Ruthenium-catalyzed alder ene type reactions. A formal synthesis of alternaric acid

Article abstract of DOI:10.1021/ja981540n

Alternaric acid, a nanomolar fungal germination inhibitor, is typified by a 1,4-diene, consisting of a terminal methylene and an (E)-l,2- disubstituted alkene. A new strategy for the synthesis of natural products containing such functionality stems from the development of a ruthenium- catalyzed addition of terminal alkenes with terminal alkynes. The alkyne substrate, 4-pentynoic acid, is commercially available or can be prepared in two steps by alkylation of tert-butyl acetate. The alkene substrate is prepared from commercially available (S)-2-methyl-l-butanol. This synthesis involves formation of a geometically defined trisubstituted alkene by involving Pd-catalyzed cross-coupling and asymmetric dihydroxylation. The ruthenium-catalyzed coupling proceeds best in the absence of alcohol protecting groups to maximize regioselectivity. The examples of this addition illustrated herein help elucidate some of the important factors controlling regioselectivity. They also illustrate the excellent chemoselectivity. The acyclic unit of alternaric acid, which is simply coupled to a dihydropyrone fragment to complete the synthesis, is available in only 11 steps and 27% overall yield compared to the one extant synthesis also starting from (S)-2- methyl-1-butanol which proceeds in 26 steps and 0.003% overall yield. This new reaction provides a powerful tool in streamlining this synthesis and should prove more generally useful.

Full text of DOI:10.1021/ja981540n

9228
J. Am. Chem. Soc. 1998, 120, 9228-9236
Ruthenium-Catalyzed Alder Ene Type Reactions. A Formal Synthesis
of Alternaric Acid
Barry M. Trost,* Gary D. Probst, and Andreas Schoop
Contribution from the Department of Chemistry, Stanford UniVersity, Stanford, California 94305-5080
ReceiVed May 4, 1998
Abstract: Alternaric acid, a nanomolar fungal germination inhibitor, is typified by a 1,4-diene, consisting of
a terminal methylene and an (E)-1,2-disubstituted alkene. A new strategy for the synthesis of natural products
containing such functionality stems from the development of a ruthenium-catalyzed addition of terminal alkenes
with terminal alkynes. The alkyne substrate, 4-pentynoic acid, is commercially available or can be prepared
in two steps by alkylation of tert-butyl acetate. The alkene substrate is prepared from commercially available
(S)-2-methyl-1-butanol. This synthesis involves formation of a geometically defined trisubstituted alkene by
involving Pd-catalyzed cross-coupling and asymmetric dihydroxylation. The ruthenium-catalyzed coupling
proceeds best in the absence of alcohol protecting groups to maximize regioselectivity. The examples of this
addition illustrated herein help elucidate some of the important factors controlling regioselectivity. They also
illustrate the excellent chemoselectivity. The acyclic unit of alternaric acid, which is simply coupled to a
dihydropyrone fragment to complete the synthesis, is available in only 11 steps and 27% overall yield compared
to the one extant synthesis also starting from (S)-2-methyl-1-butanol which proceeds in 26 steps and 0.003%
overall yield. This new reaction provides a powerful tool in streamlining this synthesis and should prove
more generally useful.
A number of biologically interesting natural products possess
Scheme 1. Retrosynthetic Analyses of Alternaric Acid
1
the diene fragment I whose access normally involves multistep
olefination protocols. A particularly attractive approach to such
natural products derives from an equivalent of an Alder ene
type reaction between a terminal alkene II and a terminal alkyne
2
III. While such intermolecular reactions, with unactivated
substrates, do not proceed thermally, we have developed a Ru-
catalyzed process that effects such reactions. To explore the
scope and limitations of this process as well as its ability to
facilitate syntheses of natural products, we chose alternaric acid
(
1) as our target (Scheme 1). This natural product, initially
3,4
isolated in 1949 from Alternaria solani, exhibits extreme
specificity for inhibition of germination in certain strains of fungi
at 100 nM concentrations.5 Some fungal strains exhibit stunting
of their hyphae at 5-10 nM concentrations. In addition to
antifungal activities, alternaric acid displays selective phytotoxic
(
1) For a recent class of interesting antitumor macrolides bearing this
unit, the amphidinolides, see: Kobayashi, J.; Ishibashi, M. Chem. ReV. 1993,
3, 1753. Ishibashi, M.; Kobayashi, J. Heterocycles 1997, 543. Also, see:
9
Ishibashi, M.; Takahashi, M.; Kobayashi, J. Tetrahedron 1997, 53, 7827.
Kobayashi, J.; Yamaguchi, N.; Ishibashi, M. J. Org. Chem. 1994, 59, 4698.
(2) (a) Trost, B. M.; Indolese, A. F.; Muller, T.; Treptow, B. J. Am.
Chem. Soc. 1995, 117, 615. (b) Trost, B. M.; Martinez, J. A.; Kulawiec, R.
J.; Indolese, A. J. Am. Chem. Soc. 1993, 115, 10402. (c) Trost, B. M.;
Indolese, A. J. Am. Chem. Soc. 1993, 115, 4361.
activities as well.6 This effect derives from an increased rate
of transpiration compared to water uptake resulting in dehydra-
tion.
(3) Brian, P. W.; Curtis, P. J.; Hemming, H. G.; Unwin, C. H.; Wright,
J. M. Nature 1949, 164, 534. For structural studies, see: Tabuchi, H.;
Ichihara, A. Tetrahedron Lett. 1992, 33, 4933. Bartels-Keith, J. R. Chem.
Soc. 1960, 1662, 860. Grove, J. F. J. Chem. Soc. 1952, 4056.
The magnitude of the challenge is illustrated by the one extant
synthesis which required 29 steps with an overall yield of less
7
(
4) Also, see: Tabuchi, H.; Ichihara, A. J. Chem. Soc., Perkin Trans. 1
994, 125.
5) Brian, P. W.; Curtis, P. J.; Hemming, H. G.; Jefferys, E. G.; Unwin,
C. H.; Wright, J. M. J. Gen. Microbiol. 1951, 5, 619.
than 0.001%. The development of our Ru-catalyzed Alder ene
1
(
(6) Brian, P. W.; Elson, G. W.; Hemming, H. G.; Wright, J. M. Ann.
Appl. Biol. 1952, 39, 308.
S0002-7863(98)01540-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/29/1998

Ru-Catalyzed Alder Ene Type Reactions
J. Am. Chem. Soc., Vol. 120, No. 36, 1998 9229
type reaction suggested two convergent strategies. Path a of
Scheme 1 envisions linking the two halves 2 and 3 by creating
the branched 1,4-diene unit using our Ru reaction. The latter
half 3 would derive from standard C-acylation of the dihydro-
pyrone. The pyrone 5 requires just three steps from com-
mercially available poly (R)-3-hydroxybutanoate and tert-butyl
acetate.7 Path b follows the strategy of Ichihara et al. Their
are catalyst inhibitors. These model studies focused our efforts
on the synthetic strategy of path b rather than path a of Scheme
1.
Synthesis of Alkene Partner
,8
The commercial availability of the alcohol 12 in enantio-
9
synthesis of the diene fragment 6 employed the Julia olefination
merically pure form led us to consider olefination of the
and required 26 steps with an overall yield of 0.003%. The
existence of the Ru-catalyzed reaction suggests the ene 2 and
yne 4 as precursors. Either path ultimately requires the same
three fragments, 2, 4, and 5. If the Ru-catalyzed reaction is
successful, the conciseness of the route depends on the efficiency
of the synthesis of 2.
1
1
corresponding aldehyde 13 followed by asymmetric dihy-
droxylation to create the three stereogenic centers of 2. As
1
2
shown in eq 3, the reaction of the ylide 14 gave the desired
E-alkene 15 as the only geometric isomer in a rather slow
reaction that required 3.5 days in refluxing chloroform. Un-
Model Studies
To probe which of the two strategies would prove most
promising, we performed several model studies. Lactic acid
was converted to the allylated derivative 7 by simple alkylation.
1
0
Reacting the acid 7a with tert-butyl 4-pentynoate (4a) in the
presence of 5 mol % of the ruthenium complex 8 gave only a
small amount of the adduct 9 (R ) H, R′ ) t-C4H9) (eq 1). On
fortunately, significant racemization accompanied this reaction.13
The anion derived from the phosphonate 16 (eq 4) underwent
reaction at -78° to room temperature but gave an E:Z ratio of
the other hand, the corresponding methyl ester 7b, under
identical conditions, gives a 49% GC yield (46% isolated yield)
of the diene 9a as a single isomer. Performing the same reaction
of 7b with the acid 4b causes a lower conversion to the acid
17 and 18, depending upon the phosphonate, ranging from 1:1.2
for 16a to 1:4 for 16b or 16c. In the case of 16a, KHMDS in
THF-toluene at -78° (87% yield) was employed. For 16b,
9b. Working up the reaction by esterfication with diazomethane
14
the Roush-Masamune conditions (DBU or Hunig’s base,
led to a 20% isolated yield of diester 9c.
LiCl, CH3CN, room temperature, 28-58% yields) were used.
The yields in these reactions were not optimized since the ratios
were deemed unsatisfactory.
These results suggest that carboxylic acids can inhibit the
reaction but that free hydroxyl groups are well tolerated. To
probe this point further, we prepared alkyne 10 by standard
acylation of dihydropyrone 5 with acid 4b as a substrate. With
An alternative synthesis involved oxidation and in situ
olefination with the parent Wittig reagent as shown in eq 5. In
our hands, oxidation of the alcohol was best performed using
1
0 mol % of complex 8 as catalyst, the reaction gave a 21%
yield of the desired adduct 11 as a single regioisomer (eq 2).
1
5
the Moffatt-Swern method. Addition of the stabilized parent
Wittig reagent gave the alkene 19 without racemization as shown
1
6
by comparison to the literature and by subsequent analysis of
17
2
4. Bromination-dehydrobromination gave nearly a quantita-
tive yield of the vinyl bromide 20 as a single geometric isomer.
1
8
Stille cross-coupling generated 17b as a single geometric
isomer whose geometry is readily established by NMR spec-
Again, only a low conversion can be realized. It appears that
relatively acidic substrates such as carboxylic acids and
acyldihydropyrones that can generate good coordinating anions
(
11) Anelli, P. L.; Montanari, F.; Quici, S. Org. Synth. 1990, 69, 212.
12) Prepared in 67% yield by alkylation of methyl (triphenylphospho-
(
ranylidene)acetate with propargyl bromide in ethyl acetate at reflux.
(13) Determined by reduction of the ester of 15 (LAH) to form the
primary alcohol and formation of the O-methylmandelate ester (DCC,
DMAP) which showed a 1.2:1 ratio of diastereomers.
(14) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfield, A. P.;
Masamune, S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183.
(15) Mancuso, A. J.; Juang, S. L.; Swern D. J. Org. Chem. 1978, 41,
2480.
(
7) Tabuchi, H.; Hamamoto, T.; Miki, S.; Tejima, T.; Ichihara, A. J.
Org. Chem. 1994, 59, 4749. Tabuchi, H.; Hamamoto, T.; Miki, S.; Tejima,
T.; Ichihara, A. Tetrahedron Lett. 1993, 34, 2327. For review, see: Ichihara,
A. J. Synth. Org. Chem. Jpn. 1995, 53, 975.
(
8) Hoffman, R. W.; Dressely, S. Angew. Chem., Int. Ed. Engl. 1986,
5, 189.
9) Kocienski, P. Phosphorus Sulfur 1985, 24, 97. Simpkins, N. S.
Sulphones in Organic Synthesis; Pergamon Press: New York, 1993.
10) Atkinson, R. S.; Grimshire, M. J. J. Chem. Soc., Perkin Trans. 1
986, 1215.
2
(
(16) Baker, R.; Head, J. C.; Swain, C. J., J. Chem. Soc., Perkin Trans.
1 1988, 85. Guerriero, A.; D’Ambrosio, M.; Cuomo, V.; Vanzanella, F.;
Pietra, F. HelV. Chim. Acta 1989, 72, 438.
(
1

9230 J. Am. Chem. Soc., Vol. 120, No. 36, 1998
Trost et al.
17a
troscopy since both geometric isomers were available by the
Emmons-Wadsworth-Horner reaction.
These substrates were synthesized to test the chemoselectivity
of the osmium-catalyzed dihydroxylations. Surprisingly, enyne
15 was inert toward all catalytic dihydroxylations, both achiral
and chiral. The literature suggests that trisubstituted alkenes
bearing a carbonyl group are readily dihydroxylated; whereas,
accompanied by desilylation by addition of pyridine‚hydrofluoride.
The primary alcohol completed its role as an alkene surrogate
and was eliminated to form the monosubstituted alkene 25 via
the Grieco method. Acid solvolysis removed the acetonide
to provide the desired alkene substrate 2. In this way, substrate
19
monosubstituted alkenes are not. However, dihydroxylation
of 17b led only to reaction at the nonconjugated alkene.
Reduction of the electron-withdrawing nature of the ester by
conversion to a carboxylate salt had no effect on this chemose-
lectivity.
21
2
is available in 8-steps from commercially available alcohol.
In examining the sequence, the employment of the acetonide
The fact that 19 undergoes asymmetric dihydroxylation
without difficulty and that such diols can also be accessed by
aldol strategies led us to examine Wittig and Claisen rearrange-
ments as shown in eqs 6 and 7, respectively. In neither case
were the desired products observed.
was to maintain differentiation among the three hydroxyl groups
present in 26. However, chemoselective substitution of the
primary alcohol to form the selenide should be feasible in the
presence of the secondary and tertiary alcohols. Indeed,
exposing the triol to the standard conditions proceeds smoothly
to give the diol 2 in 77% overall yield (eq 10). The net result
is eliminating one step and providing alkene 2 in only seven
steps from commercially available alcohol 12 although the
overall yield of 45-50% was virtually unchanged.
A straightforward resolution of this impasse used a primary
alcohol as a surrogate for the monosubstituted alkene. Suzuki
cross-coupling of the organoborane derived from the tert-
butyldimethylsilyl ether of allyl alcohol 21 with vinyl bromide
The alkyne partner 4b was originally synthesized from
propargylated malonate by hydrolysis and decarboxylation. A
convenient alternative involving fewer steps proved to be the
direct alkylation of the enolate of tert-butyl acetate with
propargyl bromide (eq 11) to give the tert-butyl ester 4a in 66%
yield, which was easily solvolyzed (TFA, CH2Cl2, 74%) to form
the acid 4b. This acid is also commercially available. The
methyl ester 4c was prepared by esterification of the acid with
diazomethane.
2
0
2
0 gave 22 as a single geometric isomer in 88% isolated yield
(eq 8). Use of cesium carbonate as base or DMF as solvent
then
one pot
gave poor results. Asymmetric dihydroxylation proceeded
uneventfully to give an 89% yield of diol 23 with a diastere-
omeric excess (de) of 98% (eq 9). Acetonide formation was
(17) Converting 19 derived from both (S)-12 and racemic 12 to 24 allows
ready analysis of the diastereomeric purity which, since the asymmetric
dihydroxylation proceeds with 98% ee and translates into the enantiopurity
of 19.
Ru-Catalyzed Coupling
(
18) Stille, J. K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508.
Initial attempts to perform the Alder ene reaction focused on
the protected form of the ene component 25 with the tert-butyl
ester 4a. Heating an approximately 1:1 mixture of the two
substrates with 10 mol % of complex 8 in 1:1 DMF-water at
(
19) Kolb, H. C.; VanNieuwenhze M. S.; Sharpless, K. B. Chem. ReV.
1
1
994, 94, 2483. Andersson, P. G.; Sharpless, K. B. J. Am. Chem. Soc. 1993,
15, 7047.
(20) Oh-e, T.; Miyaura, N.; Suzuki, A. J. Org. Chem. 1993, 58, 2201.
Johnson, C. R.; Braun, M. P. J. Am. Chem. Soc. 1993, 115, 11014. Miyaura,
N.; Ishiyama, T.; Sasaki, H.; Ishikawa, M.; Suzuki, A. J. Am. Chem. Soc.
(21) Grieco, P. A.; Gilman, S.; Nishizawa, M. J. Org. Chem. 1976, 41,
1
989, 111, 314.
1485.

Ru-Catalyzed Alder Ene Type Reactions
J. Am. Chem. Soc., Vol. 120, No. 36, 1998 9231
Table 1. Ru-Catalyzed Reaction of Acetonide Ene 25 and Alkynes 4a
a
All reactions were run at 0.1 M under nitrogen. Determined by H NMR spectroscopy. c 1:1 ratio. b 1:3 ratio. e Reaction performed at 7-8
b
1
f
kbar at room temperature. CpRu(Ph
3
P)(2-methallyl) was used at catalyst.
1
1
00 °C gave a 26% yield of a 1.6:1 ratio of 27a:28a (see Table
, entry 1). Replacing DMF-water with 1:1 methanol:water
conditions similar to entry 6 of Table 1, except that pure
methanol was employed (see Table 2, entry 1), a nearly
quantitative yield (based upon recovered starting material, brsm)
of the adducts was obtained. Furthermore, the branched-to-
linear ratio improved to 6.1. The problem that remained was
conversion. In all cases, significant quantities of unreacted
starting material remained although it was not always recovered.
In those cases where it was recovered, the yields based upon
recovered starting material were always high and frequently
quantitative. Adding a second batch of catalyst after 15 min
increased the yield to 49% and the 27a:28a ratio to 2.2 (Table
1
1
, entry 2). The reaction appears to be sensitive to pH (Table
, entries 2 and 4-6). The yield increases with increasing
acidity. Stronger acids were not explored because of the
assumed sensitivity of the product to acid, but may be generally
beneficial. The effect of chloride ion was explored by the
addition of strong (Table 1, entries 7 and 8) and moderate (Table
1
, entry 9) chloride scavengers. Except for silver sulfate, these
(Table 2, entry 2) nearly doubled the conversion and retained
did not influence the reaction. The effect of silver sulfate may
be related to pH since addition of a pH 7 phosphate buffer to
the reaction of entry 4 saw the yield plummet to 16%. Thus,
under the best reaction conditions (Table 1, entries 6 and 8), a
the excellent yield. On the other hand, initiating the reaction
with twice the amount of catalyst did not significantly improve
the turnover (Table 2, entry 3 vs 1). Addition of indium triflate
improved the branched-to-linear ratio but at the expense of
conversion (Table 2, entry 4). A significant difference occurred
upon performing the reaction under high pressure at ambient
temperature (Table 2, entry 5) which nearly doubled the
branched-to-linear ratio. The yield was high based upon total
mass recovery but was complicated by the fact that some
transesterification occurred to produce the methyl ester 29c. The
methyl ester 4c participated in a fashion similar to the tert-
butyl ester 4a as shown in entries 6 and 7, Table 2. The slightly
higher regioselectivity may derive from a steric effect.
5
8% yield of the desired ene type adducts were obtained but,
surprisingly, with disappointing regioselectivity.
Because of the rather polar hydroxylic nature of the solvent,
we hypothesized that the highly lipophilic substrate 25 may be
adopting conformations that attenuated the regioselectivity. The
use of alcoholic solvents suggested that the free diol itself 2
would be an acceptable substrate. Table 2 summarizes the
results of its ruthenium-catalyzed reaction with various 4-pen-
tynoate acceptors to give the adducts 29 and 30. Using

9232 J. Am. Chem. Soc., Vol. 120, No. 36, 1998
Trost et al.
Table 2. Ru-Catalyzed Reaction of Diol Ene 2 and Alkynes 4a
entry
1
2
alkyne
ratio 4a:2
8 (%)
12.0
2 32 × 10.9
10.8
10.9
10.1
10.9
10.6
12.3
14.5
15.4
11.7
18.5
16.7
additive
temp.
time (h)
isolated yield (%)
ratio 29:30
4a
4a
4a
4a
4a
4c
4c
4d
4d
4d
4d
4d
4d
4d
1.06
1.05
1.20
1.22
1.14
1.52
1.00
0.96
1.02
1.03
1.17
1.11
1.20
1.36
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
4
4
4
4
4
4
4
4
4
4
4
4
4
4
PF g
6
reflux
reflux
reflux
reflux
rt
reflux
reflux
reflux
rt
rt
rt
rt
rt
2
2
2
2
24
2
2
2.5
2
1
1
35(99)d
60(100)d
41
6.1
4.9
6.7
8.8
12.5
7.4
7.0
4.9
6.7
6.9
5.4
7.7
8.9
5.2
c
g
g
g
g
f
PF
6
PF
6
PF
6
PF
6
PF
6
PF
6
PF
6
PF
6
PF
6
PF
6
PF
6
PF
6
PF
6
3
4
5
+ In(OSO
2
CF
)
3 3
25
e
65(100)
25(79)
33
6
7
8
g
52 d
h
9
47(90)
58(92)d
48
i
1
1
1
1
1
0
1
2
3
4
i
+ CF
3
CO
2
H
i,j
j,k
k,l
1
1.5
1.5
39(55)
51
rt
33
a
All reactions were run at 0.1 M in methanol. b Determined by H NMR spectroscopy. c The reaction was initially charged with 10.9 mol % 8
1
d
and no NH
4
PF
6
; after 15 min, an additional 10.9 mol % 8 and 40 mol % NH
4
PF
6
were added. Yields in parentheses based upon recovered starting
e
f
g
h
material. Reaction performed at 8-9 kbar. 5.9 equiv relative to catalyst used. 2.2 equiv relative to catalyst used. Reaction performed at 5
kbar. Reaction performed at 7-8 kbar. Reaction run in 1:1 acetone:methanol. Reaction performed at 11-13 kbar. Reaction performed in acetone.
i
j
k
l
Having both the methyl and tert-butyl esters of the acyclic
fragment available, we only require the acid 27b to complete
the formal synthesis since the coupling with the dihydropyrone
and hydrolysis that completes the synthesis has been previously
accomplished in only three steps. Formation of the acetonides
(eq 11). The SEM ester did not survive the ruthenium-catalyzed
addition. On the other hand, the fmoc ester 4d did participate
(see Table 1, entries 12-14, and Table 2, entries 8-14).
Studies with the acetonide alkene 25 under standard condi-
tions (Table 1, entry 12) gave a low conversion resulting in a
low isolated yield of the desired adducts 27d and 28d.
Interestingly, the branched-to-linear ratio improved by about a
factor of 2. Given our earlier success with high pressure, we
examined its effect here (Table 1, entry 13). While the
conversion improved, even though we were operating at room
temperature, the regioselectivity was unchanged.
2
7a and 27b proceeds uneventfully under standard conditions
(eq 12). Surprisingly, neither could be hydrolyzed in our hands
The diol alkene 2 proved to be a more satisfactory substrate
in the reactions of the fmoc ester 4d as in all other cases. An
excellent yield at about 50% conversion is obtained under almost
all conditions. The bulky fmoc ester does appear to have a
slight effect on regioselectivity causing it to diminish somewhat
(Table 2, entry 8). The reaction, when performed under high
pressure, proceeds at ambient temperature (Table 2, entries 9
and 10). While the regioselectivity improved somewhat, the
effect was much smaller than that seen with the tert-butyl ester
(Table 2, entry 5). Increasing acidity by adding a trace of
trifluoroacetic acid decreased the selectivity (Table 2, entry 11).
Decreasing the polarity of the solvent by using a 1:1 mixture
of acetone-methanol caused a small increase in regioselectivity
to the desired acid 27b. In the case of the tert-butyl ester, acid
conditions led to decomposition. Cleavage of the methyl ester
by either base or by dealkylation suffered from low chemose-
lectivity between the two esters.
We, therefore, examined the ruthenium-catalyzed addition of
4-pentynoic acid with the alkene partner. The extraordinary
polarity of a diol carboxylic acid led us to examine the reaction
of the acetonide 25 with this acid to form 27b directly. As
shown in Table 1, entry 10, the reaction proceeded in somewhat
lower yield than the tert-butyl ester (cf entry 6) and the
regioselectivity was still poor. Running at high pressure (Table
(Table 2, entry 12) which further increased by raising the
pressure to about 13 kbar (Table 2, entry 13). On the other
hand, using only acetone as solvent caused the regioselectivity
to diminish (Table 2, entry 14).
1
, entry 11) did not have a favorable effect on the reaction.
Since the substrates 27a and 27c showed sensitivity to acid
and the lack of chemoselectivity with nucleophilic base, we
considered the use of an ester that could be removed selectively
with a nonnucleophilic base. For this purpose, we chose the
trimethylsilylethyl (SEM) and 9-fluorenylmethyl (fmoc) esters.
Coupling 4-pentynoic acid with 2-trimethylsilylethanol and
With the fmoc ester of the intact acyclic fragment of alternaric
acid in hand, the corresponding acetonide 27d (eq 12) was
formed in standard fashion, setting the stage for the final step,
the deesterification. Subjecting the fmoc ester 27d to piperidine
in methylene chloride led to smooth cleavage to the free acid
27b in nearly quantitative yield (eq 13).
9
-fluorenylmethanol with DCC and DMAP gave the corre-
sponding esters 4e and 4d in 85% and 71% yields respectively

Ru-Catalyzed Alder Ene Type Reactions
J. Am. Chem. Soc., Vol. 120, No. 36, 1998 9233
chagrin, the answer is that excellent chemoselectivity resides
in the opposite direction of our requirement. This result raises
the question of whether a ligand environment can be found for
osmium to impart more nucleophilicity and thereby reverse such
a chemoselectivity. That the trisubstituted double bond bearing
the ester functionality can participate in the asymmetric dihy-
droxylation is illustrated by the reaction of 22, which proceeds
in good yields and with excellent reagent controlled diastereo-
selectivity.
It is interesting to note that the acetonide is not required for
the synthesis of the acyclic fragment. The chemoselectivity of
the dehydration sequence (eq 10) and the ruthenium-catalyzed
addition of alkene 2 with alkyne 4 do not require such protection.
Thus, the acyclic fragment is actually available in only nine
linear steps. Indeed, one of the strengths of the ruthenium
reaction is its chemoselectivity, and the substrates examined
herein illustrate that point.
Scheme 2 outlines the mechanistic rationale for the ruthenium-
catalyzed addition. The equivalent of a triply coordinatively
unsaturated ruthenium cationic complex is generated by the
ionization of the chloride and the removal of the COD ligand
by a [2 + 2 + 2] cycloaddition with the alkyne. The result of
this activation event is to consume an amount of the alkyne
equivalent to the amount of the ruthenium complex. Thus, the
alkyne:alkene ratios depicted in the tables must be adjusted to
take into account the consumption of the alkyne. For example,
in Table 1, entry 1, the adjusted 4:25 ratio is 0.94 and the alkyne
is the limiting substrate. The regioselectivity is determined by
the ratio of ruthenacycle 33 (leading to branched product) to
ruthenacycle 34 (leading to linear product). Normally, regioi-
somer 33 dominates, presumably because it minimizes steric
congestion around the ruthenium compared to 34. On the other
hand, in the present case, the ester is poised to occupy the open
coordination site on ruthenium otherwise occupied by solvent
Discussion
New methodology provides an opportunity to consider
synthetic strategies that formerly did not exist. The examination
of the structural fragment I present in natural products normally
led to a retrosynthetic analysis based upon olefination protocols
such as the Wittig, Peterson, and Julia olefinations. Application
of such methods suffers from the need to juggle the functionality
in order to make it compatible and the requirement of multiple
steps with dienes such as I. The synthesis of alternaric acid
highlights this point. The creation of the ruthenium-catalyzed
reaction led to a synthesis of the acyclic unit in 10 linear steps
and 38% overall yield from commercially available (S)-2-
methyl-1-butanol and the fmoc ester of 4-pentynoic acid, the
latter available in one step from the commercially available acid
in 71% unoptimized yield. Thus, from commercially available
materials, a total of 11 steps and 27% overall yield were required
as a result of the availability of the ruthenium-catalyzed reaction.
In fact, we can cut one step from our sequence by simply adding
triethylamine to the in situ generated dibromide to make it a
total of 10 steps from commercially available materials or nine
linear steps. The fact that the yield for the one-step conversion
of alkene to vinyl bromide was in the 70-80% range compared
to quantitative for the two-step process led us to favor the latter.
It contrasts quite favorably with the previously recorded
synthesis (vide supra) which also employed (S)-2-methyl-2-
butanol as one starting material but relied upon established
methodologies. Our route requires only about one-third the
number of steps.
(or some other external ligand) in 33 and 34 as depicted in 35.
Further reduction of the number of linear steps were thwarted
by racemization during olefination and by the reverse chemose-
lectivity in the dihydroxylation of the diene 20b. To reduce
the length of the longest linear sequence, we explored the use
of alkylated stabilized olefination agents. With the Wittig
reagent 14 (eq 3), addition to (S)-2-methylbutanal gave nearly
racemic product. Since the addition of ylide 14 to the aldehyde
was very slow, we believe that its low nucleophilicity led to
the domination of an acid-base reaction which effects racem-
ization. Using the less stabilized phosphonate anions might
resolve this problem. Unfortunately, the enolates from 16a-c
led only to E-Z olefinic mixtures. Since we suspected that
other alkyl-bearing olefinating agents would show similar
behavior, they were not pursued. Work by the Sharpless group
established that monosubstituted alkenes react at a significantly
This competitive internal coordination compromises the intrinsic
preference for branched product leading to a low selectivity for
the reactions with acetonide 25. Removal of the acetonide then
leads to the prospect of internal coordination with the func-
tionality present in the alkene. Either hydroxyl group (via 36
or 37) or even the ester (via 38) may compete with the alkyne
ester for the open coordination site on ruthenium, thereby
restoring the selectivity for the branched product. Since ester
coordination as in 38 is also possible in the acetonide 25, this
motif may seem less likely although steric hindrance imposed
by the acetonide may account for the failure of the methoxy-
carbonyl group to effectively participate in this case. Increasing
pressure should increase participation by the oxygens as depicted
in 36-38. Since it is reasonable to expect that transition states
involving such coordination would have lower volumes of
activation, an increased selectivity for the branched product
19
slower rate than trisubstituted alkenes. Further, they report
that R,â-unsaturated esters are excellent substrates.22 While the
electron-withdrawing nature of the ester function decreases the
reactivity of the trisubstituted double bond, the question of
whether the loss in reactivity would make it slower than a
monosubstituted double bond was not answered. To our
(22) Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.;
Hartung, J.; Jeong, K.; Kwang, H.; Morikawa, K.; Wang, Z.; Xu, D.; Zhang,
X. J. Org. Chem. 1992, 57, 2768.

9234 J. Am. Chem. Soc., Vol. 120, No. 36, 1998
Trost et al.
Scheme 2. Mechanistic Rationale of Ru-Catalyzed Addition
would be expected and is observed. In support of this
interpretation, high-pressure had no effect on the regioselectivity
of the addition of the acetonide 25 which precludes such
coordination. Exploitation of these coordination effects con-
stitutes an important direction for further study.
butyl 4-pentynoate (4a) in 2.5 mL of water and 1.5 mL of methanol
was added 7.8 mg (0.025 mmol) of CpRu(COD)Cl in 1 mL of methanol.
The mixture was heated at 70 °C for 8 h. Water (5 mL) was added,
and the mixture was extracted with ether (3 × 15 mL). The organic
phase was washed with brine (3 × 2 mL), dried (MgSO
4
), and
evaporated in vacuo. Chromatography (ethyl acetate:hexane ) 1:4)
of the residue yielded 68 mg (46%) of the product 9a. IR (neat): 3512,
The ruthenium-catalyzed additions of alkenes and alkynes
provide a powerful new paradigm for the construction of
complex organic molecules. The example of alternaric acid is
an excellent illustration of its potential impact. The strength
of the methodology stems to a considerable extent from its
extraordinary chemoselectivity. Except for basic amines (which
can be “protected” as an amide or by protonation), and
phosphines, no other groups have yet been found to be
incompatible. Divalent sulfur, notorious as a catalyst poison,
-1
1
1
(
3
731, 1647, 1452, 1437 cm
td, J ) 6.9, 15.4 Hz, 1H), 5.64 (d, J ) 15.4 Hz, 1H), 4.75 (s, 2H),
.78 (s, 3H), 2.75 (d, J ) 6.9 Hz, 2H), 2.38-2.24 (m, 4H), 1.48 (s,
3
. H NMR (300 MHz, CDCl ): δ 5.84
13
3H), 1.43 (s, 9H). C NMR (300 MHz, CDCl ): δ 176.23, 172.50,
3
146.30, 133.36, 128.17, 110.61, 80.25, 74.26, 53.00, 38.91, 33.77, 30.99,
26 5
28.08, 25.94. Anal. Calcd for C16H P : C, 64.41; H, 8.78. Found:
C, 64.42; H, 8.54.
(E)-(6R,8′R)-5,6-Dihydro-4-hydroxy-6-methyl-3-(8-(methoxycar-
bonyl)-4-methylene-8-hydroxy-6-nonenol)pyrane-2-one (11). To a
solution of 64.5 mg (0.31 mmol) of alkene 7b and 44.6 mg (0.31 mmol)
of compound 10 in 1.5 mL of water and 1.5 mL of methanol was added
23
does not interfere. The compatibility with aqueous media also
bodes well for the prospect of employing just water as solvent
with appropriate substrates, thereby making this reaction more
environmentally benign. The ability of the reaction to form
the new 1,2-disubstituted double bond adjacent to a quaternary
center is noteworthy since the reaction has shown extreme
sensitivity to steric factors with respect to substitution on the
alkene. While many aspects of the reaction remain to be
explored, the current study helps establish its utility even in its
current stage of development.
9
.6 mg (0.031 mmol) of CpRu(COD)Cl, and the mixture was heated
at 70 °C for 1 h. Brine (3 mL) was added, and the mixture was
extracted with ether (4 × 3 mL). The organic phase was dried (MgSO
4
)
and evaporated in vacuo. After purification by chromatography (ethyl
acetate:hexane ) 1:1 and dichloromethane:methanol ) 95:5) 23 mg
28
(
(
21%) of product 11 was obtained. [R]D ) 24.8 (c ) 2.0, CHCl
3
). IR
-1
1
neat): 3506, 1737, 1715, 1573, 1558, 1454 cm
. H NMR (300 MHz,
CDCl ): δ 5.85 (td, J ) 6.8, 15.4 Hz, 1H), 5.66 (d, J ) 15.4 Hz, 1H),
3
4
.80-4.77 (m, 2H), 4.58-4.47 (m, 1H), 3.78 (s, 3H), 3.31 (s, 1H),
3.27-3.07 (m, 2H), 2.79 (m, 2H), 2.67-2.63 (m, 2H), 2.35 (m, 2H),
.49 (s, 3H), 1.46 (d, J ) 6.4 Hz, 3H). 13C NMR (300 MHz, CDCl
):
δ 203.81, 194.50, 176.20, 164.22, 146.03, 133.54, 128.08, 111.28,
03.13, 74.30, 70.32, 52.98, 39.15, 38.79, 37.01, 30.58, 25.89, 20.59.
Anal. Calcd for C18 C, 61.35; H, 6.87; MW, 352.1522.
Found: C, 61.10; H, 6.96; MW, 352.1518.
2Z,4S)-Methyl-2-bromo-4-methylhex-2-enoate (20a). Bromine
0.07 mL, 1.36 mmol) was slowly added dropwise to 19 (142 mg, 0.999
mmol) in 2 mL of methylene chloride at 0 °C. After 2 h of stirring,
the solution was diluted with saturated sodium thiosulfate, extracted
with diethyl ether, the combined organic extracts were dried over
magnesium sulfate, filtered, and concentrated in vacuo. The residue
was diluted with 2 mL of methylene chloride. Triethylamine (0.70
mL, 5.02 mmol) was then added, and the reaction was stirred for 24 h.
The heterogeneous mixture was concentrated under reduced pressure,
and the residue was flash chromatographed with 49:1 pentane:diethyl
Experimental Section
1
3
Reactions were generally conducted under a positive pressure of dry
nitrogen within flame-dried glassware. Reactions were sealed with red
rubber septa and magnetically stirred. THF and diethyl ether were
distilled from sodium/benzophenone ketyl prior to use. Methylene
chloride and acetonitrile were distilled from calcium hydride prior to
use. Methanol was distilled from magnesium methoxide prior to use.
Common reagents and materials were purchased from commercial
sources and purified by recrystallization or distillation. Anhydrous
solvents and reaction mixtures were transferred by oven-dried syringe
or cannula. Flash chromatography employed ICN silica gel (Kiesselgel
1
24 7
H O :
(
(
6
0, 230-400 mesh). Analytical TLC was performed with 0.2 mm
coated commercial silica plates (E. Merck, DC-Platten Kieselgel 60
254). Melting points were determined on a Thomas-Hoover oil bath
F
apparatus and were not corrected. Analytical gas chromatography was
performed on a Varian star 3600 gas chromatograph with a 10 m ×
ether as the eluant to yield 218 mg (99% yield) of a clear liquid, R )
f
0
.25 mm poly(dimethylsiloxane) column.
28
0
.42 (19:1 hexanes:ethyl acetate), [R]D ) +17.88 (c ) 10.0, toluene).
When this reaction was performed using 5.92 g (41.6 mmol) of 19,
.6 mL (50.5 mmol) of bromine in 84 mL of methylene chloride
tert-Butyl (E)-8-Hydroxy-8-(methoxycarbonyl)-4-methylene-6-
nonenoate (9a). To a solution of 72.1 mg (0.5 mmol) of methyl
-hydroxy-2-methyl-4-pentenoate (7b) and 77.1 mg (0.5 mmol) of tert-
2
2
followed by 11.7 mL (83.9 mmol) of triethylamine in 210 mL of
(23) Trost, B. M.; Shi, Z.-P. J. Am. Chem. Soc. 1994, 116, 7459.
methylene chloride, 7.09 g (77% yield) of 20a was obtained. IR (thin

Ru-Catalyzed Alder Ene Type Reactions
J. Am. Chem. Soc., Vol. 120, No. 36, 1998 9235
film): 1735, 1623, 1459, 1436 cm-1. 1H NMR (300 MHz, CDCl
): δ
(4S,5R,1′S)-2,2-Dimethyl-4-(methoxycarbonyl)-5-(1′-methylpro-
pyl)-4-(prop-2-enyl)-1,3-dioxolane (25). Tri-n-butylphosphine (3.4
mL, 13.6 mmol) was added to a solution of 24 (757.5 mg, 2.76 mmol)
and o-nitrophenylselenocyante (3.08 g, 13.6 mmol) in 28 mL of THF,
and the solution immediately turned dark brown. After 12 h of stirring,
sodium bicarbonate (1.1429 g, 13.6 mmol) was added followed by the
addition of 30% hydrogen peroxide (2.8 mL, 27.4 mmol). After 8 h
of stirring, the heterogeneous mixture was diluted with 10% aqueous
hydrochloric acid, extracted with diethyl ether, the combined organic
extracts were dried over magnesium sulfate, filtered, and concentrated
under reduced pressure. The residue was flash chromatographed with
19:1 pentane:diethyl ether as the eluant to yield 640 mg (90%) of a
3
7
1
.08 (d, J ) 9.6 Hz, 1H), 3.83 (s, 3H), 2.67 (m, 1H), 1.44 (m, 2H),
13
.05 (d, J ) 6.9 Hz, 3H), 0.90 (t, J ) 7.5 Hz, 3H). C NMR (300
): δ 164.83, 153.28, 116.12, 54.67, 39.95, 30.23, 19.83,
3.09. Anal. Calcd for C : C, 43.46; H, 5.93. Found: C,
3.39; H, 6.00.
2E,4S)-Methyl-2-(3-tert-butyldimethylsiloxypropyl)-4-methylhex-
-enoate (22). Borane dimethyl sulfide (10 M, 200 µL, 2.00 mmol)
MHz, CDCl
1
4
3
8 2
H13BrO
(
2
was added to a solution of 1,5-cyclooctadiene (245 µL, 2.00 mmol) in
mL of THF at 0 °C. The reaction was subsequently heated to reflux
4
for 2.5 h. After cooling to room temperature, 21 (336.2 mg, 1.95 mmol)
was added and the reaction was stirred for 22 h at which time tribasic
potassium phosphate (641.8 mg, 3.02 mmol) and bis-1,1′-(diphe-
nylphosphino)ferrocenepalladium(II) chloride (36.6 mg, 0.0500 mmol)
was added. The mixture was heated to reflux during which time it
became dark. Vinyl bromide 20a (206.9 mg, 0.936 mmol) was added,
and the mixture was heated at reflux for 13 h. The mixture was diluted
with water, extracted with diethyl ether, the combined organic extracts
were dried over magnesium sulfate, filtered, and concentrated in vacuo.
The residue was flashed chromatographed with 49:1 pentane:diethyl
23
clear liquid, R ) 0.50 (9:1 hexanes:ethyl acetate), [R] ) -54.97 (c
f
D
-
1
1
) 3.8, CH Cl ). IR (thin film): 1740, 1642, 1457, 1437 cm
2
2
.
H
NMR (300 MHz, CDCl ): δ 5.84 (m, 1H), 5.13 (m, 1H), 5.09 (s, 1H),
3
3.91 (d, J ) 9.0 Hz, 1H), 3.74 (s, 3H), 2.65 (dd, J ) 13.8 Hz, 7.2 Hz,
1H), 2.35 (dd, J ) 13.8 Hz, 6.9 Hz, 1H), 1.75 (m, 1H), 1.54 (m, 2H),
1.47 (s, 3H), 1.44 (s, 3H), 1.06 (m, 1H), 1.03 (d, J ) 6.6 Hz, 3H),
13
0.88 (t, J ) 7.2 Hz, 3H). C NMR (300 MHz, CDCl ): δ 173.33,
3
133.01, 118.68, 108.84, 85.04, 84.89, 52.35, 37.05, 33.84, 27.79, 25.13,
16.12, 10.69. Anal. Calcd for C H O : C, 65.60; H, 9.44. Found:
ether as the eluant to yield 257.7 mg (88%) of a clear liquid, R
f
) 0.44
2
14
24
4
2
8
(
19:1 hexanes:ethyl acetate), [R] ) +17.87 (c ) 10.0, CH Cl ).
2
C, 65.71; H, 9.19.
D
Performing this reaction with 5.84 g (26.4 mmol) of 20a, 9.10 g (52.8
mmol) of 21, 53.0 mmol of 9-BBN-H, and 0.98 g (1.34 mmol) of bis-
(
4S,5R,6S)-4,5-Dihydroxy-4-(methoxycarbonyl)-6-methyl-1-
octene (2). Method A: Compound 25 (0.68 g, 2.65 mmol) was
dissolved in 4 mL of dichloromethane, 4 mL of trifluoroacetic acid,
and 0.3 mL of water. After 10 h of stirring, the solution was
concentrated under reduced pressure. The residue was flash chromato-
graphed with 4:1 pentane:diethyl ether as the eluant to yield 0.57 g
1
,1-(diphenylphosphinol)ferrocene in 106 mL of THF gave 6.54 g (79%
-
1 1
yield) of 22. IR (thin film): 1717, 1645, 1462, 1436, 1253 cm . H
NMR (300 MHz, CDCl ): δ 6.52 (d, J ) 10.2 Hz, 1H), 3.72 (s, 3H),
.60 (t, J ) 6.3 Hz, 2H), 2.44 (m, 1H), 2.34 (t, J ) 7.5 Hz, 2H), 1.60
m, 2H), 1.36 (m, 2H), 0.99 (d, J ) 6.9 Hz, 3H), 0.90 (s, 9H), 0.85 (t,
3
3
(
(
99%) of a clear liquid.
13
J ) 7.5 Hz, 3H), 0.05 (s, 6H). C NMR (300 MHz, CDCl
3
): δ 170.27,
Method B: Pyridine‚hydrofluoride (0.55 mL) was added to a solution
1
1
2
50.35, 132.19, 64.15, 53.02, 36.04, 34.24, 31.08, 27.31, 24.78, 21.41,
9.66, 13.33, -3.94. HRMS: calcd for C16
of 23 (366.8 mg, 1.05 mmol) in 11 mL of THF. After 14 h of stirring,
the solution was concentrated on to silica gel in vacuo and passed
through a plug of silica with 2:3 hexanes:ethyl acetate as the eluant.
The pure triol was diluted with 10 mL of THF, and o-nitrophenylse-
lenocyanate (681.1 mg, 3.00 mmol) was added followed by the addition
of tri-n-butylphosphine (0.75 mL, 3.01 mmol). After 1 h stirring,
sodium bicarbonate (2.52 g, 30.0 mmol) and 30% peroxide hydrogen
+
H O
31 3
3
Si (M - CH )
99.2043, found 299.2051.
(
2R,3S,4S)-Methyl 2-(3-tert-Butyldimethylsiloxypropyl)-2,3-di-
hydroxy-4-methylhexanoate (23). A heterogeneous mixture of potas-
sium carbonate (8.74 g, 63.2 mmol), potassium ferricyanide (20.61 g,
6
2
2.6 mmol), (DHQD) PHAL (1.55 g, 1.99 mmol), 4% osmium
tetraoxide (5.4 mL, 0.817 mmol) in water, methanesulfonamide (1.99
g, 20.9 mmol), and 22 (6.54 g, 20.8 mmol) in tert-butyl alcohol (105
mL) and water (105 mL) was stirred for 6 days at 0 °C. The
heterogeneous mixture was diluted with saturated aqueous sodium
dithionite, stirred until the solution became homogeneous, extracted
with diethyl ether, the combined organic extracts were dried over
magnesium sulfate, filtered, and concentrated under reduced pressure.
The residue was flash chromatographed with 1:1 pentane:diethyl ether
(
3.1 mL, 30.3 mmol) was added. After 14 h of stirring, the
heterogeneous mixture was diluted with 10% aqueous hydrochloric acid,
extracted with diethyl ether, the combined organic extracts were dried
over magnesium sulfate, filtered, and concentrated under reduced
pressure. The residue was flash chromatographed with 4:1 pentane:
diethyl ether as the eluant to yield 176 mg (77% overall) of a clear
23
liquid, R
CH Cl ). IR (thin film): 3505, 1736, 1642, 1460, 1440 cm
(300 MHz, CDCl ): δ 5.71 (m, 1H), 5.13 (d, J ) 3.9 Hz, 1H), 5.08 (s,
H), 3.8 (d, J ) 1.8 Hz, 1H), 3.79 (s, 3H), 2.41 (d, J ) 7.5 Hz, 2H
exchangeable protons), 1.70 (dsextet, J ) 6.9, 1.8 Hz, 1H), 1.44 (m,
f
) 0.46 (7:3 hexanes:ethyl acetate), [R]D ) -10.16 (c ) 6.3,
-
1 1
2
2
. H NMR
as the eluant to yield 6.44 g (89%) of a clear liquid, R
f
) 0.51 (7:3
). IR (thin
): δ
2
8
3
hexanes:ethyl acetate), [R] ) +3.46 (c ) 5.0, CH
2
Cl
2
D
1
-
1
1
film): 3506, 1736, 1463, 1445 cm
.
H NMR (300 MHz, CDCl
3
3
1
.80 (s, 3H), 3.75 (s, 1H), 3.61 (m, 2H), 2.14 (d, J ) 10.8 Hz, 1H),
13
1
1
1
3
H), 1.35 (m, 1H), 0.92 (m, 6H). C NMR (300 MHz, CDCl ): δ
76.19, 131.88, 119.34, 80.80, 75.94, 52.99, 40.26, 35.05, 28.25, 12.66,
1.83. HRMS: calcd for C11
.65 (m, 4H), 1.35 (m, 3H), 0.91 (m, 6H), 0.89 (s, 9H), 0.04 (s, 6H).
1
3
C NMR (300 MHz, CDCl
5.09, 32.28, 28.28, 27.03, 25.80, 18.19, 12.69, 11.86, -5.47. Anal.
Si: C, 58.58; H, 10.41. Found: C, 58.62; H, 10.20.
4S,5R,1′S)-2,2-Dimethyl-4-(3-hydroxypropyl)-4-(methoxycarbo-
3
): δ 176.91, 80.80, 76.43, 62.97, 52.98,
+
H O
21 4
(MH ) 217.1440, found 217.1440.
3
(6E,8R,9R,10S)-tert-Butyl 8,9-(Isopropylidenedioxy)-8-(methoxy-
36 5
Calcd for C17H O
carbonyl)-10-methyl-4-methylene-6-dodecenoate (27a). From 25:
Methanol (0.50 mL) and water (0.50 mL) was added to a mixture of
(
nyl)-5-(1′-methylpropyl)-1,3-dioxolane (24). CSA (60.9 mg, 262
µmol) was added to a solution of 23 (0.41 g, 1.18 mmol) and 2,2-
dimethoxypropane (1.45 mL, 11.8 mmol) in 11.8 mL of acetone. After
4
(
5
a (18.1 mg, 117 umol), 25 (25.4 mg, 99.1 mmol), CpRu(COD)Cl
3.2 mg, 10.3 umol), and ammonuim hexafluorophosphate (9.1 mg,
5.8 umol), and the resulting mixture was heated to reflux. After 9 h,
the solution was concentrated in vacuo, and the residue was flash
chromatographed (9:1 pentane:diethyl ether) to yield 24 mg (58%) of
a clear liquid.
2
4 h of stirring, pyridine:hydrofluoride (0.60 mL) was added at 0 °C,
and the reaction was stirred for 1 h. The solution was diluted with
water, extracted with diethyl ether, the combined organic extracts were
dried over magnesium sulfate, filtered, and concentrated under reduced
pressure. The residue was flash chromatographed with 1:1 pentane:
From 29a: CSA (7.0 mg, 30.1 umol) was added to a solution of
29a (96.3 mg, 260 umol) and 2,2-dimethoxypropane (0.32 mL, 2.60
mmol) in 2.6 mL of acetone. After 16 h of stirring, the solution was
concentrated in vacuo, and the residue was flash chromatographed with
9:1 pentane:diethyl ether as the eluant to yield 96 mg (90%) of a clear
diethyl ether as the eluant to yield 0.31 g (96%) of a clear liquid, R )
f
2
D
3
0
(
.41 (1:1 hexanes:ethyl acetate), [R] ) 31.13 (c ) 10.0, CH
2
Cl
H NMR (300
): δ 3.87 (d, J ) 9.0 Hz, 1H), 3.76 (s, 3H), 3.64 (m, 2H),
2
). IR
-
1
1
thin film): 3450, 2965, 2937, 2878, 1740, 1455 cm .
MHz, CDCl
3
23
1
0
1
2
2
.99 (m, 1H), 1.60 (m, 7H), 1.45 (s, 6H), 1.01 (d, J ) 6.6 Hz, 3H),
liquid, R
CH Cl ). IR (thin film): 2971, 2934, 2878, 1732, 1647, 1457, 1436
cm . 1H NMR (300 MHz, CDCl
): δ 5.90 (dt, J ) 15.3 Hz, 6.9 Hz,
f
) 0.48 (4:1 hexanes:ethyl acetate), [R]D ) +31.5 (c ) 2.7,
1
3
.86 (t, J ) 7.2 Hz, 3H). C NMR (300 MHz, CDCl
3
): δ 173.85,
2
-
2
1
08.75, 85.42, 85.03, 62.88, 52.53, 33.72, 28.47, 27.87, 27.12, 25.20,
3
+
4.87, 16.13, 10.61. HRMS: calcd for C13
H O
23 5
(M - CH
3
)
1H), 5.66 (d, J ) 15.3 Hz, 1H), 4.78 (s, 1H), 4.77 (s, 1H), 4.11 (d, J
) 7.5 Hz, 1H), 3.76 (s, 3H), 3.66 (s, 3H), 2.81 (d, J ) 6.9 Hz, 2H),
59.1546, found 259.1548.

9236 J. Am. Chem. Soc., Vol. 120, No. 36, 1998
Trost et al.
2
1
.46 (m, 2H), 2.33 (m, 2H), 1.65 (m, 1H), 1.51 (s, 3H), 1.47 (m, 1H),
(6E,8R,9R,10S)-9-Fluorenylmethyl 8,9-(Isopropylidenedioxy)-8-
(methoxycarbonyl)-10-methyl-4-methylene-6-dodecenoate (27d). CSA
(1.4 mg, 6.03 umol) was added to a solution of 29d (34.9 mg, 70.1
umol) and 2,2-DMP (0.09 mL, 732 umol) in acetone (0.71 mL). After
48 h of stirring, the solution was concentrated under reduced pressure
and the residue was flash chromatographed with 9:1 pentane:diethyl
f
.43 (s, 3H), 1.14 (m, 1H), 0.97 (d, J ) 6.6 Hz, 3H), 0.89 (t, J ) 7.2
1
3
Hz, 3H). C NMR (300 MHz, CDCl ): δ 173.73, 173.39, 146.21,
3
130.32, 127.78, 110.96, 109.09, 85.41, 84.94, 52.62, 51.53, 39.15, 34.98,
3
C
2.32, 30.82, 27.26, 25.92, 24.95, 15.46, 11.01. HRMS: calcd for
+
21
H
(
O
35 4
(M - CO
2
CH
3
) 353.1965, found 353.1964.
6E,8R,9R,10S)-tert-Butyl 8,9-Dihydroxyl-8-(methoxycarbonyl)-
0-methyl-4-methylene-6-dodecenoate (29a). Methanol (11.0 mL)
was added to a mixture of 4a (172.9 mg, 1.12 mmol), 2 (241.9 mg,
.12 mmol), CpRu(COD)Cl (34.8 mg, 112 umol), and ammonuim
ether as the eluant to yield 37 mg (97%) of a clear liquid, R ) 0.54
2
D
3
1
(
(
4:1 hexanes:ethyl acetate), [R] ) +23.43 (c ) 2.10, CH
2
Cl
H NMR (300 MHz,
): δ 7.77 (d, J ) 7.5 Hz, 2H), 7.59 (d, J ) 7.5 Hz, 2H), 7.41
m, 2H), 7.32 (dt, J ) 1.2 Hz, 7.5 Hz, 2H), 5.92 (m, 1H), 5.67 (d, J )
2
). IR
-1
1
thin film): 1736, 1647, 1610, 1478, 145 cm
.
1
CDCl
(
3
hexafluorophosphate (36.5 mg, 224 umol), and the resulting mixture
was heated to reflux. After 2 h, the solution was concentrated in vacuo
and the residue was flash chromatographed with 4:1 pentane:diethyl
ether as the eluant to yield 168 mg (45%) of a white solid, mp ) 65
1
4
5.3 Hz, 1H), 4.79 (s, 1H), 4.77 (s, 1H), 4.39 (d, J ) 7.2 Hz, 2H),
.20 (d, J ) 7.2 Hz, 1H), 4.12 (d, J ) 7.8 Hz, 1H), 3.75 (s, 3H), 2.82
(d, J ) 6.9 Hz, 2H), 2.54 (m, 2H), 2.34 (m, 2H), 1.65 (m, 2H), 1.52
2
D
3
°
1
C, [R] ) +35.01 (c ) 1.5, CH
2
Cl
H NMR (300 MHz, CDCl
15.3 Hz, 7.2 Hz, 1H), 5.55 (dd, J ) 1.2 Hz, 15.3 Hz, 1H), 4.76 (s,
H), 4.74 (s, 1H), 3.92 (dd, J ) 2.1 Hz, 11.1 Hz, 1H), 3.83 (s, 3H),
.57 (s, 1H, D O exchangeable), 2.78 (d, J ) 7.2 Hz, 2H), 2.35 (m,
H), 2.27 (m, 2H), 2.07 (d, J ) 11.1 Hz, 1H, D O exchangeable), 1.66
2
). IR (thin film): 3549, 1728,
(s, 3H), 1.43 (s, 3H), 1.15 (m, 1H), 0.97 (d, J ) 6.6 Hz, 3H), 0.88 (t,
-
1
1
648, 1459, 1437 cm
.
3
): δ 5.99 (dt, J
13
J ) 7.2 Hz, 3H). C NMR (300 MHz, CDCl
3
): δ 174.88, 174.71,
)
1
1
2
47.65, 145.44, 142.93, 131.76, 129.37, 128.70, 126.61, 121.61, 112.51,
1
3
2
10.62, 86.93, 86.44, 67.82, 54.14, 48.28, 40.64, 36.48, 33.97, 32.30,
2
8.79, 27.43, 26.46, 16.99, 12.54. HRMS: calcd for C32
H O
37 6
(M -
+
2
CH
3
) 517.2591, found 517.2590.
1
3
(
m, 1H), 1.42 (s, 9H), 1.25 (m, 1H), 0.89 (m, 6H). C NMR (300
MHz, CDCl ): δ 175.65, 172.59, 146.41, 129.61, 129.13, 110.71, 81.32,
0.26, 75.92, 53.41, 38.93, 35.30, 33.62, 31.01, 28.20, 28.00, 12.74,
1.71. Anal. Calcd for C20 : C, 64.84; H, 9.25. Found: C, 64.93;
(6E,8R,9R, 10S)-8,9-(Isopropylidenedioxy)-8-(methoxycarbonyl)-
3
10-methyl-4-methylene-6-dodecen-1-oic Acid (27b). Piperidine (0.02
8
1
mL, 202 µmol) was added to a solution of 29d (36.5 mg, 68.5 µmol)
in methylene chloride (0.70 mL). After 48 h of stirring, the solution
was diluted with 10% HCl, extracted with methylene chloride, the
combined organic extracts were dried over magnesium sulfate, filtered,
and concentrated under reduced pressure. The residue was flash
chromatographed with 99:1 methylene chloride:methanol as the eluant
then with 9:1 methylene chloride:methanol as the eluant to yield 23
34 6
H O
H, 9.46.
6E,8R,9R,10S)-9-Fluorenylmethyl 8,9-Dihydroxy-8-methoxycar-
bonyl-10-methyl-4-methylene-6-dodecenoate. Method A. Methanol
2.0 mL) was added to a mixture of 4d (57.5 mg, 208 µmol), 2 (47.0
(
(
mg, 217 µmol), CpRu(COD)Cl (7.1 mg, 22.9 µmol), and ammonium
hexafluorophosphate (10.4 mg, 63.8 µmol), and the mixture was heated
to reflux. After 2.5 h, the solution was concentrated in vacuo, and the
residue was flash chromatographed with 1:1 pentane:diethyl ether as
the eluant to yield 48 mg (52%) of a clear liquid.
Method B. Compounds 4d (35.8 mg, 130 µmol), 2 (23.3 mg, 108
µmol), CpRu(COD)Cl (6.2 mg, 20.0 µmol), and ammonium hexafluo-
rophosphate (6.9 mg, 42.3 µmol) were sequentially added to plastic
tube sealed with a glass rod on one end. Acetone (0.55 mL) and
methanol (0.55 mL) were added, and the open end of the plastic tube
was sealed with a glass rod by using a heat gun. The tube was placed
in a high-pressure apparatus, and the pressure was gradually increased
to 13 kbar over a 15 min period. After 1.5 h, the pressure fell to 11
kbar, the tube was removed, the contents were concentrated in vacuo,
and the residue was flash chromatographed with 1:1 pentane:diethyl
mg (96% yield) of a film. The spectral properties are in agreement
with those previously recorded.7 [R]D ) +25.6 (c ) 1.16, CH
27
Cl
H NMR
): δ 5.90 (m, 1H), 5.67 (d, J ) 15.3 Hz, 1H), 4.81
s, 1H), 4.79 (s, 1H), 4.12 (d, J ) 7.8 Hz, 1H), 3.76 (s, 3H), 2.82 (d,
2
2
).
-
1
1
IR (thin film): 3200, 1739, 1712, 1648, 1455, 1436 cm
.
(300 MHz, CDCl
3
(
J ) 6.9 Hz, 2H), 2.51 (m, 2H), 2.33 (t, J ) 7.5 Hz, 2H), 1.64 (m, 1H),
.52 (s, 3H), 1.50 (m, 1H), 1.43 (s, 3H), 1.12 (m, 1H), 0.97 (d, J ) 6.6
1
13
Hz, 3H), 0.88 (t, J ) 7.5 Hz, 3H). C NMR (500 MHz, CDCl
1
5
3
): δ
78.44, 173.18, 145.74, 130.13, 127.73, 111.02, 109.04, 85.39, 84.88,
2.67, 39.26, 35.03, 32.16, 30.49, 27.31, 25.97, 25.02, 15.56, 11.10.
) 0.40 (19:1 methylene chloride:methanol).
TLC R
f
Acknowledgment. We thank the National Institutes of
Health, General Medical Sciences Institute, and the National
Science Foundation for their generous support of our programs.
A.S. was supported by a NATO Grant. Mass spectra were
provided by the Mass Spectrometry Facility, University of San
Francisco, supported by the NIH Division of Research Re-
sources.
ether as the eluant to yield 27 mg (51%) of a clear liquid, R
f
) 0.26
Cl ). IR
H NMR (300 MHz,
): δ 7.77 (d, J ) 7.5 Hz, 2H), 7.59 (d, J ) 7.5 Hz, 2H), 7.41
m, 2H), 7.32 (dt, J ) 1.2 Hz, 7.5 Hz, 2H), 6.00 (m, 1H), 5.56 (d, J )
2
D
3
(
7:3 hexanes:ethyl acetate), [R] ) +24.12 (c ) 3.05, CH
2
2
-
1
1
(thin film): 3505, 1736, 1647, 1450 cm
.
CDCl
(
1
3
5.3 Hz, 1H), 4.76 (s, 2H), 4.40 (d, J ) 7.2 Hz, 2H), 4.20 (t, J ) 7.2
Hz, 1H), 3.92 (dd, J ) 11.1 Hz, 1.8 Hz, 1H), 3.81 (s, 3H), 3.60 (s,
H), 2.78 (d, J ) 7.2 Hz, 2H), 2.53 (m, 2H), 2.32 (m, 2H), 2.11 (d, J
11.1 Hz, 1H), 1.65 (m, 1H), 1.43 (m, 1H), 1.30 (m, 1H), 0.88 (m,
1
)
Supporting Information Available: Experimental proce-
dures for preparation of 4d and 9c (1 page, print/PDF). See
any current masthead page for ordering information and Web
access instructions.
13
6
1
7
1
4
3
H). C NMR (300 MHz, CDCl ): δ 175.60, 173.19, 146.11, 143.91,
41.43, 129.42, 129.29, 127.89, 127.19, 125.07, 120.12, 110.93, 81.22,
5.81, 66.31, 53.44, 46.78, 39.96, 35.31, 32.42, 30.71, 28.20, 12.74,
+
1.72. HRMS: calcd for C30
75.2486.
H O
35 5
(M - OH) 475.2486, found
JA981540N