Anodic cyclization reactions: probing the chemistry of N,O-ketene acetal derived radical cations

Article abstract of DOI:10.1016/j.tet.2006.04.009

The chemical reactivity of radical cations derived from N,O-ketene acetals has been examined and compared with the reactivity of radical cations derived from both ketene dithioacetals and enol ethers. Synthetically, the N,O-ketene acetal radical cations lead to more efficient cyclization reactions than either the ketene dithioacetal or enol ether derived radical cations. Cyclic voltammetry experiments using allylsilane trapping groups show that the efficiency of these cyclizations is not due to the N,O-ketene acetal radical cations being more reactive but rather more stable to decomposition. Finally, cyclizations using chiral oxizolidinones were examined.

Full text of DOI:10.1016/j.tet.2006.04.009

Tetrahedron 62 (2006) 6536–6550
Anodic cyclization reactions: probing the chemistry of
N,O-ketene acetal derived radical cations
*
Yung-tzung Huang and Kevin D. Moeller
Department of Chemistry, Washington University in St. Louis, St. Louis, MO 63130, USA
Received 27 June 2005; accepted 21 September 2005
Available online 4 May 2006
Abstract—The chemical reactivity of radical cations derived from N,O-ketene acetals has been examined and compared with the reactivity of
radical cations derived from both ketene dithioacetals and enol ethers. Synthetically, the N,O-ketene acetal radical cations lead to more
efficient cyclization reactions than either the ketene dithioacetal or enol ether derived radical cations. Cyclic voltammetry experiments using
allylsilane trapping groups show that the efficiency of these cyclizations is not due to the N,O-ketene acetal radical cations being more reactive
but rather more stable to decomposition. Finally, cyclizations using chiral oxizolidinones were examined.
Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction
from ketene dithioacetal oxidation. For example, the differ-
ence in stereoselectivity obtained for the oxidation of 3a and
3b is most consistent with the oxidation of 3b leading to a
reaction controlled by thermodynamics. This hint that the
enol ether and ketene dithioacetal reactions might be gov-
erned by different mechanistic concerns suggests that while
the reactions look very similar, the radical cation inter-
mediates might be quite different.
Intramolecular anodic olefin coupling reactions using ketene
dithioacetal initiating groups often afford high levels of stereo-
selectivity.1–4 Consider the reactions illustrated in Scheme 1.
In both examples, the reaction originating from oxidation of
the ketene dithioacetal was far more selective than the anal-
ogous reaction originating from oxidation of the enol ether.
While the stereochemistry of reactions originating from enol
ether oxidation can often be rationalized by kinetic consider-
ations,5 the same cannot be said for reactions originating
It did not take long to identify other significant differences
between enol ether and ketene dithioacetal derived cycliza-
tions. Enol ethers undergo efficient oxidative cyclizations
with less reactive allylsilane trapping groups while ketene di-
thioacetals do not (Scheme 2).6 This is especially true for cy-
clizations utilizing trisubstituted allylsilane trapping groups.
The oxidation of 7b led to a product that was so messy that
the cyclic product (8b) could not be isolated in pure form.
These observations led to a working model for the cycliza-
tions that treated the radical cation derived from oxidation
of a ketene dithioacetal as being less reactive than the radical
cation derived from oxidation of an enol ether.
RVC anode
Pt cathode
0.03 M Et NOTS
4
Me
Me
Me
X
Y
O
X
O
H
Me
30% MeOH/THF
2,6-lutidine
8 mA/ 2 F/mole
Y
MeO
1a. X=OMe, Y=H
1b. X=Y= -S(CH2)3S-
2a. 74% (3:1 t/c)
2b. 83% (only trans)
RVC anode
Pt cathode
X
X
( )n
OMe
( )n
OMe
Y
0.1 M LiClO4
RVC anode
X
X
MeO
50% MeOH/ THF
2,6-lutidine
28.6 mA/ 2.2 F/mole
Y
MeO
Pt wire cathode
0.1 M Et4NOTs
MeO
Y
Y
CH2TMS
30-50% MeOH/THF
2,6-lutidine
8.0 mA, 2.2 F/mole
CH2
3a. X=OMe, Y=H
3b. X=Y= -S(CH2)3S-
4a. 82% (1:1 ratio of isomers)
4b. 75% (95:5 ratio of isomers)
R
R
Scheme 1.
5a. X=H, Y=OMe, R=H
5b. X=H, Y=OMe, R=CH3
6a. 86%
6b. 66%
8a. 55%
Keywords: Anodic electrochemistry; Oxidative cyclizations; Radical cat-
ions; Ketene acetals.
* Corresponding author. Tel.: +1 314 935 4270; fax: +1 314 935 4481;
e-mail: moeller@wustl.edu
7a. X=Y= S(CH2)3S, R=H
8b. (20-30%)
7b. X=Y= S(CH2)3S, R=CH3
Scheme 2.
0040–4020/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2006.04.009

Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
6537
The suggestion that the reactivity of a radical cation can be
altered by varying the substituents bound to it is very intrigu-
ing. If a change from a methoxy substituent to two thioether
groups caused the radical cation to be less reactive, then pos-
sibly other changes would lead to a more reactive radical cat-
ion. Such a discovery would be extremely useful. Consider
the retrosynthetic analysis of scopadulcic acid B illustrated
in Scheme 3.7 The key step in this proposed synthesis would
involve an anodic cyclization of 11 to form the bridged bi-
cyclic intermediate 12. A hydroboration/oxidation sequence
would then convert the olefin to an aldehyde for use in an in-
tramolecular aldol condensation. The strength of the route is
that the oxidative cyclization reaction would lead to a prod-
uct having all of the carbons needed for completing the third
ring. However, the proposal is badly flawed in that oxidative
cyclization reactions using less reactive allylsilane trapping
groups are normally not compatible with the simultaneous
formation of six-membered rings and quaternary centers.6b,8
But is this always the case? Are there more reactive radical
cations that would allow for this transformation? With these
questions in mind, we began to expand our exploration of
oxidative cyclization reactions by diversifying the nature
of the radical cation intermediate.
asymmetric alkylation and aldol reactions. Their availability
suggested the possibility of studying asymmetric oxidative
cyclization reactions (Scheme 4).
Initially, the very reactive enol ether trapping group was se-
lected for the cyclizations.9 This was done in order to probe
the compatibility of the N,O-ketene acetal groups with the
anodic oxidations using substrates that had the best possible
chance for cyclization. The substrates were synthesized as
outlined in Scheme 5. Both five- and six-membered ring pre-
cursors were made by nearly identical routes. The only dif-
ference was that the six-membered ring precursor required
a one carbon chain extension. The yields in the scheme for
steps e–h are given for the synthesis of the five-membered
ring substrate. Yields for the six-membered ring substrate
are included in Section 7.
O
OH
a,b
O
(
)
n
OMe
17a. n=1
16
c,d
17b. n=2
e-g
O
O
TIPSO
O
h
O
N
N
O
O
O
X
(
)
(
)
n
n
O
X
O
H
O
OMe
OMe
H
H
19a. n=1
18a. n=1
H
OCOPh
19b. n=2
18b. n=2
CO2H
Scopadulcic acid B
12
13
Scheme 5. Reagents: (a) Dibal-H, THF/CH2Cl2 (1:1), ꢀ78 ꢁC–0 ꢁC,
15 min; (b) Ph3PCH2OMeCl, s-BuLi, THF, ꢀ78 ꢁC–rt, 12 h, 90% (two
steps); (c) 4 N HCl/acetone (1:1), 0 ꢁC–rt, 1 h; (d) Ph3PCH2OMeCl,
s-BuLi, THF, ꢀ78 ꢁC–rt, 3 h, 70% (two steps); (e) PDC, DMF, 0 ꢁC–rt,
12 h, 44%; (f) Et3N, Pivaloyl chloride, THF, ꢀ20 ꢁC, 2 h; (g) LiCl,
2-Oxazolidinone, ꢀ20 ꢁC–rt, 6 h, 62% (two steps) and (h) i. LDA, THF,
ꢀ78 ꢁC, 1 h, ii. TIPSOTf, ꢀ78 ꢁC–rt, 5 h, 87%.
anodic
oxidation
X
X
X
X
MeO
MeO
O
9
10
11
TMS
The electrolyses of 19a and 19b were conducted in an
undivided cell using constant current conditions (8 mA/
2.2 F/mol), a reticulated vitreous carbon (RVC) anode and
cathode, 2,6-lutidine as a proton scavenger, and a 0.1 M
tetraethylammonium tosylate electrolyte solution.10 The
solvent for the reaction was either methanol or a mixture
of methanol and THF as indicated in Table 1. The overall
process was neutral since acid is generated at the anode
and base (from the reduction of methanol) is generated at
the cathode. Hence the added proton scavenger played no
Scheme 3.
2. N,O-Ketene acetal substrates—an initial look
The first question that needed to be addressed was whether
the lower efficiency of the dithioketene acetal derived cycli-
zations with allylsilane trapping groups was representative
of all ketene acetal groups or specifically a result of the
dithiane moiety. For this reason, we decided to build an
oxazolidone based ketene acetal substrate (Scheme 4).
Oxazolidinone ketene acetals were selected for this effort
because they are stable and can be isolated by chromato-
graphy. This enabled the use of clean substrates for the elec-
trolysis reactions. In addition, oxazolidinone ketene acetals
possessing a stereogenic atom have been utilized in
Table 1
O
TIPSO
O
O
RVC anode
0.1 M Et4NOTs
N
N
O
O
( )n
( )n
2,6-lutidine
8 mA/ 2.2 F/mole
OMe
OMe
OMe
19a. n=1
19b. n=2
20a. n=1
20b. n=2
O
O
R3SiO
O
Substrate
Solvent
MeOH
4:1 MeOH/THF
MeOH
4:1 MeOH/THF
Product
Yield (%)
Isomer ratio
N
O
anodic
N
O
*
*
(
)
n
*
(
)
n
*
R
19a
19a
19b
19b
20a
20a
20b
20b
74
87
63a
65
2:1
3:2
1:2
1:2
oxidation
?
R
14
15
TMS
a
7% recovered starting material.
Scheme 4.

6538
Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
net role in the reaction. However, due to the acid sensitivity
of ketene acetal, the presence of 2,6-lutidine was important
for maintaining the neutrality of the reaction in the region
immediately surrounding the anode. Using these conditions,
both reactions led to good yields of cyclized product. For the
oxidation of 19a, a 74% isolated yield of 20a was obtained
when methanol was used as solvent. The reaction benefited
from the use of THF as a cosolvent. An 87% isolated yield of
cyclized product was obtained with the addition of 20% THF
as a cosolvent. Like previous enol ether–enol ether coupling
reactions,9 the cyclization did not lead to the formation of
products in stereoselective fashion. In this case, the ratio
of stereoisomers of the newly formed C–C bond ranged
from 2:1 to 3:2. The major isomer was not assigned since
the isomers could not be separated and the NMR signals
for the methine protons overlap.
six-membered ring substrate was made by doing a one
carbon chain extension of the aldehyde generated from
Dibal-H reduction of the 3-caprolactone. The yields given
in Scheme 6 for steps f–i are for the five-membered ring sub-
strate. Yields for the six-membered ring case are again given
in the Section 7.
Once synthesized, the allylsilane substrates were electro-
lyzed using the same conditions described above (Table 2).
From the start, it was obvious that cyclization reactions orig-
inating from the oxidation of N,O-ketene acetals proceed far
better than do cyclizations originating from the oxidation of
a ketene dithioacetal. Electrolysis of 23a in 4:1 MeOH/THF
led to a 70% isolated yield of cyclized product 24a. The use
of dichloromethane as the cosolvent led to a 74% yield of cy-
clized product along with 7% recovered starting material.
Surprisingly, cyclized product was formed even in the
absence of a cosolvent. This was a surprise because enol
ether–allylsilane coupling reactions require a cosolvent.6a
Without it, they predominately form products derived from
methanol trapping of the radical cation prior to cyclization.
The fact that the N,O-ketene acetal reactions do not require
a cosolvent suggests that N,O-ketene acetal derived radical
cations undergo the cyclizations more efficiently than not
only ketene dithioacetal radical cations but also the more
reactive enol ether radical cations. While it was not clear
that the differences observed were really due to changes in
radical cation reactivity (vide infra), it was clear that the con-
clusions made with respect to ketene dithioacetal radical
cations cannot be universally applied to all ketene acetal
substrates.
The oxidation of 19b in methanol solvent led to a 63% iso-
lated yield of 20b in a 1:2 ratio of cis/trans isomers. With the
use of the THF cosolvent, a 65% unoptimized yield of prod-
uct was obtained having the same 1:2 ratio of cis/trans
isomers. The poor diastereoselectivity obtained for the cycli-
zation of 20b was consistent with the stereochemical
outcome of earlier anodic cyclizations using enol ether de-
rived radical cations and enol ether trapping groups.9
3. N,O-Ketene acetal radical cations and allylsilane
trapping groups
With the oxidations of 19a and 19b leading to good yields of
cyclic product, it was time to determine if a radical cation de-
rived from an N,O-ketene acetal would be reactive enough to
cyclize with a less efficient allylsilane trapping group. To
this end, cyclization substrates 23a and 23b were synthe-
sized as outlined in Scheme 6. These syntheses paralleled
the syntheses of 19a and 19b. The ylide used for generating
the trisubstituted allylsilane was prepared in situ from ethyl-
triphenylphosphonium bromide according to the known lit-
erature procedure.11 As in the earlier synthesis of 19b, the
While the oxidation of 23a led nicely to cyclized products, it
was not stereoselective. In each case, a 3:2 ratio of cis and
trans-products was obtained. This ratio was consistent with
the mixtures observed for enol ether–allylsilane coupling
reactions leading to five-membered rings (substrates 5a
and 5b in Scheme 2).6a In both cases, one would expect a
reaction controlled by thermodynamics to place the groups
on the ring trans to each other. Hence, the formation of
a stereochemical mixture was most consistent with the
reactions being controlled by kinetics and the result of
five-membered ring transition states not having well defined
O
OH
R2
a,b or c
O
(
)
n
R1
Table 2
16
21a. n=1, R1=Me, R2=CH2TMS
17a. n=1, R1=H, R2=OMe
21b. n=2, R1=Me, R2=CH2TMS
O
TIPSO
O
O
RVC anode
d,e
N
0.1 M Et4NOTs
N
O
O
( )n
( )n
2,6-lutidine
8 mA/ 2.2 F/mole
f-h
TMS
O
O
O
O
23a. n=1
23b. n=2
24a. n=1
24b. n=2
i
N
N
O
O
(
)
n
(
)
n
TMS
TMS
Substrate Solvent
Product Yield (%) Isomer ratio
23a. n=1
23b. n=2
23a
23a
23a
23b
23b
23b
23cc
MeOH
4:1 MeOH/THF
1:4 MeOH/CH2Cl2 24a
MeOH
4:1 MeOH/THF
1:4 MeOH/CH2Cl2 24b
4:1 MeOH/THF 24b
24a
24a
51
70
3:2
3:2
3:2
trans
trans
trans
trans
22a. n=1
22b. n=2
74a
54
Scheme 6. Reagents: (a) Dibal-H, THF/CH2Cl2 (1:1), ꢀ78 ꢁC–0 ꢁC,
15 min; (b) Ph3PC(CH3)CH2TMS, s-BuLi, THF, ꢀ78 ꢁC–rt, 65% (two
steps); (c) Ph3PCH2OMeCl, s-BuLi, THF, ꢀ78 ꢁC–rt, 12 h, 90% (two
steps); (d) 4 N HCl/acetone (1:1), 0 ꢁC–rt, 1 h; (e) Ph3PC(CH3)CH2TMS,
s-BuLi, THF, ꢀ78 ꢁC–rt, 12 h, 58% (two steps); (f) PDC, DMF, 0 ꢁC–rt,
12 h. 49%; (g) Et3N, Pivaloyl chloride, THF, ꢀ20 ꢁC, 2 h; (h) LiCl,
2-Oxazolidinone, ꢀ20 ꢁC–rt, 6 h, 91% (two steps) and (i) i. LDA, THF,
ꢀ78 ꢁC, 1 h, ii. TIPSOTf, ꢀ78 ꢁC–rt, 5 h, 83%.
24b
24b
71
54b
60
a
7% recovered starting material.
10 F/mol.
TBS enol ether.
b
c

Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
6539
pseudoequatorial and pseudoaxial positions. The oxidation
of 23b supported the suggestion that cyclizations originating
from the radical cation of an N,O-ketene acetal was under
kinetic control. Cyclizations leading to six-membered rings
have well defined pseudoaxial and pseudoequatorial posi-
tions in their transition states. Since both allylsilane5 and
ketene acetal4a groups are known to prefer pseudoequatorial
positions in anodic cyclizations, a kinetically controlled re-
action resulting from the oxidation of 23b was expected to
afford predominately trans-product (Scheme 7). This was
the case. Electrolysis of 23b using 4:1 MeOH/THF as sol-
vent led to a 71% isolated yield of the trans six-membered
ring product (Table 2). Only a small amount (<5%) of a
second cyclized product (assumed to be the cis-isomer)
was observed in the 1H NMR spectrum of the crude reaction
material. The stereochemistry of the major product was
was dependent upon the length of the tether connecting the
olefins. Substrates expected to give rise to faster cyclizations
afforded lower potentials.12 For example, a bis methoxyenol
ether substrate leading to six-membered ring formation (27)
had an oxidation potential 100 mV lower than that of a mono
methoxyenol ether (26) whereas an analogous bis methoxy-
enol ether substrate leading to five-membered ring formation
(28) had an oxidation potential 200 mV lower than that of
a mono methoxyenol ether (Scheme 8). A five-membered
ring precursor (30) showed a drop in potential of about
150 mV relative to its corresponding mono enol ether (29)
even when the reaction required the formation of a quater-
nary carbon (Scheme 9). These observations are consistent
with an oxidation reaction that is followed by a cyclization
fast enough to influence the concentration of the radical cat-
ion at the electrode surface (Scheme 9). In this scenario, the
oxidation reaction (k1) leads to a radical cation that un-
dergoes either the transfer of an electron back to the electrode
(kꢀ1) or cyclization (k2). If one assumes a steady state con-
centration of the radical cation, then solving for this concen-
tration and plugging the resulting expression into the Nernst
equation affords an equation relating the observed potential
to the rate of cyclization. Since E0 in this equation is positive
and everything following the minus sign is positive, a faster
cyclization will lead to a lower value for Eobs. Hence, the
value measured for Eobs provides a method for measuring
the relative rates of various cyclizations.
1
assigned using the coupling patterns observed in the H
NMR spectrum for the ring methine protons. Both protons
show the triplet of doublets pattern consistent with the
protons being axial and trans to each other.
.
+
O
TIPSO
O
O
H
N
N
O
OTIPS
TMS
H
TMS
23b
25
OMe
O
O
OMe
H
OMe
N
O
H
26. Ep/2 = +1.40 V (Ag/AgCl)
27. Ep/2 = +1.30 V (Ag/AgCl)
24b
Scheme 7.
OMe
OMe
The oxidation of 23b using dichloromethane as the cosol-
vent led to a much less efficient cyclization (10 F/mol of
charge was needed to obtain 54% of the cyclized product).
However, the cyclization still proceeded without the use of
a cosolvent in spite of the slower six-membered ring forma-
tion. With methanol as solvent, the oxidation led to a 57%
isolated yield of the trans-product.
28. Ep/2 = +1.20 V (Ag/AgCl)
Scheme 8.
MeO
MeO
OMe
Finally, the compatibility of the anodic oxidation with the
less stable TBS enol ether was examined. The oxidation of
23c in 4:1 MeOH/THF (all other conditions the same) led
to a 60% isolated yield of the trans material. In this reaction,
a significant amount of imide byproduct from methanolysis
of the ketene acetal was observed.
30. Ep/2=+1.01V (Ag/AgCl)
29. Ep/2=+1.16V (Ag/AgCl)
A shift in potential?
k1
k2
[radical cation]
Product
SM
k-1
Using steady state kinetics,...
At this point it was clear that radical cations derived from the
N,O-ketene acetals lead to more efficient cyclizations than
either radical cations derived from enol ethers or radical cat-
ions derived from ketene dithioacetals. Our working model
suggested that this results from the N,O-ketene acetals being
more reactive than their ketene dithioacetal and enol ether
counterparts. But was this really true?
k1[SM]
[radical cation] =
(k-1+k2)
...and then the Nernst equation
(k-1+k2)
[SM]
[radical cation]
Eobs = Eo - RT/nF ln
= Eo - RT/nF ln
k1
Scheme 9.
4. Cyclic voltammetry—measuring relative rates
A number of years ago,9 we observed that the potential mea-
sured by cyclic voltammetry for a bis enol ether substrate
An alternative explanation for the drop in potentials that in-
vokes an associative electron transfer where oxidation and

6540
Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
cyclization occur in a concerted fashion can be viewed as
a limiting example of this general scheme. In this case, the
more the overlap between the olefins during the electron-
transfer step, the lower is the oxidation potential observed.
Of course, the extent of overlap between the olefins during
the electron-transfer step is directly related to the rate of
cyclization. Therefore, the value of Eobs would again
provide a method for measuring the relative rates of various
cyclizations.
from oxidation of enol ethers 32a and 32b, an observation
that was consistent with our working model for the reactions.
Following these measurements, attention was turned to the
N,O-ketene acetal derived cyclizations. Surprisingly, the
drop in potential measured for substrates 23a and 23b was
not significantly different than the drop in potential mea-
sured for the ketene dithioacetal substrates 34a and 34b.
For 23b, a 40 mV drop in potential was measured relative
to 35, while for 23a an 80 mV drop in potential relative to
35 was observed. In neither case was the drop in potential
close to what was observed for the enol ether derived radical
cations. This result indicated that the most efficient (N,O-
ketene acetal) and least efficient (ketene dithioacetal) cycli-
zation reactions from a yield perspective proceeded roughly
at the same rate, whereas, the cyclization with an intermedi-
ate efficiency (enol ether) proceeded much faster than either
of the others. Clearly, our working model for the cyclizations
was not correct. The yield of the electrolyses did not depend
solely on the rate of the cyclization reaction.
With this in mind, we undertook a CV study of the ketene
acetal based substrates in order to determine the relative re-
activity of the radical cation intermediates towards allyl-
silane trapping groups. In order to establish a baseline for
these studies, the shifts in potential associated with the cou-
pling of an enol ether to an allylsilane trapping group were
examined (Table 3). The parent enol ether 31 was measured
to have a potential (Ep/2) of +1.44 V versus Ag/AgCl13 using
a Pt anode, a sweep rate of 25 mV/s, and a 0.1 M LiClO4 in
acetonitrile electrolyte solution. A substrate concentration of
0.025 M was used. All of the potentials listed in Table 3 were
obtained using identical conditions. For the six-membered
ring enol ether/allylsilane precursor (32b), a drop in poten-
tial of 100 mV relative to the parent enol ether 31 was ob-
served. For the five-membered ring precursor (32a) an
additional 80 mV drop in potential (a 180 mV drop for the
substrate relative to the parent enol ether 31) was observed.
The smaller drop in potential for the six-membered ring
precursor 32b confirmed that the cyclization derived from
this substrate proceeds more slowly than the cyclization
originating from the five-membered ring precursor 32a.
So what does control the yield of an anodic olefin coupling
reaction? Typically, when the reactions give a low yield of
the desired cyclic product they form products that are de-
rived from either trapping of the initially formed radical
cation with solvent, proton elimination from the uncyclized
radical cation, or polymerization. Hence, the yield of cyclic
product obtained depends on a partitioning between intra-
molecular trapping of the initially formed radical cation ver-
sus solvent trapping, elimination, and polymerization. If
higher yielding cyclizations are not faster with respect to
the intramolecular reaction, then the radical cation inter-
mediates involved must be more stable with respect to the
decomposition reactions. In other words, switching from the
ketene dithioacetal to the N,O-ketene acetal did not lead to
a more reactive radical cation, but rather a more stable one.
Next, attention was turned towards the cyclizations originat-
ing from oxidation of a ketene dithioacetal. In this case, the
potential measured for the six-membered ring cyclization
substrate (34b) was only 30 mV lower than the potential
measured for the parent ketene dithioacetal 33. The five-
membered ring substrate (34a) had a potential only 50 mV
lower than the parent ketene dithioacetal 33. Clearly, these
cyclizations were slower than the cyclizations resulting
It is tempting to speculate that the differences in cyclization
efficiency observed for the reactions above result from the
ability of the substituents on the radical cation to stabilize
a cation (Scheme 10). It is known that enol ether radical
cations react with allyl- and vinylsilanes in a ‘radical-like’
fashion meaning that the reaction leads to a radical at the
terminating end of the cyclization.14 At the same time, two
of the three decomposition pathways (methanol trapping
and elimination) are clearly ‘cation-like’ processes and the
third leads to a polymethoxylated polymer that may also
be ‘cation-like’ in origin. Do substituents on the radical cat-
ion that stabilize a cation reduce the rate of decomposition
and afford more time for the ‘radical-type’ cyclization that
leads to the desired product? Efforts are underway to begin
addressing this question with more electron-rich ketene
acetal derivatives.
Table 3
Ag/AgCl reference
electrode Pt anode,
Potential
V
Potential V
(Ep/2
)
25 mV/s 0.1 M LiClO4 (Ep/2
in CH3CN 0.025 M
[substrate]
)
32a n¼1 +1.26
32b n¼2 +1.34
34a n¼1 +1.10
OMe
TMS
2
n
n
+1.44
+1.15
OMe
31
S
2
S
S
S
34b n¼2 +1.12
23a n¼1 +1.07
23b n¼2 +1.11
.
.
.
+
+
+
S
OR
OR
TMS
O
33
OTIPS
R
R
R
S
N
O
O
OTIPS
O
n
O
N
N
O
n
+1.15
35
( )
n
Oxygen nucleophiles compete
with olefin trapping
Olefin trapping strongly favored
over oxygen nucleophiles
TMS
Scheme 10.

Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
6541
5. Chiral auxiliaries
when the isopropyl substituted oxazolidinone was used
(40b). The TBS based ketene acetals were significantly
less stable to the electrolysis conditions than ones possessing
the TIPS group. However, they did allow us to assess the
utility of the chiral auxiliaries for effecting asymmetric
anodic olefin coupling reactions.
The degree of stereoselectivity obtained for the cyclization
originating from 23b provided an excellent opportunity to
probe the potential for asymmetric anodic olefin coupling re-
actions. Two key, contradictory literature precedents formed
the backdrop for this effort. First, Evans and co-workers
demonstrated that chiral oxazolidinones are very useful
auxiliaries for inducing asymmetry into both aldol and alkyl-
ation reactions using silyl based N,O-ketene acetal deriva-
tives directly analogous to substrates 23a and 23b.15,16 In
these reactions, the chiral auxiliaries adopt a conformation
that minimizes the dipole of the substrate in the transition
state leading to product. Second, Sibi and co-workers have
shown that chiral oxazolidinone groups are not useful
auxiliaries for inducing asymmetry into radical reactions
that originate from a radical on the carbon alpha to the imide
carbonyl.17,18 In these cases, rotation around the bond con-
necting the nitrogen of the chiral oxazolidinone to the car-
bonyl of the substrate interferes with the auxiliaries ability
to place the sterically bulky directing group on a single
face of the substrate. For a radical cation initiated cycliza-
tion, the problem can be summarized using the transition
state structures as illustrated in Scheme 11. Transition state
36 represents the preferred transition state (relative to 37) if
the radical cation cyclization behaves in a fashion analogous
to the aldol and alkylation reactions studied by Evans. In this
picture, the large allylsilane group would approach the face
of the radical cation opposite to the sterically bulky substit-
uent on the oxazolidinone. Transition states 36 and 38 repre-
sent the two preferred transition states if the radical cation
cyclization behaves in a fashion analogous to the radical re-
actions studied by Sibi. In this case, rotation around the bond
connecting the nitrogen of the oxazolidinone and C1 of the
substrate leads to two conformations that place the bulky
R group of the chiral oxazolidinone on opposite faces of
the radical cation. Approach of the allylsilane moiety
away from the R group would then lead to a mixture of
diastereomers.
O
OH
O
a,b, c or d
N
O
TMS
R
21b
TBSO
O
TMS
39a. R = CH2Ph
N
39b. R = CH(Me)2
O
e
R
40a. R = CH2Ph
40b. R = CH(Me)2
TMS
Scheme 12. Reagents: (a) PDC, DMF, 0 ꢁC, 12 h, 65%; (b) Et3N, Pivaloyl
chloride, THF, ꢀ20 ꢁC, 2 h; (c) LiCl, (4S,5R)-(ꢀ)-4-Methyl-5-phenyl-2-
oxazolidinone, ꢀ20 ꢁC–rt, 6 h, 94%; (d) LiCl, (S)-(ꢀ)-4-Isopropyl-2-oxazoli-
dinone, ꢀ20 ꢁC–rt, 6 h, 85% and (e) i. LDA, THF, ꢀ78 ꢁC, 30–90 min. ii.
TBSOTf, ꢀ78 ꢁC–rt, 6–8 h, 57% (with 39% 39a) for 40a, 88% for 40b.
The oxidation of 40a was conducted using the same constant
current electrolysis conditions as described earlier (Table 4).
Unfortunately, only poor yields of the cyclized product
could be obtained. In the best case, a 20% yield of product
41a was obtained. For this product, only the two trans-dia-
stereomers were observed. These diastereomers were sepa-
rated, characterized, and then their experimental ratio (2:1)
determined by integration of the proton NMR spectrum
taken for the crude reaction product. Efforts to optimize the
yield for the reaction were halted when it became clear that
the chiral auxiliary was not effective for inducing asymme-
try into the oxidative cyclization.
The oxidation of 40a suggested that the chiral auxiliary
rotated and that the radical cation behaved in a fashion anal-
ogous to the radical chemistry of Sibi and co-workers rather
.
.
.
R
+
+
+
TIPSO
OTIPS
TMS
vs.
OTIPS
TMS
Table 4
N
O
O
vs.
R
O
TIPSO
O
O
N
N
RVC anode
O
TMS
O
O
N
R
R
38
N
0.1 M Et4NOTs
O
37
O
O
+
39a and 39b
36
Scheme 11.
2,6-lutidine
8 mA
R
TMS
40a. R = CH2Ph
40b. R = CH(Me)2
To test which of these two scenarios best describes the radi-
cal cation cyclizations, electrolysis substrates 40a and 40b
were prepared (Scheme 12). The syntheses began with the
formation of imides 39a and 39b in a fashion identical to
that used in Scheme 7. However, formation of the ketene
acetal moiety needed for the oxidation proved to be more
challenging. With the extra steric bulk of the chiral auxiliary,
efforts to synthesize the triisopropylsilyl ketene acetal were
not successful. In each case, the starting imide was recovered
without silylation. Fortunately, using the same reaction con-
ditions, the less hindered tert-butyldimethylsilyl ketene ace-
tal could be made in a 57% isolated yield (along with 39% of
the recovered starting material) when the benzyl substituted
oxazolidinone was used (40a) and an 88% isolated yield
41a. R = CH2Ph (trans product)
41b. R = CH(Me)2 (trans product)
Substrate Solvent
# (F/mol) Recovered 41 39 Diast.
40 (%)
ratio
40a
40a
40a
4:1 MeOH/THF
MeOH
1:4 MeOH/CH2Cl2 2.2
2.2
2.2
0
0
37
20 45 2:1
15 53 2:1
10 2:1
6
40b
40b
40b
40b
40b
40b
4:1 MeOH/THF 2.2
10
27
20 20 1:1.5
20 13 2:1
4:1 MeOH/CH3CN 2.2
1:1 MeOH/CH3CN 2.2
1:4 MeOH/CH2Cl2 2.2
1:4 MeOH/CH2Cl2 4.7
1:1 MeOH/CH2Cl2 3.0
11
26
<8
<5
26
25
42
40
8
8
6
7
2:1
1:1.5
1:1.5
2:1

6542
Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
than the alkylation chemistry of Evans and co-workers. But
what if the poor selectivity obtained from 40a was simply
a result of the benzyl group being too small to be effective?
The use of the larger isopropyl group in 40b was studied in
order to address this issue. If the poor diastereoselectivity
obtained from the oxidation of 40a was due to the benzyl
group being too small, then the use of the larger isopropyl
group in 40b should improve the diastereoselectivity.
However, if the poor diastereoselectivity obtained from the
oxidation of 40a was due to rotation of the chiral auxiliary,
then the use of the larger inducing group would not help.
The electrolyses of 40b is summarized in Table 4. While al-
terations in the solvent used and amount of current employed
helped to improve the yield of the cyclizations, the diastereo-
selectivity of the reaction remained poor in all cases.
the solution became clear. The resulting mixture was filtered
and concentrated in vacuo. The crude product was diluted
with 200 mL of ether and dried using anhydrous MgSO4. It
was then filtered and concentrated in vacuo the second
time. The crude product was then taken up in 60 mL of anhy-
drous THF and used directly in the following steps. (The
hemiacetal should not be left concentrated in order to avoid
polymerization).
A flame-dried 500-mL round-bottom flask under an argon
atmosphere was charged with a stirred suspension of
(methoxymethyl)triphenylphosphonium chloride (34.985 g,
100 mmol) in 220 mL of anhydrous THF. The reaction
was cooled to ꢀ78 ꢁC, a 1.4 M sec-butyllithium solution
in cyclohexane (110 mL, 154 mmol) was added dropwise,
and the resulting dark brown mixture allowed to warm
from ꢀ78 ꢁC to room temperature over 3 h. When complete,
the reaction was recooled to ꢀ78 ꢁC and the crude hemi-
acetal synthesized above in 60 mL of anhydrous THF added
via cannula. The mixture was allowed to warm to room tem-
perature and stirred overnight. In the morning, the reaction
was quenched with 300 mL of satd NaHCO3. The aqueous
phase was then extracted with ether (6ꢂ300 mL) and the
combined organic layers dried over MgSO4, filtered and
concentrated in vacuo. At this point 150 mL of 1:3 ether/
hexane was added to the crude product and the mixture
stirred for 4 h. (This will precipitate most of triphenylphos-
phine and triphenylphosphine oxide.) The mixture was fil-
tered and concentrated in vacuo for the second time. The
crude product was chromatographed through a silica gel
column (crude/gel¼1/15w20) with 2–3% Et3N in 1:6
EtOAc/hexane to afford the desired product 17a (6.488 g,
45 mmol, 93%, over 2 steps) as a yellow oily liquid. The
In the end, neither the benzyl nor the isopropyl directing
group induces a significant degree of asymmetry into the
process, a result consistent with rotation of the chiral auxil-
iary and the radical cation cyclizations being analogous to
the corresponding radical reactions.
6. Conclusions
Our work with N,O-ketene acetal derived radical cations has
led to two major findings. First, the use of an N,O-ketene
acetal initiating group for anodic olefin coupling reactions
leads to more efficient carbon–carbon bond formation than
the use of either an enol ether or a ketene dithioacetal initi-
ating group. The higher yield of cyclic product in these reac-
tions is not due to the cyclizations being faster but rather the
radical cation intermediates being more stable to decompo-
sition. Knowing that a change in the substituents on a radical
cation can dramatically improve the yield of subsequent cy-
clization, current efforts are aimed at identifying ketene
acetal groups that will allow for the simultaneous formation
of both quaternary carbons and six-membered rings.19
1
spectral data were as follows: H NMR (CDCl3/300 MHz)
d 6.28 (dt, J¼12.6, 1.2 Hz, trans H8, 0.56H), 5.87 (dt,
J¼6.3, 1.4 Hz, cis H8, 0.44H), 4.72 (dt, J¼12.6, 7.4 Hz,
trans H7, 0.61H), 4.34 (td, J¼7.4, 6 Hz, cis H7, 0.39H),
3.63 and 3.63 (t and t, J¼6.6 and 6.8 Hz, H2, 2H), 3.58 (s,
cis H9, 1.29H), 3.50 (s, trans H9, 1.71H), 2.13–2.02 (m,
cis H6, 0.88H), 1.98–1.88 (m, trans H6, 1.12H), 1.70 (br s,
H1, 1H), 1.63–1.51(m, H3, 2H), 1.44–1.30 (m, H4 and H5,
4H); 13C NMR (CDCl3/300 MHz) d 147.2, 146.3, 106.9,
103.1, 63.1, 63.1, 59.6, 56.0, 32.8, 30.7, 29.7, 27.8, 25.4,
25.3, 23.9; IR (neat/NaCl) 3342, 3033, 2998, 2931, 2856,
Second, the experiment using chiral oxazolidinones demon-
strated that the radical cation cyclizations are analogous to
asymmetric radical reactions in that rotation of the chiral
auxiliary precludes useful levels of induction. This observa-
tion will be critical for designing future asymmetric anodic
olefin coupling reactions that would appear to require the
use of either a Lewis acid to stop rotation of the auxiliary18
or a C2-symmetric chiral auxiliary.
1656, 1462, 1391, 1208, 1110, 1054, 933, 736 cmꢀ1
;
LRMS (EI) 144 ([M]+, 4), 97 (10), 72 (100); HRMS (EI)
m/z calculated for C8H16O2 [M]+ 144.1150, found 144.1155.
7.2. 3-(7-Methoxy-hept-6-enoyl)-oxazolidin-2-one (18a)
7. Experimental
To a flamed-dried 100-mL round-bottom flask under an
argon atmosphere were added 17a (1.503 g, 10.4 mmol)
and 60 mL of anhydrous DMF. The reaction was cooled to
0 ꢁC and the solution was treated with pyridinium dichro-
mate (PDC) (15.806 g, 42.0 mmol). The temperature of mix-
ture was allowed to warm to room temperature and stirred
overnight. The reaction was quenched with 600 mL of water,
the aqueous layer extracted with ether (10ꢂ150 mL), and the
combined organic layers were dried over anhydrous MgSO4,
filtered and concentrated in vacuo. The crude product was
eluted with 1:4 ether/hexane through a silica gel column
(crude/gel¼1:25) to afford the acid product (0.722 g,
4.6 mmol, 44%) as a colorless oily liquid.
7.1. 7-Methoxy-hept-6-en-1-ol (17a)
To a flame-dried 500-mL round-bottom flask under an argon
atmospherewere added 3-caprolactone (5.597 g, 48.5 mmol)
and 300 mL of anhydrous solvent (CH2Cl2/THF¼1:1). The
reaction was cooled to ꢀ78 ꢁC and 1 M DIBAL solution in
toluene (50 mL, 50 mmol) was added dropwise. The reaction
was stirred at ꢀ78 ꢁC for 15 min. The reaction was then
allowed to warm to room temperature, quenched with
200 mL of ether and 60 mL of H2O, and stirred until it be-
came a translucent white gel. The solution was further diluted
with 200 mL of ether, Celite 545 added, and then stirred until

Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
6543
A solution of the acid synthesized above (0.47 g, 3.0 mmol)
and 20 mL of anhydrous THF was placed in a 100-mL
round-bottom flask under an argon atmosphere. After
cooling the reaction to ꢀ20 ꢁC, anhydrous Et3N (2.0 mL,
14.3 mmol) was added dropwise followed by trimethyl-
acetyl chloride (0.44 mL, 3.5 mmol). The resulting white
suspension solution was stirred at ꢀ20 ꢁC. After 2 h, oven-
dried lithium chloride (0.155 g, 3.6 mmol) and 2-oxazoli-
done (0.325 g, 3.7 mmol) were added, the reaction allowed
to warm slowly to room temperature, and the room temper-
ature solution stirred for 8 h. After this point, the reaction
was concentrated to about 10% of its initial volume. The res-
idue was filtered through a pad of silica gel with 3% Et3N in
ether as eluant. After concentration in vacuo, the crude prod-
uct was chromatographed through a silica gel column
(crude/gel¼1:30) with 2–3% Et3N in 1:4 EtOAc/hexane as
eluant in order to afford the desired imide 18a (0.416 g,
1.8 mmol, 62%) as a colorless oily liquid. The spectral
2.16–2.02 (m of H4 with cis H6 buried, 2.44H), 1.94 (dtd,
J¼7.4, 7.2, 1.1 Hz, trans H6, 1.56H), 1.41 (p, J¼7.4 Hz,
H5, 2H), 1.25–1.07 (m of H11 with d of H10 at 1.11,
J¼5.4 Hz, 21H); 13C NMR (CDCl3/300 MHz) d 156.0,
147.4, 146.5, 140.1, 106.6, 105.0, 104.8, 102.7, 61.8, 59.6,
56.0, 46.6, 30.6, 29.7, 27.6, 25.3, 25.0, 23.8, 18.0, 13.2; IR
(neat/NaCl) 3518, 3055, 2944, 2868, 1770, 1681, 1463,
1398, 1268, 1210, 1109, 1040, 933, 883, 685 cmꢀ1; LRMS
(EI) 383 (M+, 2), 368 ([MꢀCH3]+, 10), 340 ([M
ꢀC(CH3)2]+, 100), 244 (92), 200 (40); HRMS (EI) m/z
calculated for C20H37NO4Si [M]+ 383.2492, found
383.2489.
7.4. 3-(2-Dimethoxymethyl-cyclopentanecarbonyl)-
oxazolidin-2-one (20a)
Under an argon atmosphere, a solution of oven-dried tetra-
ethylammonium tosylate (0.453 g, 1.5 mmol), ketene acetal
19a (0.094 g, 0.25 mmol), and 2,6-lutidine (0.22 mL,
1.5 mmol) in 15 mL of anhydrous solution (THF/
MeOH¼1:4) was placed in a flame-dried 25-mL three-
necked round-bottom flask equipped with a reticulated vitre-
ous carbon (RVC) anode and a RVC cathode. The reaction
1
data were as follows: H NMR (CDCl3/300 MHz) d 6.29 (d
with fine coupling, J¼12.6 Hz, trans H8, 0.81H), 5.88 (dt,
J¼6.0, 1.4 Hz, cis H8, 0.19H), 4.71 (dt, J¼12.6, 7.4 Hz,
trans H7, 0.79H), 4.42 (t, J¼8.1 Hz, H1, 2H), 4.32 (apparent
td, J¼7.2, 6.0 Hz, cis H7, 0.21H), 4.02 (t, J¼8.1 Hz, H2,
2H), 3.57 (s, cis H9, 0.57H), 3.49 (s, trans H9, 2.43H),
2.91 (t, J¼7.5 Hz, H3, 2H), 2.13–2.03 (m, cis H6, 0.42H),
2.01–1.91 (m, trans H6, 1.58H), 1.73–1.60 (m, H4, 2H),
1.47–1.36 (m, H5, 2H); 13C NMR (CDCl3/300 MHz)
d 173.5, 173.4, 153.6, 147.3, 146.4, 106.3, 102.5, 62.1,
59.5, 55.9, 42.5, 34.9, 30.1, 29.2, 27.4, 23.8, 23.6, 23.5; IR
(neat/NaCl) 3540, 3380, 2930, 2855, 1773, 1698, 1457,
1389, 1209, 1039, 937, 760 cmꢀ1; LRMS (EI) 209 (13),
140 (100); HRMS (EI) m/z calculated for C11H17NO4 [M]+
227.1158, found 227.1157.
˚
was electrolyzed at a constant current of 8 mA until 2.2 F/
mol of charge was passed. The solution was concentrated
to 10% of its original volume and then diluted with 80 mL
of pure ether to precipitate tetraethylammonium tosylate.
The reaction was then dried over sodium sulfate for 2–3 h,
filtered and concentrated in vacuo. The crude product was
chromatographed through a silica gel column (crude/
gel¼1:30) using 2–3% Et3N in 3:1 ether/hexane as eluant
in order to afford the desired product 20a (0.055 g,
0.21 mmol, 87%) as a colorless oily liquid. The spectral
1
data were as follows: H NMR (CDCl3/300 MHz) d 4.34
7.3. 3-(7-Methoxy-1-triisopropylsilanyloxy-hepta-1,6-
dienyl)-oxazolidin-2-one (19a)
(t, J¼7.8 Hz, H1, 0.7H), 4.33 (t, J¼7.8 Hz, H1, 1.3H),
4.21 (d, J¼9.3 Hz, H8, 0.65H), 4.15 (d, J¼7.8 Hz, H8,
0.35H), 4.05–3.81 (m of H2 with H3 buried, 3H), 3.23 (s,
H9, 1.05H), 3.21 (s, H9, 1.05H), 3.19 (s, H9, 1.95H), 3.14
(s, H9, 1.95H), 2.79–2.61 (m, H7, 1H), 2.12–1.40 (m, H4,
H5 and H6, 6H); 13C NMR (CDCl3/300 MHz) d 176.5,
175.0, 153.5, 153.5, 107.4, 104.3, 61.9, 61.8, 54.3, 53.4,
52.2, 51.5, 46.1, 44.6, 44.3, 43.7, 43.0, 42.9, 31.9, 29.3,
28.2, 28.0, 25.3, 24.4; IR (neat/NaCl) 3535, 3372, 2956,
2831, 1770, 1694, 1524, 1480, 1454, 1385, 1223, 1045,
973, 926, 761, 706 cmꢀ1; LRMS (EI) 226 (43), 166 (63),
138 (100); HRMS (FAB) m/z calculated for C12H19NO5
[M+Li]+ 264.1423, found 264.1424.
Into a flamed-dried 50-mL round-bottom flask under an
argon atmosphere were placed freshly distilled diisopropyl
amine (0.23 mL, 1.6 mmol) and 5 mL of anhydrous THF.
The reaction was cooled to ꢀ0 ꢁC, 1.6 M n-butyllithium
(1 mL, 1.6 mmol) was added dropwise, and then the mixture
stirred for 30 min. The reaction was then cooled to ꢀ78 ꢁC,
a solution of imide 18a (0.299 g, 1.3 mmol) in 5 mL of
anhydrous THF was added via a cannula, and the resulting
solution stirred for 20 min. While the reaction was still at
ꢀ78 ꢁC, triisopropylsilyl trifluoromethanesulfonate (0.42 mL,
1.6 mmol) was added. This mixture was allowed to slowly
warm to room temperature and stirred for 5 h. The reaction
mixture was concentrated to 10% of its original volume and
the residue was filtered through a pad of silica gel with 3%
Et3N in ether as eluant. After concentration in vacuo, the
crude product was chromatographed through a silica gel
column (crude/gel¼1:25) using 2–3% Et3N in 2:3 ether/
hexane as eluant in order to afford the desired ketene acetal
19a (0.438 g, 1.1 mmol, 87%) as a colorless oily liquid. The
7.5. 8-Methoxy-oct-7-en-1-ol (17b)
At 0 ꢁC, 240 mL of 1 N HCl/acetone (1:1) was added to
a 500-mL round-bottom flask containing enol ether 17a
(8.694 g, 60.3 mmol). The reaction was allowed to warm
to room temperature and stirred for 1 h (TLC). The reaction
was recooled to 0 ꢁC and the pH of the reaction was adjusted
to 6w8 using 10% NaOH and satd NaHCO3. After excess
NaCl was added, the reaction was extracted with ether
(6ꢂ200 mL). The combined organic layers were dried
over anhydrous MgSO4, filtered and concentrated in vacuo.
The crude aldehyde was used directly in the following step
without purification. This crude aldehyde should be diluted
with solvent (CH2Cl2 or ether) or it will polymerize during
storage.
1
spectral data were as follows: H NMR (CDCl3/300 MHz)
d 6.27 (d with fine coupling, J¼12.6 Hz, trans H8, 0.78H),
5.86 (dt, J¼6.0, 1.5 Hz, cis H8, 0.22H), 4.69 (dt, J¼13.5,
7.2 Hz, trans H7, 0.78H), 4.66 (t, J¼7.2 Hz, H3, 0.22H),
4.65 (t, H3, J¼7.2 Hz, 0.78H), 4.31 (dd of H1 with cis
H7 buried, J¼8.8, 7.1 Hz, 2.22H), 3.77 (dd, J¼8.8, 7.1 Hz,
H2, 2H), 3.56 (s, cis H9, 0.66H), 3.49 (s, trans H9, 2.34H),

6544
Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
To a flamed-dried 1000-mL round-bottom flask under an
argon atmosphere was added a stirred suspension of
(methoxymethyl)triphenylphosphonium chloride (42.667 g,
120.7 mmol) in 400 mL of anhydrous THF. The mixture
was cooled to ꢀ78 ꢁC, 1.4 M sec-butyllithium solution in
cyclohexane (115 mL, 161 mmol) was added dropwise,
and the resulting dark brown mixture allowed to warm
from ꢀ78 ꢁC to room temperature over 3 h. The reaction
was recooled to ꢀ78 ꢁC and the crude aldehyde synthesized
above in 60 mL of anhydrous THF added via a cannula. The
mixture was allowed to warm to room temperature and
stirred overnight. The reaction was quenched with 450 mL
of satd NaHCO3. The aqueous phase was extracted with
ether (6ꢂ300 mL). The combined organic layers were dried
over MgSO4 and concentrated in vacuo. At this point the
crude reaction mixture was taken up in 150 mL of 1:3
ether/hexane and allowed to stand for 4 h. This procedure
precipitates most of the Ph3PO. The mixture was filtered
and concentrated in vacuo for the second time. The crude
product was chromatographed through a silica gel column
(crude/gel¼1:15w20) with 2–3% Et3N in 1:8 EtOAc/hexane
as eluant in order to afford the desired product 17b (6.658 g,
42.1 mmol, 70%, over 2 steps) as a yellow oily liquid. The
7.2 Hz, trans H8, 0.75H), 4.41 (t, J¼8.1 Hz, H1, 2H), 4.32
(apparent td, J¼7.5, 6.3 Hz, cis H8, 0.25H), 4.02 (t,
J¼8.1 Hz, H2, 2H), 3.57 (s, cis H10, 0.58H), 3.50 (s, trans
H10, 2.42H), 2.91 (t, J¼7.5 Hz, H3, 2H), 2.11–2.01 (m,
cis H7, 0.4H), 1.98–1.87 (m, trans H7, 1.60H), 1.73–
1.59(m, H4, 2H), 1.44–1.32 (m, H5 and H6, 4H); 13C
NMR (CDCl3/300 MHz) d 173.7, 173.7, 153.7, 147.2,
146.3, 106.8, 103.0, 62.2, 59.6, 56.0, 42.7, 35.2, 30.6,
29.6, 28.8, 28.6, 27.6, 24.2, 24.2, 23.8; IR (neat/NaCl)
3538, 2993, 2930, 2855, 1781, 1700, 1655, 1388, 1208,
1115, 1040, 935, 761 cmꢀ1; LRMS (EI) 242 ([M+H]+
(22), 209 (17), 176 (19), 154 (96), 142 (37), 129 (61), 112
(100); HRMS (EI) m/z calculated for C12H19NO4 [M+H]+
242.1392, found 242.1401.
7.7. 3-(8-Methoxy-1-triisopropylsilanyloxy-octa-1,7-
dienyl)-oxazolidin-2-one (19b)
The triisopropylsilyl enol ether was prepared using the
method described as above for the synthesis of ketene acetal
19a. In this case, the reaction utilized freshly distilled diiso-
propyl amine (0.82 mL, 5.9 mmol) and 40 mL anhydrous
THF in 100-mL round-bottom flask along with 2.5 M
n-butyllithium (2.1 mL, 5.3 mmol), a solution of imide 18b
(1.003 g, 4.2 mmol) in 20 mL of anhydrous THF, and triiso-
propylsilyl trifluoromethanesulfonate (1.4 mL, 5.1 mmol).
The crude product was chromatographed through a silica
gel column (crude/gel¼1:25) using 2–3% Et3N in 4:5
ether/hexane as eluant in order to afford the desired ketene
acetal 19b (1.314 g, 3.3 mmol, 80%) as a colorless oily liq-
1
spectral data were as follows: H NMR (CDCl3/300 MHz)
d 6.28 (dt, J¼12.6, 1.5 Hz, trans H9, 0.56H), 5.87 (dt,
J¼6.3, 1.4 Hz, cis H9, 0.44H), 4.72 (dt, J¼12.6, 7.4 Hz,
trans H8, 0.58H), 4.34 (td, J¼7.4, 1.4 Hz, cis H7, 0.43H),
3.63 and 3.63 (t and t, J¼6.6 and 6.8 Hz, H2, 2H), 3.58 (s,
cis H10, 0.43H), 3.50 (s, trans H9, 0.57H), 2.15–2.02 (br s,
H1, 1H), 2.12–2.0 (m, cis H7, H), 1.97–1.87 (m, trans H7,
0.9H), 1.63–1.51(m, H3, 1.1H), 1.63–1.49 (m, H3, 2H),
1.41–1.28 (m, H4, H5 and H6, 6H); 13C NMR (CDCl3/
300 MHz) d 147.0, 146.1, 107.1, 103.2, 63.0, 63.0, 59.6,
56.0, 32.8, 30.8, 29.8, 29.1, 28.9, 27.7, 25.7, 25.7, 23.8; IR
(neat/NaCl) 3343, 3033, 2997, 2929, 2855, 1656, 1463,
1390, 1208, 1108, 1056, 934, 737 cmꢀ1; LRMS (EI) 158
([M]+, 20), 143 (100), 125 (80); HRMS (EI) m/z calculated
for C9H18NO2 [M]+ 158.1307, found 158.1318.
1
uid. The spectral data were as follows: H NMR (CDCl3/
300 MHz) d 6.26 (d, J¼12.6 Hz, trans H9, 0.82H), 5.85 (ap-
parent d, J¼6.3 Hz, cis H9, 0.18H), 4.70 (dt, J¼12.6, 7.4 Hz,
trans H8, 0.82H), 4.64 (t, J¼7.4 Hz, H3, 1H), 4.30 (dd of H1
with cis H8 buried, J¼8.5, 7.6 Hz, 2.18H), 3.77 (dd, J¼8.5,
7.6 Hz, H2, 2H), 3.56 (s, cis H10, 0.54H), 3.48 (s, trans H10,
2.46H), 2.16–1.99 (m of H4 with cis H7 buried, 2.36H),
1.96–1.84 (m, trans H7, 1.64H), 1.50–1.28 (m, H5 and H6,
4H), 1.25–1.07 (m of H12 with d of H11 at 1.10,
J¼6.0 Hz, 21H); 13C NMR (CDCl3/300 MHz) d 156.0,
147.2, 146.2, 140.0, 107.0, 105.1, 105.0, 103.1, 61.8, 59.6,
56.0, 46.6, 30.7, 29.7, 29.2, 28.9, 27.7, 25.5, 25.5, 23.8,
18.0, 13.2; IR (neat/NaCl) 3517, 3055, 2944, 2867, 1767,
1678, 1655, 1464, 1399, 1277, 1209, 1131, 1040, 883,
686 cmꢀ1; LRMS (EI) 397 (M+, 2), 382 ([MꢀCH3]+, 12),
354 ([MꢀC(CH3)2]+, 100), 244 (97), 200 (89), 128 (68);
HRMS (EI) m/z calculated for C21H39NO4Si [M]+
397.2648, found 397.2765 for [M]+ and 382.2424 for
[MꢀCH3]+.
7.6. 3-(8-Methoxy-oct-7-enoyl)-oxazolidin-2-one (18b)
The acid substrate was prepared using the same method for
the synthesis of imide 18a as described above using 17b
(3.083 g, 19.5 mmol), 200 mL of anhydrous DMF, and
PDC (29.932 g, 78.0 mmol). The crude product was chro-
matographed through a silica gel column (crude/gel¼1:25)
using 1:2 ether/hexane as eluant in order to afford the acid
product (1.47 g, 8.5 mmol, 44%) as a colorless oily liquid.
The coupling reaction was accomplished using the same
method as described above for the synthesis of imide 18a.
In this experiment was used the acid made above (1.11 g,
6.4 mmol) in 65 mL of anhydrous THF, anhydrous Et3N
(3.6 mL, 25.8 mmol), trimethylacetyl chloride (0.8 mL,
6.4 mmol), oven-dried lithium chloride (0.2649 g,
6.4 mmol) and 2-oxazolidone (0.574 g, 6.5 mmol). The
crude product was chromatographed through a silica gel col-
umn (crude/gel¼1:30) using 2–3% Et3N in 3:4 ether/hexane
as eluant in order to afford the desired imide 18b (1.003 g,
4.2 mmol, 65%) as a colorless oily liquid. The spectral
7.8. 3-(2-Dimethoxymethyl-cyclohexanecarbonyl)-
oxazolidin-2-one (20b)
The cyclization was conducted using the same procedure
described above for the synthesis of 20a. This experiment
utilized oven-dried tetraethylammonium tosylate (0.607 g,
2.0 mmol), ketene acetal 19b (0.096 g, 0.02 mmol), 20 mL
1:4 THF/MeOH, 2,6-lutidine (0.22 mL, mmol), an RVC
˚
anode and cathode, a constant current of 8 mA, and 2.2 F/
mol of charge. The crude product was chromatographed
through a silica gel column (crude/gel¼1:30) using 2–3%
Et3N in 3:5 EtOAc/hexane as eluant in order to afford the de-
sired product 20b (0.041 g, 0.015 mmol, 63%) as a colorless
1
data were as follows: H NMR (CDCl3/300 MHz) d 6.27
(d with fine coupling, J¼12.6 Hz, trans H9, 0.80H), 5.87
(dt, J¼6.3, 1.2 Hz, cis H9, 0.20H), 4.71 (dt, J¼12.6,

Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
6545
1
oily liquid. The spectral data were as follows: H NMR
(CDCl3/300 MHz) d 4.51–4.33 (m of H1 with cis H9 buried,
2.27H), 4.13 (d, J¼7.2 Hz, trans H9, 0.73H), 4.09–4.89 (m,
H2, 2H), 3.85–3.61 (m, cis and trans H3, 1H), 3.29 (s, cis
H10, 0.81H), 3.24 (s, trans H10, 2.19H), 3.23 (s, cis H10,
0.81H), 3.20 (s, trans H10, 2.19H), 2.26–2.04 (m, H8, 1H),
1.98–0.94 (m, H4, H5, H6 and H7, 8H); 13C NMR
(CDCl3/300 MHz) d 177.1, 175.3, 153.8, 153.3, 107.5,
104.4, 61.9, 54.2, 53.9, 53.0, 52.6, 51.8, 43.2, 43.0, 41.8,
41.7, 40.3, 39.9, 30.3, 29.8, 27.1, 26.1, 25.6, 25.4, 25.3,
25.0, 23.7, 23.5; IR (neat/NaCl) 3585, 3534, 3373, 2930,
2857, 1789, 1770, 1728, 1694,1480, 1454, 1385, 1325,
1262, 1223, 1046, 985, 943, 761, 705 cmꢀ1; LRMS (EI)
240 ([MꢀOCH3]+, 19), 208 (9), 185 (23), 180 (40), 152
(59), 125 (59), 93 (53), 81 (44), 75 (100); HRMS (EI) m/z
calculated for C13H21NO5 [M+] 271.1420, found 240.1231
for [MꢀOCH3]+.
d 133.2, 132.9, 122.7, 122.3, 63.2, 63.2, 33.0, 33.0, 30.1,
30.0, 30.0, 28.7, 28.3, 26.4, 25.8, 25.5, 23.4, 18.8, ꢀ0.5,
ꢀ1.1; IR (neat/NaCl) 3337, 2931, 2857, 1659, 1436, 1416,
1376, 1248, 1162, 1055, 840, 755, 693 cmꢀ1; LRMS (EI)
214 ([M]+, 100), 199 (8), 143 (47), 129 (88); HRMS (EI)
m/z calculated for C12H26OSi [M]+ 214.1753, found
214.1747.
7.10. 3-(7-Methyl-8-trimethylsilanyl-oct-6-enoyl)-
oxazolidin-2-one (22a)
The acid substrate was prepared using the same method
as described above during the synthesis of imide 18a. In
this case, the experiment used 21a (3.236 g, 15.1 mmol),
100 mL of anhydrous DMF in a 250 mL flame-dried
round-bottom flask, and PDC (22.32 g, 59.3 mmol). The
crude product was chromatographed through a silica gel
column (crude/gel¼1:25) using 1:2 ether/hexane as eluant
in order to afford the acid product (1.6805 g, 7.4 mmol,
49%) as a colorless oily liquid.
7.9. 7-Methyl-8-trimethylsilanyl-oct-6-en-1-ol (21a)
The hemiacetal from the reduction of 3-caprolactone was
prepared using the same method as for the synthesis of
enol ether (17a) as described above.
The coupling reaction was also performed using the same
method as described earlier for the synthesis of imide 18a.
In this experiment, the acid made in the preceding paragraph
(1.384 g, 6.1 mmol), 65 mL of anhydrous THF in flamed-
dried 100-mL round-bottom flask, Et3N (3.4 mL,
24.3 mmol), trimethylacetyl chloride (0.75 mL, 6.1 mmol),
oven-dried lithium chloride (0.258 g, 6.2 mmol), and 2-oxa-
zolidone (0.548 g, 6.2 mmol) were used. The crude product
was chromatographed through a silica gel column (crude/
gel¼1:30) with 2–3% Et3N in 1:1 ether/hexane as eluant
in order to afford the desired imide 22a (1.634 g,
5.5 mmol, 91%) as a colorless oily liquid. The spectral
A flamed-dried 500-mL round-bottom flask under an argon
atmosphere was charged with a suspension of (ethyl)triphe-
nylphosphonium bromide (18.567 g, 49.5 mmol) in 150 mL
of anhydrous THF. The mixture was cooled to ꢀ78 ꢁC,
a 1.4 M sec-butyllithium solution in cyclohexane (37 mL,
51.8 mmol) was added dropwise, and the resulting dark
brown mixture allowed to warm from ꢀ78 ꢁC to room tem-
perature over a period of 3 h. The reaction was recooled to
ꢀ78 ꢁC, iodomethyl-trimethylsilane (7.2 mL, 50.2 mmol)
was added dropwise, and the mixture was allowed to warm
to room temperature. After 3 h, the mixture became a cloudy
red-brown solution. To this mixture was added a 1.6 M
n-butyllithium in hexane (32 mL, 51.2 mmol) in a dropwise
fashion. The resulting dark brown mixture was allowed to
warm from ꢀ78 ꢁC to room temperature over a period of
3 h. The reaction was recooled to ꢀ78 ꢁC and then the crude
hemiacetal (2.4 g, 20.7 mmol) in 40 mL of anhydrous THF
added via a cannula. The mixture was allowed to warm to
room temperature and stirred overnight. The reaction was
then quenched with 300 mL of satd NaHCO3. The layers
were separated and the aqueous phase was extracted with
ether (6ꢂ300 mL). The combined organic layers were dried
over anhydrous MgSO4, filtered and concentrated in vacuo.
The crude product was then taken up in 150 mL of 1:3 ether/
hexane and allowed to stand for 4 h in order to precipitate
most of the triphenylphosphine and triphenylphosphine
oxide present in the crude. This mixture was filtered and con-
centrated in vacuo for the second time. The crude product
was chromatographed through a silica gel column (crude/
gel¼1:25) using 2–3% Et3N in 1:2 ether/hexane in order
to afford the desired product 21a (2.793 g, 13.0 mmol,
63%) as a colorless oily liquid. The spectral data were as fol-
lows: 1H NMR (CDCl3/300 MHz) d 5.03–4.89 (m, H7, 1H),
3.63 and 3.62 (t and t, J¼6.3 and 6.3 Hz, H2, 2H), 2.03–1.85
(m, H6, 2H), 1.68–1.64 (m, allylic coupling, H8, 1.65H),
1.59–1.56 (m, allylic coupling, H8, 1.35H), 1.65–1.51 (m,
H1 and H3, 3H), 1.49 (s, H9, 1.26H), 1.45 (s, H9, 0.74H),
1.41–1.38 (m, H4 and H5, 4H), 0.02 (s, H10, 5.35H),
ꢀ0.01 (s, H10, 3.65H); 13C NMR (CDCl3/300 MHz)
1
data were as follows: H NMR (CDCl3/300 MHz) d 5.01–
4.88 (m, H7, 1H), 4.94 (t, J¼8.0 Hz, H1, 2H), 4.00 (t,
J¼8.0 Hz, H2, 2H), 2.90 (t, J¼7.8 Hz, H3, 2H), 2.04–1.87
(m, H6, 2H), 1.71–1.59 (m, H4, 2H), 1.67–1.62 (m, allylic
coupling, H8, 2.18H), 1.53–1.55 (m, allylic coupling, H8,
0.82H), 1.48 (s, H9, 1.38H), 1.44 (s, H9, 0.62H), 1.45–
1.31 (m, H5, 2H), 0.01 (s, H10, 5.78H), ꢀ0.02 (s, H10,
3.22H); 13C NMR (CDCl3/300 MHz) d 173.7, 153.7,
133.4, 133.1, 122.2, 121.9, 62.2, 42.7, 35.2, 35.2, 30.0,
29.8, 29.6, 28.4, 28.1, 26.4, 24.2, 24.1, 23.4, 18.8, ꢀ0.5,
ꢀ1.1; IR (neat/NaCl) 3546, 3384, 2951, 1782, 1700, 1478,
1387, 1246, 1110, 1040, 953, 851, 760, 695 cmꢀ1; LRMS
(EI) 297 ([M]+, 2), 195 (8), 160 (100); HRMS (EI) m/z cal-
culated for C15H27NO3Si [M]+ 297.1760, found 297.1765.
7.11. 3-(7-Methyl-1-triisopropylsilanyloxy-8-trimethyl-
silanyl-octa-1,6-dienyl)-oxazolidin-2-one (23a)
The TIPS silyl enol ether was prepared using the same
method as described earlier for the synthesis of ketene acetal
19a. This experiment used freshly distilled diisopropyl
amine (0.65 mL, 4.6 mmol) and 20 mL of anhydrous THF
in a 100-mL round-bottom flask, a 1.6 M n-butyllithium in
hexane solution (2.9 mL, 4.6 mmol), imide 22a (1.234 g,
4.2 mmol), 16 mL of anhydrous THF, and triisopropylsilyl
trifluoromethanesulfonate (1.3 mL, 4.7 mmol) in a 50-mL
pear-bottom flask. The crude product was chromatographed
through a silica gel column (crude/gel¼1:25) with 2–3%
Et3N in 1:2 ether/hexane as eluant in order to afford the
desired ketene acetal 23a (1.567 g, 3.5 mmol, 83%) as a

6546
Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
1
colorless oily liquid. The spectral data were as follows: H
NMR (CDCl3/300 MHz) d 5.00–4.88 (m, H7, 1H), 4.65
and 4.64 (t and t, J¼7.2 and 7.2 Hz, H3, 1H), 4.29 (dd,
J¼8.7, 7.2 Hz, H1, 2H), 3.75 (dd, J¼8.7, 7.2 Hz, H2, 2H),
2.09 (dt, J¼7.5, 7.5 Hz, H4, 2H), 2.02–1.84 (m, H6, 2H),
1.64 (m, allylic coupling, H8, 2.04H), 1.56 (s, H8, 0.96H),
1.48 (s, H9, 1.36H), 1.43 (s, H9, 0.64H), 1.37 (p,
J¼7.5 Hz, H5, 2H), 1.10 (m of H12 with d of H11 at 1.06,
J¼5.7 Hz, 21H), 0.00 (s, H10, 6.12H), ꢀ0.03 (s, H10,
2.88H); 13C NMR (CDCl3/300 MHz) d 155.9, 140.0,
133.3, 133.0, 122.4, 122.1, 105.0, 61.8, 46.6, 30.1, 30.0,
29.8, 28.4, 28.2, 26.4, 25.4, 25.4, 23.3, 18.0, 13.2, ꢀ0.6,
ꢀ1.1; IR (neat/NaCl) 3519, 2946, 2867, 1770, 1681, 1464,
1398, 1248, 1054, 883, 856, 687 cmꢀ1; LRMS (EI) 453
(M+, 5), 410 ([MꢀCH(CH3)2]+, 67), 366 (14), 244 (100),
210 (36), 200 (51), 128 (59); HRMS (EI) m/z calculated
for C24H47NO3Si2 [M+] 453.3095, found 453.3097.
500-mL round-bottom flask. The crude aldehyde was used
directly in the following step without purification. The alde-
hyde was diluted with solvent (CH2Cl2 or ether) in order to
avoid polymerization.
The allylsilane alcohol 21b was prepared using the same
method as described earlier for the synthesis of 21a using
ethyltriphenylphosphonium bromide (23.538 g, 62.8 mmol)
and 250 mL of anhydrous THF in a 500 mL flame-dried
round-bottom flask, a 1.4 M sec-butyllithium solution in
cyclohexane (48 mL, 67.2 mmol), iodomethyl-trimethyl-
silane (9.4 mL, 62.7 mmol), a 1.6 M n-butyllithium solution
in hexane (41 mL, 65.6 mmol), and the crude aldehyde syn-
thesized above in 40 mL of anhydrous THF. The crude prod-
uct was chromatographed through a silica gel column (crude/
gel¼1:25) using 2–3% Et3N in 1:1.5 ether/hexane as eluant
in order to afford the desired product 21b (3.433 g,
15 mmol, 59%, over 2 steps) as an almost colorless oily liq-
1
7.12. 3-(2-Isopropenyl-cyclopentanecarbonyl)-
oxazolidin-2-one (24a)
uid. The spectral data were as follows: H NMR (CDCl3/
300 MHz) d 5.02–4.89 (m, H8, 1H), 3.63 and 3.61 (apparent
overlap of t and t, J¼6.3 and 6.6 Hz, H2, 2H), 2.02–1.83 (m,
H7, 2H), 1.68–1.64 (m, allylic coupling, H9, 1.34H), 1.66–
1.52 (m, H1 and H3, 3H), 1.59–1.56 (m, allylic coupling,
H9, 1.63H), 1.49 (s, H10, 0.46H), 1.45 (s, H10, 0.54H),
1.40–1.25 (m, H4, H5 and H6, 6H), 0.02 (s, H11,
4.15H), ꢀ0.01 (s, H11, 4.85H); 13C NMR (CDCl3/
300 MHz) d 133.0, 132.7, 122.8, 122.5, 63.1, 33.0, 32.9,
30.3, 30.1, 30.0, 29.5, 29.2, 28.7, 28.2, 26.4, 25.9, 25.9,
23.3, 18.8, ꢀ0.5, ꢀ1.1; IR (neat/NaCl) 3326, 2929,
2855, 1658, 1437, 1416, 1376, 1248, 1162, 1056, 856,
756, 694 cmꢀ1; LRMS (EI) 228 ([M]+, 84), 143 (100),
129 (73); HRMS (EI) m/z calculated for C13H28OSi
[M]+ 228.1909, found 228.1922.
The anodic cyclization was conducted using the same
procedure as described above for the synthesis of 20a.
The experiment used oven-dried tetraethylammonium
tosylate (0.517 g, 1.7 mmol), ketene acetal 23a (0.225 g,
0.05 mmol), 17 mL 1:4 THF/MeOH, 2,6-lutidine (0.36 mL,
3.1 mmol), an RVC anode and cathode, a constant current
˚
of 8 mA, and 2.2 F/mol of charge. The crude product was
chromatographed through a silica gel column (crude/
gel¼1:35) with 2–3% Et3N in 1:1 ether/hexane as eluant
in order to afford the desired product 24a (0.083 g,
0.037 mmol, 75%) as white solid, and 0.016 g of starting ma-
terial 23a was recovered (0.03 mmol, 7%). The spectral data
were as follows: Isomer 1: 1H NMR (CDCl3/300 MHz)
d 4.72 and 4.70 (s and s, H9, 2H), 4.41–4.16 (m, H1 and
H3, 3H), 4.05–3.81 (m, H2, 2H), 3.01 (dt, approx. J¼8.3,
8.3 Hz, H7, 1H), 2.13–1.97, 1.95–1.76 and 1.64–1.52 (m,
H4, H5 and H6, 6H), 1.67 (s, H8, 3H); 13C NMR (CDCl3/
300 MHz) d 175.0, 153.5, 146.2, 111.7, 62.0, 50.0, 46.8,
43.0, 30.4, 29.1, 24.7, 22.1; IR (neat/NaCl) 3508, 3363,
3083, 2957, 2868, 1778, 1690, 1644, 1477, 1388, 1362,
1249, 1226, 1107, 1038, 914, 894, 699 cmꢀ1; LRMS (EI)
223 (M+, 9), 182 (69), 136 (100), 121 (49), 108 (53), 93
(69), 88 (69); HRMS (EI) m/z calculated for C12H17NO3
[M+] 223.1208, found 223.1199. Isomer 2: 1H NMR
(CDCl3/300 MHz) d 4.71 and 4.68 (s and s, H9, 2H), 4.38
(t, J¼8.3 Hz, H1, 2H), 4.00 (t of H2 with H3 buried,
J¼8.3 Hz, 3H), 3.00 (dt, J¼9.8, 8.9 Hz, H7, 1H), 2.28–
2.10, 2.01–1.87 and 1.83–1.46 (m of H4, H5 and H6 with s
of H8 at 1.70, 9H); 13C NMR (CDCl3/300 MHz) d 176.4,
153.4, 146.7, 110.3, 62.0, 50.8, 46.5, 43.0, 31.7, 31.2, 24.9,
20.6; IR (neat/NaCl) 3538, 3075, 2979, 2953, 2871, 1795,
1772, 1682, 1645, 1485, 1386, 1362, 1223, 1041, 899,
757 cmꢀ1; LRMS (EI) 223 (M+, 16), 136 (100), 121 (100),
109 (80), 108 (73), 93 (100), 88 (66); HRMS (EI) m/z calcu-
lated for C12H17NO3 [M+] 223.1208, found 223.1212.
7.14. 3-(8-Methyl-9-trimethylsilanyl-non-7-enoyl)-
oxazolidin-2-one (22b)
The acid intermediate was prepared using the same method
as described for the synthesis of acid 18a. This experiment
used 21b (2.175 g, 9.5 mmol), 50 mL of anhydrous DMF,
and PDC (12.434 g, 33 mmol). The reaction was quenched
with 500 mL of water, the layers separated, and the aqueous
layer extracted with ether (12ꢂ150 mL). The combined
organic layers were dried over anhydrous MgSO4, filtered
and concentrated in vacuo. The crude product was chromato-
graphed through a silica gel column (crude/gel¼1:25) using
1:6 EtOAc/hexane as eluant to afford the acid product
(1.484 g, 6.1 mmol, 64%) as an almost colorless oily liquid.
The imide intermediate was prepared using the procedure as
described earlier for the synthesis of imide 18a. This exper-
iment used 0.146 g (6.0 mmol) of the acid synthesized in
the preceding paragraph in 20 mL of anhydrous THF, anhy-
drous Et3N (0.4 mL, 2.9 mmol), trimethylacetyl chloride
(0.08 mL, 6.4 mmol), oven-dried lithium chloride (0.03 g,
7.2 mmol), and 2-oxazolidone (0.06 g, 6.7 mmol). The
crude product was chromatographed through a silica gel col-
umn (crude/gel¼1:30) using 2–3% Et3N in 3:4 ether/hexane
as eluant in order to afford the desired imide 22b (0.171 g,
5.8 mmol, 96%) as a colorless oily liquid. The spectral
7.13. 8-Methyl-9-trimethylsilanyl-non-7-en-1-ol (21b)
The aldehyde from the hydrolysis of enol ether 17a was pre-
pared using the same method for the synthesis of 17b as de-
scribed above. This experiment used enol ether 17a (3.671 g,
25.5 mmol) and 300 mL of 4 N HCl/acetone (1:1) in a
1
data were as follows: H NMR (CDCl3/300 MHz) d 5.01–
4.87 (m, H8, 1H), 4.40 (t, J¼8.1 Hz, H1, 2H), 4.00 (t,
J¼8.0 Hz, H2, 2H), 2.90 (apparent t, J¼7.5 Hz, H3, 2H),

Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
6547
2.02–1.82 (m, H7, 2H), 1.72–1.58 (m, H4, 2H), 1.67–1.62
(m, allylic coupling, H9, 1.50H), 1.58–1.55 (m, allylic cou-
pling, H9, 1.50H), 1.48 (s, H10, 1H), 1.44 (s, H10, 1H), 1.42–
1.29 (m, H5 and H6, 4H), 0.01 (s, H11, 4.50H), ꢀ0.02 (s,
H11, 4.50H); 13C NMR (CDCl3/300 MHz) d 173.7, 153.7,
133.1, 132.8, 122.6, 122.3, 62.2, 42.7, 35.3, 35.3, 30.0,
30.0, 29.9, 29.1, 28.9, 28.5, 28.1, 26.4, 24.4, 24.4, 23.3,
18.8, ꢀ0.5, ꢀ1.1; IR (neat/NaCl) 3545, 3385m, 2952,
2929, 2855, 1782, 1700, 1479, 1387, 1247, 1224, 1040,
857, 760, 696 cmꢀ1; LRMS (EI) 311 ([M]+, 5), 160 (100),
144 (11); HRMS (EI) m/z calculated for C16H29NO3Si
[M]+ 311.1917, found 311.1929.
H9, 3H), 1.46–1.13 (m, H5 and H6, 4H); 13C NMR
(CDCl3/300 MHz) d 176.3, 153.5, 149.3, 109.8, 61.9, 47.0,
45.1, 42.8, 31.9, 30.5, 26.2, 25.7, 21.5; IR (neat/NaCl)
3080, 2938, 2912, 2850, 1788, 1766, 1689, 1648, 1475,
1437, 1384, 1360, 1270, 1193, 1038, 942, 909, 759 cmꢀ1
;
LRMS (EI) 237 (2), 150 (100), 135 (28); HRMS (EI) m/z cal-
culated for C13H19NO3 [M]+ 237.1365, found 237.1362.
7.17. 3-[1-(tert-Butyl-dimethyl-silanyloxy)-8-methyl-9-
trimethylsilanyl-ona-1,7-dienyl]-oxazolidin-2-one (23c)
The tert-butyldimethylsilyl enol ether was prepared using the
same method as described above for the synthesis of ketene
acetal 19a. In this experiment, freshly distilled diisopropyl
amine (0.22 mL, 1.6 mmol) in 20 mL anhydrous THF was
used along with a 2.5 M solution of n-butyllithium in hexane
(0.63 mL, 1.6 mmol), imide 22b (0.457 g, 1.5 mmol) in
10 mL of anhydrous THF, and tert-butyldimethylsilyl tri-
fluoromethanesulfonate (0.38 mL, 1.6 mmol). The crude
product was chromatographed through a silica gel column
(crude/gel¼1:25) using 2–3% Et3N in 1:2 ether/hexane as
eluant in order to afford the desired product 23c (0.347 g,
0.8 mmol, 56%) as an almost colorless oily liquid. The spec-
tral datawere as follows: 1H NMR (CDCl3/300 MHz) d 5.02–
4.89 (m, H8, 1H), 4.75 (t, J¼7.2 Hz, H3, 0.35H), 4.75 (t,
J¼7.2 Hz, H3, 0.65H), 4.32 (dd, J¼8.7, 7.4 Hz, H1, 2H),
3.74 (dd, J¼8.7, 7.4 Hz, H2, 2H), 2.07 (dt, J¼7.2, 7.2, H4,
2H), 2.02–1.84 (m, H7, 2H), 1.67 (m, allylic coupling, H9,
1.05H), 1.58 (m, allylic coupling, H9, 1.95H), 1.50 (s, H10,
0.7H), 1.46 (s, H10, 1.3H), 1.43–1.23 (m, H5 and H6, 4H),
0.97 (s, H13, 9H), 0.18 (s, H12, 6H), 0.023 (s, H11,
3.15H), 0.002 (s, H11, 5.85H); 13C NMR (CDCl3/
300 MHz) d 155.9, 139.1, 133.1, 132.8, 122.8, 122.5,
106.7, 106.7, 61.9, 45.6, 30.1, 30.0, 23.0, 29.5, 29.2, 28.6,
28.2, 26.5, 25.8, 25.6, 25.5, 23.4, 18.8, 18.2, 14.3, ꢀ0.5,
ꢀ1.0, ꢀ4.5; IR (neat/NaCl) 2954, 2930, 2857, 1766, 1683,
1472, 1402, 1249, 1041, 841, 783 cmꢀ1; LRMS (EI) 425
([M]+, 1), 410 ([MꢀCH3]+, 12), 368 ([MꢀC(CH3)2]+, 42),
202 (100), 160 (90); HRMS (EI) m/z calculated for
C22H43NO3Si2 [M]+ 425.2782, found 425.2773.
7.15. 3-(8-Methyl-1-triisopropylsilanyloxy-9-trimethyl-
silanyl-nona-1,7-dienyl)-oxazolidin-2-one (23b)
The triisopropylsilyl enol ether was prepared using the same
method as described earlier for the synthesis of ketene acetal
19a. This experiment used freshly distilled diisopropyl
amine (0.33 mL, 2.4 mmol) in 15 mL of anhydrous THF,
a 1.6 M n-butyllithium (1.5 mL, 2.4 mmol) in hexane solu-
tion, imide 22b (0.669 g, 2.1 mmol) in 10 mL anhydrous
THF, and triisopropylsilyl trifluoromethanesulfonate
(1.3 mL, 4.7 mmol). The crude product was chromato-
graphed through a silica gel column (crude/gel¼1:25) using
2–3% Et3N in 1:2 ether/hexane as eluant in order to afford
the desired ketene acetal 23b (0.814 g, 1.7 mmol, 81%) as
an almost colorless oily liquid. The spectral data were as fol-
lows: 1H NMR (CDCl3/300 MHz) d 4.99–4.87 (m, H8, 1H),
4.63 (t, J¼7.2 Hz, H3, 1H), 4.29 (dd, J¼8.4, 7.1 Hz, H1,
2H), 3.75 (dd, J¼8.4, 7.1 Hz, H2, 2H), 2.08 (dt, J¼7.2,
7.0 Hz, H4, 2H), 1.99–1.83 (m, H7, 2H), 1.64 (m, allylic
coupling, H9, 1.5H), 1.55 (s, H9, 1.5H), 1.47 (s, H10, 1H),
1.43 (s, H10, 1H), 1.40–1.26 (m, H5 and H6, 4H), 1.23–
1.06 (m of H13 with d of H12 at 1.09, J¼6 Hz, 21H), 0.00
(s, H11, 4.5H), ꢀ0.03 (s, H11, 4.5H); 13C NMR (CDCl3/
300 MHz) d 155.9, 139.9, 132.9, 132.6, 122.7, 122.4,
105.1, 105.1, 61.8, 46.6, 30.0, 29.9, 29.4, 29.1, 28.5, 28.1,
26.4, 25.6, 25.5, 23.3, 17.9, 13.1, ꢀ0.6, ꢀ1.1; IR (neat/
NaCl) 2946, 2868, 1769, 1680, 1464, 1399, 1368, 1279,
1248, 1061, 1040, 883, 856, 687 cmꢀ1; LRMS (EI) 467
(M+, 11), 452 ([MꢀCH3]+, 5), 424 ([MꢀCH(CH3)2]+, 52),
244 (100), 200 (56), 128.1 (37); HRMS (EI) m/z calculated
for C25H49NO3Si2 [M]+ 467.3251, found 467.3256.
7.18. (S)-(L)-4-Benzyl-3-(8-methyl-9-trimethylsilanyl-
non-7-enoyl)-oxazolidin-2-one (39a)
The imide intermediate 39a was prepared using the proce-
dure as described earlier for the synthesis of imides 18a
and 22b. This experiment used the acid derived from 21b
(see above) in 90 mL of anhydrous THF, anhydrous Et3N
(5.0 mL, 35.9 mmol), trimethylacetyl chloride (1.1 mL,
8.8 mmol), oven-dried lithium chloride (0.345 g,
8.4 mmol), and (S)-(ꢀ)-4-benzyl-2-oxazolidinone (1.491 g,
8.3 mmol). The crude product was chromatographed
through a silica gel column (crude/gel¼1:30) with 2–3%
Et3N in 1:4 ether/hexane as eluant in order to afford the de-
sired imide 39a (2.729 g, 6.8 mmol, 82%) as a colorless oily
liquid. The spectral data were as follows: 1H NMR (CDCl3/
300 MHz) d 7.35–7.15 (m, H4, H40, H5, H50, H6, 5H), 5.01–
4.88 (m, H12, 1H), 4.69–4.59 (m, H2, 1H), 4.21–4.10 (m, H1
and H10, 2H), 3.27 (dd, J¼13.2, 3.2 Hz, H3, 1H), 3.02–2.80
(m, H7, 2H), 2.74 (dd, J¼13.2, 9.6 Hz, H30, 1H), 2.03–1.83
(m, H11, 2H), 1.72–1.60 (m, H8, 2H), 1.68–1.61 and 1.58–
1.54 (m and m, allylic coupling, H13, 3H), 1.48 and 1.43 (s
and s, H14, 2H), 1.42–1.26 (m, H9 and H10, 4H), 0.00 and
7.16. 3-(2-Isopropenyl-cyclohexanecarbonyl)-
oxazolidin-2-one (24b)
The anodic cyclization was conducted using the same proce-
dure as described for the synthesis of 20a. This experiment
used oven-dried tetraethylammonium tosylate (0.459 g,
1.5 mmol), ketene acetal 23b (0.091 g, 0.19 mmol), 15 mL
of 1:4 THF/MeOH, 2,6-lutidine (0.16 mL, 1.4 mmol), an
˚
RVC anode and cathode, a constant current of 8 mA, and
2.2 F/mol of charge. The crude product was chromato-
graphed through a silica gel column (crude/gel¼1:35) using
2–3% Et3N in 1:1 ether/hexane as eluant in order to afford the
desired product 24b (0.033 g, 0.14 mmol, 71%) as white
1
solid. The spectral data were as follows: H NMR (CDCl3/
300 MHz) d 4.66 and 4.66 (s and s, H10, 2H), 4.40–4.32 (ap-
parent t, J¼8.1 Hz, H1, 2H), 3.96 (t, J¼8.1 Hz, H2, 2H), 3.88
(td, J¼11.2, 3.6 Hz, H3, 1H), 2.38 (td, J¼11.2, 3.2 Hz, H8,
1H), 2.02–1.94 and 1.86–1.73 (m, H4 and H7, 4H), 1.71 (s,

6548
Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
0.03 (s and s, H15, 9H); 13C NMR (CDCl3/300 MHz)
d 173.4, 153.5, 135.5, 133.0, 132.8, 129.5, 129.0, 127.4,
122.6, 122.3, 66.2, 55.2, 38.0, 35.6, 30.0, 29.9, 29.8, 29.0,
28.9, 28.5, 28.1, 26.4, 24.4, 24.3, 23.3, 18.7, ꢀ0.6, ꢀ1.1;
IR (neat/NaCl) 3550, 3385, 2930, 2856, 1790, 1770, 1704,
1704, 1698, 1604, 1454, 1385, 1246, 1210, 1107, 1029,
875, 701 cmꢀ1; LRMS (EI) 401 ([M]+, 10), 250 (100), 73
(59); HRMS (EI) m/z calculated for C23H35NO3Si [M]+
401.2386, found 401.2381.
ether/hexane as eluant. The spectral data were as follows:
Isomer 1 (trans): H NMR (CDCl3/600 MHz) d 7.35–7.18
1
(m, H4, H40, H5, H50 and H6, 5H), 4.70–4.67 (m, allylic cou-
pling, H14 and H14, 2H), 4.67–4.61 (m, H2, 1H), 4.17–4.12
(m, H1 and H10, 2H), 3.84 (dt, J¼3.3, 11.0 Hz, H7, 1H), 3.22
(dd, J¼2.7, 13.5 Hz, H3, 1H), 2.74 (dd, J¼9.6, 13.2 Hz, H30,
1H), 2.41 (dt, J¼3.3, 11.0 Hz, H12, 1H), 2.11–2.05 (m, H9,
1H), 1.88–1.77 (m, H8, H10, and H11, 3H), 1.73 (s, H13,
3H), 1.46–1.30 (m, H80, H90, and H100, 3H), 1.28–1.19 (m,
H110, 1H); 13C NMR (CDCl3/600 MHz) d 176.1, 153.4,
149.5, 135.5, 129.7, 129.1, 127.5, 109.8, 66.2, 55.3, 46.8,
45.4, 38.1, 32.0, 30.6, 26.6, 26.3, 25.8, 21.7. Isomer 2
7.19. (S)-(L)-4-Benzyl-3-[1-(tert-butyl-dimethyl-silanyl-
oxy)-8-methyl-9-trimethylsilanyl-nona-1,7-dienyl]-
oxazolidin-2-one (40a)
1
(trans): H NMR (CDCl3/600 MHz) d 7.34–7.19 (m, H4,
H40, H5, H50, and H6, 5H), 4.79–4.77 and 4.75–4.72 (m,
allylic coupling, H14 and H140, 2H), 4.68–4.63 (m, H2,
1H), 4.16–4.09 (m, H1 and H10, 2H), 3.96 (dt, J¼3.2,
11.0 Hz, H7, 1H), 3.23 (dd, J¼3.3, 13.5 Hz, H3, 1H), 2.62
(dd, J¼9.9, 13.5 Hz, H30, 1H), 2.46 (dt, J¼3.2, 11.0 Hz,
H12, 1H), 2.00–1.95 (m, H8, 1H), 1.87–1.82 (m, H11,
1H), 1.82–1.78 (m, H9 and H10, 2H), 1.77–1.75 (m, allylic
coupling, H13, 3H), 1.47–1.23 (m, H80, H90, H100, H110,
4H); 13C NMR (CDCl3/600 MHz) d 176.3, 153.5, 149.0,
135.7, 129.7, 129.1, 127.4, 110.5, 65.9, 44.4, 47.5, 45.0,
38.1, 31.9, 30.6, 26.2, 25.7, 21.5.
The tert-butyldimethylsilyl enol ether 40a was prepared us-
ing the same method initially as described above for the syn-
thesis of ketene acetal 19a above. In this experiment, freshly
distilled diisopropyl amine (0.18 mL, 1.3 mmol) in 20 mL
anhydrous THF was used along with a 1.6 M n-butyllithium
(0.86 mL, 1.4 mmol), imide 39a (0.461 g, 1.2 mmol) in
20 mL of anhydrous THF, and tert-butyldimethylsilyl
trifluoromethanesulfonate (0.35 mL, 1.5 mmol). The crude
product was chromatographed through a silica gel column
(crude/gel¼1:25) using 2–3% Et3N in 1:4 ether/hexane as
eluant in order to afford the desired ketene acetal 40a
(0.335 g, 0.65 mmol, 57%) as a colorless oily liquid and
starting material 40a (w0.18 g, 0.45 mmol, 39%). The spec-
tral data were as follows: 1H NMR (CDCl3/300 MHz)
d 7.36–7.12 (m, H4, H40, H5, H50, H6, 5H), 5.04–4.90 (m,
H12, 1H), 4.86 (dd, J¼6.3, 6.3 Hz, H7, 1H), 4.24–4.10 (m,
H1 and H10, 2H), 4.08–3.97 (m, H2, 1H), 3.2 (dd, J¼13.5,
3.3 Hz, H3, 1H), 2.61 (dd, J¼13.5, 9.6 Hz, H30, 1H), 2.32–
1.85 (m, H8 and H11, 4H), 1.69–1.63 (m, allylic coupling,
H13, 1.59H), 1.61–1.56 (m, allylic coupling, H13, 1.41H),
1.48 and 1.43 (s and s, H14, 2H), 1.48–1.30 (m, H9 and
H10, 4H), 1.00 (s, H17, 9H), 0.23 and 0.17 (s and s, H16
and H160, 6H), 0.0 and ꢀ0.02 (s and s, H15, 9H); 13C
NMR (CDCl3/300 MHz) d 155.5, 137.2, 135.8, 132.9,
132.7, 129.5, 129.0, 127.2, 122.6, 122.3, 109.9, 66.7, 56.3,
38.3, 38.2, 30.0, 29.9, 29.8, 29.3, 29.1, 28.4, 28.0, 26.3,
25.7, 25.6, 25.5, 23.3, 18.7, 18.1, ꢀ0.6, ꢀ1.1, ꢀ4.4, ꢀ4.7;
IR (neat/NaCl) 3520, 3087, 3063, 3068, 2928, 2857, 1770,
1682, 1644, 1585, 1497, 1454, 1393, 1248, 1122, 1090,
1043, 838, 699 cmꢀ1; LRMS (EI) 401 ([M]+, 10), 250
(100), 73 (59); HRMS (EI) m/z calculated for
C23H35NO3Si [M]+ 401.2386, found 401.2381.
7.21. (S)-(L)-4-Isopropyl-3-(8-methyl-9-trimethyl-
silanyl-non-7-enoyl)-oxazolidin-2-one (39b)
The imide intermediate 39b was prepared using the proce-
dure as described earlier for the synthesis of imide 18a and
22b. This experiment used the acid derived from alcohol
21b (0.481 g, 2.0 mmol) in 30 mL of anhydrous THF, anhy-
drous Et3N (1.2 mL, 8.6 mmol), trimethylacetyl chloride
(0.26 mL, 2.1 mmol), oven-dried lithium chloride (0.095 g,
2.3 mmol), and (S)-(ꢀ)-4-benzyl-2-oxazolidinone (0.276 g,
2.1 mmol). The crude product was chromatographed
through a silica gel column (crude/gel¼1:30) with 2–3%
Et3N in 1:5 ether/hexane as eluent in order to afford the de-
sired imide 39b (0.599 g, 1.7 mmol, 85%) as a colorless oily
liquid. The spectral data were as follows: 1H NMR (CDCl3/
300 MHz) d 5.02–4.88 (m, H10, 1H), 4.47–4.39 (m, H2,
1H), 4.26 (apparent t, J¼9.0 Hz, H1, 1H), 4.19 (dd, J¼9.0,
3.3 Hz, H10, 1H), 3.05–2.78 (m, H5, 2H), 2.37 (m, H3,
1H), 2.03–1.82 (m, H9, 2H), 1.72–1.59 (m, H6, 2H), 1.68–
1.63 (m, allylic coupling, H11, 1.61H), 1.59–1.56 (m, allylic
coupling, H11, 1.39H), 1.49 (s, H12, 1.04H), 1.45 (s, H12,
0.96H), 1.42–1.29 (m, H7 and H8, 4H), 0.91 and 0.87
(d and d, J¼7.2, 6.9 Hz, H4 and H40, 6H), 0.01 and ꢀ0.01
(s and s, H13, 9H); 13C NMR (CDCl3/300 MHz) d 173.2,
154.0, 132.8, 132.5, 122.5, 122.2, 63.3, 58.3, 35.4,
29.8, 29.8, 29.7, 28.9, 28.7, 28.4, 28.3, 28.0, 26.2, 24.5,
24.4, 18.6, 17.9, 14.7, ꢀ0.7, ꢀ1.3; IR (neat/NaCl) 2957,
7.20. (S)-(L)-4-Benzyl-3-(2-isopropenyl-cyclohexane-
carbonyl)-oxazolidin-2-one (41a)
The anodic cyclization was conducted using the same proce-
dure as described for the synthesis of 20a. This experiment
used oven-dried tetraethylammonium tosylate (0.458 g,
1.5 mmol), ketene acetal 40a (0.0959 g, 0.19 mmol),
15 mL MeOH, 2,6-lutidine (0.14 mL, 1.2 mmol), an RVC
2876, 2856, 1785, 1703, 1386, 1247, 1206, 858 cmꢀ1
;
LRMS (EI) 353 ([M]+, 6), 202 (100), 158 (9); HRMS (EI)
m/z calculated for C19H35NO3Si [M]+ 353.2386, found
353.2384.
˚
anode and cathode, a constant current of 8 mA, and
2.2 F/mol of charge. The crude product was chromato-
graphed (crude/gel¼1:30) through a silica gel column using
2–3% Et3N in 1:4 ether/hexane as eluant in order to afford
the desired product 41a (0.008 g, 0.0023 mmol, 13%) as
a colorless oil and imide 39a (0.039 g, 0.1 mmol, 53%). A
second chromatography (crude/gel¼1:60) was performed
to further separate two isomers using 2–3% Et3N in 1:4
7.22. 3-[1-(tert-Butyl-dimethyl-silanyloxy)-8-methyl-9-
trimethyl-silanyl-nona-1,7-dienyl]-(S)-(L)-4-isopropyl-
oxazolidin-2-one (40b)
The tert-butyldimethylsilyl enol ether 40b was prepared
using the same method as described earlier for the synthesis

Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
6549
of ketene acetal 19a. In this experiment, freshly distilled di-
isopropyl amine (0.2 mL, 1.4 mmol) in 30 mL of anhydrous
THF was used along with 1.6 M n-butyllithium (0.87 mL,
1.4 mmol), imide 39b (0.363 g, 1.0 mmol) in 20 mL of an-
hydrous THF, and tert-butyldimethylsilyl trifluoromethane-
sulfonate (0.34 mL, 1.5 mmol). The crude product was
chromatographed through a silica gel column (crude/
gel¼1:25) using 2–3% Et3N in 1:4 ether/hexane as eluant
in order to afford the desired ketene acetal 40b (0.423 g,
0.9 mmol, 88%) as a colorless oily liquid. The spectral
1.86–1.76 (m, H6 and H9, 4H), 1.75–1.72 (m, allylic cou-
pling, H11, 3H), 1.48–1.18 (m, H7 and H8, 4H), 0.88 and
0.82 (d and d, J¼6.9, 6.8 Hz, H4 and H40, 6H); 13C NMR
(CDCl3/300 MHz) d 176.1, 154.0, 148.9, 110.3, 63.0, 58.6,
31.9, 30.5, 28.5, 27.1, 26.2, 25.6, 21.3, 18.1; IR (neat/
NaCl) 3536, 3380, 3074, 2929, 1772, 1700, 1645, 1487,
1448, 1385, 1300, 1259, 1204, 1060, 894, 777, 711 cmꢀ1
;
LRMS (EI) 279 (M+, 4), 200 (12), 186 (100), 150 (71),
100 (32), 73 (57); HRMS (EI) m/z calculated for
C16H25NO3 [M+] 279.1834, found 279.1839.
1
data were as follows: H NMR (CDCl3/300 MHz) d 5.02–
4.88 (m, H10, 1H), 4.83–4.75 (m, H5, 1H), 4.23 (apparent
t, J¼9.0 Hz, H1, 1H), 4.07 (dd, J¼9.3, 8.7 Hz, H10, 1H),
3.98–3.90 (m, H2, 1H), 2.22–1.84 (m, H3, H6 and H5,
5H), 1.68–1.63 and 1.59–1.55 (m and m, allylic coupling,
H11, 3H), 1.49 and 1.45 (s and s, H12, 2H), 1.43–1.24 (m,
H7 and H8, 4H), 0.96 (s, H15, 9H), 0.92 and 0.90 (d and
d, J¼3.3, 3.0 Hz, H4 and H40, 6H), 0.2 and 0.15 (s and s,
H14 and H140, 6H), 0.02 and ꢀ0.01 (s and s, H13, 9H);
13C NMR (CDCl3/300 MHz) d 156.1, 137.7, 133.1, 132.8,
122.8, 122.5, 109.6, 63.3, 59.1, 30.0, 29.9, 29.4, 29.2,
28.9, 28.5, 28.1, 26.5, 25.8, 25.6, 25.5, 23.4, 18.8,
18.2, 17.8, 15.2, ꢀ0.5, ꢀ1.0, ꢀ4.3, ꢀ4.6; IR (neat/NaCl)
2956, 2930, 2857, 1768, 1683, 1472, 1463, 1401, 1248,
1053, 841, 784 cmꢀ1; LRMS (EI) 467 (M+, 11), 452
([MꢀCH3]+, 25), 410 (100), 244 (75), 202 (75); HRMS
(EI) m/z calculated for C25H49NO3Si2 [M+] 467.3251, found
467.3248.
Acknowledgements
We thank the National Science Foundation (CHE-9023698)
for their generous support of our work. We also gratefully
acknowledge the Washington University High Resolution
NMR facility, partially supported by NIH grants RR02004,
RR05018, and RR07155, and the Washington University
Mass Spectrometry Resource Center, partially supported
by NIHRR00954, for their assistance.
References and notes
1. For a review of anodic cyclization reactions see: Moeller, K. D.
Tetrahedron 2000, 56, 9527.
2. For alternative approaches to the oxidation of ketene acetal
derivatives see: (a) Snider, B. B.; Shi, B.; Quickley, C. A.
Tetrahedron 2000, 56, 10127; (b) Ryter, K.; Livinghouse, T.
J. Am. Chem. Soc. 1998, 120, 2658.
7.23. 3-(2-Isopropenyl-cyclohexanecarbonyl)-(S)-(L)-4-
isopropyl-oxazolidin-2-one (41b)
The anodic cyclization was conducted using the same proce-
dure as described for the synthesis of 20a. This experiment
used oven-dried tetraethylammonium tosylate (0.6240 g,
2.1 mmol), ketene acetal 40b (0.1052 g, 0.23 mmol),
20 mL of a 1:1 CH2Cl2/MeOH mixture, 2,6-lutidine
(0.2 mL, 1.33 mmol), an RVC anode and cathode, a constant
3. For a review on the use of ketene dithioacetals in synthesis see:
Kolb, M. Synthesis 1990, 171.
4. (a) For the stereochemical consequences of anodic cyclizations
utilizing ketene dithioacetal groups see: Sun, Y.; Liu, B.; Kao,
J.; d’Avignon, D. A.; Moeller, K. D. Org. Lett. 2001, 3, 1729;
(b) For a total synthesis using a ketene dithioacetal oxidation
see: Liu, B.; Duan, S.; Sutterer, A. C.; Moeller, K. D. J. Am.
Chem. Soc. 2002, 124, 10101.
5. For an example using an allylsilane terminating group see:
Frey, D. A.; Reddy, S. H. K.; Moeller, K. D. J. Org. Chem.
1999, 64, 2805.
6. (a) Hudson, C. M.; Marzabadi, M. R.; Moeller, K. D.; New,
D. G. J. Am. Chem. Soc. 1991, 113, 7372; (b) Tinao-
Wooldridge, L. V.; Moeller, K. D.; Hudson, C. M. J. Org.
Chem. 1994, 59, 2381.
7. For previous syntheses of scopadulcic acid B see: (a) Ziegler,
F. E.; Wallace, O. B. J. Org. Chem. 1995, 60, 3626; (b)
Overman, L. E.; Ricca, D. J.; Tran, V. D. J. Am. Chem. Soc.
1993, 115, 2042; For other related synthetic efforts see: (c)
Robichaud, A. J.; Meyers, A. I. J. Org. Chem. 1991, 56,
2607; (d) Abelman, M. M.; Overman, L. E. J. Am. Chem.
Soc. 1988, 110, 2328; (e) Kucera, D. J.; O’Connor, S. J.;
Overman, L. E. J. Org. Chem. 1993, 58, 5304.
˚
current of 8 mA, and 2.2 F/mol of charge. The crude product
was chromatographed through a silica gel column (crude/
gel¼1:40) using 2–3% Et3N in 1:6 ether/hexane as eluant
in order to afford the desired product 41b (0.0105 g of
Isomer 1, 0.038 mmol, 17%, and 0.0159 g of Isomer 2,
0.057 mmol, 25%) as colorless oil. The spectral data were
1
as follows: Isomer 1 (trans): H NMR (CDCl3/300 MHz)
d 4.70–4.65 (m, allylic coupling, H12, 2H), 4.46–4.39 (m,
H2, 1H), 4.27–4.15 (m, H1 and H10, 2H), 3.90 (dt, J¼3.6,
11.1 Hz, H5, 1H), 2.38 (dt, J¼3.0, 13 Hz, H10, 1H), 2.32
(dqq, J¼1.2, 7.1, 7.1 Hz, H3, 1H), 2.11–2.02 and 1.88–
1.76 (m, H6 and H9, 4H), 1.74–1.71 (m, allylic coupling,
H11, 3H), 1.52–1.15 (m, H7 and H8, 4H), 0.90 and 0.86 (d
and d, J¼7.2, 6.9 Hz, H4 and H40, 6H); 13C NMR (CDCl3/
300 MHz) d 176.0, 154.0, 149.6, 109.6, 63.5, 58.2, 46.6,
45.6, 32.1, 31.0, 29.9, 28.8, 26.3, 25.9, 21.7, 18.1; IR
(neat/NaCl) 3537, 3384, 3076, 2929, 1779, 1698, 1644,
1487, 1447, 1384, 1300, 1203, 1090, 944, 892, 708 cmꢀ1
;
8. Reddy, S. H. K.; Chiba, K.; Sun, Y.; Moeller, K. D. Tetrahedron
2001, 57, 5183.
LRMS (EI) 279 (M+, 7), 242 (10), 150 (100); HRMS (EI)
m/z calculated for C16H25NO3 [M+] 279.1834, found
279.1825. Isomer 2 (trans): H NMR (CDCl3/300 MHz)
9. For the initial study using enol ether trapping groups see:
Moeller, K. D.; Tinao, L. V. J. Am. Chem. Soc. 1992, 113, 1033.
10. The oxidations were conducted using a Model 630 coulometer,
a Model 410 potentiostatic controller, and a Model 420A power
supply purchased from The Electrosynthesis Co., Inc. For
a simple setup to try electrochemical reactions using a battery
1
d 4.77–4.73 and 4.70–4.66 (m, allylic coupling, H12, 2H),
4.47–4.40 (m, H2, 1H), 4.26–4.13 (m, H1 and H10, 2H),
3.99 (dt, J¼3.6, 11.1 Hz, H5, 1H), 2.41 (dt, J¼3.2,
11.4 Hz, H10, 1H), 2.27 (m, H3, 1H), 2.00–1.90 and

6550
Y. Huang, K. D. Moeller / Tetrahedron 62 (2006) 6536–6550
as a power supply see: Frey, D. A.; Wu, N.; Moeller, K. D.
Tetrahedron Lett. 1996, 37, 8317.
(i) Gibson, C. L.; Gillon, K.; Cook, S. Tetrahedron Lett.
´
1998, 39, 6733; (j) de Parrodi, C. A.; Clara-Sosa, A.; Perez,
´
11. (a) Seyfreth, D.; Worsthorn, K. R.; Mammarella, R. E. J. Org.
Chem. 1977, 42, 3104; (b) Flemming, I.; Paterson, I. Synthesis
1979, 445.
12. For more recent examples of potential shifts being used to
measure relative rates of cyclization see Ref. 8 as well as
Sperry, J. B.; Wright, D. L. J. Am. Chem. Soc. 2005, 127,
8034.
13. A model RE-5B Ag/AgCl reference electrode from BAS was
used. The electrode uses a AgCl coated Ag wire immersed in
3 M NaCl. The reference electrodes were calibrated using the
ferrocene/ferrocium cation redox cycle.
14. For allylsilanes see Ref. 5. For vinylsilanes see Hudson, C. M.;
Moeller, K. D. J. Am. Chem. Soc. 1994, 116, 3347.
15. For reviews see: (a) Moon, K. B.; Williams, S. F.; Masumunes,
S. The Aldol Reaction: Group III Enolates; Comprehensive
Organic Synthesis; Trost, B. M., Ed.; Pergamon: New York,
NY, 1991; Vol. 2, Chapter 1.7, pp 239–275; (b) Caine, D.
Alkylation of Enols and Enolates; Comprehensive Organic
Synthesis; Trost, B. M., Ed.; Pergamon: New York, NY,
1991; Vol. 3; See in particular page 35–54; (c) Evans, D. A.;
Nelson, J. V.; Taber, T. R. Top. Stereochem. 1982, 12, 1–115;
(d) Evans, D. A. Stereoselective Alkylation Reactions of
Chiral Metal Enolates; Asymmetric Synthesis; Morison, J. D.,
Ed.; 1984; Vol. 3, pp 1–110; (e) Ager, D. J.; Prakash, I.;
Schaad, D. R. Chem. Rev. 1996, 96, 835–875; (f) Evans,
D. A.; Ennis, M. D.; Mathre, D. J. J. Am. Chem. Soc. 1982,
104, 1737; (g) Evans, S. A.; Wu, L. D.; Wiener, J. J. M.;
Johnson, J. S.; Ripin, D. H. B.; Tedrow, J. S. J. Org. Chem.
1999, 64, 6411; (h) Bull, S. D.; Davies, S. G.; Jones, S.;
Sanganee, H. J. J. Chem. Soc., Perkin Trans. 1 1999, 387;
L.; Quintero, L.; Maran˜on, B.; Toscano, R. A.; Avin˜a, J. A.;
Rojas-Lima, Sl; Juaristi, E. Tetrahedron: Asymmetry 2001,
12, 69; (k) Abdel-Magid, A.; Pridgen, L. N.; Eggleston,
D. S.; Lantos, I. J. Am. Chem. Soc. 1986, 108, 4595–4602.
16. For examples with no added Lewis Acid to chelate the enolate
oxygen and oxazolidinone carbonyl see: Heathcock, C. H. The
Aldol Addition Reaction. Asymmetric Synthesis; Morison, J. D.,
Ed.; 1984; Vol. 3, pp 110–212.
17. Mukund, P. S.; Jianguo, J. Angew. Chem. Int. Ed. 1996, 35,
190–912.
18. High de’s for the radical reactions can be obtained when
a Lewis Acid is added to complex the carbonyl alpha to the
radical and the carbonyl of the oxazolidinone. The complexa-
tion stops rotation of the chiral auxiliary. (a) Mukund, P. S.;
Rheault, T. R. J. Am. Chem. Soc. 2000, 122, 8873; (b)
Mukund, P. S.; Jasperse, C. P.; Jianguo, J. J. Am. Chem. Soc.
1995, 117, 10779; (c) Mukund, P. S.; Porter, N. A. Acc. Chem.
Res. 1999, 32, 163; (d) Mukund, P. S.; Jianguo, J.; Sausker,
J. B.; Jasperse, C. P. J. Am. Chem. Soc. 1999, 121, 7517; (e)
Mukund, P. S.; Jianguo, J. J. Am. Chem. Soc. 1996, 118,
3063; (f) Mukund, P. S.; Jianguo, J. J. Org. Chem. 1996, 61,
6090.
19. The use of an oxazolidine based ketene acetal proved difficult
because of problems encountered during the synthesis of the re-
quired tetrasubstituted ketene acetal substrate. With substituted
oxizolidinones and a substituent on the carbon alpha to the
imide, the yield of silyl enol ether formation was very low.
When no substituent was placed on the oxazolidinone, attempts
to generate the enolate led to deprotonation and decomposition
of the oxazolidinone.