Asymmetric synthesis of stegobinone via boronic ester chemistry

Article abstract of DOI:10.1021/ja960345a

Highly stereoselective asymmetric boronic ester chemistry has been used to install all three chiral centers in a convergent synthesis of highly pure stegobinone, the epimerically labile pheromone of the drugstore beetle, Stegobium paniceum, and the furniture beetle, Anobium punctatum. Asymmetric centers were installed via the reaction of (dichloromethyl)lithium with 1,2-dicyclohexylethane- 1,2-diol boronic esters. The synthetic strategy utilizes a common (α-chloroalkyl)boronic ester intermediate as the source of both segments and all of the asymmetry of the target molecule. The two segments are joined by an aldol condensation and converted to stegobiol, a minor component of the S. paniceum pheromone and presumably the biogenetic precursor of stegobinone. Stegobiol is stable and easily purified, and is easily converted to pure stegobinone in a single oxidation step.

Full text of DOI:10.1021/ja960345a

4560
J. Am. Chem. Soc. 1996, 118, 4560-4566
Asymmetric Synthesis of Stegobinone via Boronic Ester
Chemistry
Donald S. Matteson,* Hon-Wah Man, and Oliver C. Ho
Contribution from the Department of Chemistry, Washington State UniVersity,
Pullman, Washington 99164-4630
ReceiVed February 1, 1996X
Abstract: Highly stereoselective asymmetric boronic ester chemistry has been used to install all three chiral centers
in a convergent synthesis of highly pure stegobinone, the epimerically labile pheromone of the drugstore beetle,
Stegobium paniceum, and the furniture beetle, Anobium punctatum. Asymmetric centers were installed via the reaction
of (dichloromethyl)lithium with 1,2-dicyclohexylethane-1,2-diol boronic esters. The synthetic strategy utilizes a
common (R-chloroalkyl)boronic ester intermediate as the source of both segments and all of the asymmetry of the
target molecule. The two segments are joined by an aldol condensation and converted to stegobiol, a minor component
of the S. paniceum pheromone and presumably the biogenetic precursor of stegobinone. Stegobiol is stable and
easily purified, and is easily converted to pure stegobinone in a single oxidation step.
Introduction
activity.7 Material synthesized by Mori and Ebata via another
route was similarly impure.8 The 1′-epimer, (2S,3R,1′S)-
stegobinone (2), is strongly repellent to S. paniceum at the level
of a few percent,7 and impure 1 is highly prone to epimerization
on standing for a few days.7,9 Synthetic racemate of 1 was also
reported to be repellent.9
(2S,3R,1′S,2′S)-Stegobiol (3) has also been identified as a
component of the natural S. paniceum pheromone at the 5%
level.10 Natural 3 isolated as an oil by gas chromatography
(and not necessarily free from all 1, Vide infra) was found to
be attractive to S. paniceum, though the response of the insects
was described as slow.10 Synthetic 3 has been reported as an
oil.11 A significant synthesis of serricorole, the 2-ethyl homolog
of 3, was reported just prior to our preliminary communication.12
Outstanding stereocontrol has been demonstrated in the chain
extension of chiral boronic esters via reaction with (dichlorom-
ethyl)lithium.1,2 Diastereoselection exceeding 500:1 has been
achieved for the installation of a pair of adjacent asymmetric
centers, and it has been inferred that the enantiomeric purity of
the major diastereomer is then within a few parts per million
of being equal to that of the chiral director.2 Such precise
stereocontrol is usually not necessary for synthetic purposes,
provided that minor amounts of stereoisomers can be removed
from the product or are not detrimental to its use. In order to
demonstrate the unique utility of this boron chemistry, we sought
a synthetic target where such precise stereocontrol would be
significant.3
(2S,3R,1′R)-Stegobinone (1), the epimerically labile phero-
mone of the drugstore beetle, Stegobium paniceum,4,5 and the
furniture beetle, Anobium punctatum,6 has proved to be a
particularly elusive synthetic challenge, even though it has a
relatively simple structure. Both species are economically
important pests.
Results
General Strategy. Because the natural pheromone of S.
paniceum contains both stegobinone (1) and stegobiol (3), a
common route to both compounds was sought. The selected
synthetic strategy involved initial synthesis and purification of
the stable secondary alcohol 3, which could be oxidized to the
labile diketone 1 in diastereomerically and enantiomerically pure
form. It was hoped that if 1 was initially free from diastere-
omers, especially the troublesome isomer 2, it could be purified
rapidly to a stable crystalline solid before significant isomer-
ization took place.
Ultrahigh stereoselection has been observed in reactions of
diol boronic esters of C2 symmetry with (dichloromethyl)lithium
followed by nucleophilic substitution of the resulting (R-
chloroalkyl)boronic esters.2 (R,R)-1,2-Dicyclohexylethane-1,2-
diol [“(R,R)-DICHED”] was chosen for the present work
because it is easily prepared by rhodium catalyzed hydrogenation
Pure 1 isolated from S. paniceum was crystalline, mp 52.5-
53.5 °C.4 The first synthesis by Hoffmann and co-workers
proved which stereoisomer was the natural pheromone, but their
product was impure and had only a small fraction of the natural
X Abstract published in AdVance ACS Abstracts, May 1, 1996.
(1) (a) Matteson, D. S.; Sadhu, K. M.; Peterson, M. L. J. Am. Chem.
Soc. 1986, 108, 812-819. (b) Matteson, D. S. Chem. ReV. 1989, 89, 1535-
1551. (c) Matteson, D. S.; Kandil, A. A.; Soundararajan, R. J. Am. Chem.
Soc. 1990, 112, 3964-3969.
(7) (a) Hoffmann, R. W.; Ladner, W.; Steinbach, K.; Massa, W.; Schmidt,
R.; Snatzke, G. Chem. Ber. 1981, 114, 2786-2801. (b) Hoffmann, R. W.;
Ladner, W. Tetrahedron Lett. 1979, 4653-4656.
(8) Mori, K.; Ebata, T. Tetrahedron 1986, 42, 4413-4420.
(9) Kodama, H.; Mochizuki, K.; Kohno, M.; Ohnishi, A.; Kuwahara, Y.
J. Chem. Ecol. 1987, 13, 1859-1869
(2) Tripathy, P. B.; Matteson, D. S. Synthesis 1990, 200-206.
(3) Preliminary communication: Matteson, D. S.; Man, H.-W. J. Org.
Chem. 1993, 58, 6545-6547.
(10) Kodama, H.; Ono, M.; Kohno, M.; Ohnishi, A. J. Chem. Ecol. 1987,
13, 1871-1879.
(11) Mori, K.; Ebata, T. Tetrahedron 1986, 42, 4685-4689.
(12) (a) Oppolzer, W.; Rodriguez, I. HelV. Chim. Acta 1993, 76, 1275-
1281; (b) Oppolzer, W.; Rodriguez, I. HelV. Chim. Acta 1993, 76, 1282-
1291.
(4) Kuwahara, Y.; Fukami, H.; Howard, R.; Ishii, S.; Matsumura, F.;
Burkholder, W. E. Tetrahedron 1978, 34, 1769-1774.
(5) Burkholder, W. E.; Wang, Y. Personal communication, 1994.
(6) White, P. R.; Birch, M. C. J. Chem. Ecol. 1987, 13, 1695-1706.
S0002-7863(96)00345-9 CCC: $12.00 © 1996 American Chemical Society

Asymmetric Synthesis of Stegobinone
J. Am. Chem. Soc., Vol. 118, No. 19, 1996 4561
of (R,R)-1,2-diphenylethane-1,2-diol,13,15 which in turn is avail-
able in large quantities from the catalytic asymmetric dihy-
droxylation of trans-stilbene.14 1,2-Diphenylethane-1,2-diol as
chiral director has not yielded satisfactory diastereoselection.16
The highest stereoselectivity measured rigorously has been
achieved with another chiral director of C2 symmetry, 2,5-
dimethyl-3,4-hexanediol, which requires a more laborious
preparation than DICHED.2
Preferred Route. The synthetic route utilizes chloro boronic
ester 8 as a common intermediate for making both segments of
stegobiol (3). From a practical standpoint, it is significant that
DICHED ethylboronate (4) can be purified by vacuum distil-
lation, and none of the intermediates 5-8 have to be purified
beyond removal of solvents or inorganic salts before proceeding
to the next step. Intermediate chloro boronic ester 8 contains
all of the stereocenters of stegobiol. The first true purification
is distillation of aldehyde 9, which constitutes one of the two
segments to be joined.
control and freshly distilled, aldehyde 9 contained ∼0.5-1%
epimer as judged from 1H-NMR data. The relevant NMR
curves have been published previously.3 Although this falls
far short of the diastereomeric purity achieved with compounds
that are not stereolabile,2 it has proved sufficient for our
purposes.
The second segment is the keto boronic ester 15a. Its
synthesis from chloro boronic ester 8 began with methylation
to 10, followed by debenzylation to γ-hydroxy boronic ester
11. Basic hydrolysis of 11 separated the ether soluble DICHED
and impurities and yielded an aqueous solution of the oxaboro-
lane salt 12. Acidification of 12 yielded the oxaborolane in its
dimeric anhydride form 13, which with an equivalent amount
of pinacol resulted in an equilibrium mixture containing 14
together with unchanged 13, pinacol, and according to mass
spectral analysis, the species C20H40O4B2 [13 + pinacol -
H2O].17 The mixture was oxidized by pyridinium dichromate
to the stable keto boronic ester 15a. It is again significant from
a practical standpoint that none of the intermediates between 4
and 15a require more than partial purification. Most of the
impurities are extracted by ether from the aqueous solution of
12, and 15a is obtained sufficiently pure for further use with
no purification except removal of solvent under vacuum.
Enolic epimerization of aldehyde 9 (∼10% on silica, ∼50%
on alumina) made purification by chromatography impossible.
When generated from chloro boronic ester 8 with proper pH
For connection to 9, 15a must be converted to a suitable
enolate. Of several boron enolates tested, the most satisfactory
proved to be 16a, which was prepared from 15a and
9-borabicyclo[3.3.1]nonane 9-triflate at -78 °C. Dialkylboron
halides required higher temperatures and did not lead to efficient
enolate formation. This is the only step in the reaction sequence
that has not been satisfactorily achieved at -40 °C or above.
The initially formed aldol 17a was not isolated but oxidized to
the corresponding ketone 18a, which on peroxidic deboronation
followed by acidification yielded O-benzylstegobiol (19).
O-Benzylstegobiol (19) is easily purified by chromatography.
Cleavage of the benzyl group by catalytic hydrogenation over
palladium resulted in a few percent of hydrogenation of the
carbonscarbon double bond, and a surprisingly mild alternative
cleavage with ∼25% methanesulfonic acid in chloroform at 25
°C is recommended for the conversion to stegobiol (3). It is
(13) Hoffmann, R. W.; Ditrich, K.; Ko¨ster, G.; Stu¨rmer, R. Chem. Ber.
1989, 122, 1783-1789.
(14) (a) Jacobsen, E. N.; Marko´, I.; Mungall, W. S.; Schro¨der, G.;
Sharpless, K. B. J. Am. Chem. Soc. 1988, 110, 1968-1970. (b) Sharpless,
K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.; Hartung, J.; Jeong,
K.-S.; Kwong, H.-L.; Morikawa, K.; Wang, Z.-M., Xu, D.; Zhang, X.-L. J.
Org. Chem. 1992, 57, 2768-2771. (c) Wang, Z.-M.; Sharpless, K. B. J.
Org. Chem. 1994, 59, 8302-8303. (d) A potassium osmate sample
consisting of larger crystals, black in appearance, failed to dissolve under
the conditions described in (c), and hydroxylation failed. Addition of the
osmate as a 1% aqueous solution over a period of ∼15 min resulted in
satisfactory hydroxylation of the stilbene in ∼24 h. We thank Rajendra
Prasad Singh for these observations.
(15) (R,R)-1,2-Diphenyl-1,2-ethanediol (50 g) was dissolved in ethyl
acetate and washed with aqueous sulfuric acid to remove traces of alkaloidal
catalyst poison and then concentrated and hydrogenated at 1 atm over 5%
Rh/Al2O3 (2 g) in methanol (600 mL) containing 1% water and 1% acetic
acid over a period of 2-4 weeks. We thank Dr. G. D. Schaumberg, Sonoma
State University, for developing the hydrogenation conditions. Improved
hydrogenation conditions currently under development by W. C. Hiscox
will be reported elsewhere.
(17) Plausible structures include [C7H14O]B-O-CMe2CMe2-O-
B[OC7H14], where [C7H14O]B- represents an oxaborolane unit, or
[C7H14O]B-OCHEtCHMeCHMeB[O2C2Me4], an oxaborolane-hydroxy
boronic ester-pinacol linkage. Either of these would equilibrate rapidly
with pinacol, water, 13, and 14.
(16) This finding has been replicated in our laboratory by G. D.
Schaumberg.

4562 J. Am. Chem. Soc., Vol. 118, No. 19, 1996
Matteson et al.
the problem of DICHED oxidation, at the cost of some
lengthening of the synthesis. Silylation of 20 to 21a was
straightforward, but if sufficient care was not taken to remove
all traces of acid from 21a before reductive debenzylation to
21b, a significant amount of desilylation to useless 3-methyl-
2,4-hexanediol was observed. The remainder of the route to
O-benzylstegobiol (19) was straightforward.
important that 3 be purified by chromatography before oxidation
to stegobinone (1) with N-methylmorpholine N-oxide catalyzed
by tetrapropylammonium perruthenate.18 Stegobinone (1) pre-
pared in this manner readily crystallizes after rapid chromatog-
raphy over a short silica column and concentration of the
solution, and recrystallization from pentane effectively removes
the small amounts of impurities that remain.
Alternative Routes. Leaving the relatively fragile carbons
boron bond of γ-benzyloxy boronic ester 10 in place in
subsequent intermediates, after it is no longer needed for
construction of asymmetric carbons, carries the obvious liability
of potential side reactions, especially during oxidative proce-
dures. These problems were indeed encountered in a closely
related route tested earlier. Oxidation of γ-hydroxy boronic
ester 11 to the corresponding ketone 15b with pyridinium
Before finding the successful procedures described above,
we attempted a route via an intermediate of the type used by
Mori and Ebata,8 a 3-methyl-4-(ketohexyl) 2-methyl-3-(silyly-
loxy)pentanoate. We were unable in several attempts to close
the six-membered ring via ketone enolate attack at the ester
carbonyl group under conditions similar to those described by
Mori and Ebata,8 and Oppolzer’s method12 had not yet been
reported. This work will be reported elsewhere.
Discussion
Properties of Stegobinone and Stegobiol. Tests carried out
by Burkholder and Wang have shown that pure synthetic
stegobinone (1) is highly attractive to S. paniceum.5 Crystalline
1 is stable indefinitely at -10 °C and for several months at
room temperature. A sample stored at room temperature for
more than a year became visibly gummy, acquired a different
and stronger odor, and showed a few percent of epimer 2 by
1H NMR analysis.
Ours was the first reported crystalline sample of stegobiol
(3). Crystallization is difficult to initiate in spite of the melting
point of 74 °C, and we have been unsuccessful in attempts to
purify 3 by recrystallization. Although oily 3 isolated from the
insects has been reported to be an attractant,10 pure synthetic 3
has been found to have no detectable attractant activity.5 The
presence of 5% 3 in the natural pheromone is sufficient to assure
the presence of a liquid phase and consequent significant
epimerization of attractant 1 to repellent 2 over the course of a
few hours.
dichromate, N-methylmorpholine N-oxide, and tetrapropylam-
monium perruthenate, or Swern’s method, was accompanied
by significant competing oxidation of the chiral director
DICHED. NMR evidence indicated that 1,2-dicyclohexyl-2-
hydroxyethanone was the major byproduct.
Even with the loss, the overall yield of 15b from 4 was 38%.
However, after aldol condensation of the dibutylboron enolate
from 15b with aldehyde 9, the resulting alcohol intermediate
17b (DICHED analog of 17a, not illustrated) had to be similarly
oxidized to 18b, and a second round of DICHED oxidation
occurred.
Oxidation of the boronic ester function of 10 with hydrogen
peroxide to the hydroxyl function of 20 successfully avoided
(18) Griffith, W. P.; Ley, S. V.; Whitcombe, G. P.; White, A. D. J. Chem.
Soc., Chem. Commun. 1987, 1625.
(19) (a) Paterson, I.; Lister, M. A.; McClure, C. K. Tetrahedron Lett.
1986, 27. 4787-4790. (b) Paterson, I.; McClure, C. K. Tetrahedron Lett.
1987, 28. 1229-1232.

Asymmetric Synthesis of Stegobinone
J. Am. Chem. Soc., Vol. 118, No. 19, 1996 4563
Experimental Section
Comparison of Synthetic Routes. It is evident that high
enantioselectivity and diastereoselectivity were important to the
success of our synthesis of stegobinone (1). The problems with
the previous syntheses7,8 appeared to result from the presence
of diastereomeric or enantiomeric impurities coupled with the
strong tendency of 1 to epimerize during attempted purification.
Recrystallization of 1 from pentane is successful only if 1 is
already pure enough to crystallize easily.
General Data. NMR spectra were recorded on a Bruker 300 MHz
or Varian VXR-500s spectrometer. Melting points were determined
in open-ended capillaries using a Thomas-Hoover melting point
apparatus and are uncorrected. Tetrahydrofuran, 1,2-dimethoxyethane,
and diethyl ether were distilled from benzophenone ketyl under argon
prior to use. Sure Seal bottles of dimethyl sulfoxide, dimethylforma-
mide, Grignard reagents, and lithium diisopropylamide solutions were
purchased from the Aldrich Chemical Co. Other chemicals were
reagent grade. Flash chromatography was performed on Merck silica
gel 60, 230-400 mesh. Elemental analyses were preformed by Desert
Analytics, Tucson, AZ.
The boronic ester chemistry utilized for establishment of the
asymmetry is the same for all of the synthetic routes described.
Somewhat higher yields of (2S,3S)-2-methyl-3-(benzyloxy)-
pentanal (9) were obtained when the chain extensions were
carried out with preformed (dichloromethyl)lithium [65%
based on DICHED ethylboronate (4)] than when the in situ
dichloromethane/LDA method was used (55%). In view of
the possible variations in purity of small lots of LDA or butyl-
lithium from commercial sources20 and the generally comparable
yields seen in previous work with the two procedures,1 this
difference is not necessarily real, and even if it is, it is
offset by the practical consideration that generation of preformed
(dichloromethyl)lithium requires cooling to -100 °C, but
the in situ method is successful at -40 °C and is easy to scale
up.
[4R-(4r,5â)]-4,5-Dicyclohexyl-2-ethyl-1,3,2-dioxaborolane (4).
Method A: From Dibutyl Ethylboronate and DICHED.20 (1R,2R)-
1,2-Dicyclohexylethane-1,2-diol (R,R-DICHED) (46 g, 203 mmol) was
added to a solution of dibutyl ethylboronate (37.9 g, 203 mmol) in
hexanes (500 mL) at room temperature. The solution was stirred for
1 h. Distillation yielded 4: 52.5 g, 98%; bp 85-87 °C (1 Torr); 300
MHz 1H NMR (CDCl3) δ 0.90-1.28 and 1.50-1.77 (m, 22), 0.78 (q,
J ) 8.1 Hz, 2), 0.96 (t, J ) 7.3 Hz, 3), 3.81-3.83 (m, 2); 75 MHz 13
C
NMR (CDCl3) δ 2.51 (br), 7.91, 25.88, 26.01, 26.44, 27.29, 28.27,
42.99, 83.18; HRMS calcd for C16H29BO2 (M+) 264.2261, found
264.2253. Anal. Calcd for C16H29BO2: C, 72.73; H, 11.06; B, 4.09.
Found: C, 72.86; H, 10.86; B, 4.31.
Method B: From DICHED, Triisopropyl Borate, and Ethyl-
magnesium Chloride. [R-(R*,R*)]-1,2-Dicyclohexylethane-1,2-diol
(67.54 g, 298 mmol) was dried by addition of cyclohexane (500 mL)
and distillation of the cyclohexane-water azeotrope. The residue was
treated with THF (200 mL) and triisopropyl borate (61.8 g, 328 mmol)
at room temperature. THF and 2-propanol were distilled under vacuum.
THF (200 mL) was again added, the solution was cooled to -78 °C,
and ethylmagnesium chloride (150 mL, 2 M, 300 mmol) was added
dropwise. The solution was allowed to warm to room temperature and
kept for 12 h. Hydrochloric acid (1 M, ∼20 mL) was added in small
portions to the mixture until the initial exothermic reaction and
effervescence subsided. (A small amount of unchanged Grignard
reagent may have been destroyed by this procedure.) The mixture was
concentrated under vacuum and then treated with diethyl ether (600
mL) and hydrochloric acid (1 M, 200 mL). The ether solution was
washed with water (200 mL) and dried over magnesium sulfate.
Distillation yielded 4; 59.4 g, 75%; bp 125-130 °C (0.5 Torr).
[4R-[2(R*),4r,5â]]-4,5-Dicyclohexyl-2-[1-(phenylmethoxy)propyl]-
1,3,2-dioxaborolane (6). To a solution of [4R-(4R,5â)]-4,5-dicyclo-
hexyl-2-ethyl-1,3,2-dioxaborolane (4) (54 g, 204 mmol) and dichlo-
romethane (52 g, 610 mmol) in THF (300 mL) was added LDA (120
mL, 2 M, 240 mmol) at -40 °C via cannula. After 10 min, zinc
chloride (55.5 g, 408 mmol), which had been fused before use, was
added to the solution. After 30 min, the ice bath was removed. The
solution was allowed to warm to room temperature and kept for 2 h to
form [4R-(2S*,4R,5â)]-2-(1-chloropropyl)-4,5-dicyclohexyl-1,3,2-di-
oxaborolane (5). NMR analysis of a sample of 5 obtained by
concentration of the crude solution showed that the reaction was
complete.23 The solution of crude 5 was concentrated under vacuum
to remove excess dichloromethane. THF (300 mL) was added to the
mixture, which was then added dropwise via additional funnel to a
solution of sodium benzyl oxide in THF and DMSO at 0 °C. [The
sodium benzyl oxide solution was prepared as follows. To a solution
of benzyl alcohol (26 g, 240 mmol) in 100 mL of THF and 300 mL of
DMSO was added sodium hydride (9 g, 225 mmol) (60% dispersion
in mineral oil) at room temperature, and the mixture was stirred
overnight.] The solution was allowed to warm to room temperature
and stirred for 48 h. The solvent was distilled at 1 atm. Hexanes (1000
milliliters) and aqueous ammonium chloride (500 mL) were added to
the mixture. Sufficient hydrochloric acid (a few mL) was added to
ensure that the solution was acidic to avoid any possible hydrolysis.
In spite of the detour via oxaborolane 13, the yield of ketone
intermediate 15a (39% from 4 via the in situ method) appears
to be higher than that of the alternate route intermediate 15b
(38% from 4 via the higher yielding preformed (dichlorometh-
yl)lithium). The major advantage of the route via 15a is that
the chiral director DICHED is recovered at the earliest possible
point in the procedure and without oxidative loss. (DICHED
is also readily recoverable from the byproduct boronic ester
fractions, as well as from the preparation of 9, in the general
manner previously described for 1,2-diisopropylethane-1,2-
diol.21 ) It is also helpful that purification of the sodium salt 12
by aqueous extraction and acid regeneration of 13 gets rid of
the various minor boronic ester and other organic impurities,
because only the γ-hydroxy boronic ester hydrolyzes to any
significant extent and is extracted into the aqueous phase. The
overall yield of stegobinone (1) from DICHED ethylboronate
(4) was ∼12.5%.22
Ketone 22 described in our preliminary communication3 has
been obtained in up to 29% yield from 4, via preformed
(dichloromethyl)lithium. It appears that the aldol condensation
of 22 and subsequent steps are more efficient than those of 15a,
and the net yield of stegobinone from 4 was the same within
experimental error.22 Considering the problems we had with
acid sensitivity of the silyl protecting group of 21a, we consider
the route via 15a to be more dependable.
(20) During our first successful synthesis of stegobinone, ∼5% of
contaminants having butoxy in place of benzyloxy persisted from 6 on
throughout the synthesis until debenzylation of 19 to 3 provided a readily
purifiable intermediate. It is unclear whether the butoxide source was
oxidized butyllithium or dibutyl ethylboronate not separated from 4 by
distillation. A recent simple preparation of dimethyl ethylboronate from
ethylene, boron trichloride, triethylsilane, and methanol (Soundararajan, R.;
Matteson, D. S. Organometallics 1995, 14, 4157-4166) presumably makes
the use of dibutyl ethylboronate for this purpose obsolete, though the
preparation 4 from this source has not yet been tested. Commercial LDA
does not contain lithium butoxide.
(21) Matteson, D. S.; Tripathy, P. B.; Sarkar, A.; Sadhu, K. M. J. Am.
Chem. Soc. 1989, 111, 4399-4402.
(22) To produce 1 mol of 9 required 1.82 mol of 4, and 1 mol of 15a
required 2.56 mol of 4. From 1 mol each of 9 and 15a, the yield of
O-benzylstegobiol (19) would be 0.367 mol, and conversion to stegobinone
(1) was 75%. Thus, the overall yield of 1 from 4 was 12.5%. The route via
22 was estimated to be ∼14%, but this required the higher yielding
preformed LiCHCl2, and includes a higher but small-scale yield of 1 from
19, ∼80%, only observed once.
(23) A sample of 5 obtained by concentration of the crude solution
obtained from a run in which preformed (dichloromethyl)lithium was used
(no LDA, see the supporting information) showed the following: 300 MHz
1H NMR (CDCl3) δ 0.9-1.4 and 1.55-1.98 (m, 24), 1.02 (t, J ) 7.3 Hz,
3), 3.40 (dd J ) 6.3 and 8.0 Hz, 1), 3.92-3.96 (m, 2); 75 MHz 13C NMR
(CDCl3) δ 11.92, 25.70, 25.90, 26.33, 27.18, 27.53, 28.11, 42.81, 45 (br),
84.00; HRMS calcd for C17H30O2BCl (M+) 312.2027, found 312.2048.

4564 J. Am. Chem. Soc., Vol. 118, No. 19, 1996
Matteson et al.
The organic solution was washed with water (6 × 200 mL) and brine
(100 mL). Concentration in a rotary evaporator gave crude [4R-
[2(R*),4R,5â]]-4,5-dicyclohexyl-2-[1-(phenylmethoxy)propyl]-1,3,2-di-
oxaborolane (6) (75 g), which was used in the next step without further
purification. An analytical sample of 6 was obtained by chromatog-
raphy with considerable loss due to decomposition on the column: 300
g), which was used in the next steps without further purification: 300
MHz 1H NMR (CDCl3) δ 0.84-1.82 (m, 24), 0.93 (t, J ) 7.4 Hz, 3,),
1.01 (d, J ) 6.7 Hz, 3) 2.15-2.25 (m, 1), 3.36-3.42 (m, 1), 3.88-
3.93 (m, 2), 4.00 (d, J ) 4.4 Hz), 4.56 (br s, 2), 7.25-7.41 (m, 5); 75
MHz 13C NMR (CDCl3) δ 8.29, 13.30, 22.49, 25.82, 25.92, 26.39,
27.48, 28.28, 38.73, 42.85, 47 (br), 71.72, 80.08, 84.16, 127.48, 127.92,
128.31, 138.78; HRMS calcd for C27H42O3BCl (M+) 460.2915, found
460.2875.
1
MHz H NMR (CDCl3) δ 0.90-1.42 and 1.58-1.84 (m, 24), 1.01 (t,
J ) 7.3 Hz, 3), 3.24 (t, J ) 6.7 Hz, 1), 3.91-3.96 (m, 2), 4.53 (AB,
J ) 12 Hz, 1), 4.61 (AB, J ) 12 Hz, 1), 7.22-7.41 (m, 5); 75 MHz
13C NMR (CDCl3) δ 10.87, 24.33, 25.77, 25.89, 26.31, 27.24, 28.14,
42.81, 69 (br), 71.87, 83.43, 127.11, 127.66, 128.02, 139.12; HRMS
calcd for C24H37O3B (M+) 384.2836, found 384.2802. Anal. Calcd
for C24H37BO3: C, 74.00; H, 9.70. Found: C, 74.78; H, 9.40.
[S-(R*,R*)]-2-Methyl-3-(phenylmethoxy)pentanal (9). To a solu-
tion of crude [4R-[2(1S*,2S*,3S*),4R,5â]]-2-[1-chloro-2-methyl-3-
(phenylmethoxy)pentyl]-4,5-dicyclohexyl-1,3,2-dioxaborolane (8) (33
g) in 300 mL of THF and 150 mL of phosphate buffer (pH 8) was
added hydrogen peroxide (30 mL, 30%) at 0 °C. After the addition of
hydrogen peroxide, sodium carbonate (20 g, 188 mmol) was added to
keep the pH at 8-9. After 0.5 h, the cold bath was removed. The pH
of the aqueous solution was checked frequently. Acid is produced in
the reaction, and if the pH falls below 8, the oxidation stops. Sodium
carbonate was added as necessary to keep the pH at 8-9, and the
reaction was complete within 3 h. Sodium iodide (3 g, 20 mmol) was
added at 0 °C. The iodine was reduced by sodium thiosulfate (24.8 g,
100 mmol). The THF solution was separated from aqueous solution.
The aqueous layer was extracted with ether (2 × 600 mL). The
combined organic layer was washed with sodium thiosulfate solution
(10%, 100 mL) and dried over magnesium sulfate. Vacuum distillation
of the solvent gave a residue of [4R-(4R,5â)]-4,5-dicyclohexyl-2-
hydroxy-1,3,2-dioxaborolane and [S-(R*,R*)]-2-methyl-3-(phenyl-
methoxy)pentanal (9). Bulb to bulb distillation of the residue yielded
9 containing a small amount of benzyl alcohol: 8.1 g [55.3% overall
yield via method B from [4R-(4R,5â)]-4,5-dicyclohexyl-2-ethyl-1,3,2-
dioxaborolane (4); via method A on one-fourth the described scale,
65.4% of 9 was obtained, based on 4],; bp 94-95 °C (0.4 Torr); 300
[4R-[2(1S*,2S*),4r,5â]]-4,5-Dicyclohexyl-2-[1-methyl-2-(phenyl-
methoxy)butyl]-1,3,2-dioxaborolane (7). To a solution of [4R-
(2R*,4R,5â)]-4,5-dicyclohexyl-2-[1-(phenylmethoxy)propyl]-1,3,2-di-
oxaborolane (6) (75 g) and dichloromethane (52 g, 610 mmol) in THF
(300 mL) was added LDA (120 mL, 2 M, 240 mmol) at -40 °C via
cannula. After 10 min, zinc chloride (69.36 g, 510 mmol), which was
fused before use, was added to the solution. The resulting solution of
[4R-[2(1S*,2S*)4R,5â]]-2-(1-chloro-2-(phenylmethoxy)butyl]-4,5-dicy-
clohexyl-1,3,2-dioxaborolane was allowed to warm to room temperature
and kept for 2 h. The solution was cooled to 0 °C, and methylmag-
nesium chloride (204 mL, 2.5 M, 510 mmol) was added dropwise. The
solution was allowed to warm to room temperature and kept for 36 h.
Aqueous ammonium chloride (20 mL) was added to the mixture. Some
gas (methane) was liberated. The solvent was removed by vacuum
distillation. Pentane (1000 mL) and aqueous ammonium chloride (500
mL) were added to the mixture. The organic solution was washed
with ammonium chloride (300 mL) followed by water (3 × 300 mL)
and dried over magnesium sulfate. Removal of the solvent by vacuum
distillation yielded a mixture containing a small amount of [4R-(4R,5â)]-
4,5-dicyclohexyl-2-methyl-1,3,2-dioxaborolane with the major product
[4R-[2(1S*,2S*)4R,5â]]-2-[1-methyl-2-(phenylmethoxy)butyl]-4,5-dicy-
clohexyl-1,3,2-dioxaborolane (7) (70.5 g). The crude material was used
in the next step without further purification. A partially purified
analytical sample was obtained by flash chromatography on silica with
1:30 ether/hexane: 300 MHz 1H NMR (CDCl3) δ 0.92-1.38 and 1.48-
1.82 (m, 25), 0.93 (t, J ) 7.4 Hz, 3), 0.99 (d, J ) 7.47 Hz, 3), 3.49
(ddd, J ) 4.5, 5.7, 6.8 Hz, 1), 3.80-3.85 (m, 2), 4.48 (AB, J ) 12 Hz,
1), 4.54 (AB, J ) 12 Hz, 1), 7.24-7.37 (m, 5); 75 MHz 13C NMR
(CDCl3) δ 9.86, 10.65, 20 (br), 24.61, 25.86, 25.98, 26.44, 27.52, 28.29,
42.99, 70.72, 82.84, 83.25, 127.03, 127.35, 128.11, 139.46. The
analysis was not satisfactory. Anal. Calcd for C26H41BO3: C, 75.72;
H, 10.02; B, 2.62. Found: C, 72.44; H, 9.72; B, 2.63.
[4R-[2(1S*,2S*),4r,5â]]-2-[1-Chloro-2-(phenylmethoxy)butyl]-
4,5-dicyclohexyl-1,3,2-dioxaborolane. A small sample of this com-
pound (chloro boronic ester intermediate between 6 and 7) was
concentrated for NMR and mass spectral analysis: 300 MHz 1H NMR
(CDCl3) δ 0.85-1.76 (m, 24), 0.96 (t, J ) 7.4 Hz, 3), 3.64-3.72 (m,
2), 3.93-3.96 (m, 2), 4.62 (AB, J ) 11.4 Hz, 1), 4.69 (AB, J ) 11.4
Hz, 1), 7.23-7.40 (m, 5); 125 MHz 13C NMR (CDCl3) δ 9.86, 24.78,
25.75, 25.87, 26.32, 27.34, 28.15, 42.78, 45 (br), 72.40, 81.77, 84.15,
127.36, 127.48, 128.18, 138.59; HRMS calcd for C25H38O3BCl (M+)
432.2602, found 432.2641.
[4R-[2(1S*,2S*,3S*),4r,5â]]-2-[1-Chloro-2-methyl-3-(phenyl-
methoxy)pentyl]-4,5-dicyclohexyl-1,3,2-dioxaborolane (8). A solu-
tion of crude [4R-[2(1S*,2S*),4R,5â]]-4,5-dicyclohexyl-2-[1-methyl-
2-(phenylmethoxy)butyl]-1,3,2-dioxaborolane (7) (23.5 g), and dichlo-
romethane (17 g, 200 mmol) in THF (100 mL) was treated with LDA
(40 mL, 2 M, 80 mmol) at -40 °C via cannula. After 10 min, zinc
chloride (23.3 g, 171 mmol), which was fused before use, was added
to the solution. The solution was allowed to warm to room temperature
and kept for 18 h. The solvent was distilled. The mixture was treated
with hexanes (200 mL) and aqueous ammonium chloride (100 mL).
The organic solution was washed with ammonium chloride (100 mL)
and dried over magnesium sulfate. The solution was filtered through
a short pad of magnesium sulfate to remove zinc chloride. The
magnesium sulfate pad was washed with 10% ether in pentane for
complete product recovery. Concentration in a rotary evaporator gave
a mixture containing [4R-[2(1S*,2S,3S*)4R,5â]-2-[1-chloro-2-methyl-
3-(phenylmethoxy)pentyl]-4,5-dicyclohexyl-1,3,2-dioxaborolane (8) (22
1
MHz H NMR (CDCl3) δ 0.97 (t, J ) 7.4 Hz, 3), 1.08 (d, J ) 7 Hz,
3), 1.55-1.8 (m, 2), 2.69 (dp, J ) 2.3, 7 Hz, 1), 3.64-3.69 (m, 1),
4.48 (AB, J ) 11.5 Hz, 1), 4.59 (AB, J ) 11.5 Hz, 1), 7.25-7.35 (m,
5), 9.75 (d, J ) 2.3 Hz, 1);.75 MHz 13C NMR (CDCl3) δ 8.80, 10.09,
23.37, 48.90, 71.50, 80.39, 127.66, 127.69, 128.36, 138.16, 204.69;
HRMS calcd for C13H18O2 (M+) 206.1307, found 206.1317. Anal.
Calcd for C13H18O2: C, 75.69 ; H, 8.80. Found: C, 75.17; H, 8.84.
The amount of epimer in 9 prepared in this manner was estimated by
1H NMR to be 0.5-1.0%.3
Epimer of 9: [R-(R*,S*)]-2-Methyl-3-(phenylmethoxy)pentanal.
Chromatography of samples of 9 on silica or alumina resulted in ∼10%
and ∼50% (R*,S*)-epimer formation, respectively, based on additional
1
1
absorptions in the H NMR; 300 MHz H NMR (CDCl3) δ 0.96 (t, J
) 7 Hz, 3), 1.14 (d, J ) 7 Hz, 3), 1.55-1.8 (m, 2), 2.59 (ddq, J ) 1.0,
3.9, 7.0 Hz, 1), 3.77 (dt, J ) 3.9, 6.6 Hz, 1), 4.52 (AB, J ) 11.5 Hz,
1), 4.54 (AB, J ) 11.5 Hz, 1), 7.25-7.35 (m, 5), 9.75 (d, J ) 1.1 Hz);
75 MHz 13C NMR (CDCl3) δ 8.01, 10.20, 24.33, 49.26, 71.67, 79.80,
127.66, 127.69, 128.36, 138.16, 204.74.
[4R-[2(1S*,2R*,3S*),4r,5â]]-4,5-Dicyclohexyl-2-[1,2-dimethyl-3-
(phenylmethoxy)pentyl]-1,3,2-dioxaborolane (10). Crude [4R-
[2(1S*,2S*),4R,5â]]-4,5-dicyclohexyl-2-[1-methyl-2-(phenylmethoxy)-
butyl]-1,3,2-dioxaborolane (7) (47 g) was converted to [4R-
[2(1S*,2S*,3S*),4R,5â]]-2-[1-chloro-2-methyl-3-(phenylmethoxy)pentyl]-
4,5-dicyclohexyl-1,3,2-dioxaborolane (8) by the method described
above. After the reaction mixture had been treated with zinc chloride
and kept at 25 °C for 18 h, the solution was cooled to 0 °C and
methylmagnesium chloride (136 mL, 2.5 M, 340 mmol) was added
dropwise. The solution was allowed to warm to room temperature and
kept for 36 h. Aqueous ammonium chloride (20 mL) was added to
the mixture. Some gas (methane) was liberated. The solvent was
removed by vacuum distillation. To the mixture were added pentane
(800 mL) and saturated aqueous ammonium chloride (400 mL). The
organic solution was washed with ammonium chloride solution (300
mL) followed by 1 M hydrochloric acid (2 × 300 mL) and brine (400
mL) and dried over magnesium sulfate. Concentration in a rotary
evaporator gave an oil (52 g) containing [4R-[2(1S*,2R*,3S*),4R,5â]]-
4,5-dicyclohexyl-2-[1,2-dimethyl-3-(phenylmethoxy)pentyl]-1,3,2-di-
oxaborolane (10) and [4R-(4R,5â)]-4,5-dicyclohexyl-2-methyl-1,3,2-
dioxaborolane. The crude 10 was stirred with Raney nickel (15 g) in
ethyl acetate (400 mL) to remove traces of DMSO before it was used
in the next step. A sample of crude 10 was purified with flash column

Asymmetric Synthesis of Stegobinone
J. Am. Chem. Soc., Vol. 118, No. 19, 1996 4565
chromatography (silica, 2% ether/hexanes): 300 MHz 1H NMR (CDCl3)
δ 0.88-1.85 (m, 26), 0.89 (d, J ) 6.9 Hz, 3), 0.92 (t, J ) 7.3 Hz, 3),
0.99 (d, J ) 7.5 Hz, 3), 3.38 (dt, J ) 3.6, 6.8 Hz, 1), 3.76-3.81 (m,
2), 4.47 (AB, J ) 11.6 Hz), 4.57 (AB, J ) 11.6 Hz), 7.22-7.39 (m,
5); 75 MHz 13C NMR (CDCl3) δ 9.51, 14.02, 14.39, 18 (br), 22.23,
25.88, 25.98, 26.46, 27.67, 28.49, 38.45, 43.04, 71.36, 82.93, 83.21,
127.19, 127.69, 128.17, 139.46; HRMS calcd for C28H45O3B (M+)
440.3462, found 440.3492. Anal. Calcd for C28H45BO3: C, 76.35;
H, 10.30; B, 2.45. Found: C, 76.49; H, 10.33; B, 2.58.
evaporator gave [R-(R*,S*)]-2-(1,2-dimethyl-3-oxopentyl)-4,4,5,5-tet-
ramethyl-1,3,2-dioxaborolane (15a) as an oil: 13 g, 54 mmol [39.7%
overall yield from [4R-(4R,5â)-4,5-dicyclohexyl-2-ethyl-1,3,2-diox-
aborolane (4)]; 1H NMR (CDCl3) δ 0.90 (d, J ) 7.3, 3 Hz, 3), 1.03 (t,
J ) 7.2 Hz, 3), 1.11 (d, J ) 7.0 Hz, 3, 1.24 (s, 12), 2.3-2.5 (m, 3),
2.62 (dq, J ) 7.1, 7.2 Hz, 1); 75 MHz 13C NMR (CDCl3) δ 7.69, 13.07,
16.01, 24.64, 24.73, 34.35, 48.63, 82.97, 215.53; HRMS calcd for
C13H25BO3 (M+) 240.1897, found 240.1874. Anal. Calcd for C13H25-
BO3: C, 65.02; H 10.49 B, 4.50. Found: C, 64.83; H, 10.20, B, 4.16.
[4R-[2(1S*,2R*,3S*),4r,5â]]-4,5-Dicyclohexyl-2-(3-hydroxy-1,2-
dimethylpentyl)-1,3,2-dioxaborolane (11). A solution of crude [4R-
[2(1S*,2R*,3S*),4R,5â]]-4,5-dicyclohexyl-2-[1,2-dimethyl-3-(phenyl-
methoxy)pentyl]-1,3,2-dioxaborolane (10) (52 g) in ethyl acetate (400
mL) was stirred with palladium on charcoal catalyst (8 g, 10%) under
1 atm of hydrogen until TLC or GC analysis showed no 10 remaining.
The mixture was filtered through a pad of Celite. Concentration in a
rotary evaporator gave a mixture of impurities [4R-(4R,5â)]-4,5-
dicyclohexyl-2-(1-methylethyl)-1,3,2-dioxaborolane and [4R-(4R,5â)]-
4,5-dicyclohexyl-2-methyl-1,3,2-dioxaborolane, together with the major
product [4R-[2(1S*,2R*,3S*),4R,5â]]-4,5-dicyclohexyl-2-(3-hydroxy-
1,2-dimethylpentyl)-1,3,2-dioxaborolane (11) (39.8 g), which was used
in the next step without further purification: 300 MHz 1H NMR
(CDCl3) δ 0.85-1.81 (m, 35), 2.10-2.25 (br, 1 H), 3.29-3.36 (m, 1),
3.80-3.84 (m, 2); 75 MHz 13C NMR (CDCl3) δ 10.37, 13.12, 15.43,
25.87, 25.97, 26.43, 27.61, 27.77, 28.42, 41.63, 42.90, 76.58, 83.36;
HRMS calcd for C21H37O2B (M+ - H2O) 332.2886, found 332.2903.
O-Benzylstegobiol {[2S-[2r,3r,6(1R*,2R*)]]-2,3-Dihydro-2,3,5-
trimethyl-6-[1-methyl-2-(phenylmethoxy)butyl]-4H-pyran-4-one} (19)
via 9-Borabicyclo[3.3.1]nonane 9-Triflate and 15a. A solution of
[R-(R*,S*)]-2-(1,2-dimethyl-3-oxopentyl)-4,4,5,5-tetramethyl-1,3,2-di-
oxaborolane (15a) (4.25 g, 17.7 mmol) in dichloromethane (60 mL)
was added dropwise at -78 °C to a mixture of 9-borabicyclo[3.3.1]-
nonane-9-triflate (0.5 M, 53 mL, 26.5 mmol; 11B NMR δ 64.7 (br)
and N,N-diisopropylethylamine (5.3 mL, 30 mmol) in dichloromethane
(60 mL). The 11B NMR spectrum of the 9-borabicyclononane 9-triflate
and N,N-diisopropylethylamine mixture showed a broad peak at -0.4
ppm. The reaction mixture was stirred at -78 °C for 2 h. A sample
was withdrawn at this time, and the 11B NMR spectrum was taken at
-20 °C. There were three major peaks at δ 0.9, 33.9, and 58.7. The
peak at δ 0.9 disappeared after 4 h at 0 °C. After the initial period at
-78 °C, the reaction mixture, which presumably contained [R-(R*,S*)]-
2-[1,2-dimethyl-3-[(9-borabicyclo[3.3.1]non-9-yl)oxy]-3-pentenyl]-
4,4,5,5-tetramethyl-1,3,2-dioxaborolane (16a), was placed in an ice-
water bath for 24 h. The mixture was cooled to -78 °C, and a solution
of aldehyde 9 (3.65 g, 17.7 mmol) in 20 mL of dichloromethane was
added via syringe. The solution was allowed to warm to 0 °C and
kept for 24 h. Ether (200 mL), 2-aminoethanol (1 g), and aqueous
ammonium chloride (100 mL) were added to the mixture. The aqueous
layer was extracted with ether (3 × 100 mL). The combined organic
layer was washed with brine (100 mL) and dried over magnesium
sulfate. Concentration in a rotary evaporator yielded a product (12.64
g) containing keto boronic ester 15a, aldehyde 9, and aldol condensation
product [1S-[1R*,2S*,4(R* and/or S*),5(R* and/or S*),6S*,7R*]]-2-
[5-hydroxy-1,2,4,6-tetramethyl-3-oxo-7-(phenylmethoxy)nonyl]-4,4,5,5-
tetramethyl-1,3,2-dioxaborolane (17a). The solution of the crude
mixture in dichloromethane (120 mL) was stirred with pyridinium
dichromate (39.9 g, 106 mmol) and molecular sieves (4 Å, 16 g) at
room temperature for 12 h. Celite was added to the mixture and stirring
was continued for 30 min. The mixture was filtered through a short
pad of Florisil. The solid was washed with ethyl acetate (200 mL).
Concentration in a rotary evaporator gave a mixture of keto boronate
14, aldehyde 9, and [1S-[1R*,2S*,4(R* and/or S*),6R*,7R*]]-4,4,5,5-
tetramethyl-2-[1,2,4,6-tetramethyl-3,5-dioxo-7-(phenylmethoxy)nonyl]-
1,3,2-dioxaborolane (18a) (9.26 g). To a solution of the above mixture
in 140 mL of THF and 70 mL of phosphate buffer (pH 8) was added
hydrogen peroxide (6.8 mL, 30%) at 0 °C. After 0.5 h, the cold bath
was removed. After 3 h at room temperature, the solution was treated
with sodium iodide and sodium thiosulfate. The THF solution was
separated from the aqueous solution. The aqueous layer was extracted
with ether (2 × 200 mL). The combined organic layer was washed
with sodium thiosulfate solution (10%, 300 mL) and dried over
magnesium sulfate. Concentration in a rotary evaporator gave a mixture
of deboronated product (7 g). A solution of the above mixture in
chloroform (32 mL) was stirred with trifluoroacetic acid (10.7 mL) at
[3S-[2(3'R*,4'S*,5'R*),3r,4â,5â)]]-2,2′-Oxybis(5-ethyl-3,4-dimethyl-
1,2-oxaborolane) (13). A solution of crude [4R-[2(1S*,2R*,3S*),4R,5â]]-
4,5-dicyclohexyl-2-(3-hydroxy-1,2-dimethylpentyl)-1,3,2-dioxaboro-
lane (11) (39.8 g) in diethyl ether (400 mL) was stirred with sodium
hydroxide (1 M, 400 mL) at room temperature for 8 h. At this point
the ratio of free (R,R)-1,2-dicyclohexyl-1,2-ethanediol to unhydrolyzed
[4R-[2(1S*,2R*,3S*),4R,5â]]-4,5-dicyclohexyl-2-(3-hydroxy-1,2-dim-
ethylpentyl)-1,3,2-dioxaborolane (11) in the ether phase was 4:1 as
determined by GC and proton NMR analysis. The 1,2-dicyclohexy-
lethanediol was removed by separating the organic layer and evaporating
diethyl ether. The oily residue was redissolved in diethyl ether (300
mL) and returned to the sodium hydroxide aqueous layer for another
8 h until GC or NMR showed no 11 left. The sodium hydroxide
aqueous layer, which presumably contained sodium [3S-(3R,4R,5â)]-
2,2-dihydroxy-3,4-dimethyl-5-ethyl-1,2-oxaborolate24 (12), was acidified
with hydrochloric acid to pH ∼2-3. The solution was extracted with
ethyl acetate (2 × 400 mL) and dried over magnesium sulfate.
Concentration in a rotary evaporator gave pure [3S-[2(3'R*,4'S*,5'R*),-
(3R,4â,5â)]-2,2′-oxybis(5-ethyl-3,4-dimethyl-1,2-oxaborolane) (13): 8.5
g [47% overall yield from [4R-(4R,5â)-4,5-dicyclohexyl-2-ethyl-1,3,2-
1
dioxaborolane (4)]; H NMR (CDCl3) δ 0.91 (d, J ) 7.9 Hz, 3), 0.93
(d, J ) 7.0 Hz, 3), 0.98 (t, J ) 7.3 Hz, 3), 1.40 (m, 2), 1.61 (m, 1),
1.98 (m, 1), 3.76 (dt, J ) 4.1, 7.3 Hz, 1); 75 MHz 13C NMR (CDCl3)
δ 8.18, 10.25, 13.89, 20.55 (br), 27.88, 40.43, 86.78; HRMS calcd for
C14H28B2O3 (M+) 266.2225, found 266.2196.
[R-(R*,S*)]-2-(1,2-Dimethyl-3-oxopentyl)-4,4,5,5-tetramethyl-1,3,2-
dioxaborolane (15a). A solution of [S-(3R,4R,5â)]-2,2′-oxybis(3,4-
dimethyl-5-ethyl-1,2-oxaborolane) (13) (8.5 g, 31.9 mmol) and pinacol
(11.3 g, 95.8 mmol) in diethyl ether (150 mL) was stirred at room
temperature for 1 h. Concentration in a rotary evaporator gave a
mixture of four different compounds, presumably including [1S-
(1R*,2S*,3R*)]-2-(3-hydroxy-1,2-dimethylpentyl)-4,4,5,5-tetramethyl-
1,3,2-dioxaborolane (14). One of the constituents had the composition
(13 + pinacol - H2O) [HRMS calcd for C20H40B2O4 (M+) 366.3113,
found 366.3088]. A solution of the foregoing mixture and pyridinium
dichromate (108 g) in dichloromethane (300 mL) was stirred with
molecular sieves (40 g) at room temperature for 3 h. Celite (30 g) and
silica gel (30 g) were added to the mixture, and stirring was continued
for 30 min. The mixture was filtered though a short pad of Florisil.
The solid was washed with ethyl acetate (2 × 200 mL) and then
concentrated in a rotary evaporator and purified by redissolving in
pentane (300 mL) and washing with water (2 × 200 mL). The pentane
solution was dried over magnesium sulfate. Concentration in a rotary
1
room temperature. The reaction was complete by H NMR analysis
after 50 min. Ether (200 mL) was added to the solution. The solution
was washed with sodium bicarbonate (saturated 100 mL) and dried
over magnesium sulfate. Concentration in a rotary evaporator and
separation of product by flash column chromatography (silica, 20%
ether/hexanes) gave O-benzylstegobiol (19): 2.06 g, 6.5 mmol (36.7%
from 15a); solidifies in the freezer but melts below 20 °C; 300 MHz
1H NMR (CDCl3) δ 0.97 (t, J ) 7.4 Hz, 3), 1.01 (d, J ) 7.3 Hz, 3),
1.06 (d, J ) 7 Hz, 3), 1.25 (d, J ) 6.6 Hz, 3), 1.43-1.59 (m, 1),
1.7-1.8 (m, 1), 1.77 (s, 3), 2.25 (dq, J ) 3.4, 7.3 Hz, 1), 3.02 (dq, J
) 7, 8.8 Hz, 1), 3.56 (ddd, J ) 3.5, 6.2, 8.8 Hz, 1), 4.25 (dq, J ) 3.4,
6.6 Hz, 1), 4.41 (AB, J ) 11.5 Hz, 1), 4.52 (AB, J ) 11.5 Hz, 1),
7.25-7.38 (m, 5); 13C NMR δ 8.47, 9.11, 9.45, 13.69, 16.05, 23.31,
39.28, 43.66, 71.85, 76.03, 80.92, 108.58, 127.45, 127.50, 128.29,
(24) Chemical Abstracts name: sodium [T-4-[3S-(3R*,4S*,5R*)]]-dihy-
droxy[4-methyl-3-hexanolato(2-)-C5,O3]borate(1-).

4566 J. Am. Chem. Soc., Vol. 118, No. 19, 1996
Matteson et al.
138.67, 173.69, 197.58;25 HRMS calcd for C20H28O3 (M+) 316.2038,
found 316.1999. Anal. Calcd for C20H28O3: C, 75.91; H, 8.92.
Found: C, 75.39; H, 8.89.
) 6.9 Hz, 3), 1.32 (d, J ) 6.6 Hz, 3), 1.4-1.5 (m, 1), 1.73 (s, 3), 2.32
(dq, J ) 3.4, 7.3 Hz, 1), 2.77 (dq, J ) 5.9, 6.9 Hz, 1), 3.66-3.72 (m,
1), 4.42 (dq, J ) 3.4, 6.6 Hz, 1).
[2S-[2r,3â,6(1R*,2R*)]]-2,3-Dihydro-6-[1-methyl-2-(phenyl-
methoxy)butyl]-2,3,5-trimethyl-4H-pyran-4-one (3-Epimer of O-
Benzylstegobiol). When 9-bromo-9-borabicyclo[3.3.1]nonane was used
as the enolization reagent, the 3-epimer was found in the aldol reaction
mixture and was separated by flash column chromatography (silica,
[2S-[2r,3r,6(1S*)]]-2,3-Dihydro-6-[1-methyl-2-oxobutyl]-2,3,5-tri-
methyl-4H-pyran-4-one (Stegobinone) (1). A mixture of stegobiol
(3) (0.70 g, 3 mmol), tetrapropylammonium perruthenate (TPAP) (120
mg, 0.34 mmol), N-methylmorpholine N-oxide (0.76 g, 6.5 mmol), and
molecular sieves (8 g) in 40 mL of dichloromethane was stirred at
room temperature for 1 h. The mixture was separated by flash
chromatography on a short column of silica with 4:1 dichloromethane/
ether to yield (2S,3R,1′R)-stegobinone (1) as an oil, which crystallized
after removal of solvent under vacuum and chilling to -20 °C (0.65 g,
2.66 mmol, 89%), mp 45-48 °C. Recrystallization from pentane
removed the ∼2% of unknown impurity indicated by a 1H-NMR
multiplet at δ 4.55 and yielded pure stegobinone (1): mp 51.5-52.5
°C (uncorrected) (lit.4 mp 52.5-53.5 °C); 500 MHz 1H NMR (CDCl3)
δ 1.03 (d, J ) 7.3), 1.05 (t, J ) 7.3), 1.28 (d, J ) 6.6), 1.30 (d, J )
7.0), 1.788 (s), 2.32-2.55 (m), 3.63 (q, J ) 7.0), 4.45 (dq, J ) 3.54,
1
20% ether/hexanes): 300 MHz H NMR (CDCl3) δ 0.96 (t, J ) 7.4
Hz, 3), 1.06 (d, J ) 7.0 Hz, 6), 1.34 (d, J ) 6.3 Hz, 3), 1.43-1.59 (m,
1), 1.7-1.8 (m, 1), 1.77 (s, 3), 2.25 (m, 1), 3.03 (dq, J ) 7, 8.55 Hz,
1), 3.54 (ddd, J ) 3.5, 6.3, 8.55 Hz, 1), 3.79 (dq, J ) 6.3, 12.36 Hz,
1), 4.41 (AB, J ) 11.5 Hz, 1), 4.51 (AB, J ) 11.5 Hz, 1), 7.25-7.38
(m, 5); 75 MHz 13C NMR (CDCl3) δ 8.53, 9.28, 10.79, 13.46, 19.18,
23.20, 39.20, 44.74, 71.74, 78.88, 80.97, 109.50, 127.45, 127.54, 128.28,
138.70, 173.39, 195.29; HRMS calcd for C20H28O3 (M+) 316.2038,
found 316.2011.
[2S-[2r,3r,6(1R*,2R*)]]-2,3-Dihydro-6-[2-hydroxy-1-methylbutyl]-
2,3,5-trimethyl-4H-pyran-4-one (Stegobiol) (3). Method A: With
Palladium Hydroxide. O-Benzylstegobiol (19) (1.1 g, 3.48 mmol)
and 20% palladium hydroxide on charcoal (30 mg) in ethanol (8 mL)
and cyclohexene (4 mL) were stirred under reflux for 12 h. The mixture
was filtered through a pad of Celite, concentrated in a rotary evaporator,
and purified by flash column chromatography (silica, 40% ethyl acetate/
hexanes) to yield (2S,3R,1′S,2′S)-stegobiol (3) (0.7 g, 86%). Stegobiol
(3) was crystallized after complete removal of solvent: mp 73-74.2
°C (sealed tube, uncorrected); [R]25D -118.7° ((7°) (c ) 0.107, CHCl3)
1
6.6) [lit.8 500 MHz H NMR (CDCl3) δ 1.04 (d, J ) 7), 1.06 (t, J )
7), 1.29 (d, J ) 7), 1.31 (d, J ) 7), 1.79 (s), 2.3-2.5 (m), 3.62 (q),
4.55 (dq, J ) 3.5, 7)]; 75 MHz 13C NMR (CDCl3) δ 7.87, 9.37, 9.40,
12.77, 15.68, 33.87, 43.66, 49.15, 77.17, 109.41, 168.99, 197.13, 207.60
[lit.4,7 25 MHz 13C NMR similar ((0.1 δ)]. Pure stegobinone stored
in the freezer for 9 months showed no measurable amount (<1%) of
1
epimer by 500 MHz H NMR. After storage for ∼1 year at 20-25
°C, the crystals became visibly contaminated with liquid, the odor
changed perceptibly, and the 1H NMR indicated several percent of the
1′-epimer.
[lit.10 [R]23 -98.3° (c ) 0.06, CHCl3); lit.11 -110° ((6°) (c ) 0.42,
D
1
CHCl3)]; 500 MHz H NMR (CDCl3) δ 0.99 (t, J ) 7.3 Hz, 3), 1.02
Epimerization of Stegobinone (1) to (2S,3R,1′S)-epi-Stegobinone
(2). An NMR sample of stegobinone (1) in CDCl3 treated with
triethylamine showed a few percent of epi-stegobinone (2) after 1 day
and a ∼1:1 mixture of 1 and 2 after 3 months. Data for 2: 500 MHz
1H NMR (CDCl3) δ 1.05 (d, J ) 7.3 Hz), 1.07 (t, J ) 7.3 Hz), 1.29 (d,
J ) 6.9 Hz), 1.30 (d, J ) 6.6 Hz), 1.80 (s), 2.3-2.6 (m), 3.65 (q, J )
6.9 Hz), 4.43 (dq, J ) 3.4d, 6.6q Hz) [lit.26 500 MHz 1H NMR (CDCl3)
δ 1.05 (d, J ) 7.3 Hz, 3), 1.07 (t, J ) 7.3 Hz, 3), 1.29 (d, J ) 7.1 Hz,
3), 1.30 (d, J ) 6.7 Hz, 3), 1.80 (s, 3), 2.31-2.56 (m, 3), 3.65 (q, J )
7.1 Hz, 3), 4.43 (dq, J ) 3.4, 6.7 Hz, 1)].
(d, J ) 7.3 Hz, 3), 1.16 (d, J ) 7.1 Hz, 3), 1.31 (d, J ) 6.6 Hz, 3),
1.36-1.44 (m, 1), 1.52-1.60 (m, 1), 1.73 (s, 3), 1.93 (br s, 1), 2.36
(dq, J ) 3.4, 7.3 Hz, 1), 2.84 (dq, J ) 6.6, 7.1 Hz, 1), 3.56 (ddd, J )
3.7, 6.4, 8.6 Hz, 1), 4.48 (dq, J ) 3.7, 6.6 Hz, 1) [lit.4 1H NMR δ 1.00
(t), 1.04 (d), 1.18 (d, J ) 7.1 Hz), 1.33 (d), 1.42 (m), 1.60 (ddq, J )
14.0, 7.4 Hz), 1.75 (s), 2.38 (dq, J ) 7.3 Hz), 2.86 (dq, J ) 6.8 Hz),
3.58 (ddd, J ) 3.8, 8.5 Hz), 4.49 (dq, J ) 3.4, 6.6 Hz)]; 75 MHz 13C
NMR (CDCl3) δ 9.22, 9.43, 10.08, 14.72, 15.89, 28.25, 40.85, 43.67,
75.32, 76.60, 109.29, 172.76, 197.1; HRMS calcd for C13H22O3 (M+)
226.1569, found 226.1557. From the 1H NMR, no 1′-epimer was
observed, even though the 13C-satellite was observed. Since we had
no reference NMR spectrum of the 2-epimer, we cannot rule out the
presence of some 2-epimer. Anal. Calcd for C13H22O3: C, 68.99; H,
9.80. Found: C 68.90; H, 9.66.
Acknowledgment. This work was supported by NSF Grants
CHE-8922672 and CHE-9303074 and a grant from the Wash-
ington Technology Center. Support of the WSU NMR Center
by NIH Grant RR 0631401, NSF Grant CHE-9115282, and
Battelle Pacific Northwest Laboratories Contract No. 12-097718-
A-L2 is also gratefully acknowledged.
Method B: With Methanesulfonic Acid. A mixture of O-
benzylstegobiol (19) (1.71 g, 5.41 mmol) and methanesulfonic acid
(13.6 mL) in chloroform (30 mL) was stirred for 1 h. The solution
was poured onto ice (45 mL). Ethyl acetate (150 mL) was added. The
organic layer was washed with sodium bicarbonate (saturated, 30 mL)
and brine (30 mL) and dried over MgSO4. Removal of solvent and
separation of product by flash column chromatography (silica, 25%
ethyl acetate/hexanes) yielded (2S,3R,1′S,2′S)-stegobiol (3) (1.03 g,
84%).
Supporting Information Available: Text describing the
experimental details for alternative preparations of compounds
6, 7, 8, 10, and 19, for preparations of 15b, and for the
alternative route to 19 via 20-25 (13 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, can be ordered
from the ACS, and can be downloaded from the Internet; see
any current masthead page for ordering information and Internet
access instructions.
[2S-[2r,3r,6(1S*,2R*)]]-2,3-Dihydro-6-[2-hydroxy-1-methylbutyl]-
2,3,5-trimethyl-4H-pyran-4-one ((1′)-Epimer of Stegobiol). When
an impure sample of the aldehyde 9 was used, chromatography yielded
1
an impure sample of the 1′-epimer of stegobiol: 300 MHz H NMR
(CDCl3) δ 0.97 (t, J ) 7.4 Hz, 3), 1.04 (d, J ) 7.3 Hz, 3), 1.18 (d, J
(25) In an early run, a sample of 19 known to contain (1′R)-epimer
showed additional peaks in the 75 MHz 13C NMR at δ 8.88, 9.32, 14.95,
16.24, 24.71, 40.05, 43.80, 72.52, 77.04, 81.78, 108.40, 127.57, 127.71,
128.34, and 173.30.
JA960345A
(26) Ebata, T.; Mori, K. Agric. Biol. Chem. 1990, 54, 527-530.