Intramolecular anodic olefin coupling reactions: The use of allylsilane coupling partners with allylic alkoxy groups

Article abstract of DOI:10.1021/jo982280v

Intramolecular anodic olefin coupling reactions having an alkoxy substituent on the allylic carbon of an allylsilane moiety have been studied. These substrates were examined as part of an effort to determine the compatibility of the anodic olefin coupling reaction with the presence of very acid-sensitive functional groups and the construction of functionalized five-membered rings. In the initial experiment reported, an enol ether moiety was coupled to a trisubstituted allylsilane to afford a five-membered ring product without loss of the allylic alkoxy group. The reaction was stereoselective and could be used to generate a five-membered ring with three contiguous asymmetric centers. The stereochemical outcome of the reaction was best explained by a "pseudoequatorial" alkoxy group in the transition state; an argument that implied the reaction was under kinetic control. This suggestion was tested with the use of two electrolysis substrates that led to identical products through different transition states. The two substrates led to much different product ratios, proving that the reactions were not controlled by thermodynamics but rather governed by kinetic control. This observation was opposite to the conclusion reached with earlier vinylsilane-based substrates. Finally, the reactions were shown to be compatible with the presence of the alkoxy group even when challenged to form a fused bicyclic ring skeleton and a quaternary carbon. As in the initial case, the reaction to form a quaternary carbon was also highly stereoselective.

Full text of DOI:10.1021/jo982280v

J . Org. Chem. 1999, 64, 2805-2813
2805
In tr a m olecu la r An od ic Olefin Cou p lin g Rea ction s: Th e Use of
Allylsila n e Cou p lin g P a r tn er s w ith Allylic Alk oxy Gr ou p s
Dean A. Frey, S. Hari Krishna Reddy, and Kevin D. Moeller*
Department of Chemistry, Washington University, St. Louis, Missouri 63130
Received November 17, 1998
Intramolecular anodic olefin coupling reactions having an alkoxy substituent on the allylic carbon
of an allylsilane moiety have been studied. These substrates were examined as part of an effort to
determine the compatibility of the anodic olefin coupling reaction with the presence of very acid-
sensitive functional groups and the construction of functionalized five-membered rings. In the initial
experiment reported, an enol ether moiety was coupled to a trisubstituted allylsilane to afford a
five-membered ring product without loss of the allylic alkoxy group. The reaction was stereoselective
and could be used to generate a five-membered ring with three contiguous asymmetric centers.
The stereochemical outcome of the reaction was best explained by a “pseudoequatorial” alkoxy
group in the transition state; an argument that implied the reaction was under kinetic control.
This suggestion was tested with the use of two electrolysis substrates that led to identical products
through different transition states. The two substrates led to much different product ratios, proving
that the reactions were not controlled by thermodynamics but rather governed by kinetic control.
This observation was opposite to the conclusion reached with earlier vinylsilane-based substrates.
Finally, the reactions were shown to be compatible with the presence of the alkoxy group even
when challenged to form a fused bicyclic ring skeleton and a quaternary carbon. As in the initial
case, the reaction to form a quaternary carbon was also highly stereoselective.
In tr od u ction
Sch em e 1
Anodic electrochemistry can provide an excellent means
for studying the chemistry of reactive radical cation
1
intermediates. For example, we have been using elec-
trochemistry to generate radical cations from both alkyl
and silyl enol ethers.2 These intermediates can be
trapped in an intramolecular fashion with simple alkyl
olefins, enol ethers, styrenes, allylsilanes, vinylsilanes,
and electron-rich aromatic rings. A general reaction
scheme for the coupling of two electron-rich olefins is
illustrated in Scheme 1. Reactions of this type are of
interest because they are oxidative cyclization reactions
and, therefore, lead to the formation of new ring skeletons
without giving up the functionality used to initiate the
reaction. However, while the reactions have proven useful
for forming fused and bridged bicyclic rings and generat-
-6
ing quaternary carbons in simple model systems, their
potential for constructing complex, highly functionalized
molecules has not yet been explored.
(1) For reviews, see: (a) Yoshida, K. Electrooxidation in Organic
Chemistry: The Role of Cation Radicals as Synthetic Intermediates;
J ohn Wiley and Sons: New York, 1984; pp 136-151. (b) Torii, S.
Electroorganic Synthesis: Methods and Applications: Part IsOxidations;
VCH: Deerfield Beach, FL, 1985. (c) For intramolecular reactions
see: Moeller, K. D. Top. Curr. Chem. 1997, 185, 50.
With these things in mind, the syntheses of several
natural products have been undertaken. One such effort
is focusing on the synthesis of crinipellin B (3). The
crinipellins are a family of tetraquinane natural products
(2) For selected examples, see: (a) New, D. G.; Tesfai, Z.; Moeller,
7
that were isolated from the fungus Crinipellis stipitaria.
K. D. J . Org. Chem. 1996, 61, 1578. (b) Hudson, C. M.; Moeller, K. D.
J . Am. Chem. Soc. 1994, 59, 2381. (c) Tinao-Wooldridge, L. V.; Moeller,
K. D.; Hudson, C. M. J . Org. Chem. 1994, 59, 2381. (d) Moeller, K. D.;
Tinao, L. V. J . Am. Chem. Soc. 1992, 114, 1033. (e) Hudson, C. M.;
Marzabadi, M. R.; Moeller, K. D.; New, D. G. J . Am. Chem. Soc. 1991,
They have a unique ring skeleton that has made them a
8
popular synthetic target. At the heart of the angularly
1
13, 7372.
3) For an extensive list of references concerning chemically based
(7) Anke, T.; Heim, J .; Knoch, F.; Mocek, U.; Steffan, B.; Steglich,
W. Angew. Chem., Int. Ed. Engl. 1985, 24, 109. (b) Kupka, J .; Anke,
T.; Oberwinkler, F.; Schramm, G.; Steglich, W. J . Antibiot. 1979, 32,
130.
(
oxidative cyclization reactions, see ref 1c and ref 3 therein, as well as:
Snider, B. B. Chem. Rev. 1996, 96, 339.
(
4) For more recent oxidative cyclization reactions, see: (a) J ahn,
(8) For a total synthesis of crinipillen B, see: (a) Piers, E.; Renaud,
J . J . Org. Chem. 1993, 25, 11-13. For prior synthetic efforts, see: (b)
Curran, D. P.; Sisko, J . Yeske, P. E.; Liu, H. Pure Appl. Chem. 1993,
65, 1153. (c) Mehta, G.; Rao, K. S.; Reddy, M. S. J . Chem. Soc., Perkin
Trans. 1 1991, 693. (d) Schwartz, C. E.; Curran, D. P. J . Am. Chem.
Soc. 1990, 112, 9272. (e) Mehta, G.; Rao, K. S.; Reddy, M. S.
Tetrahedron Lett. 1988, 29, 5025. (f) Mehta, G.; Rao, K. S. J . Chem.
Soc., Chem. Commun. 1987, 1578.
U.; Hartmann, P. J . Chem. Soc., Chem. Commun. 1998 209. (b) Ryter,
K.; Livinghouse, T. J . Am. Chem. Soc. 1998, 120, 2658. (c) Schmittel,
M.; Burghart, A.; Malisch, W.; Reising, J .; S o¨ llner, R. J . Org. Chem.
1
998, 63, 396.
(
(
5) Reddy, S. H. K.; Moeller, K. D. Tetrahedron Lett. 1998, 39, 8027.
6) Frey, D. A.; Wu, N.; Moeller, K. D. Tetrahedron Lett. 1996, 37,
8
317.
1
0.1021/jo982280v CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/26/1999

2
806 J . Org. Chem., Vol. 64, No. 8, 1999
Frey et al.
Sch em e 3
Sch em e 4
F igu r e 1.
Sch em e 2
3
). The alcohol was then protected as the tert-butyldi-
methylsilyl ether. The yield of this protection step was
low due to the competitive elimination of the allylic
alkoxy group. When tert-butyldimethylsilyl triflate was
used as the reagent for the protection, only the corre-
sponding diene product was obtained.
The oxidation of substrate 9 (Scheme 4) at a reticulated
vitreous carbon anode11 using constant current electroly-
sis conditions, a platinum cathode, a lithium perchlorate
in 1:1 methanol/THF electrolyte solution, and 2,6-lutidine
as a proton scavenger led to the formation of an 83%
isolated yield of cyclized product.12 No evidence for a
product resulting from elimination of the allylic alkoxy
group was obtained. Interestingly, the cyclization reac-
tion generated only two of the four possible diastereo-
meric products. Evidence that the two diastereomers
were isomeric at the carbon bearing the acetal was gained
by hydrolysis of the acetals and epimerization of the
resulting aldehydes with DBU to afford a single product.
This suggestion was confirmed by 2D-NMR.
Because of the ease with which 10a and 10b could be
characterized by NMR, and the difficulty encountered
with their separation, the products obtained from the
oxidation of 9 were characterized as a mixture. To this
end, the complete connectivity for both 10a and 10b was
assigned with the use of a COSY experiment. The
stereochemistry for each isomer was then determined
with the use of a NOESY experiment. The spectra
obtained from these experiments are included in the
Supporting Information.
fused portion of this skeleton are three contiguous
quaternary centers. It is tempting to suggest the use of
an intramolecular anodic olefin coupling reaction for
constructing the central carbon in this sequence. For
example, in the retrosynthetic analysis for crinipellin B
outlined in Scheme 2, a sequential anodic olefin coupling
reaction-intramolecular aldol condensation strategy would
be used to assemble the angularly fused portion of the
ring skeleton. The olefin coupling reaction would be
initiated by the oxidation of an enol ether and terminated
with an allylsilane group (5 to 4). While such a proposed
cyclization appears reasonable, it requires the presence
of an oxygen on the allylic carbon of the allylsilane. Would
the fact that electrolysis reactions can be done under
neutral conditions allow an anodic olefin coupling reac-
tion to tolerate the presence of such a group or would
the electrolysis reaction simply lead to elimination of the
alkoxy group and formation of a diene side product
(Figure 1)?
To address this question, we began a study to examine
The key NOE cross-peaks used to assign the stereo-
chemistry of 10a and 10b are highlighted in Figure 2.
For the major isomer, 10a , the large volume integrals
the compatibility of intramolecular anodic olefin coupling
reactions with the presence of an allylic alkoxy group.9
1 7 1
for the interactions between H and H and H and the
In itia l Stu d ies
allylic methyl provided evidence that methine proton H
1
To begin this work, substrate 9 was synthesized by
adding the vinyl anion generated from propargyltri-
(10) Negishi, E. Pure Appl. Chem. 1981, 53.
(11) Reticulated virteous carbon anodes were purchased from The
Electrosynthesis Co., Inc., 72 Ward Road, Lancaster, NY 14086-9779.
methylsilane10 to the aldehyde derived from 7 (Scheme
(
12) The initial electrolysis conditions employed were identical to
(
9) For a preliminary account of this work, see: Frey D. A.; Marx,
those used in previous cyclizations utilizing allylsilane terminating
groups.
2
b,e
J . A.; Moeller, K. D. Electrochim. Acta 1997, 42, 1967.

Intramolecular Anodic Olefin Coupling Reactions
J . Org. Chem., Vol. 64, No. 8, 1999 2807
Sch em e 5
F igu r e 2. Key NOE interactions for 10a and 10b.
reactions, the initially generated radical cation would be
protected from methanol trapping providing greater time
for the desired cyclization to occur. However, in the case
of 8 the alcohol group was covalently attached to the
substrate, dragged into the double layer, and available
for side reactions involving the radical cation intermedi-
ates generated at the electrode surface. The net result
was that the desired carbon-carbon bond-forming reac-
tion did not proceed.
and the vinyl substituent were on the same face of the
molecule. This indicated that the relationship between
the ether and the vinyl substituent was trans. The
presence of a cross-peak between H
a signal between H and H , the absence of a signal
between H and H , and the fact that the volume integral
for the trans-relationship between H and H was small
relative to the larger volume integral for the cross-peak
between H and H all indicated that the relationship
7 9
and H , the lack of
2
9
3
7
1
2
2
3
Previous cyclic voltammetry studies involving intramo-
lecular anodic olefin coupling substrates lend support to
this argument. These studies showed that the potential
measured for a olefin coupling substrate was dependent
on the length of the chain between the two olefins being
coupled;2d the faster the cyclization reaction, the lower
the potential that was measured. This observation was
consistent with either a concerted reaction where the
cyclization occurred as the initial electron was being
removed or a stepwise reaction where the cyclization was
fast enough to drain the initially formed radical cation
away from the electrode surface. In either case, the cyclic
voltammetry evidence suggested that the cyclization
reaction occurred at or near the electrode surface and,
hence, in the double layer.
between the vinyl substituent and the dimethoxymethyl
group was cis.
This assignment was also supported by the relatively
small volume integral for the cross-peak between H
2
and
H
3
in the minor isomer 10b. This observation suggested
that in the minor isomer a trans relationship existed
between the vinyl group and the dimethoxymethyl sub-
stituent, a conclusion that was supported by large volume
integrals for the cross-peak between H
2 9
and H and
between H and H . In both cases, the ring methine
3
7
proton was on the same face as the neighboring substitu-
ent, indicating that the subtituents were trans to each
other. For 10b, the relationship between the alkoxy
substituent and the vinyl group was again trans. Evi-
dence for this was based upon the presence of a cross-
peak between H
cross-peak between H
1
and the allylic methyl group and a
and H , indicating that these two
1
3
P r obin g th e Qu estion of Ster eoch em istr ysA
methine protons were on the same side of the molecule.
All of these interactions combined to make it clear that
Closer Look
1
0b had all trans stereochemistry and that 10a and 10b
Although the formation of products with control over
relative stereochemistry was not important for the cri-
nipellin synthesis (in the synthesis, the angularly fused
tricyclic ring skeleton will be used to control the relative
stereochemistry of the centers), the possibility that anodic
olefin coupling reactions could be used to stereoselectively
generate cyclic systems with contiguous asymmetric
centers was very intriguing. The formation of products
differed only at the center bearing the dimethoxy acetal
substituent.
It is interesting to note that the oxidation of the alcohol
precursor to 9 (8) did not lead to a similar cyclization
reaction. Instead, a complex mixture of products was
formed. Clearly, the presence of the free hydroxyl group
interfered with the cyclization. This was surprising since
all of the previous intramolecular anodic olefin coupling
reactions utilized methanol as either the solvent or a
cosolvent. Why was the presence of alcohol in these
reactions not a problem? One potential explanation for
this apparent contradiction is that the cyclization reac-
tions happen in the double layer surrounding the elec-
trode surface.13 The double layer is highly ordered and
serves to lower the effective concentration of solvent near
the electrode surface. This protects the reactive inter-
mediates generated at the electrode surface from solvent
trapping and provides greater time for intramolecular
reactions.14 For intramolecular anodic olefin coupling
with a trans relationship between the ether and vinyl
1,3
substituents was attributed to the presence of an A
-
interaction in the transition state (11b) leading to the
cis product (Scheme 5). This explanation assumed that
the reactions were under kinetic control and that the
product outcome was governed by the difference in
transition-state energies. While this argument seemed
reasonable, previous cyclization reactions utilizing enol
ether and vinylsilane coupling partners were best ex-
plained by suggesting that the reactions were reversible
2b
and under thermodynamic control. One could invoke a
similar argument here and justify the formation of
products with a trans relationship between the alkoxy
group and the vinyl substituent based upon thermody-
namic grounds. This explanation was not favored because
of the major product having the vinyl group and
dimethoxymethyl substituent cis to each other, but it
could not be ruled out.
(13) For discussions of the double layer, see ref 1a p 13 and: Fry,
A. J . Synthetic Organic Electrochemistry, 2nd ed.; J ohn Wiley and
Sons: New York; 1989; p 37.
(14) For the discussion of a related example involving the dimer-
ization of acrylonitrile, see: Organic Electrochemistry: An Introduction
and a Guide, 2nd ed.; Baizer, M. M., Lund, H., Eds.; Marcel Dekker:
New York, 1983; p 394.

2
808 J . Org. Chem., Vol. 64, No. 8, 1999
Frey et al.
Sch em e 6
F igu r e 3.
Sch em e 7
F igu r e 4. Key NOE interactions for 14a and 14b.
Sch em e 8
For this reason, the disubstituted allylsilane substrates
1
2 and 13 were examined (Scheme 6). For the trans-
allylsilane substrate 12, neither the transition state
leading to the trans product 14 nor the transition state
1
,3
leading to the cis product 15 would have an A
-
1
,3
interaction. If the A -interaction did control the reaction,
then the selectivity observed in the initial cyclization
reaction would be lost. For the cis-allylsilane substrate
Sch em e 9
1
3, the transition state leading to the cis product would
1
,3
again be higher in energy due to the presence of the A
-
interaction. If the reaction were under kinetic control and
governed by the A1 -interaction, then this reaction would
be selective like the initial cyclization. However, if the
reactions were governed by thermodynamics, then both
the trans-allylsilane substrate 12 and the cis-allylsilane
substrate 13 would afford the same stereochemical
results since both substrates lead to the same products.
The trans-disubstituted allylsilane substrate 12 was
synthesized by substituting Dibal-H for the trimethyl-
aluminum reagent used in Scheme 3. All other aspects
of the synthesis were identical. Although the yield of the
overall process was very low (ca. 3% over the three steps),
enough of substrate 12 was obtained for exploring the
electrolysis reaction. The synthesis was not attempted a
second time because of the lack of selectivity observed
in the cyclization (Scheme 7). To this end, substrate 12
was oxidized at a reticulated vitreous carbon anode using
the conditions described above for the oxidation of 9. In
this case, a 54% unoptimized yield of the cyclized
products was obtained. No significant degree of selectivity
was observed, and the reaction led to a complex mixture
of diastereomeric products. The proton NMR of the
mixture was consistent with the presence of all four
possible diastereomers. The oxidation of 12 was not
optimized.
,3
allylsilane substrate 13 was again oxidized using the
conditions described above for 9 (Scheme 9). In this case,
four diastereomeric products were obtained in a 74%
overall yield. The two major products 14a and 14b were
formed in a 2:1 ratio. The ratio of the two major products
(14a plus 14b) to the two minor products (15a plus 15b)
was 8:1. The product ratios were determined by integra-
tion of the acetal methine region of the proton NMR
spectrum obtained for the crude reaction mixture. As
with 10a and 10b, the ease with which the products could
be characterized by 2D-NMR and the difficulty associated
with separating the isomers led to characterization of 14a
and 14b as the mixture. This was again accomplished
using a combination of COSY data to assign the con-
nectivity of the molecules and NOESY data to assess the
spatial arrangements of the substituents about the ring
(Figure 4).
The cis-disubstitued allylsilane was synthesized as
illustrated in Scheme 8. In this case, the substrate was
made from a propargyl anion addition to the aldehyde
derived from 7, a selective hydrogenation of the acetylene,
and then protection of the resulting alcohol. The cis-
As illustrated, the trans relationship between the
alkoxy and vinyl substituents in 14a was made by first
establishing the relationship between H and H and then
3 2

Intramolecular Anodic Olefin Coupling Reactions
J . Org. Chem., Vol. 64, No. 8, 1999 2809
by showing that H
. The trans relationship between H
established by the interaction between H
proton H , the interaction between H and H
small size of the interaction between H and H
1
was on the same face of the ring as
and H was
and the vinyl
, and the
relative
Sch em e 10
H
3
2
3
3
6
2
8
2
3
to the same interaction in isomer 14b, which was
approximately twice as large. All of these items pointed
to H
hence, the vinyl and dimethoxymethyl groups having a
trans relationship. The NOESY cross-peak between H
and H then established H as being on the same face of
the molecule as H and therefore trans to H . Isomer 14a
was the all-trans compound.
2 3
and H being on opposite faces of the ring and,
1
3
1
3
2
The stereochemistry of the second major product 14b
was determined in a similar fashion. In this case, a cross-
Sch em e 11
peak between H
interaction between H
6
and H
8
and a large cross-peak for the
(relative to that observed
2
and H
3
for 14a ) were both indicative of a cis relationship between
the vinyl group and the dimethoxymethyl substituent.
The trans relationship between H and H was estab-
1 2
lished by comparing the small volume integral for this
interaction with the larger volume integral for the known
cis interaction between H
grals suggested that H was further away from H
is H , a situation that was consistent with a trans
2
and H
3
. These volume inte-
1
2
than
3
relationship between the alkoxy and vinyl subsituents
and not consistent with a cyclized product having all
three groups cis.
formation of five-membered rings could tolerate the
presence of an allylic alkoxy substituent, it was not clear
what would happen with a substrate that led to a less
efficient cyclization. For example, what would happen in
the cyclization proposed for the crinipellin synthesis
when the reaction was challenged with the formation of
a quaternary carbon? Would such a steric barrier reduce
the rate of cyclization to a point where elimination of the
allylic alkoxy group began to compete with the desired
reaction? This was certainly a possibility. Related but
The two minor products from the cyclization reaction
could not be separated from the major isomers and
therefore could not be unambiguously assigned. The
NMR resonances for these two minor products were
buried beneath the resonances for the major products.
However, the resonances that could be observed were
consistent with what would be expected for the remaining
two possible diastereomers from the cyclization reaction
having a cis relationship between the alkoxy and vinyl
substituents.
1
5
much slower six-membered ring cyclization reactions
did lead to a small amount of an elimination product.
1
6
Despite our inability to assign the two minor products,
it was clear that the use of a cis-disubstituted allylsilane
substrate did reintroduce the stereoselectivity into the
reaction. Both major products were trans with respect
to the alkoxy and vinyl substituents, and the yield of the
reaction was good. No such propensity for forming trans
products was observed when the trans-disubstituted
allylsilane substrate was oxidized. Therefore, it was
concluded that the stereochemical outcome of the cycliza-
tion reaction was dependent on the stereochemistry of
the starting allylsilane and that reactions using allyl-
silane terminating groups were under “kinetic control”.
Our current working hypothesis for the intramolecular
To determine the feasibility of the cyclization proposed
for the crinipillen synthesis, substrate 18 was synthe-
sized (Scheme 10). The synthesis started from the known
cyclopentanone derivative 1617 and directly paralleled the
earlier construction of substrate 9. Once again, the final
protection step of the synthesis was complicated by
elimination of the allylic alkoxy group and formation of
the corresponding diene. While the elimination reaction
did lower the yield of the silation step, it did not interfere
with the electrolysis reaction. When 18 was oxidized
using the same conditions described earlier, a 75%
isolated yield of two cyclized products was obtained
(
Scheme 11). No evidence for the elimination reaction was
anodic olefin coupling reaction is that they cyclize to
1
generate a radical at the terminating end of the reaction.2b
present in the H NMR spectrum of the crude reaction
mixture. Both of the cyclized products possesed a cis-
fused ring juncture and a trans relationship between the
alkoxy and vinyl substituents. As in the earlier studies,
the stereochemistry of the cyclized products was assigned
with the use of a 2D-NOESY experiment. The assignment
of chemical shifts for the individual resonances was again
made using a 2D-COSY experiment. The key NOESY
cross-peaks used to make the assignments are high-
lighted in Figure 5.
Kinetic vs thermodynamic control is then determined by
the rate at which the molecule undergoes the second
oxidation reaction. In the case of the vinylsilane groups,
the cyclizations lead to an R-silyl radical that is not prone
to rapid oxidation. However, in the case of an allylsilane
terminating group the cyclizations lead to â-silyl radicals
that are much more likely to oxidize quickly. Hence,
kinetic control of the reactions was observed.
A Bicyclic Su bstr a te a n d th e F or m a tion of a
Qu a ter n a r y Ca r bon
(15) For comparisons of intramolecular anodic olefin coupling reac-
tions leading to five- and six-membered rings, see refs 2c and 2e.
While the above studies demonstrated that intramo-
lecular anodic olefin coupling reactions leading to the
(
16) Unpublished results with Mr. Nicholas Wu.
(17) Molander, G. A.; Harris, C. R. J . Org. Chem. 1998, 63, 812.

2
810 J . Org. Chem., Vol. 64, No. 8, 1999
Frey et al.
Sch em e 12
F igu r e 5. Key NOE interactions for 19a and 19b.
The assigment for compound 19a was based on estab-
lishing that the methine proton R to the alkoxy group
3
(H ), the vinyl substituent, and the dimethoxymethyl
substituent were all on the same face of the molecule.
This relationship was established by the presence of
cross-peaks between H
and the allylic methyl group, H
proton (H ), H and one of the vinyl protons, and H
the allylic methyl group. A cross-peak between the acetal
proton H and the bridgehead proton H established the
3
and one of the vinyl protons, H
and the dimethoxy acetal
and
3
3
9
9
9
9
5
position of the alkoxy group in the transition state
dictated product stereochemistry.
cis stereochemistry of the ring juncture. In this isomer,
no cross-peak appeared for an interaction between the
allylic methine H
For isomer 19b, the presence of a cross-peak between
the allylic proton (H ) and the acetal methine (H ) was
2 9
and the acetal methine proton H .
Con clu sion s
2
9
Intramolecular anodic olefin coupling reactions leading
to the formation of five-membered rings have proven
compatible with the presence of an allylic alkoxy group.
No elimination of the alkoxy group was observed. The
cyclization reactions utilizing either a trisubstituted or
cis-disubstituted allylsilane-terminating olefin were
stereoselective. The stereochemistry of the five-membered
ring products could be explained with the use of a
pseudochair transition state with the allylic alkoxy group
occupying a psuedoequatiorial position. These conclusions
remained consistent when the cyclization reaction was
used to generate a fused bicyclic ring skeleton and a
quaternary carbon.
A cyclization reaction utilizing a trans-disubstituted
allylsilane as the trapping olefin led to a mixture of
diastereomeric products. No stereoselectivity was ob-
served for this reaction, indicating that the stereochem-
istry of the products was dependent on the stereochem-
istry of the initial terminating olefin. This conclusion was
indicative of a cyclization reaction that was under kinetic
control.
used to establish the trans relationship between the
dimethoxymethyl and vinyl substituents. Cross-peaks
between both the methine proton R to the alkoxy group
3 3
(H ) and the allylic methyl group and H and one of the
vinyl protons established the relationship between the
alkoxy group and the vinyl substituent as trans. The cis
stereochemistry of the ring juncture was again estab-
lished using a cross-peak between the acetal methine
5
proton and the bridgehead proton H .
The exclusive formation of products 19a and 19b
indicated that the cyclization reaction was very stereo-
selective. This conclusion was based on the fact that the
starting material (18) used for the cyclization was a
mixture of diastereomers. In addition to the two enol
ether isomers, the substrate was a mixture with respect
to the stereochemistry of alkoxy group relative to that of
the carbon R to the enol ether. Ignoring the enol ether
stereoisomers, it would appear that each of the starting
material diastereomers led to the formation of a single
cyclized product. Consider the transition state represen-
tations shown in Scheme 12. Structures 18a and 18a ′
represent the two possible “chairlike” transition states
for one diastereomer (again ignoring the enol ether
isomers), and 18b and 18b′ represent the two possible
What is clear from these studies is that the intramo-
lecular anodic olefin coupling reaction proposed for the
synthesis of the crinipellin ring skeleton is feasible. Work
aimed at exploring the application of this reaction to such
a total synthesis effort is underway.
“chairlike” transition states for the second starting mate-
rial diastereomer. For the first diastereomer, transition
Exp er im en ta l Section 18
state 18a would be expected to be of lower energy than
1
8a ′ because of both the A1,3-interaction between the
5
-Meth oxy-(E,Z)-4-p en ten -1-ol (7). To a stirred suspen-
alkoxy group and the allylic methyl and the 1,3-diaxial
interaction between the alkoxy group and the ring in
sion of 13.71 g (40.0 mmol) of methoxymethyltriphenylphos-
phonium chloride in 40 mL of tetrahydrofuran at 0 °C was
added dropwise 30.8 mL (40.0 mmol) of 1.3 M sec-butyllithium
in cyclohexane. The dark red mixture was allowed to warm to
room temperature over 1 h. In another flask, a stirred solution
of 1.72 g (20.0 mmol) 4-hydroxybutyric acid γ-lactone in 20
mL of ether at -20 °C was treated dropwise with 11.0 mL of
1
8a ′. In practice, the reaction gave rise to 19a and none
of the isomeric product 20, an observation that was
consistent with this transition state analysis. For the
second diastereomer, transition state 18b would be
expected to be of lower energy than 18b′ because of the
1
.0 M diisobutylaluminum hydride in hexanes. The reaction
1
,3
psuedoaxial alkoxy group and corresponding A -interac-
tion in 18b′. In this case, the formation of 19b and none
of the isomeric 21 again supported the notion that the
(
18) For general experimental details, see: Wong, P. L.; Moeller,
K. D. J . Am. Chem. Soc. 1993, 115, 11434.

Intramolecular Anodic Olefin Coupling Reactions
J . Org. Chem., Vol. 64, No. 8, 1999 2811
mixture was allowed to warm to room temperature. After 2 h,
the reaction was cooled to 0 °C and quenched with 20 mL of
methanol. The white slurry was diluted with 100 mL of
pentane, filtered through a plug of silica gel, and eluted with
silica gel that was slurry-packed using a 1% triethylamine/
pentane solution. Elution with a 1% triethylamine/pentane
mobile phase afforded 0.167 g (46%) of the protected product
9. The spectral data for the mixture of olefin isomers were as
1
% triethylamine in pentane. The filtered solution was dried
follows: 1H NMR (CDCl
3
/300 MHz) δ 6.27 (d, J ) 12.6 Hz, 0.6
H), 5.86 (d, J ) 6.2 Hz, 0.4 H), 4.93 (d, J ) 8.7 Hz, 1 H), 4.74
(dt, J ) 12.6 Hz, J ) 7.3 Hz, 0.6 H), 4.32 (m, 1.4 H), 3.54
and 3.47 (2s, 3 H), 1.95 (m, 2 H), 1.61 (s, 3 H), 1.47 (m, 2 H);
0.87 (s, 9 H) 0.04 (m, 9 H); 13C NMR (CDCl
/75 MHz) δ 147.4,
over MgSO , filtered, and concentrated in vacuo. The concen-
4
trated crude lactol was diluted with 10 mL of tetrahydrofuran
and then slowly cannulated into the stirring 0 °C methoxy-
methyltriphenylphosphoranylidene solution generated above.
The reaction mixture was allowed to warm to room temper-
ature. After 24 h, the reaction mixture was diluted with of
ether and quenched with brine. The layers were separated,
and the aqueous layer was extracted with ether. The combined
d
t
3
146.5, 133.2, 127.9, 106.9, 103.2, 70.1, 69.8, 59.6, 56.1, 40.6,
39.5, 30.1, 29.8, 26.1, 26.0, 24.2, 24.0, 20.5, 19.4, 18.5; IR (neat/
-
1
NaCl) 3070, 3030, 1660 cm ; HRMS (EI) calcd for C19
40 2
H O -
+
Si
2
(M ) 356.2566, found 356.2553.
4
organic extracts were dried over MgSO , concentrated in vacuo,
1-(ter t-Bu tyld im eth lsilyloxy)-3-d im eth oxym eth yl-2-(1-
and chromatographed through silica gel that was slurry-
packed using 20% ether/pentane solution containing 1% tri-
ethylamine. Elution with 20% ether/pentane solution contain-
ing 1% triethylamine afforded 1.13 g (48.7% based on the
m eth yl-1-eth en yl)cyclop en ta n e (10a a n d 10b). To 0.169
g (0.47 mmol) of 5-(tert-butyldimethlsilyloxy)-1-methoxy-7-
methyl-8-(trimethylsilyl)-1(E,Z),6(E)-octadiene (9) in 7.5 mL
of tetrahydrofuran and 7.5 mL of methanol were added 0.33
mL (2.8 mmol) of 2,6-lutidine and 1.59 g (15.0 mmol) of lithium
perchlorate. The electrodes were inserted into the flask using
a two-hole septum equipped with a nitrogen inlet. A reticulated
vitreous carbon anode and a platinum wire cathode were used.
The reaction mixture was stirred until all of the electrolyte
was dissolved in solution and then degassed by sonication for
10 min. The reaction mixture was cooled to 0 °C, and current
was passed at a constant rate of 26.8 mA and until 100.4 C
(2.2 F/mol) of electricity had been passed. The reaction mixture
was diluted with brine and ether. The layers were separated,
and the aqueous layer was extracted with ether. The combined
lactone) of the alcohol enol ether (7). The spectral data for the
1
mixture of olefin isomers were as follows: H NMR (CDCl
3
/
3
0
)
3
7
3
1
00 MHz) δ 6.28 (d, J ) 12.6 Hz, 0.9 H), 5.91 (d, J ) 6.7 Hz,
.1 H), 4.69 (dt, J ) 12.6 Hz, J ) 7.4 Hz, 0.9 H), 4.32 (q, J
6.7 Hz, 0.1 H), 3.61 (t, J ) 6.5 Hz, 2 H), 3.55 and 3.47 (2s,
d
t
q
1
3
3
H), 1.98 (q, J ) 7.3 Hz, 2 H), 1.58 (m, 2H); C NMR (CDCl /
5 MHz) δ 147.4, 102.3, 62.2, 55.9, 33.5, 24.0; IR (neat/NaCl)
-
1
248, 1675, 1670 cm ; HRMS (EI) m/z calcd for C
6
H
12
O
2
+
16.0837 (M ), found 116.0839.
5
-Hyd r oxy-1-m eth oxy-7-m eth yl-8-(tr im eth ylsilyl)-1(E,-
Z),6(E)-octa d ien e (8). To a stirred suspension of 1.71 g (7.14
mmol) of zirconocene dichloride in 10 mL of methylene chloride
was added 7.1 mL (14.3 mmol) of 2.0 M trimethylaluminum
in toluene followed by 1.60 g (14.3 mmol) of propargyl
trimethylsilane at room temperature. After 16 h, the reaction
mixture was cooled to -78 °C, and 4.3 mL (10.7 mmol) of a
4
organic extracts were dried over MgSO , concentrated in vacuo,
and chromatographed through silica gel that was slurry-
packed using a 1% triethylamine/pentane solution. Elution
with a 1% triethylamine/pentane solution afforded 0.111 g
(84%) of desired products 10a and 10b. The spectral data for
the 60:40 ratio of trans-trans and trans-cis isomers of the
2
.5M n-butyllithium solution in hexanes was added dropwise.
In a second flask, 0.81 g (7.1 mmol) of 5-methoxy-(E,Z)-4-
penten-1-ol (6) was diluted with 10 mL of methylene chloride
treated with 3.50 g of crushed 4A molecular sieves, 1.25 g (10.7
mmol) of 4-methylmorpholine N-oxide, and 0.125 g (0.357
mmol) of tetrapropylammonium perruthenate at 0 °C and
allowed to warm to room temperature. After 30 min, the black
reaction mixture was filtered through a plug of silica gel with
a 1% triethylamine/ether solution and concentrated in vacuo.
The crude aldehyde was diluted with 10 mL of methylene
chloride, slowly cannulated into the stirring -78 °C vinyl anion
solution prepared above, and allowed to warm to room tem-
perature. After 18 h, the reaction mixture was cooled to 0 °C,
quenched slowly with Rochelle’s solution, and diluted with
ether. The layers were separated, and the aqueous layer was
extracted with ether. The combined organic extracts were dried
cyclized products were as follows: 1H NMR (CDCl
3
/300 MHz)
δ 4.85 (s, 0.6 H), 4.76 (s, 0.4 H), 4.72 (s, 0.4 H), 4.59 (s, 0.6 H),
4.17 (m, 1 H), 4.13 (m, 0.6 H), 3.90 (q, J ) 6.0 Hz, 0.4 H),
3.35, 3.32, 3.27, 3.26 (4s, 6 H), 2.60 (p, J ) 8.5 Hz, 0.6 H),
2.41 (dd, J ) 4.1, 8.3 Hz, 0.6 H), 2.25 (m, 0.4 H), 2.16 (p, J )
d
8.2 Hz, 0.4 H), 2.04 (q, J ) 3.5 Hz, 0.4 H), 1.80 (m, 1 H), 1.74
and 1.71 (2s, 3 H), 1.65 (m, 0.8), 1.50 (m, 1.6 H), 0.85 (s, 9 H),
0.04 (s, 6 H); 13C NMR (CDCl
/75 MHz) δ 145.9, 144.8, 111.5,
3
108.1, 78.2, 78.1, 78.0, 77.9, 76.4, 58.1, 56.6, 54.1, 53.8, 53.5,
52.3, 43.5, 43.4, 42.8, 42.7, 34.5, 34.4, 34.0, 25.9, 25.8, 25.7,
25.0, 24.6, 23.7, 23.6, 20.9, 18.1; IR (neat/NaCl) 3080, 1647
-
1
+
cm ; HRMS (EI) calcd for C17
H
31
O
2
3
Si (M - OCH ) 283.2093,
found 283.2095. COSY and NOESY data for these products
are included in the Supporting Information.
5-(ter t-Bu tyld im eth lsilyloxy)-1-m eth oxy-8-(tr im eth yl-
silyl)-1(E,Z),6(E)-octa d ien e (12). To a stirred suspension of
2.78 g (11.6 mmol) of zirconocene dichloride in 20 mL of
methylene chloride was added 23.2 mL (23.2 mmol) of 1.0 M
diisobutylaluminum hydride in hexanes followed by 2.60 g
(11.6 mmol) of propargyl trimethylsilane at room temperature.
After 16 h, the reaction mixture was cooled to -78 °C, and
12.4 mL (17.4 mmol) of 1.4 M methyllithium in hexanes was
added dropwise. In another flask, 1.35 g (11.6 mmol) of
5-methoxy-(E,Z)-4-pentene-1-ol (7) was diluted with 20 mL of
methylene chloride treated with 5.80 g of crushed 4A molecular
sieves, and 2.05 g (17.3 mmol) of 4-methylmorpholine N-oxide
and 0.204 g (0.58 mmol) of tetrapropylammonium were per-
ruthenate added at 0 °C. The mixture was allowed to warm
to room temperature. After 30 min, the black reaction mixture
was filtered through a plug of silica gel with a 1% triethyl-
amine/ether solution and concentrated in vacuo. The crude
aldehyde was diluted with 10 mL of methylene chloride and
slowly cannulated into the stirring -78 °C vinyl anion solution
prepared above and allowed to warm to room temperature.
After 18 h, the reaction mixture was cooled to 0 °C, quenched
slowly with Rochelle’s solution, and diluted with ether. The
layers were separated, and the aqueous layer was extracted
with ether. The combined organic extracts were dried over
4
over MgSO , concentrated in vacuo, and chromatographed
through silica gel that was slurry-packed using 5% ether/
hexane solution containing 1% triethylamine. Elution with 5%
ether/hexane solution containing 1% triethylamine afforded
0
.933 g (54%) of desired product 8. The spectral data for the
1
product were as follows: H NMR (CDCl
J ) 12.6 Hz, 1 H), 4.97 (d, J ) 8.8 Hz, 1 H), 4.72 (dt, J
Hz, J ) 7.2 Hz, 1 H), 4.34 (q, 1 H), 3.48 (s, 3 H), 1.95 (m, 2
H), 1.65 (s, 3 H), 1.60-1.40 (m, 4 H); 1.22 (m, 1 H), 0.04 (s, 9
H); C NMR 147.2, 137.3, 125.9, 102.6, 68.3, 55.9, 38.8, 30.2,
2
3
/300 MHz) δ 6.27 (d,
d
) 12.6
t
1
3
-
1
3.9, 19.2; IR (neat/NaCl) 3410, 3060, 3045, 1673, 1655 cm ;
+
HRMS (IE) m/z (rel intensity) calcd for C13
42.1702, found 242.1699.
-(ter t-Bu t yld im et h lsilyloxy)-1-m et h oxy-7-m et h yl-8-
tr im eth ylsilyl)-1(E,Z),6(E)-octa d ien e (9). To a stirred
26 2
H O Si (M )
2
5
(
solution of 0.248 g (1.02 mmol) of 5-hydroxy-1-methoxy-7-
methyl-8-(trimethylsilyl)-1(E,Z),6(E)-octadiene (7) in 5 mL of
tetrahydrofuran was added 0.013 g (0.102 mmol) of 4-(di-
methylamino)pyridine, 0.18 mL (1.2 mmol) of 1,8-diazobicyclo-
[5.4.0]undec-7-ene, and a solution of 0.170 g (1.13 mmol) of
tert-butyldimethlsilyl chloride in 3 mL of tetrahydrofuran.
After 14 h, the reaction was diluted with ether and brine. The
layers were separated, and the aqueous layer was extracted
with ether. The combined organic extracts were dried over
MgSO
4
, concentrated in vacuo, and chromatographed through
4
MgSO , concentrated in vacuo, and chromatographed through
silica gel that was slurry-packed using 5% ether/hexane

2
812 J . Org. Chem., Vol. 64, No. 8, 1999
Frey et al.
solution containing 1% triethylamine. Elution with 5% ether/
hexane solution containing 1% triethylamine afforded a mix-
ture of products that was carried on through a protection step.
To a stirred mixture of approximately 0.273 g (1.20 mmol) of
through silica gel that was slurry-packed using 25% ether/
hexane solution containing 1% triethylamine. Elution with 5%
ether/hexane solution containing 1% triethylamine afforded
0.1754 g (98%) of the cis olefin. The spectral data for the
mixture of enol ether isomers are included in the Supporting
Information.
5
-hydroxy-1-methoxy-8-(trimethylsilyl)-1(E,Z),6(E)-octadiene
in 5 mL of tetrahydrofuran was added 0.014 g (0.12 mmol) of
-(dimethylamino)pyridine, 0.18 mL (1.2 mmol) of 1,8-diazo-
4
5
-(ter t-Bu tyld im eth lsilyloxy)-1-m eth oxy-8-(tr im eth yl-
bicyclo[5.4.0]undec-7-ene, and a solution of 0.180 g (1.20 mmol)
of tert-butyldimethlsilyl chloride in 3 mL of tetrahydrofuran.
After 14 h, the reaction was diluted with ether and brine. The
layers were separated, and the aqueous layer was extracted
with ether. The combined organic extracts were dried over
silyl)-1(E,Z),6(Z)-octa d ien e (13). To a stirred solution of
0
1
0
.175 g (0.77 mmol) of 5-hydroxy-1-methoxy-8-(trimethylsilyl)-
(E,Z),6(Z)-octadiene in 2 mL of tetrahydrofuran was added
.009 g (0.08 mmol) of 4-(dimethylamino)pyridine, 0.13 mL
(0.85 mmol) of 1,8-diazobicyclo[5.4.0]undec-7-ene, and a solu-
4
MgSO , concentrated in vacuo, and chromatographed through
tion of 0.138 g (0.92 mmol) of tert-butyldimethlsilyl chloride
in 3 mL of tetrahydrofuran. After 14 h, the reaction was
diluted with ether and brine. The layers were separated, and
the aqueous layer was extracted with ether. The combined
organic extracts were dried over MgSO , concentrated in vacuo,
4
and chromatographed through silica gel that was slurry-
packed using 1% triethylamine/pentane solution. Elution with
silica gel that was slurry-packed using a 1% triethylamine/
pentane solution. Elution with a 1% triethylamine/pentane
solution afforded 0.127 g (3.1%) of the protected product 12
over the two steps. The spectral data for the mixture of enol
ether isomers are included in the Supporting Information.
Th e Cycliza tion of th e tr a n s-Allylsila n e Su bstr a te 12:
1
-(ter t-Bu t yld im et h lsilyloxy)-3-d im et h oxym et h yl-2-(1-
1
% triethylamine/pentane solution afforded 0.127 g (48%) of
eth en yl)cyclop en ta n e. To 0.052 g (0.15 mmol) of 5-(tert-
butyldimethlsilyloxy)-1-methoxy-8-(trimethylsilyl)-1(E,Z),6(E)-
octadiene (12) in 2.5 mL of tetrahydrofuran and 2.5 mL of
methanol were added 0.11 mL (0.91 mmol) of 2,6-lutidine and
the protected product 13. The spectral data for the mixture of
enol ether isomers are included in the Supporting Information.
1-(ter t-Bu tyld im eth lsilyloxy)-3-d im eth oxym eth yl-2-(1-
eth en yl)cyclop en ta n e (14a a n d 14b). To 0.100 g (0.29
mmol) of 5-(tert-butyldimethylsilyloxy)-1-methoxy-8-(trimeth-
ylsilyl)-1(E,Z),6(Z)-octadiene (13) in 2.5 mL of tetrahydrofuran
and 2.5 mL of methanol were added 0.20 mL (1.7 mmol) of
2,6-lutidine and 0.531 g (5.0 mmol) of lithium perchlorate. The
electrodes were inserted into the flask using a two-hole septum
equipped with a nitrogen inlet. A reticulated vitreous carbon
anode and a platinum wire cathode were used. The reaction
mixture was stirred until the entire electrolyte was dissolved
in solution and then degassed by sonication for 10 min. The
reaction mixture was cooled to 0 °C, and current was passed
at a constant rate of 26.8 mA and until 62.0 C (2.2 F/mol) of
electricity had been passed. The reaction mixture was diluted
with brine and ether. The layers were separated, and the
aqueous layer was extracted with ether. The combined organic
0
.531 g (5.00 mmol) of lithium perchlorate. The electrodes were
inserted into the flask using a two-hole septum equipped with
a nitrogen inlet. A reticulated vitreous carbon anode and a
platinum wire cathode were used. The reaction mixture was
stirred until all of the electrolyte was dissolved in solution and
then degassed by sonication for 10 min. The reaction mixture
was cooled to 0 °C, and current was passed at a constant rate
of 26.8 mA and until 32.2 C (2.2 F/mol) of electricity had been
passed. The reaction mixture was diluted with brine and ether.
The layers were separated and the aqueous layer was ex-
tracted with ether. The combined organic extracts were dried
4
over MgSO , concentrated in vacuo, and chromatographed
through silica gel that was slurry-packed using a 1% triethyl-
amine/pentane solution. Elution with a 1% triethylamine/
pentane solution afforded 0.024 g (54%) of the cyclized product.
The spectral data for the mixture of enol ether isomers are
included in the Supporting Information.
4
extracts were dried over MgSO , concentrated in vacuo, and
chromatographed through silica gel that was slurry-packed
using a 1% triethylamine/pentane solution. Elution with a 1%
triethylamine/pentane solution afforded 0.0643 g (74%) of the
cyclized products. Four isomeric products were observed in the
proton NMR of the material. This spectrum is included in the
Supporting Information. The two major products (14a and
5
-H yd r oxy-1-m et h oxy-8-(t r im et h ylsilyl)-1(E,Z)-en -6-
octyn e. To a solution of 1.00 g (8.92 mmol) of propargyl
trimethylsilane in 10 mL of tetrahydrofuran was added 6.86
mL of 1.3 M sec-butyllithium in cyclohexane dropwise at 0 °C.
In another flask, 0.517 g (4.46 mmol) of 5-methoxy-(E,Z)-4-
penten-1-ol (7) was diluted with 10 mL of methylene chloride
treated with 2.50 g of crushed 4A molecular sieves, 0.78 g (6.7
mmol) of 4-methylmorpholine N-oxide, and 0.078 g (0.22 mmol)
of tetrapropylammonium perruthenate at 0 °C and allowed to
warm to room temperature. After 30 min, the black reaction
mixture was filtered through a plug of silica gel with a 1%
triethylamine/ether solution and concentrated in vacuo. The
crude aldehyde was diluted with 10 mL of methylene chloride,
slowly cannulated into the stirring 0 °C propargyl anion
solution prepared above, and allowed to warm to room tem-
perature. After 18 h, the reaction mixture was cooled to 0 °C,
quenched slowly with Rochelle’s solution, and diluted with
ether. The layers were separated, and the aqueous layer was
extracted with ether. The combined organic extracts were dried
1
4b) were generated in a 2:1 ratio. In addition, two minor
isomers were obtained. The ratio of the two major isomers to
the two minor isomer was 8:1. The ratio of products were
determined by integration of the acetal region of the spectrum.
The spectral data for the mixture isomers are included in the
Supporting Information.
1
-(Meth oxy-(E,Z)-m eth ylid en e)-2-(2′-h yd r oxyeth yl)cy-
clop en ta n e (17). To a stirred suspension of 1.309 g (3.82
mmol) of methoxymethyltriphenylphosphonium chloride in 5.0
mL of tetrahydrofuran at 0 °C was added 7.6 mL (3.8 mmol)
of 0.5 M potassium hexamethyldisilane in toluene dropwise.
After 1 h, the dark red mixture was treated with a solution of
0
.466 g (1.91 mmol) of 2-(2′-tert-butyldimethylslilyloxyethy)-
cyclopentanone (16) in 5 mL of tetrahydrofuran. After 48 h,
the reaction was diluted with ether and brine. The layers were
separated, and the aqueous layer was extracted with ether.
4
over MgSO , concentrated in vacuo, and chromatographed
through silica gel that was slurry-packed using 25% ether/
hexane solution containing 1% triethylamine. Elution with 5%
ether/hexane solution containing 1% triethylamine afforded
4
The combined organic extracts were dried over MgSO and
concentrated in vacuo. The crude product was diluted with 10
mL of tetrahydrofuran and treated with 0.75 mL (0.75 mmol)
of 1 M tetrabutylammonium fluoride. After 18 h, the reaction
was diluted with ether and brine. The layers were separated,
and the aqueous layer was extracted with ether. The combined
0
.635 g (63%) of product. The spectral data for the mixture of
enol ether isomers are included in the Supporting Information.
-Hyd r oxy-1-m eth oxy-8-(tr im eth ylsilyl)-1(E,Z),6(Z)-oc-
ta d ien e. To 0.176 g (0.78 mmol) of 5-hydroxy-1-methoxy-8-
trimethylsilyl)-1(E,Z)-en-6-octyne in 10 mL of methanol was
5
(
added 0.018 mL (0.16 mmol) of quinoline and 0.165 g (0.008
mmol) of 5% palladium on calcium carbonate. The reaction
flask was equipped with a hydrogen balloon, and the mixture
was stirred under a hydrogen atmosphere for 48 h. The
reaction mixture was filtered through a plug of Celite. The
filtrate was concentrated in vacuo and chromatographed
4
organic extracts were dried over MgSO , concentrated in vacuo,
and chromatographed through silica gel that was slurry-
packed using 5% ether/hexane solution containing 1% triethyl-
amine to afford 0.191 g (1.22 mmol) 64% of the desired product.
The spectral data for products are included in the Supporting
Information.

Intramolecular Anodic Olefin Coupling Reactions
-(Meth oxy-(E,Z)-m eth ylid en e)-2-(2′-h yd r oxy-4-m eth -
J . Org. Chem., Vol. 64, No. 8, 1999 2813
1
The spectral data for the mixture of diastereomers are included
in the Supporting Information.
yl-5-tr im eth ylsila n e-3-p en ten e)cyclop en ta n e. To a stirred
suspension of 1.067 g (4.46 mmol) of zirconocene dichloride in
3-(ter t-Bu tyld im eth lsilyloxy)-1-d im eth oxym eth yl-2-(1-
m eth yl-1-eth en yl)bicyclo[3.3]octa n e (19a a n d 19b). To
0.217 g (0.618 mmol) of 1-(methoxy-(E,Z)-methylidene)-2-(2′-
tert-butyldimethylsilyloxy-4-methyl-5-trimethylsilane-3-pen-
tene)cyclopentane (18) in 5.0 mL of tetrahydrofuran and 5.0
mL of methanol were added 0.43 mL (3.7 mmol) of 2,6-lutidine
and 1.063 g (10.0 mmol) of lithium perchlorate. The electrodes
were inserted into the flask using a two-hole septum equipped
with a nitrogen inlet. A reticulated vitreous carbon anode and
a platinum wire cathode were used. The reaction mixture was
stirred until all of the electrolyte was dissolved in solution and
then degassed by sonication for 10 min. The reaction mixture
was cooled to 0 °C, and current was passed at a constant rate
of 26.8 mA and until 131.1 C (2.2 F) of electricity had been
passed. The reaction mixture was diluted with brine and ether.
The layers were separated, and the aqueous layer was
extracted with ether. The combined organic extracts were dried
5
2
mL of methylene chloride was added 4.50 mL (8.9 mmol) of
.0 M trimethylaluminum in toluene followed by 1.00 g (8.92
mmol) of propargyl trimethylsilane at room temperature. After
6 h, the reaction mixture was cooled to -78 °C, and 4.3 mL
6.4 mmol) of 1.4 M methyllithium in diethyl ether was added
dropwise. In another flask, 0.696 g (4.46 mmol) of 1-(methoxy-
E,Z)-methylidene)-2-(2′-hydroxyethyl)cyclopentane (17) was
1
(
(
diluted with 5 mL of methylene chloride, treated with 2.50 g
of crushed 4A molecular sieves, 0.783 g (6.69 mmol) of
4
-methylmorpholine N-oxide, and 0.078 g (0.22 mmol) of
tetrapropylammonium perruthenate at 0 °C, and allowed to
warm to room temperature. After 30 min, the black reaction
mixture was filtered through a plug of silica gel with a 1%
triethylamine/ether solution and concentrated in vacuo. The
crude aldehyde was diluted with 10 mL of methylene chloride,
slowly cannulated into the stirring -78 °C vinyl anion solution
prepared above, and allowed to warm to room temperature.
After 20 h, the reaction mixture was cooled to 0 °C, quenched
slowly with Rochelle’s solution, and diluted with ether. The
layers were separated, and the aqueous layer was extracted
with ether. The combined organic extracts were dried over
over MgSO , concentrated in vacuo, and chromatographed
4
through silica gel that was slurry-packed using a 1% triethyl-
amine/pentane solution. Elution with a 1% triethylamine/
pentane solution afforded 0.163 g (75%) of desired product.
The two isomers were separated and characterized. The COSY
and NOESY data for 19a and 19b are included in the
Supporting Information as are the full characterization data.
4
MgSO , concentrated in vacuo, and chromatographed through
silica gel that was slurry-packed using 5% ether/hexane
solution containing 1% triethylamine. Elution with 5% ether/
hexane solution containing 1% triethylamine afforded 0.579
g (2.05 mmol) 46% of desired alcohol product. The spectral data
for the mixture of diastereomers are included in the Support-
ing Information.
Ack n ow led gm en t. We thank the National Science
Foundation (CHE-9023698) for their generous support
of this work. We also gratefully acknowledge the Wash-
ington University High Resolution NMR facility, par-
tially supported by NIH grants RR02004, RR05018, and
RR07155, and the Washington University Mass Spec-
trometry Resource Center, partially supported by NI-
HRR00954, for their assistance.
1
-(Meth oxy-(E,Z)-m eth ylid en e)-2-(2′-ter t-bu tyld im eth -
ylsilyloxy-4-m et h yl-5-t r im et h ylsila n e-3-p en t en e)cyclo-
p en ta n e (18). To a stirred solution of 0.104 g (0.367 mmol) of
1
-(methoxy-(E,Z)-methylidene)-2-(2′-hydroxy-4-methyl-5-tri-
methylsilane-3-pentene)cyclopentane in 2 mL of tetrahydro-
furan was added 0.13 mL (0.85 mmol) of 1,8-diazo-bicyclo-
[
5.4.0]undec-7-ene followed by a solution of 0.138 g (0.92 mmol)
Su p p or tin g In for m a tion Ava ila ble: The COSY and
NOESY spectral data are included for the cyclized products
10, 14, and 19) along with tables highlighting the key NOESY
interactions. In addition, the raw proton and carbon NMR data
are included for all new compounds, as well as the full
characterization data for compounds 12-19. This material is
available free of charge via the Internet at http://pubs.acs.org.
of tert-butyldimethlsilyl chloride in 3 mL of tetrahydrofuran.
After 14 h, the reaction was diluted with ether and brine. The
layers were separated, and the aqueous layer was extracted
with ether. The combined organic extracts were dried over
(
4
MgSO , concentrated in vacuo, and chromatographed through
silica gel that was slurry-packed using a 1% triethylamine/
pentane solution. Elution with a 1% triethylamine/pentane
solution afforded 0.127 g (49%) of the protected product 18.
J O982280V