75 (m, 2H),
3.61 (br, 2H), 2.22–1.90 (m, 4H), 1.39 (s, 3H), 1.26 (s,
3H), 0.88 (s, 9H), 0.08 (s, 6H); ESIMS m/z 411.3
([M+Na]+); ESIHRMS calcd for C19H36O6SiNa
([M+Na]+) 411.2173; found 411.2180.
4.1.14. Oxidation of 25 and addition of TMSC^CLi to 26
(27b). (Method A) PCC oxidation. A mixture of 25 (640 mg,
1.65 mmol), PCC (2.113 g, 10.01 mmol), and NaOAc
(1.012 g, 10.21 mmol) in dry CH2Cl2 (10 mL) was stirred
at ambient temperature for 6 h. After dilution with Et2O,
the reaction mixture was washed with satd aq NaHCO3
and dried over anhydrous MgSO4. Removal of the solvent
and column chromatography on silica gel (1:3 EtOAc/PE)
gave 26 as a colorless oil (320 mg, 0.89 mmol, 50% yield).
Data for 24a (derived from 23a): [a]2D5 +32.36 (c 0.80,
CHCl3). FTIR (film) 2958, 1783, 1732, 1462, 1162,
837 cmꢀ1
;
1H NMR (300 MHz, CDCl3) d 5.92 (dd,
J¼15.7, 4.7 Hz, 1H), 5.87 (dd, J¼15.7, 5.0 Hz, 1H), 4.95
(q, J¼7.1 Hz, 1H), 4.49 (dd, J¼7.7, 4.7 Hz, 1H), 4.10–4.00
(m, 3H), 3.85 (dd, J¼7.7, 3.6 Hz, 1H), 2.55 (dt, J¼6.3,
2.5 Hz, 2H), 2.41 (sextet, J¼7.9 Hz, 1H), 2.08–1.98 (m,
1H), 1.43 (s, 3H), 1.40 (s, 3H), 1.22 (s, 9H), 0.91 (s, 9H),
0.13 (s, 3H), 0.12 (s, 3H); 13C NMR (75 MHz) d 178.2,
176.7, 131.3, 130.3, 109.3, 80.7, 79.4, 76.8, 70.1, 65.2,
38.8, 28.4, 28.3, 27.2, 27.0, 26.7, 25.7, 18.0, ꢀ4.5, ꢀ4.7;
ESIMS m/z 493.4 ([M+Na]+); ESIHRMS calcd for
C24H42O7SiNa ([M+Na]+), 493.2592; found, 493.2573.
(Method B) IBX oxidation. A solution of 25 (792 mg,
2.04 mmol) and IBX (3.009 g, 10.75 mmol) in DMSO
(25 mL) was stirred at ambient temperature for 2 h. The mix-
ture was diluted with Et2O, washed with water and brine, and
dried over anhydrous MgSO4. Removal of the solvent and
column chromatography on silica gel (1:3 EtOAc/PE) gave
26 as a colorless oil (614 mg, 1.60 mmol, 78% yield).
Data for 24b (derived from 23b): [a]2D5 ꢀ3.1 (c 1.75, CHCl3).
FTIR (film) 2919, 1732, 1682, 1446, 1299, 1019, 998 cmꢀ1
;
(Method C) Dess–Martin oxidation. A solution of 25
(218 mg, 0.56 mmol) and Dess–Martin periodinane (784 mg,
1.85 mmol) in dry CH2Cl2 (8 mL) and dry pyridine (1 mL)
was stirred at 0 ꢁC until TLC showed disappearance of the
starting alcohol. The mixture was diluted with Et2O, washed
with water and brine, and dried over anhydrous MgSO4.
Removal of the solvent and column chromatography on
silica gel (1:3 EtOAc/PE) gave 26 as a colorless oil (90 mg,
0.23 mmol, 41% yield).
1H NMR (600 MHz, CDCl3) d 5.91 (dd, J¼15.6, 5.2 Hz,
1H), 5.87 (dd, J¼15.6, 5.4 Hz, 1H), 4.98 (dd, J¼14.1,
4.9 Hz, 1H), 4.49 (dd, J¼7.7, 5.5 Hz, 1H), 4.10–4.00 (m,
3H), 3.84 (dd, J¼7.7, 4.0 Hz, 1H), 2.56–2.50 (m, 2H),
2.42 (sextet, J¼6.4 Hz, 1H), 2.04–1.98 (m, 1H), 1.42 (s,
3H), 1.39 (s, 3H), 1.21 (s, 9H), 0.90 (s, 9H), 0.13 (s, 3H),
0.11 (s, 3H); 13C NMR (125 MHz) d 178.1, 176.5, 131.3,
130.6, 81.0, 79.3, 77.0, 70.3, 65.4, 38.8, 28.4, 28.2, 27.2,
27.0, 26.8, 25.8, 18.0, ꢀ4.48, ꢀ4.46; EIMS m/z (%) 455
(M+ꢀCH3, 1.5), 413 (3.3), 57 (100), 41 (27.5), 73 (23.9),
159 (23.7), 43 (21.3), 85 (20.8), 259 (19.8), 75 (18.6);
HRMS calcd for C23H39O7Si ([MꢀCH3]+) 455.2465; found
455.2444.
Data for 26a (derived from 25a): FTIR (film) 3501 (br),
2857, 1778, 1737, 1167 cmꢀ1
;
1H NMR (300 MHz,
CDCl3) d 9.61 (s, 1H), 5.88 (dd, J¼15.6, 5.2 Hz, 1H), 5.72
(dd, J¼16.2, 6.6 Hz, 1H), 4.93 (q, J¼6.6 Hz, 1H), 4.53 (t,
J¼7.2 Hz, 1H), 4.26 (dd, J¼4.3, 1.1 Hz, 1H), 3.99 (dd,
J¼8.2, 3.6 Hz, 1H), 2.53 (td, J¼6.7, 1.7 Hz, 2H), 2.40 (sex-
tet, J¼5.5 Hz, 1H), 2.03–1.90 (m, 1H), 1.42 (s, 6H), 0.89 (s,
9H), 0.08 (s, 7H); EIMS m/z (%) 369 (M+ꢀCH3, 4.1), 327
(5.4), 153 (100), 73 (91.1), 85 (67.5), 75 (57.6), 117 (38.3).
4.1.13. Reduction of 24 with DIBAL-H (25a and 25b). DI-
BAL-H (0.8 mL, 1.0 M, in toluene) was added to a solution
of 24a or 24b (100 mg, 0.22 mmol) in dry CH2Cl2 (5 mL)
stirred in a dry ice–acetone bath under argon. After comple-
tion of the addition, the stirring was continued at the same
temperature for 30 min before quenching the reaction with
MeOH (added dropwise). The reaction mixture was diluted
with EtOAc, washed in turn with satd aq potassium sodium
tartrate and satd aq NaHCO3, and dried over anhydrous
MgSO4. Removal of the solvent left 25a or 25b as a colorless
oil (80 mg, 0.21 mmol, 95% yield).
Data for 26b (derived from 25b): FTIR (film) 3500 (br),
2857, 1778, 1738, 1169 cmꢀ1
;
1H NMR (300 MHz,
CDCl3) d 9.6 (d, J¼1.4 Hz, 1H), 6.31 (dd, J¼17.9, 3.6 Hz,
1H), 5.88–5.75 (m, 1H), 5.67–5.56 (m, 1H), 4.52 (br q,
J¼7.7 Hz, 1H), 4.25–4.23 (m, 1H), 3.97–3.94 (m, 1H),
2.22–1.90 (m, 2H), 1.80–1.65 (m, 2H), 1.42 (s, 6H), 0.89
(s, 9H), 0.08 (s, 6H).
Data for 25a (derived from 24a): FTIR (film) 3423 (br),
2930, 1253, 1054 cmꢀ1
d 6.00–5.65 (m, 2H), 5.58 (br, 1H), 5.50 (br, 1H), 4.68 (q,
J¼6.6 Hz, 1H), 4.55–4.38 (m, 2H), 3.97–3.80 (m, 2H),
3.64 (br, 2H), 2.15–1.90 (m, 4H), 1.42 (s, 3H), 1.40 (s,
3H), 0.91 (s, 9H), 0.11 (s, 7H); ESIMS m/z 411.3
;
1H NMR (300 MHz, CDCl3)
n-BuLi (0.14 mL, 1.6 M in hexane, 0.22 mmol) was added
to a solution of TMSC^CH (32 mL, 0.21 mmol) in dry
THF (2 mL) stirred in a dry ice–acetone bath under argon.
After completion of the addition the stirring was continued
for 15 min before a solution of 26b (138 mg, 0.36 mmol)

4898
C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
in dry THF (2 mL) was introduced via a syringe. The mix-
ture was stirred at the same temperature for 3 h, then at am-
bient temperature overnight. Satd aq NH4Cl was added to
quench the reaction. The mixture was diluted with EtOAc,
washed with satd aq NaHCO3, and dried over anhydrous
MgSO4. Removal of the solvent and column chromato-
graphy on silica gel (1:1 EtOAc/PE) gave 27b as a colorless
oil (70 mg, 0.145 mmol, 40% yield): FTIR (film) 3460,
4.1.17. Oxidation of 29 and addition of TMSC^CH to 30
(31a and 31b). SO3$Py (550 mg, from Acros, 48–53%) in
dry DMSO (4 mL) was added to a solution of 29 (490 mg,
1.20 mmol) in dry CH2Cl2 (4 mL) stirred in an ice-water
bath under argon. When TLC showed completion of the ox-
idation (ca. 15 min), the bath was removed. The reaction
mixture was diluted with Et2O (150 mL), washed with satd
aq NaHCO3 (25 mLꢂ2), and dried over anhydrous
MgSO4. Removal of the solvent and chromatography on sil-
ica gel (1:10 EtOAc/PE) afforded aldehyde 30 as a colorless
1
2928, 2174, 1778, 1462, 1251, 1165, 973 cmꢀ1; H NMR
(300 MHz, CDCl3) d 6.05–5.85 (m, 2H), 5.04–4.96 (m,
1H), 4.60–4.52 (m, 2H), 4.06 (br, 1H), 3.85 (td, J¼7.4,
1.1 Hz, 1H), 3.72–3.66 (m, 1H), 2.60–2.50 (m, 2H), 2.49–
2.38 (m, 1H), 2.10–1.97 (m, 1H), 1.45 (s, 3H), 1.44 (s,
3H), 0.92 (s, 9H), 0.18 (s, 9H), 0.12 (s, 6H). ESIMS 505.2
([M+Na]+), 500.3 ([M+NH+4]); ESIHRMS calcd for
C24H42O6Si2Na ([M+Na]+) 505.24326; found 505.2412.
1
oil (401 mg, 0.98 mmol, 82% yield): H NMR (300 MHz,
CDCl3) d 9.69 (d, J¼1.4 Hz, 1H), 4.38 (dd, J¼7.4, 4.9 Hz,
1H), 4.32 (dd, J¼7.4, 4.0 Hz, 1H), 4.27 (dd, J¼3.8,
1.4 Hz, 1H), 3.88 (d, J¼4.9 Hz, 1H), 2.80–2.60 (m, 4H),
1.42 (s, 3H), 1.36 (s, 3H), 1.272 (t, J¼7.4 Hz, 3H), 1.267
(t, J¼7.4 Hz, 3H), 0.99 (t, J¼7.8 Hz, 9H), 0.67 (q,
J¼8.0 Hz, 6H).
4.1.15. TES protection of 20 (28). A solution of 20 (3.605 g,
9.90 mmol), imidazole (2.841 g, 20.71 mmol), and TESCl
(2.203 g, 14.69 mmol) in dry DMF (4 mL) was stirred at am-
bient temperature under argon overnight. The mixture was
extracted with petroleum ether (10 mL) thrice. The com-
bined petroleum ether layers were diluted with Et2O
(100 mL), washed in turn with 0.5 N HCl and satd aq
NaHCO3 before being dried over anhydrous MgSO4. Rotary
evaporation and column chromatography on silica gel (1:24
EtOAc/PE) gave 21 as a colorless oil (4.660 g, 9.53 mmol,
95% yield): [a]2D5 +61.8 (c 1.0, CHCl3). FTIR (film) 2971,
n-BuLi (2.8 mL, 1.6 M in hexane, 4.5 mmol) was added to
a solution of TMSC^CH (0.65 mL, 4.6 mmol) in dry
THF (50 mL) stirred in a dry ice–acetone bath under argon.
After completion of the addition the stirring was continued
at ꢀ78 ꢁC for 15 min before a solution of 30 (1.340 g,
3.28 mmol) in dry THF (5 mL) was introduced via a syringe.
The mixture was stirred at the same temperature for 3 h. Satd
aq NH4Cl was added to quench the reaction. The mixture
was diluted with Et2O (200 mL), washed with satd aq
NaHCO3, and dried over anhydrous MgSO4. Removal of
the solvent and column chromatography on silica gel (1:20
Et2O/PE) gave 31a (791 mg, 1.56 mmol) and 31b
(547 mg, 1.08 mmol) as colorless oils (81% total yield).
1737, 1480, 1458, 1282, 1159, 878 cmꢀ1
;
1H NMR
(300 MHz, CDCl3) d 4.32–4.25 (m, 3H), 4.10 (dd, J¼11.8,
3.6 Hz, 1H), 3.99 (m, 1H), 3.95 (d, J¼3.1 Hz, 1H), 2.82–
2.70 (m, 4H), 1.46 (s, 3H), 1.36 (s, 3H), 1.28 (t, J¼7.4 Hz,
3H), 1.26 (t, J¼7.4 Hz, 3H), 1.23 (s, 9H), 0.97 (t,
J¼7.8 Hz, 9H), 0.66 (q, J¼7.9 Hz, 6H); EIMS m/z (%)
330 (M+ꢀHCOCH (SEt)2, 3.2), 329 (3.9), 301 (2.1), 271
(24.1), 243 (9.9), 199 (10.4), 169 (10.8), 135 (51.7), 85
(29.3), 57 (100). Anal. Calcd for C23H46O5S2Si: C, 55.83;
H, 9.37. Found C, 55.97; H, 9.09.
Data for 31a (the more polar component): [a]2D0 +25.0 (c
0.82, CHCl3). FTIR (film) 3468 (br), 2958, 2174, 1456,
1250 cmꢀ1 1H NMR (300 MHz, CDCl3) d 4.56 (dd,
;
J¼4.6, 3.6 Hz, 1H), 4.43 (dd, J¼7.4, 2.4 Hz, 1H), 4.33
(dd, J¼7.4, 6.1 Hz, 1H), 4.10 (d, J¼2.4 Hz, 1H), 3.95 (dd,
J¼3.8, 5.9 Hz, 1H), 2.85–2.68 (m, 4H), 2.33 (d, J¼4.3 Hz,
1H, OH), 1.47 (s, 3H), 1.39 (s, 3H), 1.28 (t, J¼7.6 Hz,
3H), 1.27 (t, J¼7.4 Hz, 3H), 0.98 (t, J¼8.0 Hz, 9H), 0.70
(q, J¼8.0 Hz, 6H), 0.18 (s, 9H); 13C NMR (75 MHz,
CDCl3) d 109.8, 102.9, 92.3, 83.4, 79.4, 76.1, 65.7, 53.5,
27.4, 27.1, 14.6, 14.4, 6.9, 5.2, ꢀ0.2; EIMS m/z (%) 445
(2.5), 415 (3.1), 313 (23.9), 254 (23.7), 181 (96.9), 149
(49.9), 135 (40.6), 115 (68.3), 87 (100); ESIMS m/z 524.2
([M+NH4]+). Anal. Calcd for C23H46O4S2Si2: C, 54.50; H,
9.15. Found C, 54.61; H, 9.10.
4.1.16. Removal of Piv protecting group in 28 (29). DI-
BAL-H (50 mL, 1.0 M in cyclohexane) was added to a solu-
tion of 28 (10.216 g, 20.63 mmol) in dry CH2Cl2 (100 mL)
stirred in a dry ice–acetone bath under argon. After comple-
tion of the addition, the stirring was continued at the same
temperature for 30 min before quenching the reaction with
MeOH (added dropwise). Satd aq potassium sodium tartrate
(100 mL) was added. After stirring for 30 min, the mixture
was extracted with Et2O (100 mLꢂ3). The combined ethe-
real layers were washed with brine and dried over anhydrous
MgSO4. Removal of the solvent and chromatography on sil-
ica gel (1:8 EtOAc/PE) afforded 29 as a colorless oil
(7.660 g, 18.68 mmol, 91% yield): [a]2D0 +55.7 (c 1.0,
CHCl3). FTIR (film) 3490 (br), 2957, 1456, 1239, 1085,
Data for 31b (the less polar component): [a]2D0 +15.5 (c 1.03,
CHCl3). FTIR (film) 3471 (br), 2958, 2174, 1250, 742 cmꢀ1
;
1H NMR (300 MHz, CDCl3) d 4.64 (d, J¼2.7 Hz, 1H), 4.44
(dd, J¼7.3, 2.3 Hz, 1H, H-2), 4.24 (dd, J¼8.8, 7.0 Hz, 1H),
4.12 (d, J¼2.7 Hz, 1H), 3.66 (dt, J¼8.9, 3.4 Hz, 1H), 2.82–
2.68 (m, 4H), 2.55 (d, J¼3.7 Hz, 1H, OH), 1.48 (s, 3H), 1.39
(s, 3H), 1.29 (t, J¼7.4 Hz, 3H), 1.28 (t, J¼7.4 Hz, 3H), 0.99
(t, J¼7.6 Hz, 9H), 0.69 (br q, Jz8 Hz, 6H), 0.18 (s, 9H);
EIMS: m/z (%) 415 (19.9), 339 (19.0), 295 (44.3), 241
(47.3), 181 (97.2), 87 (100); ESIMS m/z 524.3
([M+NH4]+). Anal. Calcd for C23H46O4S2Si2: C, 54.50; H,
9.15. Found C, 54.70; H, 9.18.
1
976, 743 cmꢀ1; H NMR (300 MHz, CDCl3) d 4.29 (dd,
J¼7.3, 3.2 Hz, 1H), 4.20 (t, J¼7.0 Hz, 1H), 3.98 (d,
J¼3.1 Hz, 1H, H-3), 3.83 (dt, J¼6.6, 4.5 Hz, 1H), 3.71
(dd, J¼6.1, 4.2 Hz, 2H), 2.83–2.65 (m, 4H), 2.17 (t,
J¼6.4 Hz, 1H, OH), 1.46 (s, 3H), 1.38 (s, 3H), 1.27 (t,
J¼7.4 Hz, 3H), 1.26 (t, J¼7.4 Hz, 6H), 0.99 (t, J¼7.8 Hz,
9H), 0.67 (q, J¼7.5 Hz, 6H); EIMS m/z (%) 411 (M+, 1.8),
291 (86.7), 217 (73.5), 145 (69.8), 135 (62.6), 117 (86.4),
87 (87.8), 75 (62.3), 115 (100). Anal. Calcd for
C18H38O4S2Si: C, 52.64; H, 9.33. Found: C, 52.69; H, 9.19.
4.1.18. Acetylation of 31a (32). Ac2O (2 mL) and DMAP
(100 mg, 0.81 mmol) were added to a solution of 31a

C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
4899
(991 mg, 1.955 mmol) in dry Et3N (3.3 mL) and dry CH2Cl2
(7 mL) stirred at ambient temperature overnight. The solvent
was removed on a rotary evaporator and chromatography on
silica gel (1:11 Et2O/PE) afforded acetate 32 as a colorless oil
(876 mg, 1.596 mmol, 82% yield): [a]2D5 ꢀ14.7 (c 0.70,
1H NMR (300 MHz, CDCl3) d 5.62 (d, J¼2.0 Hz, 1H),
4.34 (dd, J¼6.8, 2.2 Hz, 1H), 4.27 (t, J¼7.2 Hz, 1H), 4.00
(d, J¼2.0 Hz, 1H, H-1), 3.88 (dd, J¼7.9, 2.4 Hz, 1H),
2.82–2.62 (m, 4H), 2.13 (s, 3H), 1.57 (s, 3H), 1.46 (s, 3H),
1.15–1.11 (m, 6H), 1.00 (t, J¼7.3 Hz, 9H), 0.69 (q,
J¼7.2 Hz, 6H), 0.17 (s, 9H); 13C NMR (75 MHz, CDCl3)
d 169.0, 110.1, 98.8, 92.7, 84.5, 78.4, 75.6, 67.0, 54.0,
27.3, 27.0, 25.4, 24.5, 14.5, 14.3, 6.8, 5.0, ꢀ0.4; ESIMS m/z
566.2 ([M+NH4]+); ESIHRMS calcd for C25H48O5S2Si2Na
([M+Na]+) 571.2384; found 571.2374. Anal. Calcd for
C25H48O5S2Si2: C, 54.70; H, 8.81. Found C, 55.18; H, 8.97.
a syringe over 1.5 h (using a syringe pump) to a solution
of (S)-2-methyl-CBS-oxazaborolidine (0.20 mL, 1.0 M in
toluene, 0.20 mmol) and BH3$SMe2 (0.3 mL, 2.0 M in
THF, 0.60 mmol) in dry THF (3 mL) stirred in an ice-water
bath under argon. After completion of the addition, the stir-
ring was continued at the same temperature for 1 h (when
TLC showed completion of the reduction). MeOH was
added to quench the reaction. The solvent was removed by
rotary evaporation and the residue was chromatographed
on silica gel (1:1 Et2O/PE) to give 34 as a colorless oil
(186 mg, 0.33 mmol, 61% yield, 6:1 mixture of the two epi-
mers as shown by 1H NMR), which was used in the next step.
By repeated chromatography a small pure sample of the ma-
jor isomer (expected to be the one with the configuration
drawn in the structure for 34) was obtained, from which
the following physical and spectroscopic data were col-
lected: [a]2D5 ꢀ32.3 (c 0.65, CHCl3). FTIR (film) 3487
CHCl3). FTIR (film) 2958, 2182, 1754, 1249, 743 cmꢀ1
;
(br), 2955, 2181, 1743, 1224.3, 743.7 cmꢀ1 1H NMR
;
4.1.19. Conversion of 32 into 33. NBS (114 mg,
0.64 mmol) was added to a solution of 32 (47 mg,
0.085 mmol) in 5:1 (v/v) MeCN/H2O (3 mL). After stirring
for 7 min, aq Na2SO3 was added, followed by 1:1 (v/v)
n-hexane/CH2Cl2. The phases were separated. The organic
layer was washed with satd aq NaHCO3 and dried over an-
hydrous Na2SO4. The solvent was removed by rotary evapo-
ration and the residue was dissolved in toluene (3 mL) and
treated with 9 (58 mg, 0.15 mmol) at 70–80 ꢁC (bath) until
TLC showed completion of the reaction. The solvent was
removed by rotary evaporation and the residue was chroma-
tographed on silica gel (1:5 EtOAc/PE) to give 33 as a color-
less oil (26 mg, 0.047 mmol, 55% from 32).
(300 MHz, CDCl3) d 5.89 (dd, J¼15.3, 5.0 Hz, 1H), 5.78
(dd, J¼15.8, 5.7 Hz, 1H), 5.47 (d, J¼4.4 Hz, 1H), 4.48 (t,
J¼6.1 Hz, 1H), 4.21 (br q, J¼4.5 Hz, 1H), 4.00 (dd,
J¼5.4, 4.1 Hz, 1H), 3.88 (dd, J¼7.2, 5.7 Hz, 1H), 3.68 (s,
3H), 2.48 (t, J¼7.4 Hz, 2H), 2.12 (s, 3H), 1.93 (d,
J¼4.6 Hz, 1H, OH), 1.92–1.76 (m, 2H), 1.42 (s, 3H), 1.40
(s, 3H), 0.99 (t, J¼8.12 Hz, 9H), 0.68 (q, J¼7.7 Hz, 6H),
0.18 (s, 9H); ESIMS m/z 579.3 ([M+Na]+); ESIHRMS calcd
for C27H48O8Si2Na ([M+Na]+) 579.2780; found 579.2779.
4.1.21. Lactonization of 34 (35). A solution of 34 (208 mg,
0.37 mmol) and PPTS (10 mg, 0.04 mmol) in dry toluene
(5 mL) was stirred at 40–50 ꢁC (bath) until TLC showed
completion of the reaction. The reaction mixture was diluted
with Et2O (100 mL), washed in turn with satd aq NaHCO3
and brine, and dried over anhydrous Na2SO4. Removal of
the solvent on a rotary evaporator and chromatography on sil-
ica gel (1:1 Et2O/PE) gave the lactone 35 as a colorless oil
(115 mg, 0.22 mmol, 59% yield). Data for the major isomer:
[a]2D5 ꢀ48.2 (c 1.65, CHCl3). FTIR (film) 2957, 2181, 1782,
1752, 1223, 846 cmꢀ1; 1H NMR (300 MHz, CDCl3) d 6.00–
5.85 (m, 2H), 5.50 (d, J¼2.9 Hz, 1H), 5.02–4.95 (m, 1H),
4.52 (dd, J¼6.7, 4.1 Hz, 1H), 4.02–3.90 (m, 2H), 2.57 (td,
J¼6.2, 1.6 Hz, 2H), 2.48–2.38 (m, 1H), 2.12 (s, 3H), 2.10–
2.00 (m, 1H), 1.43 (s, 3H), 1.40 (s, 3H), 0.99 (t, J¼7.5 Hz,
9H), 0.68 (q, J¼7.7 Hz, 6H), 0.17 (s, 9H); 13C NMR
(75 MHz, CDCl3) d 176.7, 169.2, 131.7, 129.3, 109.4, 99.1,
92.7, 80.6, 79.6, 74.3, 66.4, 28.4, 26.9, 20.9, 6.8, 4.9, ꢀ0.4.
ESIMS m/z 547.2 ([M+Na]+); ESIHRMS calcd for
C26H44O7Si2Na ([M+Na]+) 547.2518; found 547.2525.
Alternatively, solid I2 (1.330 g, 5.2 mmol) was added in por-
tions to a mixture of 32 (869 mg, 1.59 mmol) and powdered
NaHCO3 (890 mg, 10.59 mmol) in 5:1 (v/v) acetone/water
(15 mL) stirred in an ice-water bath. After completion of
the addition, the mixture was stirred for 15 min before the re-
action was quenched with satd aq Na2S2O3. The mixture was
diluted with Et2O (200 mL), washed in turn with satd aq
Na2S2O3, satd aq NaHCO3 and brine, and dried over anhy-
drous Na2SO4. Removal of the solvent on a rotary evaporator
left the intermediate aldehyde as a yellowish oil (558 mg,
1.26 mmol), which was dissolved in toluene (8 mL) and
treated with Wittig reagent 9 (500 mg, 1.28 mmol) at 70–
80 ꢁC (bath) until TLC showed completion of the reaction.
The solvent was removed by rotary evaporation and the resi-
due was chromatographed on silica gel (1:5 EtOAc/PE) to
give 33 as a colorless oil (581 mg, 1.05 mmol, 83% from
32): [a]2D5 ꢀ31.0 (c 1.0, CHCl3). FTIR (film) 2956, 2182,
1746, 1220, 847 cmꢀ1
;
1H NMR (300 MHz, CDCl3)
4.1.22. Removal of silyl protecting groups in 31b (36).
Powdered K2CO3 (714 mg, 5.17 mmol) was added to a solu-
tion of 31b (2.328 g, 4.60 mmol) in 2:1 (v/v) MeOH/THF
(30 mL) stirred in an ice-water bath. The stirring was contin-
ued at the same temperature for 30 min. The reaction mix-
ture was diluted with EtOAc (200 mL), washed with brine
(30 mLꢂ2), and dried over anhydrous Na2SO4. Removal
of the solvent and chromatography (1:2 EtOAc/PE) afforded
diol 36 as a colorless oil (1.397 g, 4.36 mmol, 95% yield).
d 6.97 (dd, J¼16.0, 4.2 Hz, 1H), 6.45 (dd, J¼15.6, 1.8 Hz,
1H), 5.55 (d, J¼2.2 Hz, 1H), 4.70–4.60 (m, 1H), 4.05–
3.97 (m, 2H), 3.68 (s, 3H), 2.91 (t, J¼6.8 Hz, 2H), 2.64 (t,
J¼6.7 Hz, 2H), 2.11 (s, 3H), 1.41 (s, 3H), 1.39 (s, 3H),
0.99 (t, J¼7.7 Hz, 9H), 0.68 (q, J¼7.9 Hz, 6H), 0.17 (s,
9H); 13C NMR (75 MHz, CDCl3) d 197.4, 173.0, 169.0,
144.0, 128.8, 110.4, 98.7, 93.0, 80.1, 77.0, 76.6, 74.4,
66.3, 51.6, 35.1, 27.5, 26.7, 26.4, 20.9, 6.7, 4.8, ꢀ0.5;
ESIMS m/z 572.3 ([M+NH4]+). Anal. Calcd for
C27H46O8Si2: C, 58.45; H, 8.36. Found: C, 58.35; H, 8.37.
Alternatively, a solution of 31b (1.420 g, 2.81 mmol) and
n-Bu4NF (6 mL, 1.0 M in THF, 6.0 mmol) in THF (10 mL)
was stirred at ambient temperature for 5 min. Satd aq
NH4Cl was added, followed by EtOAc. The phases were
4.1.20. CBS reduction of 33 (34). A solution of ketone 33
(300 mg, 0.54 mmol) in dry THF (4 mL) was added via

4900
C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
1
separated and the organic layer was washed, dried, and chro-
matographed as described above to give diol 36 as a colorless
oil (808 mg, 2.52 mmol, 90% yield).
1250, 745 cmꢀ1; H NMR (300 MHz, CDCl3) d 7.02 (dd,
J¼15.8, 4.5 Hz, 1H), 6.50 (dd, J¼15.9, 1.7 Hz, 1H), 4.88
(ddd, J¼7.6, 4.5, 1.7 Hz, 1H), 3.99 (d, J¼7.7 Hz, 1H),
3.75 (d, J¼9.5 Hz, 1H), 3.68 (s, 3H), 3.67 (d, J¼10 Hz,
1H), 2.94 (s, 1H, OH), 2.92 (t, J¼6.6 Hz, 2H), 2.64 (t,
J¼7.1 Hz, 2H), 1.49 (s, 3H), 1.42 (s, 3H), 0.97 (t,
J¼8.0 Hz, 9H), 0.64 (q, J¼8.0 Hz, 6H), 0.18 (s, 9H); 13C
NMR (75 MHz, CDCl3) d 197.8, 173.1, 144.6, 128.7,
109.3, 102.8, 93.4, 80.5, 76.1, 74.7, 51.6, 34.7, 27.7, 26.7,
26.3, 6.7, 6.6, 5.0, 4.6, ꢀ0.4; ESIMS m/z 530.2
([M+NH4]+); EIMS m/z (%) 497 (M+ꢀCH3, 0.8), 425
(1.5), 407 (3.6), 309 (8.4), 241 (100), 183 (48.3), 115
(87.1); EIHRMS calcd for C24H41O7Si2 (M+ꢀCH3)
497.2391; found 497.2399.
Data for 36: [a]2D0 +69.5 (c 1.66, CHCl3). FTIR (film) 3427
1
(br), 2984, 2929, 2116, 1454, 1243, 1077, 882 cmꢀ1; H
NMR (300 MHz, CDCl3) d 4.64 (ddd, J¼6.3, 4.0, 2.2 Hz,
1H, H-5), 4.43 (dd, J¼6.7, 4.0 Hz, 1H, H-2), 4.21 (dd,
J¼8.6, 6.8 Hz, 1H, H-3), 4.07 (d, J¼4.2 Hz, 1H, H-1),
3.77 (ddd, J¼8.9, 4.5, 1.3 Hz, 1H, H-4), 2.80–2.70 (m, 4H,
SCH2), 2.66 (d, J¼5.8 Hz, 1H, OH), 2.58 (d, J¼1.8 Hz,
1H, H-7), 1.49 (s, 3H), 1.41 (s, 3H), 1.289 (t, J¼7.3 Hz,
3H), 1.281 (t, J¼7.4 Hz, 3H); 13C NMR (75 MHz, CDCl3)
d 110.6, 83.5, 80.9, 78.9, 75.6, 75.0, 64.7, 53.2, 27.3, 27.1,
25.4, 25.0, 14.4, 14.3; EIMS m/z (%) 320 (M+, 3.4), 302
(1.5, M+ꢀH2O), 241 (18.4), 201 (7.9), 167 (10.7), 135
(91.9), 127 (41.5), 59(100); EIHRMS calcd for
C14H24O4S2 320.11160; found 320.11439. Anal. Calcd for
C14H24O4S2: C, 52.47; H, 7.55. Found C, 52.00; H, 8.03.
4.1.25. Tosylation of 38 and desilylation of 39 (40). p-TsCl
(252 mg, 1.31 mmol) was added to a solution of 38 (444 mg,
0.87 mmol) and DMAP (325 mg, 2.66 mmol) in dry CH2Cl2
(1 mL) and Et3N (1 mL) stirred in an ice-water bath. The
bath was removed and the mixture was stirred at ambient
temperature overnight. The reaction mixture was then di-
luted with EtOAc (80 mL), washed with satd aq NH4Cl,
and dried over anhydrous Na2SO4. Removal of the solvent
and chromatography (1:4 EtOAc/PE) afforded tosylate 39
as a yellowish oil (424 mg, 0.64 mmol, 74% yield), which
gave the following data: 1H NMR (300 MHz, CDCl3)
d 7.82 (d, J¼8.3 Hz, 2H), 7.33 (d, J¼8.3 Hz, 2H), 6.89
(dd, J¼15.9, 4.4 Hz, 1H), 6.42 (dd, J¼16.0, 1.5 Hz, 1H),
5.24 (d, J¼2.7 Hz, 1H), 4.63–4.59 (m, 1H), 4.03 (dd,
J¼7.1, 2.8 Hz, 1H), 3.84, (t, J¼7.0 Hz, 1H), 3.68 (s, 3H),
2.92 (t, J¼6.8 Hz, 2H), 2.63 (t, J¼6.5 Hz, 2H), 2.45 (s,
3H), 1.43 (s, 3H), 1.38 (s, 3H), 0.96 (t, J¼7.8 Hz, 9H),
0.67 (q, J¼7.9 Hz, 6H), 0.02 (s, 9H).
4.1.23. Conversion of diol 36 into acetonide 37. A solution
of 36 (121 mg, 0.38 mmol) and PPTS (23 mg, 0.12 mmol) in
2,2-dimethoxypropane (3 mL) was stirred at ambient tem-
perature overnight. The mixture was diluted with EtOAc
(60 mL), washed with satd aq NaHCO3 (10 mLꢂ2), and
dried over anhydrous Na2SO4. Removal of the solvent on
a rotary evaporator and chromatography on silica gel (1:20
EtOAc/PE) gave diacetonide 37 (configuration assigned on
the basis of 2D NMR, marked here using the cepacin A
atom numbering system) as a colorless oil (130 mg,
0.36 mmol, 95% yield): [a]2D0 +105.9 (c 1.0, CHCl3). FTIR
1
(film) 3274, 2986, 2115, 1455, 1227, 1066, 856 cmꢀ1; H
NMR (300 MHz, CDCl3) d 4.94 (dd, J¼5.6, 2.0 Hz, 1H,
H-10), 4.50 (dd, J¼8.8, 7.6 Hz, 1H, H-8), 4.39 (dd, J¼7.6,
2.4 Hz, 1H, H-7), 4.07 (d, J¼2.1 Hz, 1H, H-6), 4.02 (dd,
J¼8.8, 5.6 Hz, 1H, H-9), 2.79–2.69 (m, 4H, SCH2), 2.62
(d, J¼1.9 Hz, 1H, H–C^C), 1.53 (s, 3H), 1.48 (s, 3H),
1.40 (s, 3H), 1.35 (s, 3H), 1.293 (t, J¼7.4 Hz, 3H), 1.286
(t, J¼7.4 Hz, 3H); EIMS m/z (%) 360 (M+, 7.6), 345 (2.5),
299 (4.0), 287 (3.9), 241 (26.5), 167 (92.2), 135 (89.4), 81
(64.4), 43 (100). Anal. Calcd for C17H28O4S2: C, 56.63; H,
7.83. Found C, 56.49; H, 8.03.
The tosylate 39 (641 mg, 0.96 mmol) was dissolved in 2:1
(v/v) MeOH/THF (21 mL). With cooling (ice-water bath)
and stirring, powdered K2CO3 (131 mg, 0.95 mmol) was
added. The mixture was stirred at the same temperature for
25 min before being diluted with EtOAc (100 mL), washed
with brine (30 mLꢂ2), and dried over anhydrous Na2SO4.
Removal of the solvent and chromatography (1:3 EtOAc/
PE) delivered 40 as a colorless oil (523 mg, 0.88 mmol,
92% yield): [a]2D0 ꢀ21.5 (c 0.90, CHCl3). FTIR (film)
3272, 2955, 2120, 1740, 1682, 1371, 1177 cmꢀ1; 1H NMR
(300 MHz, CDCl3) d 7.83 (d, J¼8.5 Hz, 2H), 7.35 (d, J¼
8.5 Hz, 2H), 6.87 (dd, J¼15.9, 4.1 Hz, 1H), 6.41 (dd,
J¼16.1, 1.9 Hz, 1H), 5.28 (t, J¼2.3 Hz, 1H), 4.59 (ddd,
J¼8.5, 4.0, 1.5 Hz, 1H), 4.01 (dd, J¼7.8, 2.1 Hz, 1H),
3.85 (dd, J¼7.6, 7.0 Hz, 1H), 3.69 (s, 3H), 2.92 (t, J¼
6.6 Hz, 2H), 2.64 (t, J¼6.5 Hz, 2H), 2.46 (s, 3H), 2.42 (d,
J¼2.0 Hz, 1H), 1.41 (s, 3H), 1.38 (s, 3H), 0.95 (t,
J¼7.9 Hz, 9H), 0.58 (br q, J¼7.8 Hz, 6H); EIMS m/z (%)
579 (M+ꢀCH3, 3.0), 507 (2.0), 335 (18.0), 257 (100), 183
(23.6), 154 (19.5), 115 (68.3), 94 (20.5). Anal. Calcd for
C29H42O9SSi: C, 58.56; H, 7.12. Found C, 58.18; H, 7.22.
4.1.24. Conversion of 31a into 38. Solid I2 (5.737 g,
22.58 mmol) was added in portions to a mixture of 31a
(3.834 g, 7.58 mmol) and powdered NaHCO3 (4.580 g,
54.52 mmol) in 5:1 (v/v) acetone/water (70 mL) stirred in
an ice-water bath. After completion of the addition, the mix-
ture was stirred for 20 min before the reaction was quenched
with satd aq Na2S2O3. The mixture was diluted with Et2O
(300 mL), washed in turn with satd aq Na2S2O3, satd aq
NaHCO3 and brine, and dried over anhydrous Na2SO4.
Removalof thesolventona rotary evaporatorlefttheinterme-
diate aldehyde as a yellowish oil (558 mg, 1.26 mmol), which
was dissolved in toluene (50 mL) and treated with Wittig re-
agent 9 (3.115 g, 8.19 mmol) at 70–80 ꢁC (bath) until TLC
showed completion of the reaction. The solvent was removed
by rotary evaporation and the residue was chromatographed
on silica gel (1:5 EtOAc/PE) to give 38 (cis/trans¼1:4) as
a colorless oil (2.822 g, 5.51 mmol, 73% from 31a).
4.1.26. Conversion of tosylate 40 into bromoallene 41. A
solution of tosylate 40 (521 mg, 0.88 mmol) in dry THF
(5 mL, another 2 mL to assist the transfer) was added via
a syringe to a dry flask containing CuBr$SMe2 (629 mg,
3.06 mmol, recrystallized) and LiBr (250 mg, 2.87 mmol,
dried) stirred at ambient temperature under argon. The mix-
ture was heated to reflux under argon for 10 h. After being
Data for the trans-isomer (major isomer): [a]2D0 +14.7 (c
0.86, CHCl3). FTIR (film) 3470 (br), 2956, 2170, 1742,

C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
4901
cooled to ambient temperature, the reaction mixture was di-
luted with EtOAc (100 mL), washed with satd aq NH4Cl
(30 mLꢂ2), and dried over anhydrous Na2SO4. Removal
of the solvent and chromatography (1:4 EtOAc/PE) gave
bromoallene 41 as a colorless oil (271 mg, 0.54 mmol,
61% yield) along with unreacted 40 (141 mg, 0.23 mmol).
J¼15.9, 5.5 Hz, 2H), 6.43 (dd, J¼15.9, 1.5 Hz, 1H), 6.02
(dd, J¼5.9, 1.8 Hz, 1H), 5.29 (dd, J¼6.5, 5.9 Hz, 1H),
5.17 (ddd, J¼7.2, 5.6, 1.6 Hz, 1H), 4.57 (ddd, J¼7.5, 5.6,
1.4 Hz, 1H), 3.92 (dd, J¼7.5, 5.6 Hz, 1H), 3.69 (s, 3H),
2.91 (t, J¼6.6 Hz, 2H), 2.64 (t, J¼6.6 Hz, 2H), 2.46 (s,
3H), 1.41 (s, 3H), 1.39 (s, 3H).
Data for 41: [a]2D7 ꢀ53.9 (c 1.59, CHCl3). FTIR (film) 2955,
Step 2 (CBS reduction). The above obtained tosylate 43
(70 mg, 0.13 mmol) was dissolved in dry THF (2 mL) and
added via a syringe over 5 min to a solution of (S)-2-meth-
yl-CBS-oxazaborolidine (0.20 mL, 1.0 M in toluene,
0.20 mmol) and BH3$SMe2 (0.08 mL, 2.0 M in THF) in
dry THF (2 mL) stirred in an ice-water bath under argon. Af-
ter completion of the addition, the stirring was continued at
the same temperature for 10 min. MeOH was added to
quench the reaction. The solvent was removed by rotary
evaporation and the residue was chromatographed on silica
gel (2:3 EtOAc/PE) to give the intermediate alcohol as a
colorless oil (56 mg, 0.10 mmol, 77% from 43) along with
unreduced ketone 43 (13 mg, 0.023 mmol). The following
data were acquired from the intermediate alcohol: [a]D26
ꢀ31.6 (c 0.96, CHCl3). FTIR (film) 3501 (br), 2987, 1966,
2877, 1961, 1740, 1682, 1214, 744 cmꢀ1
;
1H NMR
(300 MHz, CDCl3) d 6.86 (dd, J¼15.9, 4.8 Hz, 1H), 6.45
(dd, J¼15.8, 1.4 Hz, 1H), 6.11 (dd, J¼6.0, 1.7 Hz, 1H),
5.36 (dd, J¼6.9, 5.5 Hz, 1H), 4.62 (ddd, J¼7.8, 4.6,
1.4 Hz, 1H), 4.42 (ddd, J¼8.6, 5.8, 1.5 Hz, 1H), 3.79 (dd,
J¼7.7, 5.2 Hz, 1H), 3.70 (s, 3H), 2.93 (t, J¼6.6 Hz, 2H),
2.65 (t, J¼6.7 Hz, 1H), 1.45 (s, 3H), 1.42 (s, 3H), 0.99 (t,
J¼7.9 Hz, 9H), 0.68 (q, J¼7.7 Hz, 6H); 13C NMR
(75 MHz, CDCl3) d 201.85, 197.46, 172.98, 143.34,
129.04, 110.21, 100.77, 83.39, 77.02, 74.22, 70.78, 51.61,
35.10, 27.58, 26.80, 26.69, 6.64, 4.74; EIMS m/z (%) 487
(M+ꢀCH3, 0.8), 437 (3.0), 435 (3.0), 335 (4.5), 263 (16.7),
261 (16.6), 182 (39.4), 183 (54.4), 184 (7.5), 153 (32.9),
115 (100), 87 (41.6); EIHRMS calcd for C21H32O6BrSi
(M+ꢀCH3) 487.1151; found 487.1186.
1
1736, 1372, 1245, 923 cmꢀ1; H NMR (300 MHz, CDCl3)
d 7.80 (d, J¼8.5 Hz, 2H), 7.36 (d, J¼8.4 Hz, 2H), 5.94 (dd,
J¼5.7, 1.6 Hz, 1H), 5.93 (dd, J¼15.2, 6.1 Hz, 1H), 5.71 (dd,
J¼15.4, 7.7 Hz, 1H), 5.26 (dd, J¼6.8, 5.6 Hz, 1H), 5.13
(ddd, J¼8.4, 5.8, 1.8 Hz, 1H), 4.40 (t, J¼7.9 Hz, 1H), 4.23
(m, 1H), 3.86 (dd, J¼7.9, 5.8 Hz, 1H), 3.69 (s, 3H), 2.50–
2.40 (m, 2H), 2.46 (s, 3H), 2.21 (d, J¼3.9 Hz, 1H), 2.00–1.80
(m, 2H), 1.41 (s, 3H), 1.37 (s, 3H); 13C NMR (75 MHz,
CDCl3) d 202.80, 174.25, 145.10, 138.09, 137.86, 133.7,
129.9, 127.9, 127.1, 127.0, 110.1, 95.5, 81.0, 80.9, 78.6,
78.2, 76.7, 75.2, 70.6, 70.5, 51.6, 31.6, 29.8, 27.0, 26.6, 21.6.
4.1.27. Removal of the TES group in 41 (42). n-Bu4NF
(1.4 mL, 1.0 M in THF) was added to a solution of 41
(672 mg, 1.33 mmol) in THF (10 mL) stirred in an ice-water
bath. After stirring for 6 min, satd aq NH4Cl (2 mL) was in-
troduced, followed by EtOAc (100 mL). The phases were
separated, the organic layer was washed with satd aq
NH4Cl (2 mL) and dried over anhydrous Na2SO4. Removal
of the solvent and chromatography (1:2 EtOAc/PE) afforded
42 as a colorless oil (500 mg, 1.29 mmol, 97% yield): [a]D26
+11.82 (c 1.15, CHCl3). FTIR (film) 3466, 2988, 1960, 1737,
1
1679, 1636, 1373, 1215, 1060, 858, 658 cmꢀ1; H NMR
Step 3 (lactonization). The above obtained intermediate
alcohol (144 mg, 026 mmol) was dissolved in toluene
(6 mL) and treated with PPTS (14 mg, 0.05 mmol) at 40–
50 ꢁC (bath) until TLC showed completion of the reaction.
The reaction mixture was diluted with EtOAc (50 mL),
washed in turn with satd aq NaHCO3 and brine, and dried
over anhydrous Na2SO4. Removal of the solvent and column
chromatography on silica gel (1:1 EtOAc/PE) gave lactone
44 as a colorless oil (111 mg, 0.22 mmol, 85% from the in-
termediate alcohol), from which the following data were
measured: [a]3D0 ꢀ39.4 (c 1.42, CHCl3). FTIR (film) 2987,
(300 MHz, CDCl3) d 6.86 (dd, J¼15.9, 5.2 Hz, 1H), 6.48
(dd, J¼15.9, 1.5 Hz, 1H), 6.22 (dd, J¼5.9, 2.7 Hz, 1H),
5.50 (t, J¼5.7 Hz, 1H), 4.65 (ddd, J¼7.7, 5.0, 1.5 Hz, 1H),
4.55–4.50 (m, 1H), 3.85 (dd, J¼8.0, 5.1 Hz, 1H), 3.69 (s,
3H), 2.92 (t, J¼6.6 Hz, 2H), 2.65 (t, J¼6.7 Hz, 2H), 2.36
(d, J¼4.2 Hz, 1H), 1.48 (s, 3H), 1.44 (s, 3H); 13C NMR
(75 MHz, CDCl3) d 201.3, 197.8, 173.2, 143.0, 129.4,
110.4, 100.1, 82.7, 76.9, 75.5, 68.5, 51.7, 35.3, 27.6, 26.8,
26.7; EIMS m/z (%) 373 (M+ꢀ15, 1.9), 375 (1.9), 241
(21.2), 183 (86.9), 167 (25.9), 115 (76.1), 94 (55.1), 81
(77.3); HRMS calcd for C15H18O6Br (MꢀCH3) 373.0487;
found 373.0316.
1
1966, 1776, 1597, 1456, 1370, 1215, 1095, 666 cmꢀ1; H
NMR (300 MHz, CDCl3) d 7.80 (d, J¼8.4 Hz, 2H), 7.37
(d, J¼7.7 Hz, 2H), 5.98 (dd, J¼5.7, 1.7 Hz, 1H), 5.97 (dd,
J¼16.5, 5.6 Hz, 1H), 5.80 (dd, J¼16.5, 6.5 Hz, 1H), 5.25
(t, J¼6.1 Hz, 1H), 5.12 (ddd, J¼1.4, 5.4, 8.2 Hz, 1H), 5.00
(br q, J¼6.0 Hz, 1H), 4.45 (t, J¼7.3 Hz, 1H), 3.87 (dd,
J¼7.7, 5.7 Hz, 1H), 2.60–2.37 (m, 2H), 2.46 (s, 3H), 2.10–
2.00 (m, 2H), 1.40 (s, 3H), 1.37 (s, 3H).
4.1.28. Synthesis of epoxide 46 from alcohol 42. Step 1 (to-
sylation of 42). DMAP (33 mg, 0.27 mmol) and p-TsCl
(121 mg, 0.63 mmol) were added to a solution of 42
(167 mg, 0.43 mmol) in dry CH2Cl2 (8 mL) and Et3N
(2 mL) stirred in an ice-water bath. After completion of
the addition, the mixture was stirred at ambient temperature
overnight before being diluted with EtOAc (60 mL), washed
with satd aq NH4Cl (2 mL), and dried over anhydrous
Na2SO4. Removal of the solvent and chromatography (1:2
EtOAc/PE) afforded tosylate 43 as a colorless oil (168 mg,
0.31 mmol, 72% from 42), from which the following data
were acquired: [a]2D6 ꢀ43.2 (c 0.91, CHCl3). FTIR (film)
2989, 1967, 1737, 1681, 1648, 1598, 1438, 1371, 1190,
Step 4 (hydrolysis of acetonide). The above obtained lactone
44 (111 mg, 0.22 mmol) was dissolved in dry CH2Cl2
(5 mL), to which BF3$Et2O (5 mL, 0.05 mmol) and
HS(CH2)3SH (50 mL, 0.49 mmol) were added. The mixture
was stirred at ambient temperature until TLC showed com-
pletion of the reaction. The mixture was diluted with EtOAc
(50 mL), washed with brine (10 mLꢂ2), and dried over an-
hydrous Na2SO4. Removal of the solvent and column chro-
matography on silica gel (3:1 EtOAc/PE) gave diol 45 as
671 cmꢀ1
;
1H NMR (300 MHz, CDCl3) d 7.82 (d,
J¼8.3 Hz, 2H), 7.37 (dd, J¼8.3 Hz, 2H), 6.76 (dd,

4902
C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
a colorless oil (72 mg, 0.15 mmol, 69% from 44), from
which the following data were obtained: 1H NMR
(300 MHz, CDCl3) d 7.82 (d, J¼8.0 Hz, 2H), 7.39 (d,
J¼8.1 Hz, 2H), 5.98–5.81 (m, 3H), 5.37 (dd, J¼6.7,
5.9 Hz, 1H), 5.01–4.97 (m, 2H), 4.45 (br s, 1H), 3.72–3.63
(m, 1H), 2.80–2.72 (m, 1H), 2.63–2.38 (m, 3H), 2.47 (s,
3H), 2.10–1.95 (m, 2H).
given above, except that the TMSC^CH was replaced by
TMSC^C–C^CH. Yield: 59%. Data for 52 (a yellowish-
light brown oil): [a]2D5 +45.0 (c 0.02, CHCl3). FTIR (film)
1
2956, 2201, 2104, 1950, 1743, 1251, 846 cmꢀ1; H NMR
(400 MHz, CDCl3) d 6.84 (ddd, J¼15.5, 10.0, 4.8 Hz, 1H),
6.43 (dd, J¼15.8, 5.2 Hz, 1H), 5.54 (d, J¼4.8 Hz, 1H), 4.61–
4.56 (m, 1H), 4.36–4.33 (m, 1H), 3.80–3.75 (m, 1H), 3.69 (s,
3H), 2.92 (t, J¼6.6 Hz, 2H), 2.64 (t, J¼6.6 Hz, 2H), 1.45 (s,
3H), 1.41 (s, 3H), 0.97 (t, J¼6.7 Hz, 9H), 0.64 (q, J¼8.0 Hz,
6H), 0.20 (s, 9H); 13C NMR (100 MHz, CDCl3) d 214.6,
198.0, 173.6, 143.8, 129.8, 110.9, 96.0, 90.9, 88.2, 84.2,
77.7, 76.5, 71.5, 69.5, 52.2, 35.6, 28.2, 27.4, 27.3, 27.26,
7.2, 5.3, ꢀ0.001; ESIMS m/z 567.3 ([M+Na]+); ESIHRMS
calcd for C29H44O6Si2Na ([M+Na]+) 567.2569; found
567.2577.
Step 5 (epoxidation). Powdered K2CO3 (18 mg, 0.13 mmol)
was added to a solution of diol 45 (72 mg, 0.15 mmol) in
50:1 (v/v) Et2O/H2O (5 mL) stirred at ambient temperature
until TLC showed completion of the reaction. The mixture
was diluted with EtOAc (50 mL), washed with brine
(10 mLꢂ2), and dried over anhydrous Na2SO4. Removal
of the solvent and column chromatography on silica gel
(3:1 EtOAc/PE) gave epoxide 46 (with the NMR signals as-
signed on the basis of 2D NMR) as a colorless oil (30 mg,
0.10 mmol, 77% from 45): [a]2D0 ꢀ123.5 (c 0.6, CHCl3).
FTIR (film) 3438 (br), 3050, 2990, 1957, 1770, 1458, 975,
4.1.31. Oxidation of 31a into ketone 31c. Method A. A
mixture of MnO2 (181 mg, 2.0 mmol) and 31a (62 mg,
0.12 mmol) in dry CH2Cl2 (4 mL) was heated to reflux
with stirring until TLC showed disappearance of the starting
material. The solids were filtered off (washing with CH2Cl2).
The filtrate/washings were evaporated to dryness and the
residue was chromatographed on silica gel (1:15 EtOAc/PE)
to give ketone 31c as a yellowish oil (35 mg, 0.07 mmol,
58% yield).
1
897, 657 cmꢀ1; H NMR (300 MHz, CDCl3) d 6.25 (d,
J¼5.7 Hz, 1H, H-12), 6.00–5.89 (m, 2H, H-5 and H-6),
5.16 (dd, J¼7.9, 5.6 Hz, 1H, H-10), 5.03 (m, 1H, H-4),
4.26 (m, 1H, H-7), 3.55 (dd, J¼8.1, 2.0 Hz, 1H, H-9), 3.07
(m, 1H, H-8), 2.62–2.54 (m, 2H, H-2), 2.48 (m, 1H, H-3),
2.10 (d, J¼6.9 Hz, 1H, OH), 2.05 (m, 1H, H-3); 13C NMR
(75 MHz, CDCl3) d 204.7 (C-11), 176.8 (C-1), 131.0 (C-5/
6), 130.0 (C-6/5), 98.1 (C-10), 79.5 (C-4), 74.9 (C-12),
69.7 (C-7), 61.6 (C-8), 52.0 (C-9), 28.4 (C-2), 28.2 (C-3);
EIMS m/z (%) 161 (5.1, C-12 to C-8 fragment), 159 (4.5,
C-12 to C-8 fragment), 142 (5.7), 141 (6.9, C-7 to C-1 frag-
ment), 123 (10.3), 113 (8.7), 95 (30.3), 81 (100); ESIMS m/z
323.1 ([M+Na]+); ESIHRMS calcd for C12H13O4BrNa
([M+Na]+) 322.9878; found 322.9889.
Method B. A solution of 31a (2.300 g, 4.54 mmol) and
Dess–Martin periodinane (1.900 g, 4.48 mmol) in dry
CH2Cl2 (50 mL) was stirred at ambient temperature for
20 min. The reaction mixture was worked up as usual and
chromatographed on silica gel (1:15 EtOAc/PE) to give ke-
tone 31c as a colorless oil (1.650 g, 3.27 mmol, 72% yield):
[a]2D0 +19.7 (c 1.10, CHCl3). FTIR (film) 2960, 2152, 1685,
1
1252 cmꢀ1; H NMR (300 MHz, CDCl3) d 4.51–4.45 (m,
4.1.29. Coupling of 41 with TMSC^CH. PdCl2(PPh3)2
(3.06 mg, 4.35 mmol), CuI (2.72 mg, 14.22 mmol), and
i-Pr2NH (0.26 mL) were quickly added to a solution of 41
(29 mg, 57 mmol) and TMSC^CH (11 mg, 90 mmol) in
dry de-aired EtOAc (2 mL) at ꢀ20 ꢁC under argon with pre-
caution against strong light. The mixture was stirred while
the temperature was allowed to rise to 0 ꢁC over 30 min.
The stirring was then continued at ambient temperature until
TLC showed completion of the reaction (ca. 20 min). Satd
aq NH4Cl was added, followed by Et2O. The phases were
separated. The organic layer was washed with brine and
dried over anhydrous Na2SO4. Removal of the solvent and
column chromatography on silica gel (1:4 Et2O/PE) gave
51 as a yellowish oil (24 mg, 44 mmol, 77%): [a]2D5 +50.0
(c 0.04, CHCl3). FTIR (film) 2956, 2158, 1953, 1742,
2H), 4.42–4.38 (m, 1H), 3.82–3.79 (m, 1H), 2.82–2.64 (m,
4H), 1.46 (s, 3H), 1.40 (s, 3H), 1.28 (t, J¼7.5 Hz, 6H),
0.99 (t, J¼7.9 Hz, 9H), 0.66 (q, J¼7.8 Hz, 6H), 0.27 (s,
9H); ESIMS m/z 527.2 ([M+Na]+). Anal. Calcd for
C23H44O4S2Si2: C, 54.71; H, 8.78. Found C, 55.08; H, 9.17.
4.1.32. DIBAL-H reduction of 31c into 31b. DIBAL-H
(3.5 mL, 1.0 M, 3.50 mmol) was added to a solution of
31c (1.600 g, 3.17 mmol) in dry CH2Cl2 (530 mL) stirred
at ꢀ78 ꢁC. The stirring was then continued at the same tem-
perature for 40 min. Excess DIBAL-H was destroyed by ad-
dition of MeOH. The mixture was diluted with EtOAc
(200 mL), washed in turn with satd aq potassium sodium tar-
trate and brine, and dried over anhydrous Na2SO4. Removal
of the solvent and column chromatography on silica gel
(1:20 Et2O/PE) gave 31b as a colorless oil (1.250 g,
2.48 mmol, 80% yield).
1
1250, 845 cmꢀ1; H NMR (400 MHz, CDCl3) d 6.84 (dd,
J¼15.2, 11.5, 4.6 Hz, 1H), 6.42 (d, J¼15.8 Hz, 1H), 5.54–
5.48 (m, 1.3H), 5.40 (t, J¼7.7 Hz, 0.4H), 4.60–4.56 (m,
1H), 4.34–4.28 (m, 1H), 3.80–3.76 (m, 1H), 3.69 (s, 3H),
2.91 (t, J¼6.7 Hz, 2H), 2.64 (t, J¼6.7 Hz, 2H), 1.45 (s,
3H), 1.42 (s, 3H), 1.00–0.95 (m, 9H), 0.71–0.62 (m, 6H),
0.17 (s, 9H); 13C NMR (100 MHz, CDCl3) d 212.9, 197.8,
173.3, 143.7, 129.5, 110.6, 97.4, 96.3, 95.0, 84.1, 78.7,
71.4, 52.0, 35.4, 27.9, 27.1, 27.1, 27.0, 7.0, 5.0, ꢀ0.002;
ESIMS m/z 543.2 ([M+Na]+); ESIHRMS calcd for
C27H44O6Si2Na ([M+Na]+) 543.2569; found 543.2559.
4.1.33. Conversion of 31b into 380. The procedure, yield
and the cis/trans ratio were the same as that reported for con-
verting 31a into 38, except that the 31a was replaced by 31b.
Data for the trans-isomer (major isomer): [a]2D0 ꢀ7.3 (c 0.97,
CHCl3). FTIR (film) 3497 (br), 2956, 2174, 1743, 1372,
1
1251, 845 cmꢀ1; H NMR (300 MHz, CDCl3) d 7.02 (dd,
J¼16.0, 4.4 Hz, 1H), 6.49 (dd, J¼16.1, 1.6 Hz, 1H), 4.72
(ddd, J¼5.9, 4.2, 1.9 Hz, 1H), 4.64 (d, J¼3.5 Hz, 1H),
3.78 (dd, J¼7.4, 3.6 Hz, 1H), 3.75–3.67 (m, 1H), 3.68 (s,
3H), 2.94 (t, J¼6.8 Hz, 2H), 2.67 (t, J¼6.7 Hz, 2H), 1.45
4.1.30. Coupling of 41 with TMSC^C–C^CH. The pro-
cedure was the same as that employed in the synthesis of 51

C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
4903
(s, 3H), 1.41 (s, 3H), 1.01 (t, J¼7.7 Hz, 9H), 0.699 (t,
J¼7.6 Hz, 3H), 0.691 (q, J¼7.7 Hz, 3H), 0.16 (s, 9H);
ESIMS m/z 535.3 ([M+Na]+). Anal. Calcd for
C25H44O7Si2: C, 58.56; H, 8.65. Found C, 58.27; H, 8.74.
from 53): [a]2D0 ꢀ6.7 (c 1.40, CHCl3). FTIR (film) 3275,
1
2961, 2127, 1743, 1217, 882 cmꢀ1; H NMR (300 MHz,
CDCl3) d 7.18 (s, 2H), 6.71 (dd, J¼15.8, 5.6 Hz, 1H), 6.38
(dd, J¼15.8, 1.1 Hz, 1H), 5.46–5.39 (m, 2H), 4.62 (td,
J¼6.0, 1.0 Hz, 1H), 4.09 (heptet, J¼6.7 Hz, 2H), 4.00 (t,
J¼7.5 Hz, 1H), 3.69 (s, 3H), 2.94–2.88 (m, 3H), 2.63 (t,
J¼6.4 Hz, 2H), 2.40 (d, J¼2.3 Hz, 1H), 2.07 (s, 3H), 1.43
(s, 3H), 1.42 (s, 3H), 1.29–1.24 (m, 18H); EIMS m/z (%)
619 (M+ꢀ15, 5.5), 311 (2.9), 283 (21.6), 267 (47.4), 154
(68.2), 43 (100). Anal. Calcd for C33H46O10S: C, 62.44; H,
7.30. Found C, 62.34; H, 7.48.
4.1.34. Tosylation of 380 and desilylation of 390 (400). The
tosylation procedure and yield were the same as that re-
ported above for converting 38 into 39, except that 38 was
replaced by 380. The following data were acquired from
1
the intermediate 390: H NMR (300 MHz, CDCl3) d 7.83
(d, J¼8.3 Hz, 2H), 7.33 (d, J¼8.0 Hz, 2H), 7.09 (dd,
J¼16.0, 3.8 Hz, 1H), 6.43 (dd, J¼15.9, 1.9 Hz, 1H), 4.83–
4.79 (m, 3H), 4.30 (dd, J¼7.4, 3.5 Hz, 1H), 3.68 (s, 3H),
2.90 (t, J¼6.7 Hz, 2H), 2.62 (t, J¼6.5 Hz, 2H), 2.44 (s,
3H), 1.30 (s, 3H), 1.23 (s, 3H), 0.93 (t, J¼8.0 Hz, 9H),
0.605 (q, J¼7.5 Hz, 3H), 0.597 (q, J¼8.2 Hz, 3H), 0.15 (s,
9H). ESIMS m/z 689.35 ([M+Na]+).
4.1.37. Conversion of 57 into 54. The above mentioned 54
could also be prepared from 57 (derived from 36, vide infra)
following the same procedure employed for conversion of
31a into 38 except that 31a was replaced by 57. The yield
over the two steps from 57 (deprotection of the thioacetal
in 57 and the following Wittig reaction with 9) was 82%.
The desilylation of 390 to give 400 was performed using the
same procedure for converting 39 into 40 with the same
yield.
4.1.38. Conversion of 54 into bromoallene 55. A solution
of tosylate 54 (2.600 g, 4.10 mmol) in dry THF (15 mL, an-
other 10 mL to assist the transfer) was added via a syringe to
a dry flask containing CuBr$SMe2 (629 mg, 3.06 mmol, re-
crystallized) and LiBr (250 mg, 2.87 mmol, dried) stirred at
ambient temperature under argon. The mixture was heated to
reflux under argon for 3 h. After being cooled to ambient
temperature, the reaction mixture was diluted with EtOAc
(400 mL), washed with satd aq NH4Cl (50 mLꢂ2), and dried
over anhydrous Na2SO4. Removal of the solvent and chro-
matography (1:3 EtOAc/PE) gave bromoallene 55 as a color-
less oil (1.520 g, 3.53 mmol, 86% yield): [a]2D0 ꢀ63.6 (c
0.75, CHCl3). FTIR (film) 2989, 1964, 1743, 1682, 1228,
Data for 400: FTIR (film) 3273, 2955, 2117, 1741,
1177 cmꢀ1
;
1H NMR (300 MHz, CDCl3) d 7.84 (d,
J¼8.3 Hz, 2H), 7.34 (d, J¼8.6 Hz, 2H), 7.06 (dd, J¼16.3,
2.8 Hz, 1H), 6.44 (dd, J¼16.1, 1.8 Hz, 1H), 4.86–4.80 (m,
3H), 4.30–4.24 (m, 1H), 3.69 (s, 3H), 2.92 (dt, J¼7.0,
3.1 Hz, 2H), 2.64 (t, J¼6.8 Hz, 2H), 2.53 (d, J¼1.7 Hz,
1H), 2.45 (s, 3H), 1.33 (s, 3H), 1.27 (s, 3H), 0.94 (t,
J¼8.4 Hz, 9H), 0.58 (br q, J¼7.8 Hz, 6H); EIMS m/z (%)
579 (M+ꢀCH3, 3.0), 507 (2.0), 335 (18.0), 257 (100), 183
(23.6), 154 (19.5), 115 (68.3), 94 (20.5); ESIMS m/z
617.20 ([M+Na]+). Anal. Calcd for C29H42O9SSi: C,
58.56; H, 7.12. Found C, 58.21; H, 7.02.
856 cmꢀ1
;
1H NMR (300 MHz, CDCl3) d 6.80 (dd,
J¼15.9, 5.7 Hz, 1H), 6.45 (dd, J¼15.8, 1.5 Hz, 1H), 6.18
(dd, J¼5.2, 1.6 Hz, 1H), 5.54 (td, J¼5.4, 2.1 Hz, 1H), 5.48
(t, J¼5.6 Hz, 1H), 4.54 (td, J¼5.6, 1.3 Hz, 1H), 3.99 (dd,
J¼8.0, 5.5 Hz, 1H), 3.97 (s, 3H), 2.93 (t, J¼6.8 Hz, 2H),
2.67 (t, J¼6.4 Hz, 2H), 2.13 (s, 3H), 1.46 (s, 6H); 13C
NMR (75 MHz, CDCl3) d 202.4, 197.4, 173.1, 169.4,
141.7, 130.12, 110.9, 96.4, 80.8, 77.3, 75.4, 69.3, 51.8,
35.5, 27.6, 26.9, 26.8, 26.73; ESIMS m/z 448.1
([M+NH4]+); ESIHRMS calcd for C18H23O7BrNa
([M+Na]+) 453.0519; found 453.0509. Anal. Calcd for
C18H23O7Br: C, 50.13; H, 5.38. Found C, 50.49; H 5.45.
4.1.35. Desilylation of 380 (53). A solution of 380 (848 mg,
1.66 mmol) and n-Bu4NF (3.5 mL, 1.0 M in THF,
3.5 mmol) in THF (8 mL) was stirred at 0 ꢁC for 10 min.
Satd aq NH4Cl was added, followed by EtOAc. The phases
were separated and the organic layer was washed with brine,
dried over Na2SO4, and chromatographed on silica gel (2:3
EtOAc/PE) to give diol 53 as a colorless oil (460 mg,
1.41 mmol, 85% yield): [a]2D0 ꢀ4.6 (c 1.4, CHCl3). FTIR
1
(film) 3458 (br), 2988, 2115, 1736, 1215 cmꢀ1; H NMR
(300 MHz, CDCl3) d 6.99 (dd, J¼15.6, 4.5 Hz, 1H), 6.48
(dd, J¼15.6, 1.4 Hz, 1H), 4.71 (td, J¼6.4, 1.4 Hz, 1H),
4.63 (br s, 1H), 3.92–3.85 (m, 2H), 3.38 (br d, J¼4.8 Hz,
1H), 3.13 (br d, J¼3.8 Hz, 1H), 2.94 (t, J¼6.2 Hz, 2H),
2.64 (t, J¼6.5 Hz, 2H), 2.58 (d, J¼2.5 Hz, 1H), 1.45 (s,
3H), 1.43 (s, 3H); ESIMS m/z 349.1 ([M+Na]+); ESIHRMS
calcd for C16H22O7Na ([M+Na]+) 349.1258; found
349.1271.
4.1.39. Conversion of 36 into 57. A solution of 36 (118 mg,
0.37 mmol), DMAP (66 mg, 0.54 mmol), and TPSCl
(134 mg, 0.44 mmol) in dry CH2Cl2 (6 mL) was stirred at
40 ꢁC for 6.5 h. Usual aqueous workup and chromatography
(1:8 EtOAc/PE) gave 56 as a colorless oil (185 mg, 0.3
1 mmol, 85% from 36), from which the following data
were obtained: 1H NMR (300 MHz, CDCl3) d 7.19 (s,
2H), 5.41 (t, J¼2.4 Hz, 1H), 4.40 (dd, J¼5.9, 3.7 Hz, 1H),
4.20–4.08 (m, 3H), 4.08 (dd, J¼5.1, 1.8 Hz, 1H), 4.00 (d,
J¼3.8 Hz, 1H), 2.91 (heptet, J¼6.6 Hz, 1H), 2.80–2.64
(m, 4H), 2.55 (d, J¼5.4 Hz, 1H), 2.48 (d, J¼2.1 Hz, 1H),
1.47 (s, 3H), 1.35 (s, 3H), 1.33–1.25 (m, 24H).
4.1.36. Conversion of 53 into 54. A solution of 53 (384 mg,
1.18 mmol), DMAP (287 mg, 2.35 mmol), and TPSCl
(400 mg, 1.32 mmol) in dry CH2Cl2 (14 mL) was stirred at
40 ꢁC overnight. When TLC showed completion of the
reaction, the heating bath was removed. Another portion
of DMAP (200 mg, 1.63 mmol) and Ac2O (150 mg,
1.47 mmol) were introduced. The mixture was stirred at am-
bient temperature until TLC showed completion of the reac-
tion. Usual workup and chromatography (1:4 EtOAc/PE)
afforded 54 as a colorless oil (476 mg, 0.75 mmol, 64%
The above prepared alcohol 56 (68 mg, 0.12 mmol) was dis-
solved in dry CH2Cl2 (3 mL), to which DMAP (33 mg,
0.27 mmol) and Ac2O (30 mg, 0.29 mmol) were introduced
in turn. The mixture was stirred at ambient temperature until
TLC showed completion of the reaction. Usual aqueous

4904
C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
workup and chromatography (1:10 EtOAc/PE) gave 57 as
a colorless oil (60 mg, 0.095 mmol, 79% from 56).
Na2SO4. Removal of the solvent and chromatography (2:3
EtOAc/PE) gave 60 as a colorless oil (458 mg, 1.28 mmol,
75% from 59): [a]2D0 +16.8 (c 1.20, CHCl3). FTIR (film)
1
The above two-step transformation could be better fulfilled in
a one-pot way. A solution of 36 (1.700 g, 5.31 mmol), DMAP
(1.400 g, 11.47 mmol), and TPSCl (2.700 g, 8.94 mmol,
added in portions) in dry CH2Cl2 (50 mL) was stirred at
50 ꢁC (bath temperature) for 2.5 h. After the mixture
was cooled to ambient temperature, DMAP (0.400 g,
3.27 mmol) and Ac2O (762 mg, 7.47 mmol) were added.
The mixturewas stirred at ambient temperature for 4 h. Usual
aqueous workup and chromatography (1:10 EtOAc/PE) de-
livered 57 as a colorless oil (3.131 g, 4.98 mmol, 94% over
the two steps from 36 via 56): [a]2D0 +22.7 (c 1.22, CHCl3).
3446 (br), 2987, 1960, 1771, 1181, 874 cmꢀ1; H NMR
(400 MHz, CDCl3) d 6.18 (dd, J¼5.7, 1.9 Hz, 1H), 5.95
(dd, J¼15.5, 5.6 Hz, 1H), 5.87 (dd, J¼15.6, 5.8 Hz, 1H),
5.47 (t, J¼5.5 Hz, 1H), 5.00 (q, J¼6.9 Hz, 1H), 4.51–4.42
(m, 2H), 3.81 (dd, J¼7.9, 4.7 Hz, 1H), 2.80 (br s, 1H),
2.60–2.54 (m, 2H), 2.50–2.38 (m, 1H), 2.07–1.98 (m, 1H),
1.45 (s, 3H), 1.43 (s, 3H); 13C NMR (100 MHz, CDCl3)
d 201.3, 147.0, 131.1, 130.7, 109.9, 100.1, 82.9, 79.5,
77.0, 75.4, 68.2, 28.4, 28.3, 27.0, 26.9; MALDI-MS m/z
381.1 ([M+Na]+); MALDI-HRMS calcd for C15H19BrO5
([M+Na]+) 381.0299; found 381.0291. Anal. Calcd for
C15H19BrO5: C, 50.15; H, 5.33. Found C, 49.77; H, 5.44.
1
FTIR (film) 3275, 2963, 2127, 1755, 1214 cmꢀ1; H NMR
(300 MHz, CDCl3) d 7.18 (s, 2H), 5.48 (t, J¼3.3 Hz, 1H),
5.39 (dd, J¼6.3, 2.8 Hz, 1H), 4.44–4.37 (m, 2H), 4.10 (hep-
tet, J¼6.7 Hz, 2H), 3.83 (d, J¼2.8 Hz, 1H), 2.91 (heptet,
J¼6.5 Hz, 1H), 2.80–2.60 (m, 4H), 2.45 (d, J¼2.3 Hz, 1H),
2.08 (s, 3H), 1.46 (s, 3H), 1.37 (s, 3H), 1.29–1.23 (m,
24H); EIMS m/z (%) 509 (31.4), 435 (12.45), 267 (21.6),
135 (41.1), 43 (100). Anal. Calcd for C31H48O7S3: C,
59.20; H, 7.69. Found C, 59.54; H, 7.44.
4.1.42. TBS protection of 60 and subsequent coupling
leading to 62. 2,6-Lutidine (0.34 mL, 2.92 mmol) and
TBSOTf (0.30 mL, 1.31 mmol) were added to a solution
of 60 (217 mg, 0.60 mmol) in dry CH2Cl2 (8 mL) stirred at
ꢀ78 ꢁC under argon. The bath temperature was then allowed
to rise naturally to 0 ꢁC. When TLC showed completion of
the reaction, the mixture was diluted with Et2O, washed
with brine, and dried over anhydrous Na2SO4. Removal of
the solvent and chromatography (1:2 Et2O/PE) afforded 61
as a colorless oil (228 mg, 0.48 mmol, 80% yield), from
which the following data were obtained: 1H NMR
(300 MHz, CDCl3) d 6.08 (dd, J¼5.6, 1.9 Hz, 1H), 5.90 (t,
J¼4.6 Hz, 2H), 5.34 (dd, J¼6.3, 5.7 Hz, 1H), 4.53 (dd,
J¼7.8, 5.0 Hz, 1H), 4.49–4.40 (m, 1H), 3.79–3.73 (m,
1H), 2.60–2.52 (m, 2H), 2.48–2.36 (m, 1H), 2.10–2.00 (m,
1H), 1.430 (s, 3H), 1.417 (s, 3H), 0.93 (s, 9H), 0.129 (s,
3H), 0.112 (s, 3H).
4.1.40. CBS reduction of 55 (59). A solution of ketone 55
(864 mg, 0.20 mmol) in dry THF (10 mL) was added via
a syringe (driven by a syringe pump) over 10 min to a solu-
tion of (R)-2-methyl-CBS-oxazaborolidine (2.3 mL, 1.0 M
in toluene, 2.3 mmol) and BH3$SMe2 (1.1 mL, 2.0 M in
THF) in dry THF (5 mL) stirred in an ice-water bath under
argon. After completion of the addition, the stirring was con-
tinued at the same temperature for 1 min. MeOH was added
to quench the reaction. The solvent was removed by rotary
evaporation and the residue was chromatographed on silica
gel (2:3 EtOAc/PE) to give the alcohol 59 as a colorless
oil (721 mg, 1.70 mmol, 83% yield). The de value of this re-
duction was ca. 86% by HPLC analysis.
The above prepared 61 (228 mg, 0.48 mmol) was then trans-
formed to 62 (192 mg, 0.37 mmol, 77% yield) using the
same procedure employed above for converting 41 to 51
(with 41 and TMSC^CH replaced by 60 and TMSC^C–
C^CH, respectively).
Data for 59: [a]2D0 ꢀ50.1 (c 1.05, CHCl3). FTIR (film) 3482
(br), 1964, 1739, 1229, 876 cmꢀ1 1H NMR (300 MHz,
;
CDCl3) d 6.16 (td, J¼5.4, 1.1 Hz, 1H), 5.89 (dd, J¼15.5,
5.9 Hz, 1H), 5.75 (ddd, J¼15.4, 6.9, 1.2 Hz, 1H), 5.55–
5.39 (m, 2H), 4.37 (t, J¼7.7 Hz, 1H), 4.24 (br s, 1H), 3.92
(dd, J¼8.1, 4.7 Hz, 1H), 3.70 (s, 3H), 2.48 (td, J¼7.3,
1.9 Hz, 2H), 2.12 (s, 3H), 2.13–2.05 (m, 1H), 1.99–1.77
(m, 2H), 1.45 (s, 3H), 1.43 (s, 3H); 13C NMR (75 MHz,
CDCl3) d 202.2, 174.1, 169.5, 137.6, 127.1, 109.8, 96.2,
80.8, 78.1, 74.9, 70.4, 69.1, 51.6, 31.6, 29.8, 26.9, 26.6,
20.7; MALDI-MS m/z 455.1 ([M+Na]+); MALDI-HRMS
calcd for C18H25BrO7Na ([M+Na]+) 455.0675; found
455.0686. Anal. Calcd for C18H25BrO7: C, 49.90; H, 5.82.
Found C, 50.05; H, 5.91.
Data for 62 (an essentially colorless oil, chromatography,
eluting with 1:3 Et2O/PE): [a]2D5 +25.86 (c 0.47, CHCl3).
FTIR (film) 2958, 2200, 2100, 1949, 1783, 1251,
851 cmꢀ1 1H NMR (300 MHz, CDCl3) d 6.50 (ddd,
;
J¼6.4, 2.0, 0.7 Hz, 1H), 5.92 (dd, J¼3.0, 1.8 Hz, 2H),
5.86 (dd, J¼1.5, 0.5 Hz, 1H), 5.55 (ddd, J¼8.0, 6.6,
1.8 Hz, 1H), 5.03–4.94 (m, 1H), 4.55–4.48 (m, 1H), 4.42–
4.35 (m, 1H), 3.80–3.74 (m, 1H), 2.60–2.50 (m, 2H),
2.48–2.35 (m, 1H), 2.08–1.95 (m, 1H), 1.44–1.42 (three sin-
glets, 6H), 0.92 (s, 9H), 0.22 (s, 4.5H), 0.20 (s, 4.5H), 0.17 (s,
3H), 0.14 (s, 3H); ESIMS m/z 537.3 ([M+Na]+); ESIHRMS
calcd for C28H42O5Si2Na, 537.2463; found, 537.2486.
4.1.41. Synthesis of 60 from 59. A solution of 59 (721 mg,
1.70 mmol) in 0.2 M methanolic MeONa (15 mL) was
stirred at 0 ꢁC for 25 min. Satd aq NH4Cl was then added
to quench the reaction. The reaction mixture was diluted
with EtOAc, washed with brine, and dried over anhydrous
Na2SO4. The solvent was removed on a rotary evaporator
and the residue was dissolved in toluene (20 mL) and treated
with PPTS (30 mg, 0.11 mmol) at 50–60 ꢁC with stirring for
3 h. The mixture was diluted with EtOAc, washed in turn
with satd aq NaHCO3 and brine, and dried over anhydrous
4.1.43. Synthesis of 64 from 60. The procedure was the
same as that employed for converting 41 to 51 (with 41
and TMSC^CH replaced by 60 and TMSC^C–C^CH,
respectively). Yield (chromatography using 1:2 EtOAc/
PE): 54%.
Data for 64 (an essentially colorless oil, chromatography,
eluting with 1:3 Et2O/PE): [a]2D0 ꢀ31.0 (c 0.40, CHCl3).
FTIR (film) 3448 (br), 2986, 2201, 2102, 1948, 1775,
1
1251, 845 cmꢀ1; H NMR (300 MHz, CDCl3) d 5.97 (d,

C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
4905
5. Schoerkhuber, W.; Zbiral, E. Liebigs Ann. Chem. 1980, 1455–
1469.
6. Ireland, R. E.; Countney, L.; Fitzsimmons, B. J. J. Org. Chem.
1983, 48, 5186–5198.
7. Liu, H.-J.; Yip, J.; Shia, K.-S. Tetrahedron Lett. 1997, 38,
2253–2256.
8. Kiso, M.; Hasegawa, A. Carbohydr. Res. 1976, 52, 95–101.
9. Cason, J. Org. Synth. Coll. Vol. 1955, 3, 169–171.
10. Ronald, R. C.; Wheeler, C. J. J. Org. Chem. 1983, 48, 138–139.
11. Macdonald, S. F. Can. J. Chem. 1974, 52, 3257–3258.
12. Enholm, E. J.; Trivellas, A. J. Am. Chem. Soc. 1989, 111, 6463–
6465.
J¼15.8, 5.0 Hz, 1H, part of AB system), 5.87 (d, J¼16.0,
6.2 Hz, 1H, part of AB system), 5.67–5.63 (m, 1H, H-12),
5.59 (td, J¼6.7, 1.1 Hz, 1H, H-10), 5.01 (br q, J¼7.1 Hz,
1H, H-4), 4.51 (dd, J¼8.2, 7.2 Hz, 1H, H-7), 4.46–4.43
(m, 1H, H-9), 3.81 (dd, J¼8.0, 4.1 Hz, 1H, H-8), 2.61–
2.52 (m, 2H), 2.50–2.38 (m, 1H), 2.34 (d, J¼4.1 Hz, 1H,
OH-H), 1.46 (s, 3H), 1.45 (s, 3H), 0.21 (s, 9H); 13C NMR
(75 MHz, CD3OD) d 215.8 (C-11), 180.0 (C-1), 133.0,
132.8, 111.1 (C-18), 96.6 (C-10), 90.7 (C-15), 89.1 (C-16),
84.7 (C-8), 82.1 (C-4), 79.5 (C-12), 78.0 (C-7), 76.4 (C-
13), 70.7 (C-14), 70.5 (C-9), 29.7, 29.6, 27.6, 27.5, ꢀ0.1
(with the NMR signals assigned on the basis of DEPT and
2D spectra); ESIMS m/z 423.2 ([M+Na]+); ESIHRMS calcd
for C22H28O5SiNa ([M+Na]+) 423.1598; found 423.1586.
13. White, J. D.; Jensen, M. S. J. Am. Chem. Soc. 1995, 117, 6224–
6233.
14. Dondoni, A.; Marra, A.; Merino, P. J. Am. Chem. Soc. 1994,
116, 3324–3336.
15. (a) Luche, J.-L. J. Am. Chem. Soc. 1978, 100, 2226–2227; (b)
Gemal, A. L.; Luche, J.-L. J. Am. Chem. Soc. 1981, 103, 5454–
5459.
16. Xiao, X.; Bai, D. Synlett 2001, 535–537.
17. Murugensan, V.; Pandurangan, A. Indian J. Chem. 1992, 31B,
377–378.
18. For another failure with QFC oxidation, see: Wu, Y.-K.; Shen,
X.; Yang, Y.-Q.; Hu, Q.; Huang, J.-H. J. Org. Chem. 2004, 69,
3857–3865.
19. Kutterer, K. M. K.; Just, J. Heterocycles 1999, 51, 1409–1420.
20. (a) Miyashita, M.; Yoshikoshi, A.; Grieco, P. A. J. Org. Chem.
1977, 42, 3772–3774; (b) Sterzycki, R. Synthesis 1979, 724–
725.
21. (a) Frigerio, M.; Santagostino, M. Tetrahedron Lett. 1994, 35,
8019–8022; (b) De Munari, S.; Frigerio, M.; Santagostino, M.
J. Org. Chem. 1996, 61, 9272–9279; (c) Frigerio, M.;
Santagostino, M.; Sputore, S. J. Org. Chem. 1999, 64, 4537–
4538.
22. (a) Corey, E. J.; Suggs, J. W. Tetrahedron Lett. 1975, 2647–
2650; (b) Piancatelli, G.; Scettri, A.; D’Auria, M. Synthesis
1982, 245–258.
23. (a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155–
4156; (b) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991,
113, 7277–7278; (c) Ireland, R. E.; Liu, L.-B. J. Org. Chem.
1993, 58, 2899; (d) Stevenson, P. J.; Treacy, A. B.;
Nieuwenhuyzen, M. J. Chem. Soc., Perkin Trans. 2 1997,
589–591.
4.1.44. Treatment of 64 with p-TsCl leading to 66. DMAP
(5 mg, 0.04 mmol) and p-TsCl (33 mg, 0.173 mmol) were
added to a solution of 64 (70 mg, 0.18 mmol) in dry Et3N
(0.1 mL) and CH2Cl2 (4 mL) stirred at 0 ꢁC. After comple-
tion of the addition, the mixture was stirred at the same tem-
perature for 2 h before the bath was removed. The stirring
was continued at ambient temperature until TLC showed
disappearance of the starting material. The mixture was
diluted with EtOAc, washed with brine, and dried over anhy-
drous Na2SO4. Removal of the solvent and chromatography
(1:3 EtOAc/PE) afforded 66 as a colorless oil (51 mg,
0.14 mmol, 75% yield): [a]2D0 +78.8 (c 0.90, CHCl3). FTIR
(film) 2930, 2182, 2169, 2073, 1778, 1372 cmꢀ1; UV
(EtOH) lmax¼341, 318, 298, 281, 252, 239, 220, 213,
1
198 nm; H NMR (300 MHz, CDCl3) d 6.31 (dd, J¼15.8,
5.7 Hz, 1H), 5.95–5.76 (m, 3H), 4.99 (q, J¼5.9 Hz, 1H, H-
4), 4.18–4.09 (m, 2H, H-7, 8), 2.60–2.53 (m, 2H), 2.49–
2.40 (m, 1H), 2.07–1.97 (m, 1H), 1.44 (s, 3H), 1.43 (s, 3H),
0.21 (s, 9H); 13C NMR (75 MHz, CDCl3) d 176.5, 143.4,
132.2, 128.5, 111.1, 110.2, 89.4, 87.9, 80.8, 80.7, 78.9,
75.8, 74.4, 67.1, 61.2, 28.4, 28.1, 26.9, 26.7, 0.6; ESIMS
m/z 400.2 ([M+NH4]+); ESIHRMS calcd for C22H26O4SiNa
([M+Na]+) 405.1493; found 405.1496.
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (20025207, 20272071, 20372075,
20321202, 20672129, 20621062) and the Chinese Academy
of Sciences (‘Knowledge Innovation’, KJCX2.YW.H08).
Mr. Ya-Jun Jian is thanked for assistance in correcting
typographic errors.
24. (a) Parikh, J. R.; Doering, W. E. J. Am. Chem. Soc. 1967, 89,
5505–5507; (b) Tidwell, T. T. Org. React. 1990, 39, 297–572
(a review).
25. Coery, E. J.; Erickson, B. W. J. Org. Chem. 1971, 19, 3553–
3560.
26. Armstrong, R. W.; Teegarden, B. R. J. Org. Chem. 1992, 57,
915–922.
References and notes
27. Shi, X.-X.; Wu, Q.-Q. Synth. Commun. 2000, 30, 4081–
4086.
28. (a) Loudon, G. M.; Radhakrishna, A. S.; Almond, M. R.;
Blodgett, J. K.; Boutin, R. H. J. Org. Chem. 1984, 49, 4272–
4276; (b) Stork, G.; Zhao, K. Tetrahedron Lett. 1989, 30,
287–290.
29. (a) Nicolaou, K. C.; Bunnage, M. E.; McGarry, D. G.; Shi, S.;
Somers, P. K.; Wallace, P. A.; Chu, X.-J.; Agrios, K. A.;
Gunzner, J. L.; Yang, Z. Chem.—Eur. J. 1999, 5, 599–617;
(b) Russell, G. A.; Ochrymowy, L. A. J. Org. Chem. 1969,
34, 3618–3624.
1. Parker, W. L.; Rathnum, M. L.; Seiner, V.; Trejo, W. H.;
Principe, P. A.; Sykes, R. B. J. Antibiot. 1984, 37, 431–440.
2. (a) Bu’Lock, J. D.; Jones, E. R. H.; Leeming, P. R. J. Chem. Soc.
1955, 4270–4276; (b) Bu’Lock, J. D.; Jones, E. R. H.; Leeming,
P. R.; Thompson, J. M. J. Chem. Soc. 1956, 3767–3771.
3. (a) Lowe, G. J. Chem. Soc., Chem. Commun. 1965, 411–413;
(b) Brewster, J. H. Topics in Stereochemistry; Allinger, N. L.,
Eliel, E. L., Eds.; John Wiley and Sons: New York, NY,
1967; Vol. 2, pp 1–72.
30. (a) Corey, E. J.; Bakshi, R. K.; Shibata, S. J. Am. Chem. Soc.
1987, 109, 5551–5553; (b) Corey, E. J.; Bakshi, R. K.;
4. For a communication on part of the present work, see: Tang,
C.-J.; Wu, Y.-K. Helv. Chim. Acta 2004, 87, 667–673.

4906
C.-J. Tang, Y. Wu / Tetrahedron 63 (2007) 4887–4906
34. Jones, C. E.; Kondrick, D. A.; Holmes, A. B. Org. Synth. 1987,
Shibata, S.; Chen, C.-P.; Singh, V. K. J. Am. Chem. Soc.
1987, 109, 7925–7926; (c) For a recent review: Corey,
E. J.; Helal, C. J. Angew. Chem., Int. Ed. 1998, 36, 1986–
2012.
65, 52–60.
35. (a) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett.
1975, 4467–4470; (b) Sonogashira, K. J. Organomet. Chem.
2002, 653, 46–49 (a review).
36. Andrus, M. B.; Lepore, S. D.; Turner, T. M. J. Am. Chem. Soc.
1997, 119, 12159–12169.
31. The natural cepacin A is a 1:1 mixture at this position. We
chose to do asymmetric reduction at this stage purely for
facilitating spectral assignment.
32. (a) D’Aniello, F.; Mann, A.; Taddei, M.; Wermuth, C.-G.
Tetrahedron Lett. 1994, 35, 7775–7778; (b) D’Aniello, F.;
Schoenfelder, A.; Mann, A.; Taddei, M. J. Org. Chem. 1996,
61, 9631–9639.
33. Bartik, B.; Dembinski, R.; Bartik, T.; Arif, A. M.; Gladysz,
J. A. New J. Chem. 1997, 21, 739–750.
37. We also tried to use TES in place of TBS as the protecting
group, but the removal was just as difficult as the TBS.
Interestingly, the corresponding coupling products (with the
C-9 OH either TBS- or TES-protected) from TMSC^CH
instead of TMSC^C–C^CH could be readily desilylated to
obtain the alcohol.