Synthesis of the Spatane Nucleus (cis-anti-cis-tricyclo[5.3.0.02,6]decane) using the Pauson-Khand Reaction with a Remarkable Reversal in Regioselectivity

Article abstract of DOI:10.1021/jo9708194

The Pauson-Khand reaction of cyclobutenes 3-5 with a variety of acetylenes yielded cis-anti cis-tricyclo[5.3.0.0^2,6]decanes 9-17. The unwanted regioisomers 9a and 10a were the sole products using acetylene, but there was a remarkable reversal in orientation of the cyclobutene component yielding the desired regioisomer 13b upon using (trimethylsilyl)acetylene. The importance of the allylic methyl group in the cyclobutenes in directing the regiochemical outcome was substantiated by the lack of selectivity in Pauson-Khand reactions of desmethylcyclobutene 5 with acetylene and (trimethylsilyl)acetylene. The relative unimportance of electronic control of regiochemistry was concluded from the consistent ratio of Pauson-Khand reaction products from norbornenone 22 with various acetylenes. A hypothesis rationalizing the regiochemical outcome was based on steric interactions of the allylic methyl group from the cyclobutene component with either the smaller acetylene substituent or the CO ligands on the cobalt. This steric interaction was further hypothesized to be influenced by the larger acetylene substituent sterically crowding the CO ligands on the cobalt.

Full text of DOI:10.1021/jo9708194

J . Org. Chem. 1998, 63, 1379-1389
1379
Syn th esis of th e Sp a ta n e Nu cleu s
(cis-a n ti-cis-tr icyclo[5.3.0.02,6]d eca n e) u sin g th e P a u son -Kh a n d
Rea ction w ith a Rem a r k a ble Rever sa l in Regioselectivity
Bruce A. Kowalczyk,*,† Timothy C. Smith,‡ and William G. Dauben§
Department of Chemistry, University of California, Berkeley, California 94720
Received May 7, 1997
The Pauson-Khand reaction of cyclobutenes 3-5 with a variety of acetylenes yielded cis-anti cis-
tricyclo[5.3.0.02,6]decanes 9-17. The unwanted regioisomers 9a and 10a were the sole products
using acetylene, but there was a remarkable reversal in orientation of the cyclobutene component
yielding the desired regioisomer 13b upon using (trimethylsilyl)acetylene. The importance of the
allylic methyl group in the cyclobutenes in directing the regiochemical outcome was substantiated
by the lack of selectivity in Pauson-Khand reactions of desmethylcyclobutene 5 with acetylene
and (trimethylsilyl)acetylene. The relative unimportance of electronic control of regiochemistry
was concluded from the consistent ratio of Pauson-Khand reaction products from norbornenone
22 with various acetylenes. A hypothesis rationalizing the regiochemical outcome was based on
steric interactions of the allylic methyl group from the cyclobutene component with either the smaller
acetylene substituent or the CO ligands on the cobalt. This steric interaction was further
hypothesized to be influenced by the larger acetylene substituent sterically crowding the CO ligands
on the cobalt.
Sch em e 1
In tr od u ction
The Pauson-Khand reaction, that was first reported
in 1973, has received much attention because of the
utility of this cyclopentenone formation.1 A great variety
of acetylenes and alkenes have been coupled to make
synthetically useful cyclopentenones. In many cases
there is good regiochemical and stereochemical control.
The spatane diterpenes, that are of biological interest,
have been synthetically approached by several different
methods.2 One of the main challenges in synthesizing
spatanes has been efficiently making an appropriately
functionalized cis-anti-cis-tricyclo[5.3.0.02,6]decane nucleus.
In our total synthesis of spatadiene (1) and formal
synthesis of spatol (2), we employed the retrosynthetic
and stereoselectively, on an appropriately functionalized
cyclobutene. In this paper, we report our complete
synthetic efforts at making the cis-anti-cis-tricyclo-
[5.3.0.02,6]decane nucleus using Pauson-Khand meth-
odology.
Resu lts a n d Discu ssion
Cyclobu ten e Syn th eses. Three different cyclobutenes
(3-5) were synthesized as substrates for study of the
Pauson-Khand reaction. Cyclobutene 4 was synthesized
by the route outlined in Scheme 2. The first step was a
photo [2 + 2] reaction between 3-methyl-2-cyclopentenone
(6) and 1,2-dichloroethylene to produce cyclobutane 7.
This photoreaction was done in cyclohexane with a large
excess of 1,2-dichloroethylene to minimize dimerization
of 3-methyl-2-cyclopentenone which was added in two
analysis outlined in Scheme 1.3 The most critical
step in the process was the use of a Pauson-Khand
reaction to annulate a cyclopentenone, regioselectively
† Current address: Roche Bioscience, 3401 Hillview Ave., Palo Alto,
CA 94304.
‡ Current address: Southwestern Oklahoma State University,
Department of Chemistry, 100 Campus Drive, Weatherford, OK 73096.
§ Deceased J anuary 1, 1997.
(1) Reviews: (a) Schore, N. E. Org. React. 1991, 40, 1. (b) Shore, N.
E. In Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Eds.;
Pergamon: Oxford, 1991; Vol. 5, pp 1037-1064. (c) Pauson, P. L.
Tetrahedron 1985, 41, 5855. (d) Pauson, P. L. Organometallics in
Organic Synthesis; de Meijere, A., tom Dieck, H., Eds.; Springer-
Verlag: Berlin, 1987; p 233. (e) Schore, N. E. Chem. Rev. 1988, 88,
1081.
(2) (a) Review: Salomon, R. G. Strategies and Tactics in Organic
Synthesis; Academic Press, Inc.: New York, 1991; Vol. 3, p 381. Recent
spatane synthesis work: (b) Murthi, K. K.; Salomon, R. G. Tetrahedron
Lett. 1994, 35, 517. (c) Miesch, M.; Cotte, A.; Franck-Neumann, M.
Tetrahedron Lett. 1994, 35, 7031.
(3) Dauben, W. G.; Kowalczyk, B. A. Tetrahedron Lett. 1990, 31,
635.
S0022-3263(97)00819-0 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/12/1998

1380 J . Org. Chem., Vol. 63, No. 5, 1998
Kowalczyk et al.
Sch em e 2a
a
(a) hν, 1,2-dichloroethylene, 89%; (b) ethylene glycol, PPTs,
F igu r e 1. Reaction mechanism key step.
96%; (c) sodium naphthalenide, 77%.
Alkenes 3-5 were all good substrates for the cyclopen-
tenone annulations as anticipated since cyclobutenes are
generally good substrates for the Pauson-Khand reac-
tion.6 The orientation of the acetylene component in the
adducts of table entries 3-5, 7, and 9 followed the trend
typically observed, i.e., with the larger substituent on the
initial acetylene R to the carbonyl group in the newly
formed cyclopentenone.1
Ta ble 1. Op tim iza tion of P a u son -Kh a n d Rea ction
To understand the orientation of the alkene (cy-
clobutene) component in the adducts, the postulated
mechanism must be considered.1a,b,e,6b,7 Thus, the initial
steps in the Pauson-Khand reaction are complexation
of the acetylene component with dicobalt octacarbonyl,
followed by complexation of the alkene to one cobalt atom
by a reversible dissociative process involving the loss of
a CO ligand. A key step in the mechanism is the
subsequent irreversible insertion of the cobalt-complexed
face of the alkene π bond into one of the formal cobalt-
carbon bonds of the alkyne complex (see Figure 1). This
step is believed to be both rate and product determining.
The product is formed by subsequent CO insertion,
reductive elimination of one cobalt, and decomplexation
of the second cobalt.
run
solvent
temp (°C) time (h) concn (M) yield (%)
1
2
heptane
heptane
DME
heptane
heptane
heptane
65-70
65-70
65-70
65-70
55-60
65-70
8
17
15
8
8
8
0.15
0.14
0.16
0.14
0.15
0.031
50
50
50
53
59
73
3
4a
5
6
a
n-Bu3PO (1.2 equiv).
portions to keep its relative concentration low, helping
to further reduce dimerization. A Pyrex filter was used
to cut off light of low wavelengths that would be destruc-
tive. The ketone of cyclobutane 7 was protected as a
ketal to give 8. The desired cyclobutene 4 was made by
reduction of vicinal dichloride 8 using sodium naphtha-
lenide.4 Cyclobutene 5 was made from 2-cyclopentenone
following a similar route to that for cyclobutene 4.
Cyclobutene 3 was made by a L-Selectride reduction of
ketone 7 followed by reductive elimination of the vicinal
dichloride using sodium naphthalenide. The cyclobutenol
product was mainly 3 (85%) along with a minor amount
of 3a (15%).
Two closely related previous examples of Pauson-
Khand reactions of unsymmetrically substituted cy-
clobutenes are the 18 to 19 (eq 1)6a and 20 to 21 (eq 2)6b,c
conversions. A rationale for the preferred orientation in
P a u son -Kh a n d Rea ction Op tim iza tion . The Pau-
son-Khand reaction between cyclobutene 3 and acety-
lene was used to optimize the reaction conditions (see
Table 1). The reactions were run by adding cyclobutene
3 (containing 15% 3a ) to the complex formed between
Co2(CO)8 (1.2 equiv) and acetylene and then heating the
mixture to the appropriate temperature under an atmo-
sphere of CO and acetylene to give tricyclic cyclopenten-
one 9a . The first entry was run in heptane giving a yield
of 50%. In entry 2, the reaction time was doubled, in
entry 3, the solvent was switched to dimethoxymethane,
and in entry 4, n-Bu3PO5 was added but all conditions
produced essentially the same yield of 50%. In run 5,
the reaction temperature was lowered 10 °C increasing
the yield to 59%. The most useful improvement in yield
came in run 6 when the concentration of cyclobutene 3
was lowered by a factor of 5 to 0.031 M, giving 9a in 73%
yield. The only product isolated in these Pauson-Khand
reactions was 9a derived from 3 and not 3a .
these remarkably selective reactions is based on a steric
argument. Thus, the bigger allylic substituent (methoxy
for 18 and methyl for 20) and a CO ligand on cobalt are
in a 1,3-pseudodiaxial relationship in the disfavored
regioisomeric intermediate creating an unfavorable steric
(6) (a) Bladon, P.; Khand, I. U.; Pauson, P. L. J . Chem. Res. (S) 1977,
8. (b) La Belle, B. E.; Knudsen, M. J .; Olmstead, M. M.; Hope, H.;
Yanuck, M. D.; Schore, N. E. J . Org. Chem. 1985, 50, 5215. (c)
Sampath, V.; Lund, E. C.; Knudsen, M. J .; Olmstead, M. M.; Schore,
N. E. J . Org. Chem. 1987, 52, 3595.
(7) (a) Magnus, P.; Exon, C.; Albaugh-Robertson, P. Tetrahedron
1985, 41, 5861. (b) Magnus, P.; Principe, L. M. Tetrahedron Lett. 1985,
26, 4851. (c) Krafft, M. E. J . Am. Chem. Soc. 1988, 110, 968. (d) Krafft,
M. E.; J uliano, C. A.; Scott, I. L.; Wright, C.; McEachin, M. D. J . Am.
Chem. Soc. 1991, 113, 1693. (e) Krafft, M. E.; J uliano, C. A. J . Org.
Chem. 1992, 57, 5106. (f) Krafft, M. E.; Scott, I. L.; Romero, R. H.;
Feibelmann, S.; Van Pelt, C. E. J . Am. Chem. Soc. 1993, 115, 7199.
(g) Castro, J .; Moyano, A.; Pericas, M. A.; Riera, A. Tetrahedron 1995,
51, 6541.
P a u son -Kh a n d Rea ction s. Pauson-Khand reac-
tions of cyclobutenes 3-5 with various acetylenes were
carried out, and the results are presented in Table 2.
(4) Scouten, C. G.; Barton, F. E.; Burgess, J . R.; Story, P. R.; Garst,
J . F. J . Chem. Soc., Chem. Commun. 1969, 78.
(5) Billington, D. C.; Helps, I. M.; Pauson, P. L.; Thomson, W.;
Willison, D. J . Organomet. Chem. 1988, 354, 233.

Synthesis of the Spatane Nucleus
J . Org. Chem., Vol. 63, No. 5, 1998 1381
Ta ble 2. P a u son -Kh a n d Rea ction s
methyl group and a CO ligand on cobalt is absent.
Therefore, intermediates C/D are predicted to be favored
over intermediates A/B. Intermediates C/D lead to 9b
which is, therefore, expected to be the favored product.
However, 9a is the actual product formed, presumably,
through intermediate A and/or B. As for 3, the steric
argument outlined above clearly is inadequate to properly
predict the orientation of cyclobutene addition in the
Pauson-Khand reaction of 4 with acetylene.
F igu r e 2. 1,3-Pseudodiaxial steric interaction.
We propose a modified steric argument to explain the
orientation of cyclobutene addition for the Pauson-
Khand reactions summarized in Table 2. Thus, our
model also is based on steric interactions in the inter-
mediates of Figure 3. In intermediates A/B the allylic
methyl group has a 1,3-pseudodiaxial steric interaction
with a CO ligand on cobalt. In intermediates C/D the
allylic methyl group has a 1,3-pseudodiaxial steric in-
teraction with the acetylenic substituent R (R ) H). The
1,3-pseudodiaxial interaction of the R group (R ) H) with
the allylic methyl is apparently greater than that be-
tween a CO ligand and the allylic methyl group in A/B.
Therefore, intermediates A/B are favored. Perhaps the
CO ligands for these intermediates can be accommodated
and not cause much steric interaction with the allylic
methyl group. This new steric hypothesis also rational-
izes the orientation of cyclobutene addition in the Pau-
son-Khand reaction of 4 with acetylene (Table 2, entry
2).
interaction (see Figure 2). In the favored regioisomeric
intermediate, the smaller allylic substituent and a CO
ligand on cobalt are in a 1,3-pseudodiaxial relationship.
The orientation of the alkene component in Pauson-
Khand reactions of several other bicyclic alkenes were
rationalized previously by similar steric consider-
ations.1a,b,e,6b,8
The preferential formation of cyclopentenone 9a in the
Pauson-Khand reaction of cyclobutene 3 was unexpected
in view of the steric rationale mentioned above. Thus,
the four possible reaction-determining intermediates in
the reaction of 9a with acetylene and dicobaltoctacarbo-
nyl (Table 2, entry 1) are depicted in Figure 3. In
intermediates A and B, the 1,3-pseudodiaxial interaction
between the allylic methyl group, which is clearly bigger
than the allylic hydrogen, and a CO ligand on cobalt
would be unfavorable. In intermediates C and D, the
disfavored 1,3-pseudodiaxial interaction between the
The unexpected regiochemistry of entries 1 and 2
(Table 2) prompted the investigation of 1-(trimethylsilyl)-
propyne in an effort to take advantage of the unexpected
(8) Price, M. E.; Schore, N. E. Tetrahedron Lett. 1989, 30, 5865.

1382 J . Org. Chem., Vol. 63, No. 5, 1998
Kowalczyk et al.
portion of intermediates A/B or C/D. However, we believe
the R′ group plays a critical role by sterically congesting
the CO ligands on cobalt. The larger the R′ group results
in more congested CO ligands which cannot, in turn, as
easily accommodate steric interactions with other groups.
The size of the CO ligands, therefore, becomes effectively
larger. Reevaluation of the steric interactions for inter-
mediates A/B reveals that increasing the size of the CO
ligands makes the interaction with the allylic methyl
more significant and less favorable. In intermediates C/D
the increasing effective size of the CO ligands can be more
easily accommodated with H rather than methyl at the
ring juncture. Therefore, as the acetylenic R′ group
increases in size from entry 2 through 5, intermediates
C/D become more favorable compared to intermediates
A/B corresponding with the switch from product a to b.
The TMS group in entry 5 (Table 2) had an even
greater effect than the t-Bu group in entry 4. This may
be due to the different bond distances of the TMS group,
which may more optimally interact with the CO ligands,
or the Si may have a subtle electronic reinforcing effect.
The electronic effect of the Si will be discussed later.
For entries 6 and 7 (Table 2), the acetylenes are
disubstituted which gives them both a R′ substituent and
a R substituent which can both affect the orientation of
the alkene component.9 The R′ ) Me group for entry 6
would be expected to have a similar directing effect as
the n-Pr group in entry 3, slightly favoring intermediates
A/B. However, the R ) Me group for entry 6 also plays
a role. In intermediates A/B the R ) Me group has a
1,3-pseudodiaxial steric interaction with the ring juncture
H, but in intermediates C/D it has a 1,3-pseudodiaxial
steric interaction with the angular Me which must
disfavor intermediates C/D. The greater selectivity for
14a over 14b compared to entry 3 is therefore under-
standable.
The large magnitude of the directing influence of the
R′ ) TMS group is demonstrated in entry 7. The R′ )
TMS group greatly favors intermediates C/D versus
intermediates A/B by analogy to entry 5, but the R ) Me
group’s influence is opposite, favoring intermediates A/B
versus intermediates C/D by analogy to entry 6. The
production of only 15b clearly illustrates the dominance
of the R′ ) TMS group’s influence on the orientation of
the alkene in the Pauson-Khand reaction of cyclobutene
4 with 1-(trimethylsilyl)propyne.
Cyclobutene 5 was prepared without an angular Me
group to more clearly define the role of this Me group.
In both entries 8 and 9 the low regioselectivity suggests
the ring juncture Me group is indeed extremely important
for high regioselectivity. The ketal functionality main-
tained in alkene 5 apparently has little directing influ-
ence itself.
F igu r e 3. The four possible Pauson-Khand alkene insertion
intermediates.
regiochemistry and make 15a with the methyl group in
the correct place for spatanes. However, the opposite
regioisomer 15b was produced (Table 2, entry 7). The
difficulty in predicting the regiochemistry in the Pauson-
Khand reactions of many unsymmetrical olefins points
out the delicate balance of factors involved in many
systems including this series of examples. The system-
atic investigation of the acetylenes substitution pattern
appeared critical in this series to understand the regio-
chemical outcome and to develop an efficient synthesis
of the spatane nucleus. In the literature, the effect of
disubstituted acetylenes on the regiochemical outcome
of the olefin has been investigated and will be discussed
in due course, but the systematic investigation of the
effect of monosubstituted acetylenes has not been un-
dertaken and was crucial in this study.9 The reactions
of Table 2, entries 3-5, were carried out to study the
effect of differing monosubstituted acetylenes on the
orientation of the alkene addition.
For entry 3 in Table 2, the same orientation, i.e.,
favoring regioisomer 11a , in the Pauson-Khand reaction
of olefin 4 predominated as in entries 1 and 2, but only
by a small margin over the other regioisomer 11b. In
entry 4, the opposite regioisomer 12b became the major
product instead of 12a . In entry 5, the remarkable
reversal in regiochemistry was complete with only 13b
produced. Examining the possible intermediates A/B
versus C/D, it is surprising that for entries 2 through 5
intermediates C/D became not only competitive with
intermediates A/B but became the sole intermediate for
entry 5. The only difference between these entries was
in R′ which is remote to anything on the alkene derived
While steric control arguments are usually postulated
to rationalize regioselectivities, electronic control has also
been used to rationalize product distribution.10 The
complete change in regioselectivity of entries 1-7 could
potentially be rationalized based on the electronic argu-
ment that the polarization of the cobalt-acetylene com-
plex has somehow changed based on the substitution of
the acetylene. The ability of silicon to stabilize a positive
charge on a â carbon, thus affecting the polarization, is
(10) (a) MacWhorter, S. E.; Sampath, V.; Olmstead, M. M.; Schore,
N. E. J . Org. Chem. 1988, 53, 203. (b) MacWhorter, S. E.; Schore, N.
E. J . Org. Chem. 1991, 56, 338.
(9) Krafft, M. E. Tetrahedron Lett. 1988, 29, 999.

Synthesis of the Spatane Nucleus
J . Org. Chem., Vol. 63, No. 5, 1998 1383
Ta ble 3. P a u son -Kh a n d Rea ction s of Nor bor n en on e
Ta ble 4. Selected 1H NMR Sp ectr a l Da ta of P a u son -Kh a n d Ad d u cts
1H NMR (chemical shift (ppm), multiplicity, coupling constant (Hz))
compound
H-1
7.77 dd
J ) 3.1, 5.4
7.73 dd
J ) 3.1, 5.4
7.34 s(br)
H-2
H-3
H-9
H-10
2.51 d
J ) 4.9
2.55 d
J ) 4.9
2.62 d
J ) 4.8
2.88 m
9a
6.28 d
J ) 5.3
6.23 d
J ) 5.4
-
-
-
-
3.34 m
3.14 m
10a
11a
11b
12a
12b
13b
14a
14b
15b
3.02 s(br)
-
-
-
-
-
-
-
-
7.33 s(br)
2.66 dd
J ) 3.2, 5.0
2.91 ddd
J ) 2.7, 3.0, 5.2
2.60 dd
J ) 3.1, 5.0
2.55 dd
7.26 d
-
-
2.53 d
J ) 5.2
2.77 dd
J ) 3.4, 5.0
2.93 dd
J ) 3.3,4.9
2.50 d
J ) 5.0
2.75 d
J ) 4.8
2.76 d
J ) 5.3
7.26 d
J ) 3.4
7.75 d
J ) 3.0
-
-
J ) 3.2, 4.9
2.76 s(br)
-
-
-
-
2.61 dd
J ) 3.0, 5.0
2.53 dd
-
J ) 2.9,5.2
F igu r e 4. Norbornenone Pauson-Khand reactions.
F igu r e 5. Similar reversal in regioselectivity.
suggestive that it may contribute to the great seletivity
for entries 5 and 7 in Table 2.11 One of the most clearly
defined examples of electronic control is with nor-
bornenone 22 shown in Figure 4.10 To investigate further
the possible influence of the R′ substituent on the
polarization of the cobalt-acetylene complex, Pauson-
Khand reactions of norbornenone 22 with several other
monosubstituted acetylenes (R′ ) H, t-Bu, and TMS)
were examined (see Table 3). The regioselectivity was
2.2/1 to 3.1/1 which was remarkably similar to the
literature examples in Figure 4. This seems to rule out
any switch of the polarization of the cobalt-acetylene
complex as a rational for the change in regioselectivity
demonstrated in Table 2. The only example we know of
in the literature that demonstrates a similar reversal,
although not complete, in orientation of the alkene
component as a consequence of varying the substitution
in the acetylene component of a Pauson-Khand reaction
is shown in Figure 5.12
1
1
An a lysis of H NMR Sp ectr a l Da ta . The H NMR
spectral data were a valuable tool for structural assign-
ment of Pauson-Khand adducts 9a -17b (see Table 4).
(11) (a) Gabelica, V.; Kresge, A. J . J . Am. Chem. Soc. 1996, 118,
3838. (b) Colvin, E. W. Silicon in Organic Synthesis; Butterworths:
London, 1981.
(12) (a) de Meijere, A.; Wessjohann, L. Synlett 1990, 20. (b) Reference
1a, pp 68-69.

1384 J . Org. Chem., Vol. 63, No. 5, 1998
Kowalczyk et al.
F igu r e 7.
F igu r e 6.
Sch em e 3a
The analysis for 9a was done as follows. There were two
possible cyclopentenones annulated onto cyclobutene 3
resulting in either structure 9a or 9b. The vinyl proton
at 6.28 ppm (doublet) was assigned as R to the carbonyl
(H-2) and the vinyl proton at 7.77 ppm (doublet of
doublets) as â to the carbonyl (H-1) because of their
coupling patterns. The proton at 3.34 ppm (H-9) was
coupled to the â vinyl proton, therefore, it was assigned
as next to the â vinyl carbon. The proton at 2.51 ppm
(H-10) was coupled only to the proton at 3.34 ppm;
therefore, it was assigned as next to the proton at 3.34
ppm, and structure 9b could be ruled out, because the
proton at 2.51 ppm was only a doublet. The proton-
proton coupling of protons H1, 2, 9, and 10 were deter-
mined by homonuclear decoupling experiments. Struc-
tural assignment of enone 10a was based on the same
arguments and similarity to enone 9a . The regiochem-
istry of enones 11a , 11b, 12a , 12b, 13b, all having â vinyl
protons, were assigned based on similar reasoning.
The assignment of the regiochemistry of the carbonyl
at C-1 or C-3 for enones 14a , 14b, and 15b could not be
made based on coupling of a vinyl proton to H-9 or H-10
as above. The assignment, however, could be based on
the knowledge that the proton on the cyclobutane carbon
adjacent to the â carbon of the enone was further
downfield than the cyclobutane proton on the carbon
adjacent to the carbonyl in all the compounds already
assigned (9a , 10a , 11a , 11b, 12a , 12b, 13b). Therefore,
with the assignment of which proton of the two protons
in question is adjacent to each end of the enone func-
tionality, the multiplicity of the H-9 and H-10 protons
was sufficient to assign the regiochemistry.
(a) H2, Pd/C, 100%; (b) sodium tert-amylate, Ph3CH3+I-, 80%;
a
(c) diimide, 92%; (d) Mitsunobu, 92%.
Sch em e 4a
The monosubstituted acetylenes used to make 11a ,
11b, 12a , 12b, and 13b all gave the R-substituted enones
as expected. This assignment was based on their vinyl
protons having a chemical shift of over 7.2 ppm, indicat-
ing the protons as â not R to the carbonyl based on
comparison to the chemical shifts of the vinyl protons of
9a and 10a .
The orientation of the allylic methyl group at 2.12 ppm
in enone 15b was assigned on the â carbon of the enone.
This assignment was based on comparison to enones 14a
and 14b which have the protons of the R methyl group
at 1.62 ppm and 1.75 ppm, but their â methyl protons at
1.93 and 2.02 ppm.
The structure of 17b was determined by analysis of
the proton NMR spectral data. Protons H-3, H-4, H-8,
H-9, and H-10 were assigned based on their coupling
using a COSY spectrum (see Figure 6). The H-8 proton
was a doublet of doublets, and H-4 was a multiplet which
is consistent with 17b and not 17a .
The structures of the adducts with norbornenone 22
were assigned based on the â vinyl proton NMR spectral
data. The chemical shift of the â vinyl proton for the
anti-diketones appeared approximately 0.1 ppm upfield
of that in the corresponding syn-diketone isomers.10
a
(a) H2, Pd/C, 74%; (b) (1) TsNHNH2; (2) NaBH4; (3) H+, 54%;
(c) Swern oxidation, 99%.
Supporting NOESY (nuclear Overhauser effect spectros-
copy) was carried out on Pauson-Khand adduct 24b.
Interactions were evident between H-3 and H-4 and
between H-6-endo and H-7a (see Figure 7). These
interactions are consistent with an anti-diketone 24b and
a syn-diketone 24a .
Ch em ica l Str u ctu r e P r oof. Pauson-Khand adduct
13b in a separate publication was a key intermediate in
a formal total synthesis of spatol (2) and a total synthesis
of spatadiene (1).3 This chemical proof of structure
confirmed both the regiochemical and stereochemical
assignment of this Pauson-Khand adduct.
In a similar chemical sequence to what was used for
13b, Pauson-Khand adduct 9a was converted to alcohol
29 for comparison to known alcohol 30 (see Scheme 3).
Adduct 9a was first hydrogenated using 5% palladium
on carbon as catalyst to give ketone 26. The attachment
of C-11 was achieved using a Wittig reaction under Conia
equilibrating conditions, which involved the use of so-

Synthesis of the Spatane Nucleus
J . Org. Chem., Vol. 63, No. 5, 1998 1385
(40 mL), pyridinium p-toluenesulfonate (2.75 g), and benzene
(500 mL). The mixture was refluxed overnight. The solution
was cooled, and the ethylene glycol was separated. The
benzene layer was washed with aqueous NaHCO3 (125 mL)
and aqueous NaCl (125 mL). The solution was dried over
MgSO4, filtered, and concentrated to give 8 (30.50 g, 96%) as
clear liquid: 1H NMR (250 MHz, CDCl3) δ 1.17-1.28 (singlets
(biggest at 1.23), 3), 1.38-2.73 (m, 5), 3.74-3.91 (m, 4), 3.94-
4.79 (m, 2). Anal. Calcd for C10H14O2Cl2: C, 50.65; H, 5.95;
Cl, 29.90. Found: C, 50.81; H, 5.87; Cl, 29.68.
dium tert-amylate as the base in THF, giving exocyclic
methylene 27.13 Reduction of the exocyclic double bond
of 27 was done with diimide generated from anhydrous
hydrazine and 30% hydrogen peroxide in THF/EtOH.14
The reduction gave 28 with the reduction occurring from
the top face giving the R methyl group. The stereochem-
istry of the alcohol was inverted using Mitsonubu condi-
tions to give alcohol 29.15 Alcohol 29 was compared
known alcohol 30. Thus, comparison of the spectral data
confirmed the structure assigned to alcohol 29 and,
therefore, also confirmed the assignment of Pauson-
Khand adduct 9a .
Definitive chemical proof of the structural assignment
of Pauson-Khand adduct 15b was done by converting
15b into ketone 33 (see Scheme 4). Adduct 15b was
hydrogenated with concurrent loss of the trimethylsilyl
group to give 32. On larger scale runs, ketone 31 could
be isolated in minor amounts, indicating that hydrogena-
tion occurred first followed by loss of the trimethylsilyl
group due to the acidic nature of the reaction conditions.
Ketone 33 was made by reduction of the tosylhydrazone
of 32 with NaBH4 followed by deketalization. Ketone 33
proved identical to the ketone produced by Swern oxida-
tion of alcohol 28, confirming the structural assignments
made.16
5-Met h ylsp ir o[b icyclo[3.2.0]h ep t -6-en e-2,2′-[1,3]d iox-
ola n e] (4). In a N2-swept flask equipped with a glass-coated
stirring bar were placed naphthalene (70.0 g), THF (300 mL),
and sodium ribbon (10.4 g). The solution was stirred overnight
at room temperature and stored in a refrigerator. In a
separate N2-swept flask were placed dichloro ketal 8 (29.3 g)
and THF (150 mL). The solution was placed in an ice/water
bath, and the sodium naphthalide solution was added via a
small cannula using N2 pressure until the dichloro ketal
solution remained deeply colored. After 20 min, the reaction
was quenched with methanol (3 mL), aqueous NH4Cl (30 mL),
and water (100 mL). The mixture was extracted with ethyl
ether (200 mL) three times. The combined organic layers were
dried over MgSO4, filtered, and concentrated. The residue was
purified by chromatography (6 × 20 cm column) using a solvent
gradient starting with hexanes up to 10% ethyl ether in
hexanes as the eluent. The resulting residue was further
purified by chromatography (6 × 20 cm column) using a solvent
gradient starting with n-pentane up to 20% ethyl ether in
n-pentane to give 4 (15.75 g, 77%) as a clear liquid: IR (KBr)
3040, 2960, 1455, 1440, 1340 cm-1; 1H NMR (250 MHz, CDCl3)
δ 1.30 (s, 3), 1.35-1.58 (m, 2), 1.72 (dd, 1, J ) 6.8, 12.9), 2.23
(ddd, 1, J ) 7.4, 12.9, 12.9), 2.47 (s, 1), 3.84-3.98 (m, 4), 6.07
(s, 1), 6.09 (s, 1). Anal. Calcd for C10H14O2: C, 72.26; H, 8.49.
Found: C, 72.09; H, 8.49.
Exp er im en ta l Section
All reagents were obtained from commercial suppliers. For
NMR spectra, coupling constants are reported in Hz. Silica
gel chromatography was performed using Silica Gel 60 (230-
400 mesh ASTM) from VWR Scientific. The HPLC used was
a Waters model M-6000A pump equipped with a recycle valve
and a Whatman M-9 column with 10/50 Partisil packing
operating at 4.0 mL/min. All solutions were concentrated
using a rotary evaporator followed when necessary with high
vacuum. Elemental analyses were performed by the Mi-
croanalytical Laboratory operated by the College of Chemistry,
University of California, Berkeley, CA.
6,7-Dich lor o-5-m eth ylbicyclo[3.2.0]h ep ta n -2-on e (7). A
2.1 L Hanovia reactor was thoroughly washed, treated with a
5% Me2Cl2Si in toluene, MeOH-rinsed, and oven-dried. Into
the reactor were placed a mixture of 1,2-dichloroethylene (550
mL) and cyclohexane (1550 mL) filtered (4.5 × 11 cm column)
through alumina (no. 1 neutral) and 3-methylcyclopent-2-
enone (5.05 g). The solution was degassed using N2, and a
constant N2 flow was maintained throughout the irradiation.
The solution was irradiated (250 W) through a Pyrex filter,
while stirring with a magnetic stirring bar, for 20 h. More
3-methylcyclopent-2-enone (5.11 g) was added, and the ir-
radiation was continued for another 24 h. The excess 1,2-
dichloroethylene was distilled from the mixture for reuse, and
the remaining solution was concentrated. The residue was
purified by chromatography (4.5 × 12 cm column) using 20%
ethyl acetate in hexanes as the eluent. The resulting residue
was further purified by chromatography (4.5 × 15 cm column)
using a gradiant of 10 to 20% ethyl acetate in hexanes as the
eluent to give 7 (18.06 g, 89%) as a clear liquid: IR (KBr) 2975,
1745 cm-1; 1H NMR (200 MHz, CDCl3) δ 1.39-1.51 (3 singlets,
Sp ir o[bicyclo[3.2.0]h ep t-6-en e-2,2′-[1,3]d ioxola n e] (5)
via 6,7-Dich lor obicyclo[3.2.0]h ep ta n -2-on e. A Hanovia
reactor was treated with a solution of 5% Me2SiCl2 in toluene,
rinsed with methanol, and dried. The reactor was charged
with a solution of 500 mL of cyclohexane and 100 mL of 1,2-
dichloroethylene (a mixture of cis and trans isomers). The
solution was first filtered through alumina (neutral, 4.5 × 11
cm). The solution was degassed using nitrogen, and a constant
nitrogen flow was maintained throughout the irradiation. The
2-cyclopentenone (7.0 g, 85 mmol) was added via a syringe
pump over 18 h. During this time, the solution was irradiated
(250 W) through a Pyrex filter, while stirring. When all the
2-cyclopentenone was added, the irradiation was continued for
another 18 h while the reaction was monitored by infrared
spectroscopy. The solution was concentrated under reduced
pressure. The residue was purified by column chromatography
(10% EtOAc/hexane, 7 × 13 cm) to give dichloro ketone as an
oil (11.12 g, 73%). For spectroscopic purposes, one pure
fraction was collected in the middle of the elution of the
product. Three diastereomers are present in this sample. No
effort was made to separate these isomers. The reported 1H
NMR spectrum represents a mixture of the three isomers. In
the 13C NMR spectrum, it was possible (by signal heights) to
report one diastereomer separately from the other two. IR
1
1750, 1725, 1165 cm-1; H NMR (CDCl3) δ 2.01-2.62 (m, 4),
2.87-3.43 (m, 2), 4.15-4.20 (m, 0.5), 4.37-4.52 (m, 1), 4.66
(dd, 0.5, J ) 6.75, 8.63); 13C NMR (CDCl3) first diastereomer
δ 213.26, 62.62, 56.31, 47.70, 45.05, 37.29, 25.52; second and
third diastereomers δ 215.16, 214.01, 61.69, 60.07, 58.17,
57.97, 52.61, 52.46, 46.73, 38.07, 37.86, 37.06, 23.75, 20.56.
Anal. Calcd for C7H8Cl2O: C, 46.96; H, 4.50. Found: C, 47.34;
H, 4.58.
3), 1.63-2.87 (m, 5) 4.05-4.92 (m, 2). Anal. Calcd for C8H10
-
OCl2: C, 49.77; H, 5.22; Cl, 36.72. Found: C, 49.71; H, 5.12;
Cl, 36.68.
6,7-Dich lor o-5-m eth ylsp ir o[bicyclo[3.2.0]h ep ta n e-2,2′-
[1,3]d ioxola n e] (8). In a flask equipped with a Dean-Stark
trap were placed dichloro ketone 7 (25.90 g), ethylene glycol
A 250 mL flask was charged with 8.6 g (48 mmol) of dichloro
ketone from above, 10.7 mL (192 mmol) of ethylene glycol, and
200 mL of benzene. A few milligrams of p-toluenesulfonic acid
were added, and a Dean-Stark trap was attached. The
solution was heated under reflux for 12 h, giving about 1 mL
of water in the trap. The solution was transferred into a
separatory funnel, and the lower, dark ethylene glycol layer
was drained. The benzene layer was washed with 50 mL of
(13) Conia, J . M.; Limasset, J . C. Bull. Soc. Chim. Fr. 1967, 1936.
(14) Paquette, L. A.; Ternansky, R. J .; Balogh, D. W.; Kentgen, G.
J . Am. Chem. Soc. 1983, 105, 5446.
(15) Varasi, M.; Walker, K. A. M.; Maddox, M. L. J . Org. Chem.
1987, 52, 4235.
(16) Mancuso, A. J .; Swern, D. Synthesis 1981, 165.

1386 J . Org. Chem., Vol. 63, No. 5, 1998
Kowalczyk et al.
saturated sodium bicarbonate and 50 mL of saturated brine.
The benzene solution was dried over magnesium sulfate,
filtered, and concentrated under reduced vacuum to 9.23 g of
a brown oil.
by chromatography (6.5 × 15 cm column) using a solvent
gradient starting with 30% ethyl ether in pentane up to 1/1
ethyl ether/pentane to give 3 and 3a (10.01 g, 86%, 81.5 to
18.5 ratio of 3 to 3a ) as a clear liquid: IR (KBr) 3345, 3040,
2950 cm-1; 1H NMR (250 MHz, CDCl3) δ 1.07-1.21 (m, 1), 1.21
(s, 3), 1.40-1.47 (m, 1), 1.79-1.89 (m, 2), 2.57 (s(br), 1), 2.69
(d, 1, J ) 7.0), 3.99 (ddd, 1(major isomer), J ) 7.2, 7.2, 9.3),
4.32 (ddd, 1(minor isomer)), 6.00 (s(br), 2(minor isomer)), 6.06
(m, 2(major isomer)); 13C NMR (50 MHz, CDCl3) major isomer
only δ 143.89, 131.60, 71.92, 55.60, 53.39, 31.93, 29.63, 22.87;
HRMS calcd for C8H12O: 124.0889, found 124.0883.
A 500 mL, three-necked flask containing a glass stirbar, 35
g of naphthalene, and 150 mL of tetrahydrofuran was blan-
keted with argon. Sodium (5.2 g) was added in portions over
1 min. The solution was allowed to stir overnight under argon
during which time the solution attained a dark blue color.
The ketal (7.20 g, 32.3 mmol) was added to a 500 mL flask,
and 50 mL of tetrahydrofuran was added. A nitrogen atmo-
sphere was established, and the flask was immersed in an ice
bath. The sodium dihydronaphthalide solution (∼100 mL) was
added via cannula over 15 min. The endpoint was determined
when the ketal solution remained blue. The solution was
allowed to stir 45 min, and 5 mL of methanol was added. The
solution was transferred into a separatory funnel containing
40 mL of saturated ammonium chloride. The mixture was
extracted with diethyl ether (2 × 100 mL). Then 50 mL of
water was added to the aqueous layer to dissolve all the
precipitates, and this was further extracted with 100 mL of
diethyl ether. The combined organic extracts were washed
with 40 mL of saturated brine. The organic layers were dried
over magnesium sulfate, filtered, and concentrated under
reduced pressure to afford an oil which was subjected to
column chromatography. The naphthalene was eluted with
hexane, and the product was eluted with 10% EtOAc/hexane
to give 5 as an oil (2.77 g, 57%): IR 2960, 1340, 1165, 1120,
Gen er a l P a u son -Kh a n d R ea ct ion P r oced u r e Usin g
Acetylen e. (3a â,3br,4â,6a r,6bâ)-3a ,3b,4,5,6,6a -Hexa h y-
dr o-4-h ydr oxy-6a-m eth ylcyclobu ta[1,2:3,4]dicyclopen ten -
1(6bH)-on e (9a ). In
a N2-swept flask equipped with a
magnetic stirrer were placed Co2(CO)8 (5.98 g) and n-heptane
(460 mL). Acetylene was swept slowly for 1 h over the solution
which gave off gas and turned reddish. Alcohol 3 (1.81 g, 81.5
to 18.5 mixture of alcohols 3 and 3a ) dissolved in n-heptane
(5 mL) was added to the solution, and an atmosphere of
acetylene and CO was generated over the solution. The
solution was heated to 65-70 °C for 8 h. The solution was
cooled to room temperature, and the reaction mixture was
filtered through a pad of Celite with the aid of ethyl ether (450
mL). The solution was concentrated, and the residue purified
by chromatography (4.5 × 12 cm column) using 2% MeOH in
ethyl ether as the eluent. The resulting residue was further
purified by chromatography (2.7 × 12 cm column) using 2%
MeOH in ethyl ether as the eluent to give 9a (1.54 g, 73%
based on major cyclobutene alcohol) as a clear oil: IR (KBr)
1
1030, 750 cm-1; H NMR (CDCl3) δ 1.48-1.70 (m, 3); 2.18-
2.26 (m, 1), 2.90 (d, 1, J ) 3.14), 3.23 (dd, 1, J ) 3.14, 6.62),
3.85-3.93 (m, 4), 6.03 (d, 1, J ) 2.72), 6.06 (d, 1, J ) 2.72);
13C NMR (CDCl3) δ 140.05, 136.72, 115.09, 64.84, 63.72, 52.01,
45.88, 31.28, 22.63. Anal. Calcd for C9H12O2: C, 71.02; H,
7.95. Found: C, 71.13; H, 8.05.
1
3420(br), 2970, 1690, 1580 cm-1; H NMR (250 MHz, CDCl3)
δ 1.03 (s, 3), 1.36-1.50 (m, 1), 1.75-2.11 (m, 4), 2.22 (s(br),
1), 2.51 (d, 1, J ) 4.9), 3.33-3.35 (m, 1), 4.32-4.41 (m, 1), 6.28
(d, 1, J ) 5.3), 7.77 (dd, 1, J ) 3.1, 5.4); 13C NMR (50 MHz,
CDCl3) δ 212.20 (C), 167.61 (CH), 136.03 (CH), 73.51 (CH),
51.51 (CH), 49.09 (CH), 44.04 (C), 38.50 (CH2), 34.88 (CH),
32.56 (CH2), 23.62 (CH3); HRMS calcd for C11H14O2: 178.0994,
found 178.0990.
(1r,2r,5r)- a n d (1r,2â,5r)-5-Meth ylbicyclo[3.2.0]h ep t-
6-en -2-ol (3 a n d 3a ) via (1r,2r,5r)- a n d (1r,2â,5r)-6,7-
Dich lor o-5-m et h ylb icyclo[3.2.0]h ep t a n -2-ol. In
a N2-
swept flask equipped with an addition funnel and a magnetic
stirrer were placed ketone 7 (16.85 g) and 300 mL of THF.
The solution was cooled to -78 C, and L-selectride (105 mL,
1.0 M in THF) was added dropwise via the addition funnel
over 15 min. The solution was left at -78 C for 1 h and
allowed to warm to room temperature (3 h). To quench the
reaction, MeOH (15 mL), water (15 mL), 20% aqueous NaOH
(15 mL), and dropwise 30% H2O2 (20 mL) were added with
ice/water bath surrounding the entire mixture. Most of the
THF was removed using a rotary evaporator. The residue was
diluted with ethyl ether (300 mL), and this was washed with
aqueous NaCl twice (150 mL). The solution was dried over
Na2SO4 and concentrated. The residue was kugelruhr distilled
at 6 mmHg and 140-175 °C to give a mixture of dichloro
alcohols (16.41 g, 96.4%) as a clear liquid, or the residue was
purified by chromatography (4.5 × 16 cm column) using 30%
ethyl acetate in hexanes as the eluent to give dichloro alcohols
(3a r,3bâ,6a â,6br)-1,2,3,3a ,6a ,6b-Hexa h yd r o-3a -m eth yl-
sp ir o[cyclobu ta [1,2:3,4]d icyclop en ten e-1(3bH),2′-[1,3]d i-
oxola n ]-4-on e (10a ). Following the general procedure for 9a ,
a Pauson-Khand reaction using Co2(CO)8 (1.23 g, 1.2 equiv),
n-heptane (94.9 mL), and cyclobutene 4 (0.50 g, 1.0 equiv)
dissolved in n-heptane (2 mL) was carried out at 65-70 °C
for 8 h; workup and silica gel chromatography (solvent
gradient starting from 10% ethyl ether in hexanes up to ethyl
ether, and the second column using a solvent gradient starting
with 1/1 ethyl ether/pentane up to ethyl ether) provided 10a
(0.55 g, 83%) as a clear oil: IR (KBr) 2960, 1700, 1580 cm-1
;
1H NMR (250 MHz, CDCl3) δ 1.03 (s, 3), 1.51-2.19 (m, 5), 2.55
(d, 1, J ) 4.9), 3.12-3.16 (m, 1), 3.80-3.95 (m, 4), 6.23 (d, 1,
J ) 5.4), 7.73 (dd, 1, J ) 3.1, 5.4); 13C NMR (50 MHz, CDCl3)
δ 211.30 (C), 166.02 (CH), 135.89 (CH), 117.07 (C), 64.82 (CH2),
64.12 (CH2), 50.89 (CH), 50.69 (CH), 43.48 (C), 38.55 (CH2),
38.26 (CH), 33.69 (CH2), 23.55 (CH3). Anal. Calcd for
(15.85, 93%) as a clear liquid: IR (KBr) 3420(br), 2970 cm-1
;
1H NMR (200 MHz, CDCl3) δ 1.23-1.33 (4 singlets, 3), 1.37-
2.78 (m, 5), 4.24-5.01 (m, 3).
C
13H16O3: C, 70.89; H, 7.32. Found: C, 70.68; H, 7.23.
Gen er a l P a u son -Kh a n d P r oced u r e Usin g a Liqu id
In a oven-dried, N2-swept flask equipped with a glass-coated
magnetic stirrer bar were placed naphthalene (76.9 g), THF
(300 mL), and sodium ribbon (9.6 g). The solution was stirred
overnight at room temperature to dissolve all the sodium and
stored in a refrigerator. In a separate N2-swept flask was
placed dichloro alcohols from above (18.33 g) and THF (100
mL) which was cooled to 0 °C in an ice/water bath. Using a
small cannula and N2 pressure, a portion of the sodium
naphthalide solution was transferred into the dichloro alcohol
solution until the solution remained deeply colored. The
reaction was quenched by adding MeOH (30 mL) and aqueous
ammonium chloride (20 mL). The resulting mixture was
diluted with water (100 mL) and extracted with ethyl ether
(100 mL) five times. The combined ether layers were dried
over Na2SO4, filtered, and concentrated. The residue was
purified by chromatography (6.5 × 15 cm column) using a
solvent gradient starting with pentane up to 1/1 pentane/ethyl
ether as the eluent. The resulting residue was further purified
Acetylen e. (3a r,3bâ,6a â,6br)-1,2,3,3a ,6a ,6b-Hexa h yd r o-
3a-m eth yl-5-pr opylspir o[cyclobu ta[1,2:3,4]dicyclopen ten e-
1(3bH),2′-[1,3]d ioxola n ]-4-on e (11a ) a n d (3a r,3bâ,6a â,-
6b r)-1,2,3,3a ,6a ,6b -H e x a h y d r o -3a -m e t h y l-5-p r o p y l-
sp ir o[cyclobu ta [1,2:3,4]d icyclop en ten e-1(3bH),2′-[1,3]d i-
oxola n ]-6-on e (11b). In a N2-swept flask was placed Co2(CO)8
(0.62 g, 1.2 equiv), n-octane (48 mL), and 1-pentyne (0.124 g,
1.2 equiv). After 50 min, a CO atmosphere was generated,
and cyclobutene 4 (0.2514 g, 1.0 equiv) was added. The
solution was heated at 122 °C for 5.5 h. The cooled solution
was purified by chromatography (1.8 × 13 cm column with
1.8 × 2 cm of Celite on top of the column) using a solvent
gradient starting with hexanes up to ethyl acetate. The
resulting residue was further purified by chromatography (1.8
× 14 cm column) using a solvent gradient starting with 10%
ethyl acetate in hexanes up to 30% ethyl acetate in hexanes
to give a mixture of 11a and 11b (0.2847 g, 72%, ratio of 1.32/
1.0 of 11a to 11b). The mixture was separated by HPLC (15

Synthesis of the Spatane Nucleus
J . Org. Chem., Vol. 63, No. 5, 1998 1387
mg/run, 5 recycles through the column) using 30% ethyl
acetate in hexanes as the eluent.
gradient starting with 10% ethyl acetate in hexanes up to 30%
ethyl acetate in hexanes) provided 14a and 14b (0.2897 g, 77%,
ratio of 14a /14b of 5.3/1.0). The mixture (38 mg/run) was
separated by HPLC using 30% ethyl acetate in hexanes as the
eluent.
The first fraction was concentrated to give 14a as a clear
oil: 1H NMR (250 MHz, CDCl3) δ 0.90 (s, 3), 1.51-2.21 (m, 5),
1.62 (s, 3), 1.93 (s, 3), 2.50 (d, 1, J ) 5.0), 2.76 (s(br), 1), 3.77-
3.90 (m, 4); 13C NMR (50 MHz, CDCl3) δ 210.10, 170.91, 138.19,
117.08, 64.59, 63.91, 50.99, 50.86, 44.46, 39.77, 38.30, 33.64,
The first fraction was concentrated to 11b as a clear oil:
1H NMR (250 MHz, CDCl3) δ 0.94 (t, 3, J ) 7.4), 0.99 (s, 3),
1.50-1.96 (m, 6), 2.18-2.26 (m, 3), 2.66 (dd, 1, J ) 5.0, 3.2),
2.88 (m, 1), 3.89-3.96 (m, 4), 7.33 (s(br), 1); 13C NMR (50 MHz,
CDCl3) δ 210.21, 158.12, 149.56, 117.44, 64.73, 64.11, 49.28,
48.08, 45.46, 41.90, 37.57, 33.90, 27.07, 22.94, 21.19, 13.93;
HRMS calcd for C16H22O3: 262.1570, found: 262.1571.
The second fraction was concentrated to give 11a as a clear
oil: 1H NMR (250 MHz, CDCl3) δ 0.92 (t, 3, J ) 7.2), 1.00 (s,
3), 1.47-2.25 (m, 9), 2.62 (d, 1, J ) 4.8), 3.02 (s(br), 1), 3.85-
3.93 (m, 4), 7.34 (s(br), 1); 13C NMR (50 MHz, CDCl3) δ 210.83,
158.95, 147.92, 117.24, 64.85, 64.11, 51.25, 43.45, 38.54, 35.54,
33.83, 26.79, 23.57, 20.99, 13.90; HRMS calcd for C16H22O3:
262.1570, found: 262.1570.
23.38, 14.56, 7.99; HRMS calcd for
found: 248.1417.
C15H20O3: 248.1413,
The second fraction was concentrated to give 14b as a clear
oil: 1H NMR (250 MHz, CDCl3) δ 0.96 (s, 3), 1.67-2.28 (m, 5),
1.75 (s, 3), 2.02 (s, 3), 2.61 (dd, 1, J ) 3.0, 5.0), 2.75 (d, 1, J )
4.8), 3.86-3.98 (m, 4); 13C NMR (50 MHz, CDCl3) δ 209.98,
170.94, 139.48, 117.36, 64.70, 64.06, 50.12, 48.29, 48.06, 41.23,
37.80, 33.91, 22.45, 17.76, 8.24; HRMS calcd for C15H20O3:
248.1413, found: 248.1417.
(3a r,3bâ,6a â,6br)-1,2,3,3a ,6a ,6b-Hexa h yd r o-3a -m eth yl-
5-(1,1-d im eth yleth yl)sp ir o[cyclobu ta [1,2:3,4]d icyclop en -
ten e-1(3bH),2′-[1,3]d ioxola n ]-4-on e (12a ) a n d (3a r,3bâ,-
6a â,6b r)-1,2,3,3a ,6a ,6b -H e x a h y d r o -3a -m e t h y l-5-(1,1-
d im et h ylet h yl)sp ir o[cyclob u t a [1,2:3,4]d icyclop en t en e-
1(3bH),2′-[1,3]d ioxola n ]-6-on e (12b). Following the general
procedure for 11a and 11b, a Pauson-Khand reaction using
Co2(CO)8, n-octane, tert-butylacetylene, and cyclobutene 4 was
carried out at 120-125 °C for 22 h; workup and silica gel
chromatography (started with hexane followed with 20%
EtOAc/hexane) provided 12a and 12b (78% yield, ratio of 5.3
to 1.0 of 12b to 12a ). The mixture was separated by HPLC.
The first fraction was concentrated to give 12a as a solid:
(3a r,3b â,6a â,6b r)-1,2,3,3a ,6a ,6b -H e xa h yd r o-3a ,4-d i-
m eth yl-5-(tr im eth ylsilyl)sp ir o[cyclobu ta [1,2:3,4]d icyclo-
p en ten e-1(3bH),2′-[1,3]d ioxola n ]-6-on e (15b). Following
the general procedure for 11a and 11b, a Pauson-Khand
reaction using Co2(CO)8 (0.62 g, 1.2 equiv), n-decane (47.5 mL),
1-(trimethylsilyl)propyne (0.20 g, 1.2 equiv), and cyclobutene
4 (0.25 g, 1.0 equiv) in n-decane (2 mL) was carried out at
155-164 °C for 75 h; workup and chromatography (solvent
gradient starting with 20% ethyl ether in hexanes up to 1/1
ethyl ether/hexanes) provided 15b (0.17 g, 37%) as a solid: mp
1
mp 85-87 °C; IR 2980, 1700, 1330, 1140, 1115 cm-1
:
1H NMR
55-57 °C; IR (solution cell, CCl4) 2980, 1695, 1585 cm-1'; H
(CDCl3) δ 0.98 (s, 3), 1.16 (s, 9), 1.50-1.85 (m, 4), 2.06-2.14
(m, 1), 2.53 (d, 1, J ) 5.16), 2.91 (ddd, 1, J ) 2.73, 2.98, 5.16),
3.82-3.92 (m, 4), 7.26 (d, 1, J ) 2.98); 13C NMR (CDCl3) δ
210.00, 156.92, 155.42, 117.26, 64.82, 64.06, 52.16, 51.18,
43.60, 38.57, 34.46, 33.80, 31.96, 28.23, 23.26.
NMR (250 MHz, CDCl3) δ 0.22 (s, 9), 0.97 (s, 3), 1.66-2.30
(m, 5), 2.12 (s, 3), 2.53 (dd, 1, J ) 2.9, 5.2), 2.76 (d, 1, J ) 5.3),
3.87-3.93 (m, 4); 13C NMR (50 MHz, CDCl3) δ 214.51 (C),
186.45 (C), 142.58 (C), 117.27 (C), 64.54 (CH2), 63.92 (CH2),
53.82 (CH), 48.41 (CH), 48.36 (C), 42.67 (CH), 37.75 (CH2),
33.88 (CH2), 22.27 (CH3), 20.91 (CH3), -0.62 (CH3). Anal.
Calcd for C17H26O3Si: C, 66.62; H, 8.55. Found C, 66.57; H,
8.62.
The second fraction was concentrated to give 12b as an oil:
1
IR 2980, 1700, 1320, 1110 cm-1; H NMR (CDCl3) δ 0.95 (s,
3), 1.19 (s, 9), 1.61-1.89 (m, 4), 2.17-2.26 (m, 1), 2.60 (dd, 1,
J ) 3.14, 5.02), 2.77 (dd, 1, J ) 3.42, 5.02), 3.87-3.93 (m, 4),
7.26 (d, 1, J ) 3.42); 13C NMR (CDCl3) δ 209.41, 157.14, 156.37,
117.43, 64.63, 64.04, 49.32, 48.16, 44.18, 42.92, 37.51, 33.92,
32.07, 28.35, 22.81. Anal. Calcd for C17H24O3: C, 73.88; H,
8.75. Found: C, 73.82; H, 8.93.
(3a r,3b â,6a â,6b r)-1,2,3,3a ,6a ,6b -H e x a h y d r o s p ir o -
[cyclobu ta[1,2:3,4]dicyclopen ten e-1(3bH),2′-[1,3]dioxolan ]-
4-on e (16a ) a n d (3a r,3bâ,6a â,6br)-1,2,3,3a ,6a ,6b-Hexa h y-
dr ospir o[cyclobu ta[1,2:3,4]dicyclopen ten e-1(3bH),2′-[1,3]-
d ioxola n ]-6-on e (16b). Following the general procedure for
9a , a Pauson-Khand reaction using Co2(CO)8, n-heptane,
acetylene, and cyclobutene 5 was carried out at 90-95 °C for
22 h; workup and silica gel chromatography (started with
hexane followed with 20% EtOAc/hexane) provided 16a and
16b in 63% yield. The regioisomers were subjected to HPLC,
but they did not separate well enough to be characterized
separately. The components were characterized as a mix-
(3a r,3bâ,6a â,6br)-1,2,3,3a ,6a ,6b-Hexa h yd r o-3a -m eth yl-
5-(tr im eth ylsilyl)sp ir o[cyclobu ta [1,2:3,4]d icyclop en ten e-
1(3bH),2′-[1,3]d ioxola n ]-6-on e (13b). Following the general
procedure for 11a and 11b, a Pauson-Khand reaction using
Co2(CO)8 (18.52 g, 1.2 equiv), n-octane (1424 mL), (trimethyl-
silyl)acetylene (5.31 g, 1.2 equiv), and cyclobutene 4 (7.50 g,
1.0 equiv) in n-octane (5 mL) was carried out at 112-118 °C
for 23 h; workup and silica gel chromatography (solvent
gradient starting with 20% ethyl ether in hexanes up to 1/1
ethyl ether/hexanes, and a second column using a solvent
gradient starting from 20% ethyl ether in hexanes up to 1/1
ethyl ether/hexanes) provided 13b (11.33 g, 86%) as a white
solid: mp 74-76 °C; IR (solution cell, CDCl3) 2970, 2900, 1685,
ture: IR 2980, 1700, 1355, 1110 cm-1 1H NMR (CDCl3) δ
;
1.67-2.23 (m, 5), 2.43-2.51 (m, 1.5), 2.69 (dd, 0.5, J ) 3.23,
4.74), 2.92-2.95 (m, 0.5), 3.22-3.25 (m, 0.5), 3.81-3.93 (m,
4), 6.26 (dd, 0.5, J ) 1.04, 5.43), 6.27 (dd, 0.5, J ) 0.90, 5.45),
7.70 (dd, 0.5, J ) 3.18, 5.45), 7.75 (dd, 0.5, J ) 3.15, 5.46); 13
C
NMR (CDCl3) δ 211.98, 211.14, 166.17, 165.75, 135.39, 135.34,
117.32, 117.16, 64.94, 64.77, 64.04, 48.18, 47.47, 45.16, 43.30,
42.88, 42.18, 40.70, 37.48, 32.96, 32.94, 29.42, 28.75. Anal.
Calcd for C12H14O3: C, 69.89; H, 6.84. Found: C, 69.53; H,
6.99.
1
1570 cm-1; H NMR (250 MHz, CDCl3) δ 0.15 (s, 9), 0.93 (s,
3), 1.57-2.28 (m, 5), 2.55 (dd, 1, J ) 3.2, 4.9), 2.93 (dd, 1, J )
3.3, 4.9), 3.85-3.88 (m, 4), 7.75 (d, 1, J ) 3.0); 13C NMR (50
MHz, CDCl3) δ 214.00 (C), 172.82 (CH), 150.49 (C), 117.27 (C),
64.53 (CH2), 63.93 (CH2), 49.24 (CH), 49.00 (CH), 48.36 (C),
42.08 (CH), 37.43 (CH2), 33.77 (CH2), 22.83 (CH3), -1.87 (CH3).
Anal. Calcd for C16H24O3Si: C, 65.71; H, 8.27. Found: C,
65.64; H, 8.28.
(3a r,3bâ,6a â,6br)-1,2,3,3a ,6a ,6b-Hexa h yd r o-5-(tr im eth -
ylsilyl)sp ir o[cyclobu ta [1,2:3,4]d icyclop en ten e-1(3b H),2′-
[1,3]d ioxola n ]-4-on e (17a ) a n d (3a r,3bâ,6a â,6br)-1,2,3,-
3a,6a,6b-Hexah ydr o-5-(tr im eth ylsilyl)spir o[cyclobu ta[1,2:
3,4]d icyclop en ten e-1(3bH),2′-[1,3]d ioxola n ]-6-on e (17b).
Following the general procedure for 11a and 11b, a Pauson-
Khand reaction using Co2(CO)8, n-heptane, (trimethylsilyl)-
acetylene, and cyclobutene 5 was carried out at 90-95 °C for
22 h; workup and silica gel chromatography (started with
hexane followed with 20% EtOAc/hexane) provided 17a and
17b in 80% yield. The mixture was separated by HPLC.
The first fraction was concentrated to give 17b as a solid:
mp 62-63 °C; IR 2990, 1700, 1295, 1265, 1110 cm-1; 1H NMR
(CDCl3) δ 0.15 (s, 9), 1.66-1.71 (m, 1), 1.83-1.93 (m, 2), 2.08
(dd, 1, J ) 3.06, 5.77), 2.17-2.24 (m, 1), 2.44-2.47 (m, 1), 2.66
(3a r,3bâ,6a â,6br)-1,2,3,3a ,6a ,6b-Hexa h yd r o-3a ,5,6-tr i-
m eth ylsp ir o[cyclobu ta [1,2:3,4]d icyclop en ten e-1(3bH),2′-
[1,3]d ioxola n ]-4-on e (14a ) a n d (3a r,3bâ,6a â,6br)-1,2,3,-
3a ,6a ,6b-Hexa h yd r o-3a ,4,5-tr im eth ylsp ir o[cyclobu ta [1,2:
3,4]d icyclop en ten e-1(3bH),2′-[1,3]d ioxola n ]-6-on e (14b).
Following the general procedure for 11a and 11b, a Pauson-
Khand reaction using Co2(CO)8 (0.62 g, 1.2 equiv), n-decane
(48 mL), 2-butyne (0.098 g, 1.2 equiv), and cyclobutene 4
(0.2515 g, 1.0 equiv) was carried out at 172-174 °C for 5.5 h;
workup and chromatography (solvent gradient starting with
hexanes up to ethyl acetate, second column using a solvent

1388 J . Org. Chem., Vol. 63, No. 5, 1998
Kowalczyk et al.
(dd, 1, J ) 3.06, 4.98), 2.88 (ddd, 1, J ) 2.67, 2.86, 4.98), 3.83-
3.94 (m, 4), 7.81 (d, 1, J ) 2.86); 13C NMR δ 214.41, 173.46,
147.97, 117.44, 64.69, 63.96, 46.25, 44.30, 42.94, 42.34, 33.00,
28.78, -1.90. Anal. Calcd for C15H22SiO3: C, 64.71; H, 7.96.
Found: C, 64.50; H, 8.15.
The second fraction was concentrated to give 17a as an oil:
∼2:1 17a :17b; 1H NMR (CDCl3) δ (* ) assigned to 17a , other
values are mixtures of 17a and 17b) 0.15 (s, 9), 1.66-2.22 (m,
5), 2.41-2.45 (m, 2), 3.16-3.18* (m, 1), 3.83-3.94 (m, 4), 7.75*
(d, 1, J ) 3.00); 13C NMR (CDCl3) δ (17a only) 215.21, 172.99,
148.03, 117.32, 64.94, 64.04, 48.50, 48.21, 41.75, 37.64, 32.97,
29.44, -1.91. Anal. Calcd for C15H22SiO3: C, 64.71; H, 7.96.
Found: C, 64.45; H, 8.03.
silica gel chromatography (started with hexane followed with
20% EtOAc/hexane) provided 25a and 25b in 42% yield. The
mixture was separated by HPLC.
The first fraction was concentrated to give 25b: mp 141 °C;
IR 1760, 1705, 1315, 1280, 1260 cm-1; 1H NMR (CDCl3) δ 0.17
(s, 9), 1.25-1.28 (m, 1), 1.48-1.51 (m, 1), 1.91 (dd, 1, J ) 4.33,
17.81), 2.12 (dd, 1, J ) 4.47, 17.81), 2.37 (d, 1, J ) 5.35), 2.54
(s, 1), 2.80 (s, 1), 3.03 (dd, 1, J ) 2.58, 5.35), 7.49 (d, 1, J )
2.58); 13C NMR (CDCl3) δ 215.34, 212.07, 169.07, 152.29, 52.76,
51.28, 45.79, 43.19, 37.41, 30.45, -1.90. Anal. Calcd for
C
13H18SiO2: C, 66.62; H, 7.74. Found: C, 66.61; H, 7.91.
The second fraction was concentrated to give 25a : mp 105-
106 °C; IR 1760, 1705, 1290, 1260, 1170 cm-1
:
1H NMR
(3a r,4r,7r,7a r)-4,5,7,7a -Tet r a h yd r o-4,7-m et h a n o-1H -
in d en e-1,6(3a H)-d ion e (23a ) a n d (3a r,4r,7r,7a r)-3a ,4,6,7,-
7a -Te t r a h yd r o-4,7-m e t h a n o-1H -in d e n e -1,5(4H )-d ion e
(23b). Following the general procedure for 9a , a Pauson-
Khand reaction using Co2(CO)8 (7.60 g, 22.2 mmol), heptane
(500 mL), acetylene, and norbornenone (2.00 g, 18.5 mmol) was
carried out at 90-95 °C for 22 h; workup and silica gel
chromatography (started with hexane followed with 20%
EtOAc/hexane) provided 23a and 23b as a solid (1.71 g, 57%).
The mixture was separated by HPLC.
The first fraction was concentrated to give 23b: mp 108-
109 °C; IR 2980, 1750, 1710, 1580, 1340, 1310 cm-1; 1H NMR
(CDCl3) δ 1.32-1.36 (m, 1), 1.47-1.51 (m, 1), 1.86 (dd, 1, J )
4.35, 17.84), 2.08 (dd, 1, J ) 4.50, 17.84), 2.33 (d, 1, J ) 5.19),
2.50 (s, 1), 2.75-2.76 (m, 1), 3.02-3.04 (m, 1), 6.32 (dd, 1, J )
1.58, 5.62), 7.44 (dd, 1, J ) 2.74, 5.62); 13C NMR (CDCl3) δ
214.61, 208.68, 162.30, 138.02, 51.47, 50.79, 44.63, 42.93,
37.01, 30.22. Anal. Calcd for C10H10O2: C, 74.06; H, 6.21.
Found: C, 73.81; H, 6.27.
(CDCl3) δ 0.15 (s, 9), 1.32-1.35 (m, 1), 1.46-1.50 (m, 1), 1.95
(dd, 1, J ) 4.29, 17.77), 2.17 (dd, 1, J ) 4.50, 17.77), 2.39 (d,
1, J ) 5.34), 2.59-2.60 (m, 1), 2.75 (s, 1), 2.96 (dd, 1, J ) 2.52,
5.34), 7.63 (d, 1, J ) 2.52); 13C NMR (CDCl3) δ 231.99, 210.87,
171.72, 152.40, 51.48, 50.66, 47.11, 43.57, 36.94, 30.27, -1.93.
Anal. Calcd for C13H18SiO2: C, 66.62; H, 7.74. Found: C,
66.67; H, 7.78.
(3a â,3br,4â,6a r,6bâ)-Octa h yd r o-4-h yd r oxy-6a -m eth yl-
cyclobu ta [1,2:3,4]d icyclop en ten -1(2H)-on e (26). In a flask
mounted on a shaker were placed enone 9a (1.08 g), ethyl
acetate (24 mL), and 5% palladium on carbon (0.72 g). The
flask was purged of air, and an atmosphere of hydrogen was
generated. The flask was shook for 1.5 h. The reaction
mixture was filtered through a pad of Celite and then through
a plug (1.8 × 3 cm column) of silica gel using ethyl acetate as
the eluent. The filtrate was concentrated to give 26 (1.09 g,
100%) as a clear liquid: IR (KBr) 3425(br), 2970, 1740 cm-1
;
1H NMR (250 MHz, CDCl3) δ 1.03 (s, 3), 1.47 (ddd, 1, J ) 6.9,
12.9, 12.9), 1.74 (dd, 1, J ) 6.5, 13.1), 1.84-2.58 (m, 9), 2.81-
2.84 (m, 1), 4.34 (ddd, 1, J ) 6.6, 6.6, 10.3); 13C NMR (50 MHz,
CDCl3) δ 221.93 (C), 74.30 (CH), 52.92 (CH), 43.38 (C), 38.58
(CH2), 38.30 (CH2), 32.56 (CH2), 28.21 (CH, CH2), 22.37 (CH,
CH3). Anal. Calcd for C11H16O2: C, 73.30; H, 8.95. Found:
C, 73.06; H, 8.88.
The second fraction was concentrated to give 23a , contami-
nated with about 15-20% of 23b: mp 58-59 °C, IR 1765,
1715, 1360, 1195, 1170 cm-1 1H NMR (CDCl3) δ 1.48-1.52
;
(m, 2), 1.95 (dd, 1, J ) 4.06, 17.84), 2.18 (dd, 1, J ) 4.52, 17.84),
2.41 (d, 1, J ) 5.24), 2.62-2.63 (m, 1), 2.76 (s, 1), 3.02-3.04
(m, 1), 6.37 (dd, 1, J ) 1.57, 5.66), 7.61 (dd, 1, J ) 2.65, 5.66);
13C NMR (CDCl3) δ 213.39, 207.65, 164.68, 138.48, 51.18,
49.41, 46.18, 43.46, 36.70, 30.18. Anal. Calcd for C10H10O2:
C, 74.06; H, 6.21. Found: C, 73.73; H, 6.39.
(1â,3a r,3b â,6a â,6b r)-Deca h yd r o-3a -m et h yl-4-m et h yl-
en ecyclobu ta [1,2:3,4]d icyclop en ten -1-ol (27). Into a N2-
swept flask equipped with an addition funnel were placed
methyltriphenylphosphonium iodide (5.75 g, 2.3 equiv), THF
(30 mL), and sodium tert-amylate (12.93 mL, 1 M solution in
toluene, 2.1 equiv). The solution was stirred for 15 min. Into
the ylide solution via the addition funnel was added keto
alcohol 26 (1.11 g, 1.0 equiv) dissolved in THF (50 mL) over 3
h. The reaction was stirred for 19 h. The reaction was
quenched with water (50 mL) and extracted with ethyl ether
(50 mL) five times. The combined ethyl ether layers were dried
over Na2SO4, filtered, and concentrated. The residue was
purified by chromatography (2.7 × 11.5 cm column) using 10%
ethyl acetate/90% methylene chloride as the eluent. The
resulting residue was further purified by chromatography (2.7
× 16 cm column) using 10% ethyl acetate/90% methylene
chloride as the eluent to give 27 (0.88 g, 80%) as an oil which
crystallized upon standing to white crystals: mp 61-64 °C,
(3a r,4r,7r,7a r)-4,5,7,7a -Tet r a h yd r o-2-(1,1-d im et h yl-
eth yl)-4,7-m eth a n o-1H-in d en e-1,6(3a H)-d ion e (24a ) a n d
(3a r,4r,7r,7a r)-3a ,4,6,7,7a -Tet r a h yd r o-2-(1,1-d im et h yl-
eth yl)-4,7-m eth a n o-1H-in d en e-1,5(4H)-d ion e (24b). Fol-
lowing the general procedure for 11a and 11b, a Pauson-
Khand reaction using Co2(CO)8, n-heptane, tert-butylacetylene,
and norbornenone was carried out at 90-95 °C for 22 h;
workup and silica gel chromatography (started with hexane
followed with 20% EtOAc/hexane) provided 24a and 24b in
64% yield. The mixture was separated by HPLC.
The first fraction was concentrated to give 24b: mp 133-
134 °C; IR 1755, 1710, 1370, 1330, 1220, 1165 cm-1; 1H NMR
(CDCl3) δ 1.18 (s, 9), 1.31 (d, 1, J ) 10.99), 1.48 (d, 1, J )
10.99), 1.89 (dd, 1, J ) 4.37, 17.82), 2.12 (dd, 1, J ) 4.51,
17.82), 2.37 (d, 1, J ) 5.54), 2.51 (s, 1), 2.79 (d, 1, J ) 4.51),
2.86 (dd, 1, J ) 2.82, 5.54), 7.00 (d, 1, J ) 2.82); 13C NMR
(CDCl3) δ 215.33, 207.42, 158.57, 153.20, 52.99, 51.46, 43.04,
41.13, 37.30, 32.15, 30.14, 28.26. Anal. Calcd for C14H18O2:
C, 77.03; H, 8.31. Found: C, 77.23; H, 8.73.
IR (solution cell, CCl4) 3650, 3350(br), 3090, 2960, 1665 cm-1
;
1H NMR (250 MHz, CDCl3) δ 0.88 (s, 3), 1.43-2.00 (m, 8),
2.30-2.53 (m, 3), 2.60-2.67 (m, 1), 4.27 (ddd, 1, J ) 6.9, 6.9,
9.6), 4.70 (s(br), 4.96 (s(br),1); 13C NMR (50 MHz, CDCl3) δ
154.21 (C), 107.32 (CH2), 75.05 (CH), 52.03 (CH), 51.74 (CH),
42.39 (C), 38.70 (CH2), 34.01 (CH2), 32.72 (CH2), 32.33 (CH),
32.11 (CH2), 21.70 (CH3). Anal. Calcd for C12H18O: C, 80.85;
H, 10.18. Found: C, 81.01; H, 10.16.
The second fraction was concentrated to give 24a : mp 100-
1
102 °C; IR 2980, 1755, 1710, 1270 cm-1; H NMR δ (CDCl3)
1.17 (s, 9), 1.38-1.49 (m, 2), 1.94 (dd, 1, J ) 4.28, 17.79), 2.17
(dd, 1, J ) 4.50, 17.79), 2.41 (d, 1, J ) 5.20), 2.65 (s, 1), 2.75
(s, 1), 2.81 (dd, 1, J ) 2.76, 5.20), 7.15 (d, 1, J ) 2.76); 13C
NMR (CDCl3) δ 214.05, 206.30, 159.00, 155.80, 51.56, 47.56,
45.74, 43.66, 37.04, 32.14, 30.04, 28.30. Anal. Calcd for
(1â,3a r,3b â,4r,6a â,6b r)-Deca h yd r o-3a ,4-d im et h ylcy-
clobu ta [1,2:3,4]d icyclop en ten -1-ol (28). In a 1-L round-
bottomed flask were placed olefin 27 (2.75 g), ethanol (500 mL),
THF (100 mL), and anhydrous hydrazine (20 mL). The
solution was cooled to 0 °C in an ice/water bath. Over 10 min,
30% hydrogen peroxide (66 mL) was added. After 15 min, the
cooling bath was removed, and the solution was stirred
overnight. The solution was concentrated to 40 mL. The
concentrate was diluted with aqueous NaCl (100 mL) and
extracted with CH2Cl2 (100 mL) five times. The combined
organic layers were dried over Na2SO4, filtered, and concen-
trated. The residue was purified by chromatography (4.5 ×
C
14H18O2: C, 77.03; H, 8.31. Found: C, 77.41; H, 8.59.
(3a r,4r,7r,7a r)-4,5,7,7a -Tetr a h yd r o-2-(tr im eth ylsilyl)-
4,7-m eth an o-1H-in den e-1,6(3aH)-dion e (25a) an d (3ar,4r,-
7r,7ar)-3a,4,6,7,7a-Tetr ah ydr o-2-(tr im eth ylsilyl)-4,7-m eth -
a n o-1H-in d en e-1,5(4H)-d ion e (25b). Following the general
procedure for 11a and 11b, a Pauson-Khand reaction using
Co2(CO)8, n-heptane, (trimethylsilyl)acetylene, and nor-
bornenone was carried out at 90-95 °C for 22 h; workup and

Synthesis of the Spatane Nucleus
J . Org. Chem., Vol. 63, No. 5, 1998 1389
15 cm column) with 10% ethyl acetate/90% methylene chloride
as the eluent to give 28 (2.55 g, 92%) as an oil which slowly
crystallized to white crystals: mp 59-61 °C, IR (solution cell,
(3ar,3bâ,4r,6aâ,6br)-Octah ydr o-3a,4-dim eth ylcyclobu ta-
[1,2:3,4]d icyclop en ten -1(2H)-on e (33) fr om 32. In a flask
equipped with a reflux condenser and drying tube were placed
ketone 32 (0.0841 g, 1.0 equiv), methanol (5 mL), and p-
toluenesulfonhydrazide (0.13 g, 2.0 equiv). The solution was
refluxed for 2 h. The solution was cooled, and sodium
borohydride (0.13 g) was added slowly. The solution was
refluxed for 5 h. The cooled solution was concentrated,
removing most of the methanol. To the residue was added
water (30 mL), and the mixture was extracted with ethyl ether
(30 mL) three times. The combined organic layers were dried
over MgSO4, filtered, and concentrated. The residue was
purified by chromatography (1.7 × 12 cm column) using 20%
ethyl ether in hexanes as the eluent to give a clear oil (0.0418
g). In a flask were placed a portion of the above isolated oil
(0.0180 g), THF (4 mL), water (0.5 mL), and concentrated HCl
(10 drops). After 4 h, the solution was concentrated removing
most of the THF. To the residue was added water (20 mL),
and the mixture was extracted with ethyl ether (20 mL) three
times. The combined organic layers were dried over MgSO4,
filtered, and concentrated. The residue was purified by
chromatography (1.2 × 10 cm column) using 20% ethyl ether
in hexanes as the eluent to give 33 (0.0148 g, 54%) as a clear
liquid: IR (KBr) 2960, 1745 cm-1; 1H NMR (200 MHz, CDCl3)
δ 1.17 (s, 3), 1.17 (d, 3, J ) 7.1), 1.62-2.45 (m, 11), 2.67 (ddd,
1, J ) 9.7, 9.7, 17.4); 13C NMR (50 MHz, CDCl3) δ 221.23 (C),
56.26 (CH), 50.79 (CH), 42.07 (C), 40.08 (CH), 39.22 (CH2),
38.82 (CH), 37.32 (CH2), 33.08 (CH2), 32.92 (CH2), 23.63 (CH3),
14.42 (CH3). Anal. Calcd for C12H18O: C, 80.85; H, 10.18.
Found: C, 80.53; H, 10.17.
1
CCl4) 3640, 2960 cm-1; H NMR (250 MHz, CDCl3) δ 1.11 (s,
3), 1.15 (d, 3, J ) 6.9), 1.25-1.96 (m, 10), 2.04 (dd, 1, J ) 7.0,
7.0), 2.33 (s(br), 1), 2.38-2.45 (m, 1), 4.18 (ddd, 1, J ) 7.1,
7.1, 9.4); 13C NMR (50 MHz, CDCl3) δ 74.77 (CH), 52.53 (CH),
49.94 (CH), 43.84 (C), 40.36 (CH, CH2), 33.63 (CH2), 32.95
(CH2), 32.76 (CH2), 31.52 (CH), 23.85 (CH3), 14.57 (CH3). Anal.
Calcd for C14H20O: C, 79.94; H, 11.18. Found: C, 80.05; H,
11.35.
(1r,3a r,3b â,4r,6a â,6b r)-Deca h yd r o-3a ,4-d im et h ylcy-
clobu ta [1,2:3,4]d icyclop en ten -1-ol (29). In
a N2-swept
flask were placed alcohol 28 (0.19 g, 1.0 equiv), THF (0.5 mL),
diethyl azodicarboxylate (0.249 mL, 1.5 equiv), and trifluoro-
acetic acid (0.122 mL, 1.5 equiv). To this solution was added
triphenylphosphine (0.416 g, 1.5 equiv) dissolved in THF (1.6
mL). The solution was stirred for 5 min, and sodium benzoate
(0.2364 g, 1.6 equiv) was added. After 20 h, the solution was
concentrated removing most of the THF. The flask was
equipped with
a reflux condenser, and the residue was
dissolved in methanol (2.2 mL). The solution was refluxed for
4 h. The solution was concentrated removing most of the
methanol. The residue was diluted with water (10 mL) and
extracted with methylene chloride (10 mL) five times. The
combined organic layers were dried over Na2SO4, filtered, and
concentrated. The residue was purified by chromatography
(2.8 × 16 cm column) using 10% ethyl acetate/90% methylene
chloride as the eluent to give 29 (0.1743 g, 92%) as a clear oil
which crystallized upon standing: mp 64-66 °C; IR (solution
1
cell, CCl4) 3350(br), 2950 cm-1; H NMR (250 MHz, CDCl3) δ
1.17 (d, 3, J ) 6.7), 1.25 (s, 3), 1.51-2.18 (m, 10), 4.04 (d, 1, J
) 3.5); 13C NMR (50 MHz, CDCl3) δ 79.38 (CH), 57.64 (CH),
48.34 (CH), 44.93 (C), 41.31 (CH2), 40.26 (CH), 36.94 (CH),
34.08 (CH2), 33.58 (CH2), 32.75 (CH2), 24.37 (CH3), 14.63 (CH3);
HRMS calcd for C12H20O: 180.1515, found 180.1511.
(3ar,3bâ,4r,6aâ,6br)-Octah ydr o-3a,4-dim eth ylcyclobu ta-
[1,2:3,4]d icyclop en ten -1(2H)-on e (33) fr om 28. In a N2-
swept flask cooled to -78 °C were placed methylene chloride
(10 mL) and oxalyl chloride (0.416 mL, 1.9 equiv). Dimethyl
sulfoxide (0.677 mL, 3.8 equiv) was added dropwise, and after
5 min, alcohol 28 (0.45 g, 1.0 equiv) was added dissolved in
methylene chloride (8 mL). After 20 min, triethylamine (1.66
mL, 4.8 equiv) was added. After 10 min, the solution was
taken out of the -78 °C bath for 30 min which allowed the
solution to come to room temperature. The reaction was
poured into water (50 mL) and extracted with methylene
chloride (30 mL) twice. The combined organic layers were
washed with aqueous NaCl (50 mL), dried over Na2SO4,
filtered, and concentrated. The residue was purified by
chromatography (1.7 × 13.5 cm column) using 10% ethyl
acetate in hexanes as the eluent to give 33 (0.44 g, 99%) as a
clear liquid, identical to 33 isolated above.
(1r,3a r,3b â,6r,6a â,6b r)-Deca h yd r o-3a ,6-d im et h ylcy-
clobu ta [1,2:3,4]d icyclop en ten -1-ol (30). 1H NMR (200
MHz, CDCl3) δ 0.94 (d, 3, J ) 6.0), 1.01 (s, 3), 1.23-2.18 (m,
13), 3.92 (d, 1, J ) 3.6); 13C NMR (50 MHz, CDCl3) δ 79.66,
57.92, 50.18, 44.39, 42.04, 39.63, 36.82, 34.17, 33.99, 28.06,
20.64, 13.70. Anal. Calcd for C12H20O: C, 79.94; H, 11.18.
Found: C, 79.68; H, 10.87.
(3a r,3b â,4r,6a â,6b r)-Oct a h yd r o-3a ,4-d im et h ylsp ir o-
[cyclobu ta[1,2:3,4]dicyclopen ten e-1(6H),2′-[1,3]dioxolan ]-
6-on e (32). In a flask mounted on a shaker were placed enone
15b (0.0296 g), ethyl acetate (5 mL), and 5% palladium on
carbon (0.0342 g). The flask was purged of air, and a hydrogen
atmosphere was generated. The solution was shook for 3 d.
The solution was filtered through a plug (1 × 6 cm) of silica
gel using ethyl acetate as the eluent. The solution was
concentrated, and the residue was purified by chromatography
(1 × 11 cm column) using a solvent gradient starting with 25%
ethyl ether in hexanes up to 50% ethyl ether in hexanes to
give 32 (0.0169 g, 74%) as white crystals: mp 66-67 °C; IR
Ack n ow led gm en t. Unfortunately Professor Dauben
passed away during the preparation of this manuscript.
He will be greatly missed at the University of California,
Berkeley. B.K. and T.S. greatly appreciated Professor
Dauben being our mentor, particularly his guidance and
inspiration during both our time in the Dauben group
and after, during our respective careers.
(solution cell, CCl4) 2980, 1753 cm-1 1H NMR (500 MHz,
;
CDCl3) δ 1.24 (d, 3, J ) 5.9), 1.30 (s, 3), 1.55 (ddd, 1, J ) 6.9,
13.5, 13.5), 1.67 (dd, 1, J ) 7.4, 12.6), 1.76 (s(br), 1), 1.83 (dd,
1, J ) 6.9, 13.2), 2.18 (ddd, 1, J ) 7.4, 13.6, 13.6), 2.35-2.54
(m, 5), 3.82-3.92 (m, 4); 13C NMR (50 MHz, CDCl3) δ 221.23
(C), 117.69 (C), 64.70 (CH2), 64.06 (CH2), 50.55 (CH), 46.70
(CH), 45.50 (C), 44.99 (CH2), 44.70 (CH), 40.11 (CH2), 35.58
(CH), 33.74 (CH2), 24.92 (CH3), 15.75 (CH3); HRMS calcd for
Su p p or tin g In for m a tion Ava ila ble: Copies of 1H and 13C
NMR spectra (13 pages). This material is contained in
libraries on microfiche, immediately follows this article in the
microfilm version of the journal, and can be ordered from the
ACS; see any current masthead page for ordering information.
C
14H20O3: 236.1413, found 236.1405. In a larger scale runs,
31 could be isolated in small amounts: 1H NMR (250 MHz,
CDCl3) δ 0.09 (s, 9), 1.21 (d, 3, J ) 6.9), 1.25 (s, 3), 1.32-2.57
(m, 9), 3.86-3.92 (m, 4).
J O9708194