Synthesis of L-daunosamine derivatives on the basis of the asymmetric dihydroxylation of 3-((E)-1-propenyl)-4,5-dihydroisoxazole

Article abstract of DOI:10.1021/jo962293d

Methyl L-N,O-diacetyldaunosaminide was prepared from 3-nitro-4,5-dihydroisoxazole in 8.5% overall yield. A key step in the synthesis involved the AD reaction of (E)-3-(1-propenyl)-4,5-dihydroisoxazole (2b), affording the corresponding diol in 76% yield (92% ee). A second key step involved reductive cleavage of the dihydroisoxazole 4a and subsequent N-acetylation to afford separable diastereomeric γ-(acetylamino)alcohols 7a and 8a in 62% yield (72:28, 7a/8a). Swern oxidation of 7a and subsequent methanolysis followed by acetylation provided methyl L-N,O-diacetyldaunosaminide as an anomeric mixture. The AD reactions of chiral alkenyl dihydroisoxazole 16 with (DHQ)₂-PHAL and (DHQD)₂-PHAL afforded diastereomeric diol products, isolated as the acetates 18 and 19 (98:2 and 5:95 ratios, respectively, depending on the chiral auxiliary).

Full text of DOI:10.1021/jo962293d

J . Org. Chem. 1997, 62, 3671-3677
3671
Syn th esis of L-Da u n osa m in e Der iva tives on th e Ba sis of th e
Asym m etr ic Dih yd r oxyla tion of
3-((E)-1-P r op en yl)-4,5-d ih yd r oisoxa zole
Peter A. Wade,*,† Stephen G. D’Ambrosio,† J etla Appa Rao,† Sharmila Shah-Patel,†
Damien T. Cole,† J ames K. Murray, J r.,† and Patrick J . Carroll‡
Department of Chemistry, Drexel University, Philadelphia, Pennsylvania 19104, and the Department of
Chemistry, University of Pennsylvania, Philadelphia, Pennsylvania 19104
Received December 9, 1996X
Methyl L-N,O-diacetyldaunosaminide was prepared from 3-nitro-4,5-dihydroisoxazole in 8.5% overall
yield. A key step in the synthesis involved the AD reaction of (E)-3-(1-propenyl)-4,5-dihydroisoxazole
(2b), affording the corresponding diol in 76% yield (92% ee). A second key step involved reductive
cleavage of the dihydroisoxazole 4a and subsequent N-acetylation to afford separable diastereomeric
γ-(acetylamino)alcohols 7a and 8a in 62% yield (72:28, 7a /8a ). Swern oxidation of 7a and
subsequent methanolysis followed by acetylation provided methyl L-N,O-diacetyldaunosaminide
as an anomeric mixture. The AD reactions of chiral alkenyl dihydroisoxazole 16 with (DHQ)2-
PHAL and (DHQD)2-PHAL afforded diastereomeric diol products, isolated as the acetates 18 and
19 (98:2 and 5:95 ratios, respectively, depending on the chiral auxiliary).
In tr od u ction
reactions of 16 were used to initially confirm that
3-alkenyl DHIs undergo AD reaction according to the
Sharpless model.
Several syntheses of aminosugar derivatives employing
4,5-dihydroisoxazoles (DHIs) as intermediates have been
reported.1,2 The DHI ring typically serves as a latent
γ-amino alcohol synthon in these syntheses. The key
transformation, reductive cleavage of the DHI ring,
affords the corresponding γ-amino alcohols with a dem-
onstrated diastereofacialselectivity at the C,N-double
bond. Substituents at any of three sites, the C-4 ring
atom, C-5 ring atom, and the C-3 side chain (R- and
â-positions), can control access of the reducing agent to
the DHI C,N-double bond. Thus, the main advantage of
a DHI intermediate is that it can be stereoselectively
converted to the desired γ-amino alcohol.
The synthesis of 3-(1-alkenyl)-substituted DHIs is
relatively straightforward, and we have previously shown3
that the Sharpless AD reaction4 can be applied to these
compounds to afford DHI R,â-diols. However, we had not
previously applied phthalazine-derived chiral auxiliaries
to enantioselective DHI R,â-diol synthesis. Here, we
describe the enantioselective total synthesis of L-daun-
osamine derivatives on the basis of the AD reaction of
(E)-3-(1-propenyl)-4,5-dihydroisoxazole (2b) and subse-
quent reductive cleavage of the DHI ring. We also report
experimental procedures for double asymmetric synthesis
using chiral alkenyl DHI 16 in the AD reaction. The
Resu lts a n d Discu ssion
The synthesis of 3-(1-propynyl)-4,5-dihydroisoxazole (1)
from 3-nitro-4,5-dihydroisoxazole and its conversion to
the Z- and E-isomers of 3-(1-propenyl)-4,5-dihydroisox-
azole (2a ,b) have previously been reported.1b Lindlar
reduction of alkynyl DHI 1 afforded either predominantly
(Z)-alkene 2a or (E)-alkene 2b depending on the condi-
tions used (Scheme 1). Using 5% by weight of catalyst
afforded crude alkene that was largely the Z-isomer (2a /
2b, 90:10). Using 50% by weight of the Lindlar catalyst
afforded a 50:50 mixture of the (E)-alkene 2b and the
overreduction product 3-(1-propyl)-4,5-dihydroisoxazole.
Here, we report an improved procedure for obtaining 2b.
The crude alkene, largely Z-isomer (2a /2b, 90:10), was
prepared as previously described and was isomerized to
2b, isolated in 79% overall yield and containing none of
the Z-isomer. The cis f trans isomerization was con-
ducted using catalytic iodine under sun lamp irradia-
tion: the (E)-alkene was obtained nearly pure [contami-
nated by 2% of 3-(1-propyl)-4,5-dihydroisoxazole] after
flash chromatography. Fifty-gram quantities of 2b were
readily preparable by the method, and the small amount
of overreduction product did not interfere with the
subsequent AD reaction.
The AD reaction of alkene 2b was carried out using
AD mix-R and the conditions recommended by Sharpless4
but with one substantive variation: benzenesulfonamide
was used rather than the recommended methanesulfona-
mide. This change was made because chromatographic
separation of methanesulfonamide from diol 3 was inef-
ficient. However, we were able to easily separate the less
polar benzenesulfonamide from 3. In this way, the diol
3 was obtained in 76% isolated yield and with [R]D
+23.7°.
† Drexel University.
‡ University of Pennsylvania.
X Abstract published in Advance ACS Abstracts, April 15, 1997.
(1) (a) Wade, P. A.; D’Ambrosio, S. G.; Price, D. T. J . Org. Chem.
1995, 60, 6302. (b) Wade, P. A.; Rao, J . A.; Bereznak, J . F.; Yuan, C.-
K. Tetrahedron Lett. 1989, 30, 5969. (c) See also: Wade, P. A.; Shah,
S. S.; Govindarajan, L. J . Org. Chem. 1994, 59, 7199.
(2) For pertinent reviews, see: (a) J a¨ger, V.; Mu¨ller, I.; Leibold, T.;
Hein, M.; Schwartz, M.; Fengler, M.; J araskova, L.; Pa¨tzel, M.; LeRoy,
P.-Y. J . Chem. Soc. Belg. 1994, 103, 491. (b) Hassner, A.; Murthy, K.
S. K.; Maurya, R.; Dehaen, W.; Friedman, O. Lect. Heterocycl. Chem.
1994, 687. (c) Torssell, K. B. G. In Nitrile Oxides, Nitrones, and
Nitronates in Organic Synthesis; VCH: New York, 1988. (d) Curran,
D. P. Adv. Cycloadd. 1988, 1, 129. (e) Kozikowski, A. P.; Chen, Y.-Y.
Tetrahedron 1984, 40, 2345.
Conversion of diol 3 to diastereomeric benzylidene
acetals 4a ,b was carried out in 92% yield, and these were
separated by flash chromatography. The major acetal
4a , obtained in 57% isolated yield, proved to be crystal-
(3) (a) Wade, P. A.; Cole, D. T.; D’Ambrosio, S. G. Tetrahedron Lett.
1994, 35, 53. (b) For an account of our first attempts to perform AD
reactions on alkene 2b, see ref 1b.
(4) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. Rev.
1994, 94, 2483 and references cited therein.
S0022-3263(96)02293-1 CCC: $14.00 © 1997 American Chemical Society

3672 J . Org. Chem., Vol. 62, No. 11, 1997
Wade et al.
Sch em e 1a
a
Reagents: (a) H2, Lindlar cat., quinoline, C6H6; (b) cat. I2, hν, C6H6; (c) AD mix-R, PhSO2NH2, t-BuOH/H2O; (d) PhCHdO, ZnCl2; (e)
LiBH4, Et2O, 35 °C; (f) HOCH2CH2NH2, C6H6; (g) Ac2O, NaHCO3, aqueous MeOH; (h) Swern; (i) 0.1 N HCl in MeOH; (j) Ac2O, pyridine.
line, and its repetitive recrystallization resulted in mate-
rial of increased rotation: [R]D rose from +30° to +32.0°.
Assignment of stereochemistry to the two product acetals
was not straightforward. Benzylidene acetals with a
single substituent at the 4-position or with cis-4,5 sub-
stituents are easily assigned on the basis of 1H-NMR
spectra: the isomer having the more upfield signal
attributable to the dioxolane 2-proton has the phenyl
group cis to the substituents.5 However, for 4,5-trans-
F igu r e 1. Preferred direction of hydride attack.
substituents, it is unclear which of the two groups will
γ-C-atoms under the C,N double bond (Figure 1). As-
be cis to the phenyl group. An X-ray structure determi-
suming reaction through a similar conformation, hydride
nation was necessary to make the assignment: the
was then introduced from the top, next to the R-O-atom.
methyl group proved to be cis to the phenyl group in the
major isomer, acetal 4a .22 Hydrolysis of a portion of this
material back to diol 3 gave material [R]D +25.7°. These
results, in conjunction with rotation data for the L-
daunosamine derivative obtained at the end of the
synthesis, confirmed that 4a was now optically pure, or
very nearly so, allowing us to establish that the AD
reaction had occurred with 92% ee.
This is counter to predictions based on the anti-peripla-
nar effect8 but is similar to our previous observation1b
for attack on the racemic DHI acetal 9 leading to
acosamine derivatives. For DHI acetal 9, the γ-C-atom
(methyl group) shields the bottom of the C,N-double bond.
Reductive cleavage1a of acetal 4a using lithium boro-
hydride afforded a nonseparable diastereomeric mixture
of γ-amino alcohols that were N-acetylated and separated
as the acetamide derivatives 7a and 8a . In this way, a
48% yield of 7a and a 14% yield of 8a were obtained, and
the diastereomer ratio of the original reductive cleavage
products was established as 72:28. No variability was
noted here as a function of the lithium borohydride
reagent whether freshly prepared or of commercial
origin.6 It was, however, noted that a lengthy reaction
time with ethanolamine7 was necessary for complete
removal of boron from the initially produced complex.
Diastereoselectivity for the reductive cleavage was
presumably controlled by groups on the side chain
attached at the 3-position of the DHI ring. The X-ray
structure determination established the preferred con-
formation of 4a present in the crystal lattice. This
preferred conformation placed the R-H-atom syn to the
N-atom of the ring and placed the side chain â-C- and
Models do not show similar shielding in 4a , however. We
attribute facial selectivity in 4a as arising predominantly
from the steric influence of groups attached to the R-C-
atom. The X-ray data show a crystal-state conforma-
tional preference where the OR-CR-C3-N2 dihedral
angle is -125.6 ((0.2)° and the Câ-CR-C3-N2 dihedral
angle is 117.2 ((0.3)°. The R-H-atom is located 3((2)°
above the C,N double bond. Assuming the Bu¨rgi-Dunitz9
approach trajectory of 107°, the top face of the C,N-double
bond is more open in this conformation. If the reacting
conformation in solution is similar, attack should be
preferable between the O-atom and R-H-atom, affording
the major observed product 7a . It is also quite possible
(8) Moses, S. R.; Wu, Y.-D.; Rondan, N. G.; Houk, K. N.; Schohe, R.;
J a¨ger, V.; Fronczek, F. R. J . Am. Chem. Soc., 1984, 106, 3880. Attack
anti-periplanar to the R-C,O-bond would be inconsistent with the
observed major product: Kozikowski, A. P.; Ghosh, A. K. J . Am. Chem.
Soc. 1982, 104, 5788 and references cited therein.
(9) (a) Bu¨rgi, H. B.; Dunitz, J . D.; Shefter, E. J . J . Am. Chem. Soc.
1973, 95, 5065. (b) Bu¨rgi, H. B.; Lehn, J .-M.; Wipff, G. J . Am. Chem.
Soc. 1974, 96, 1956.
(5) Assigned by analogy to 2-phenyl-4,5-cis-dimethyl-1,3-dioxalane
diastereomers: Willy, W. E.; Binsch, G.; Eliel, E. E. J . Am. Chem. Soc.,
1970, 92, 5394. Eliel, E. L.; Ko, K.-Y. Tetrahedron Lett. 1983, 24, 3547.
(6) Reductive cleavage of 3-[2-(1,3-dithianyl)]-4-(benzyloxy)-4,5-di-
hydroisoxazole using LiBH4 showed variability of from 70:30 to 95:5
in the product diastereomer ratio: ref 1a.
(7) Krishnamurthy, S.; Brown, H. C. J . Org. Chem. 1977, 42, 1197.

Synthesis of L-Daunosamine Derivatives
J . Org. Chem., Vol. 62, No. 11, 1997 3673
Sch em e 2
Sch em e 3a
a
that lithium borohydride coordinates with 4a at any of
the four heteroatoms. Coordination of lithium at the R-O-
atom might be responsible for directing hydride attack
from the top face to afford 7a preferentially.
Reagents: (a) BuLi, THF, -78 °C; (b) 3-nitro-4,5-dihydro-
isoxazole, 0-5 °C; (c) H2, Lindlar cat., quinoline; (d) cat. I2, hν,
C6H6; (e) pyridinium tosylate, MeOH; (f) Ac2O, pyridine.
â-anomer 5b. The presence of 5b can be easily dis-
cerned: the 1H NMR spectrum in deuteriochloroform
exhibits a signal at δ 3.48 attributable to the C-1 methoxy
protons and occurring downfield from the corresponding
signal of 5a (δ 3.35). The anomer separation procedure
was not suitable for large quantities of material, and we
were unable to obtain pure 5b,13 a compound that
previously had been synthesized free of 5a .
The acetamide 7b, a benzylidene diastereomer of 7a ,
was also converted to an anomeric mixture of methyl
N,O-diacetyldaunosaminide. Swern oxidation of 7b af-
forded aldehyde 6b in 85% yield. Acid-catalyzed metha-
nolysis of 6b afforded deacetalization and ensuing cy-
clization to the methyl pyranoside. This was acetylated
to afford an 83:17 anomeric mixture of 5a ,b in 75%
overall yield but in lower optical purity (91%) than the
anomeric mixture obtained from acetamide 7a .
In a preliminary communication,3a we reported the AD
reaction of chiral 3-alkenyl DHI 16, establishing that
alkenes of this type follow the Sharpless model. The
alkenyl DHI 16 was available via a three-step synthesis
using 3-nitro-4,5-dihydroisoxazole and the known14 opti-
cally active alkyne 14 as starting materials (Scheme 3).
First, alkyne 14 was converted to the corresponding
lithium acetylide using excess butyllithium. The acetyl-
ide was allowed to react with excess 3-nitro-4,5-dihy-
droisoxazole, affording the alkynyl DHI 15, obtained in
76% yield. Lindlar reduction of alkynyl DHI 15 followed
by iodine-catalyzed photochemical cis f trans isomer-
ization afforded alkenyl DHI 17 in 67% overall yield for
the two steps. Removal of the acetonide using pyri-
dinium tosylate15 followed by acetylation afforded alkenyl
DHI 16 in 94% yield from 17.
Using optically active alkenyl DHI 16 as a substrate
for the AD reaction constitutes a double asymmetric
synthesis.16 One can envision a matched pair and a
mismatched pair of stereoisomers. The (DHQD)2-
PHAL4 chiral system recommended by Sharpless et al.
would be expected to promote anti-addition: application
to 16 then should constitute the matched pair, affording
The DHI acetal 4b, the minor benzylidene diastere-
omer in which phenyl is trans to methyl, could also be
reductively cleaved with lithium borohydride: here, too,
two products, 7b and 8b, were obtained in a ratio of 77:
23, respectively, slightly higher diastereoselectivity than
for the major diastereomer. The diols were N-acetylated,
and the resulting acetamide mixture was separated,
affording 7b in 52% yield and 8b in 12% yield. However,
the acetal 4b proved to be a noncrystallizable oil so that
we were unable to increase optical purity as done by
recrystallization of the major diastereomer 4a prior to
reductive cleavage. Hydrolysis of acetal 4b provided diol
3 in 81% yield. This portion of 3 could then presumably
have been reacetalized to improve the efficiency of the
synthesis, although in actual practice this was not carried
out.
Partial autoxidation10 of acetal 4b was noted for a
sample left standing for 2 weeks (Scheme 2). Two
inseparable isomeric esters, 10 and 11, were obtained in
9% yield (37% conversion; 85:15 10/11). It is thought that
autoxidation at C-2 of the dioxolane ring occurred,
affording a hydroperoxide that was reduced to hydroxy
acetal 12. Acid-catalyzed cleavage of 12 then afforded
the esters 10 and 11. The preference for ester 10 might
arise from intramolecular hydrogen bonding in its direct
precursor, oxonium ion 13: protonation at O-3 of 12 to
afford 13 might be favored over protonation at O-1 to
afford the isomeric oxonium ion.
The synthesis of methyl N,O-diacetyldaunosaminide
(two anomers: 5a ,b) was then completed by sequential
Swern oxidation, methanolysis, and acetylation. Swern
oxidation of 7a under the standard conditions afforded
aldehyde 6a in 82% yield. The benzylidene protecting
group was removed by acid-catalyzed methanolysis,
affording the methyl pyranoside. The free hydroxyl
group of the methyl pyranoside was acetylated, and an
82:18 anomeric mixture of 5a ,b was obtained in 73%
overall yield. Flash chromatography as recommended by
J urczak et al.11 was applied to the anomeric mixture, and
pure R-anomer 5a was obtained. It is noteworthy that
the observed optical rotation ([R]D -210.2°) was some-
what higher than rotations reported in the literature.11,12
Indeed, it appears that previous so-called pure samples
of 5a were contaminated with small amounts of the
(12) Arcamone, F.; Cassinelli, G.; Francesschi, G.; Mondelli, R.;
Orezzi, P.; Penco, S. Gazz. Chim. Ital. 1970, 100, 949. Arcamone, F.;
Cassinelli, G.; Francesschi, G.; Mondelli, R. J . Am. Chem. Soc. 1964,
86, 5335.
(13) Horton, D.; Weckerle, W. Carbohydr. Res. 1975, 44, 227.
(14) Taylor, E. C.; Ray, P. S.; Darwish, I. S.; J ohnson, J . L.;
Rajagopalan, K. V. J . Am. Chem. Soc. 1989, 111, 7664.
(15) Sterzycki, R. Synthesis 1979, 724.
(16) Masamune, S.; Choy, W.; Petersen, J . S.; Sita, L. R. Angew.
Chem. Int., Ed. Engl. 1985, 24, 1.
(10) Autoxidation of acetals: (a) Suzuki, M.; Inai, T.; Matsushima,
R. Bull. Chem. Soc. J pn. 1976, 49, 1585. (b) Kuramshin, E. M.; Kulak,
L. G.; Nazarov, M. N.; Zlotsky, S. S. J . Prakt. Chem. 1989, 331, 591.
(11) J urczak, J .; Kozak, J .; Golebiowski, A.; Tetrahedron 1992, 48,
4231.

3674 J . Org. Chem., Vol. 62, No. 11, 1997
Wade et al.
tetraacetate 18 as the major product after acetylation.
This proved to be the case but necessitated an essential
modification of the recommended catalytic conditions.
Attempts to use AD mixes were unsuccessful owing to
insufficient potassium osmate. It was necessary to use
0.08 mol equiv (40 times the recommended amount) of
potassium osmate for efficient dihydroxylation of 18.
After acetylation of the crude diols, the diastereomeric
tetraacetates 18 and 19 were obtained as a 98:2 mixture,
respectively, in 82% overall yield.
the abnormally high potassium osmate requirements of
16 and 17. Dihydroxylation of alkenes is an electrophilic
reaction: low dihydroxylation rates for alkenes possess-
ing an electronegative atom (i.e., O- or N-) at the allylic
site have been noted in some cases although not in
others.4,19,20 We therefore suggest that the high potas-
sium osmate requirement of 16 and 17 is due in part to
the presence of an allylic O-atom. The adjacent C,N-
double bond would also retard an electrophilic process
at the C,C-double bond for all of the alkenyl DHIs. The
C,C-double bond of 16 and 17 is less accessible than the
C,C-double bond of alkenyl DHI 2b, probably contribut-
ing to the lower rate. Finally, although Os chelation at
the DHI ring N-atom cannot be solely responsible for the
high potassium osmate requirement, it may be partially
responsible in combination with the other factors.
Exp er im en ta l Section
Gen er a l Meth od s. Reactions were routinely run under
argon. Workup involved separation of the organic layer,
drying over anhydrous Na2SO4, filtering, and concentration
at reduced pressure. 1H NMR spectra and 13C NMR spectra
were determined in CDCl3 (TMS internal standard) on a
Bruker WM-250 instrument unless otherwise noted. 3-Nitro-
4,5-dihydroisoxazole was prepared in 53% yield from 1-bromo-
3-chloropropane by the published procedure.21 Sunlamp ir-
radiation was conducted using a 250-W GE bulb placed 2 ft
from the Pyrex reaction vessel. Analytical HPLC was con-
ducted using a Rainin Microsorb silica column. The AD mix-R
was used as purchased (Aldrich). Other routine procedures
have been previously published.1a
The configuration assigned to 18 is based on its 1H
NMR spectrum. In particular, the coupling constant J â,γ
) 8.2 Hz is consistent with the anti-introduction of
acetate to the â-C-atom. The glucose-derived pyrroles
20a -c exhibited similar coupling constants: J â,γ ) 8.6-
8.8 Hz.17 In contrast, tetraacetate 19 exhibited J â,γ ) 4.5
P r ep a r a t ion of 3-(1-P r op yn yl)-4,5-d ih yd r oisoxa zole
(1). Propyne was condensed (30 mL, 21.2 g, 0.53 mol) in a
precalibrated reaction flask cooled by dry ice. A solution of
butyllithium (120 mL, 2.5 M in hexanes; 0.3 mol of BuLi) was
cautiously added over 1 min to the flask, directly followed over
2-3 min by THF (880 mL). The cold reaction mixture was
stirred for 30 min and then allowed to warm to -25 °C.
3-Nitro-4,5-dihydroisoxazole (30.5 g, 0.27 mol) was added
dropwise, maintaining the reaction temperature below 0 °C.
The reaction mixture was stirred for 2 h at 0 to -5 °C, treated
with water (18 mL), concentrated, diluted with CH2Cl2 (750
mL), and washed with water (three 150-mL portions). Further
workup and distillation of the crude product afforded 24.8 g
(86% yield) of pure 1: bp 50-55 °C (0.1 mmHg); IR (film) 2234
Hz. It is assumed that the conformations of 18 and
20a -c are similar, presumably a regular zig-zag ar-
rangement.
The preceding results are for the matched pair of
stereoisomers and might be anticipated. Application of
the (DHQ)2-PHAL chiral auxilliary to the AD reaction
of 16, however, presented a mismatched pair where the
stereochemical outcome was not initially predictable. The
AD reaction of 16 using (DHQ)2-PHAL afforded, after
acetylation of the initially produced diols the diastereo-
meric tetraacetates 18 and 19 as a 5:95 mixture, respec-
tively, in 85% overall yield. Thus, the chiral auxilliary
strongly controlled the diastereoselectivity in this mis-
matched pair.
It is noteworthy that alkenyl DHI 2b required only the
recommended amount of potassium osmate for efficient
AD reaction, whereas alkenyl DHI 16 required 40 times
as much as the recommended amount of potassium
osmate for smooth reaction. Alkenyl DHI 17 also exhib-
ited a high potassium osmate requirement for AD reac-
tion as previously noted. At the time preliminary results
on 16 were published, it was suggested18 that the high
potassium osmate requirement might be due to Os
chelation at the DHI ring N-atom resulting in a low
catalyst turnover rate. However, the more recent results
with 2b are not in agreement with this explanation. It
now seems that several factors must be involved with
1
cm-1; H NMR δ 4.40 (t, 2H, J ) 10.3 Hz), 3.02 (t, 2H, J )
10.3 Hz), 2.05 (s, 3H); 13C NMR δ 142.8, 94.2, 68.8, 68.7, 37.6,
3.5; LRMS (EI) m/ e 109 (M+). Anal. Calcd for C6H7NO: C,
66.04; H, 6.47. Found: C, 65.91; H 6.58.
P r ep a r a t ion of (E)-3-(1-P r op en yl)-4,5-d ih yd r oisox-
a zole (2). A solution of alkyne 1 (3.04 g, 28 mmol) and
quinoline (215 mg, 0.8 mmol) in benzene (70 mL) was
transferred to a Parr hydrogenator flask containing a pre-
equilibrated (10 min) mixture of benzene (70 mL) and Lindlar
catalyst (107 mg) under hydrogen. The flask contents were
subjected to vigorous shaking under hydrogen (16 psi) for 30
min and were then filtered, concentrated, and diluted with
CH2Cl2 (100 mL). The resulting solution was sequentially
washed with 5% aqueous HCl (75 mL), 5% aqueous NaHCO3
(75 mL), and water (75 mL). Further workup afforded 2.71 g
of crude alkene (Z/ E, 90:10) contaminated with a small
amount (ca. 1-2% by 1H NMR) of 3-propyl-4,5-dihydroisox-
(19) (a) Cha, J . K.; Christ, W. J .; Kishi, Y. Tetrahedron Lett. 1983,
24, 3943. (b) Christ, W. J .; Cha, J . K.; Kishi, Y. Tetrahedron Lett. 1983,
24, 3947. (c) See also: Stork, G.; Kahn, M. Tetrahedron Lett. 1983,
24, 3951.
(20) For example: Morikawa, K.; Sharpless, K. B. Tetrahedron Lett.
1993, 34, 5575.
(21) Wade, P. A. J . Org. Chem. 1978, 43, 2020.
(22) The author has deposited atomic coordinates for 4a with the
Cambridge Crystallographic Data Centre. The coordinates can be
obtained, on request, from the Director, Cambridge Crystallographic
Data Centre, 12 Union Road, Cambridge, CB2 1EZ, UK.
(17) Go´mez-Sa´nchez, A.; Hidalgo, F.-J .; Chiara, J .-L. Carbohydr. Res.
1987, 167, 55.
(18) See ref 3a, footnote 10.

Synthesis of L-Daunosamine Derivatives
J . Org. Chem., Vol. 62, No. 11, 1997 3675
azole. A solution of the crude alkene (2.71 g, 25 mmol) and I2
(273 mg, 1 mmol) in benzene (90 mL) was stirred and
irradiated with a sunlamp for 3 h. The resulting solution was
concentrated, diluted with CH2Cl2 (150 mL), and washed with
10% aqueous sodium thiosulfate (70 mL). The aqueous layer
was extracted with CH2Cl2 (two 50-mL portions) and the
combined organic layers were worked up to give an oil that
was distilled at reduced pressure to give 2.45 g (79% yield) of
2 contaminated with a trace (ca. 1-2% by 1H NMR) of
3-propyl-4,5-dihydroisoxazole: bp 38-39 °C (0.1 mmHg): IR
The above procedure was repeated, but the crude product
was directly recrystallized from CH2Cl2/hexanes without chro-
matography to give 2.68 g (30% yield) of 4a .
Regen er a tion of Diol 3 fr om 4a a n d 4b. A mixture of
thrice-recrystallized acetal 4a (71 mg, 0.3 mmol), water (1 mL),
and acetic acid (4 mL) was stirred for 46 h at rt, and then
volatiles were removed under reduced pressure to give 48.7
mg of crude product. The crude product was purified by
preparative TLC (EtOAc) to give 38 mg (87% yield) of diol 3:
[R]23 +25.7° (c 0.71, CHCl3).
D
1
(film) 1650, 1622 cm-1; H NMR δ 6.43 (d, 1H, J ) 15.9 Hz),
Similar treatment of the oil 4b (71 mg, 0.3 mmol) gave 36
mg (81% yield) of diol 3: [R]26 +21.7° (c 1.12, CHCl3).
D
6.01 (dq, 1H, J ) 6.7, 15.8 Hz), 4.34 (t, 2H, J ) 10.0 Hz), 3.05
(t, 2H, J ) 10.0 Hz), 1.90 (dd, 3H, J ) 1.6, 6.7 Hz); 13C NMR
δ 157.0, 134.4, 120.9, 68.3, 33.7, 18.0; HRMS (EI) calcd for
C6H9NO (M+) 111.0684, found 111.0684.
Red u ctive Clea va ge of DHI 4a . A solution of 4a (1.04 g,
4.48 mmol) in diethyl ether (100 mL) was added to LiBH4 (36
mL of a 2 M THF solution, 72 mmol of LiBH4), and the
resulting solution was refluxed for 3 d. More LiBH4 (10 mL,
20 mmol of LiBH4) was added, and the solution was refluxed
for another 3 d. The reaction solution was concentrated at
reduced pressure, and CH2Cl2 (100 mL) was added. The
resulting solution was cooled to 0-5 °C, and aqueous 10%
NaH2PO4 (adjusted to pH 7.0 with NaOH, 200 mL) was
cautiously added (foaming!). The organic layer was separated,
and the aqueous layer was extracted with CH2Cl2 (two 100-
mL portions). The combined organic layers were washed with
10% brine (two 100-mL portions). Workup afforded 1.34 g of
crude boron-containing product, which was taken up in
benzene (100 mL). Ethanolamine (2 g) was added, and the
resulting mixture was stirred vigorously for 5 d. Concentra-
tion afforded a residue that was dissolved in CH2Cl2 (80 mL),
the resulting solution being washed with 10% brine (three 30-
mL portions). Further workup gave 1.19 g of boron-free crude
product that was purified by flash chromatography (CH2Cl2/
MeOH from 97:3 to 90:10). An inseparable mixture of dia-
AD Rea ction of Alk en e 2. A mixture of AD mix-R (59.6
g), t-BuOH (225 mL), and water (225 mL) was stirred for 10
min, and benzenesulfonamide (6.89 g, 43.9 mmol) was added.
After 5 min, alkene 2 (4.27 g, 38 mmol) was added, and the
resulting mixture was stirred for 26 h at rt. Anhydrous
Na2SO3 (64 g) was added and stirring continued for 1 h
followed by removal of volatiles at reduced pressure. Methanol
(100 mL) and acetone (400 mL) were added to the residue,
and the resulting slurry was stirred for 1 h. The slurry was
filtered through silica gel and the filtrate concentrated to give
13.5 g of crude product. The crude product was purified by
flash chromatography (acetone/hexanes, 50:50), affording 64
mg of 3-propyl-4,5-dihydroisoxazole (90% pure by 1H NMR)
as the most mobile product: 1H NMR δ 4.27 (t, 2H, J ) 10.0
Hz), 2.93 (t, 2H, J ) 10.0 Hz), 2.36 (t, 2H, J ) 7.5 Hz), 1.63
(sx, 2H, J ) 7.5 Hz), 0.97 (t, 3H, J ) 7.4 Hz); 13C NMR δ 158.5,
67.6, 37.1, 29.2, 19.5, 13.3. HRMS (EI) calcd for C6H11NO (M+)
113.0840, found 113.0840.
1
stereomeric γ-amino alcohols (0.77 g, 75:25 ratio by H NMR)
Further elution afforded benzenesulfonamide followed by
4.55 g of diol 3 as the least mobile fraction. The diol was
Kugelrohr distilled to furnish 4.18 g (76% yield) of 3 (pure by
was obtained. The γ-amino alcohol mixture was dissolved in
MeOH/water (70:30, 18 mL), and the solution was cooled (0-5
°C). Acetic anhydride (0.6 mL, 6.32 mmol) and NaHCO3 (110
mg, 1.32 mmol) were added, and the resulting cold solution
was stirred for 3 h. Volatiles were removed at reduced
pressure, and acetone (10 mL) was added to the residue. The
resulting mixture was filtered, and the filtrate was concen-
trated to give 1.15 g of crude product (7a /8a , 78:22 by 1H
NMR). Repetitive preparative TLC (CH2Cl2/MeOH, 95:5)
furnished 0.60 g (48% yield) of amide 7a as the less mobile
product (pure by 1H NMR). The analytical sample was
1H NMR) as a syrup: bp 110-125 °C (0.04 mmHg); [R]23
D
+23.7° (c 2.93, CHCl3); IR (film) 3378 cm-1; 1H NMR δ 4.35 (t,
2H, J ) 10.0 Hz), 4.26 (d, 1H, J ) 4.5 Hz), 3.95-4.1 (m, 1H),
2.9-3.2 (m, 2H), 2.78 (br var s, 2H), 1.28 (d, 3H, J ) 6.4 Hz);
13C NMR δ 160.0, 71.8, 68.5, 68.4, 34.7, 18.7; HRMS (FAB,
NaBr) calcd for C6H11NO3Na (M + Na+) 168.0637, found
168.0633.
P r ep a r a tion of 4,5-Dih yd r o-3-(5-m eth yl-2-p h en yl-1,3-
d ioxola n -4-yl)isoxa zole (4a ,b). A mixture of diol 3 (5.60 g,
39 mol), ZnCl2 (5.0 g, 36 mmol), and PhCHO (27 mL, 266
mmol) was stirred for 18 h. Aqueous 40% NaHSO3 (75 mL)
was added and the mixture stirred for 15 min. Volatiles were
removed at reduced pressure, and the residue was treated with
CH2Cl2 (100 mL). The resulting slurry was stirred vigorously
for 1 h and was filtered. The filtrate was concentrated to give
9.59 g of crude acetal as a mixture of two diastereomers. Flash
chromatography (hexanes/EtOAc, 93:7) afforded PhCHO as the
most mobile product followed by partially separated acetal
diastereomers. Repetitive chromatography of mixed fractions
afforded pure 4a and pure 4b. The more mobile isomer (2.18
recrystallized from CH2Cl2/hexanes: mp 124-126 °C; [R]23
D
-30.1° (c 1.06, CHCl3); IR (film) 3284, 1649 cm-1; H NMR δ
1
7.35-7.5 (m, 5H), 5.96 (br d, 1H, J ) 8.8 Hz), 5.88 (s, 1H),
4.15-4.35 (m, 2H), 3.82 (dd, 1H, J ) 4.6, 7.7 Hz), 3.7-3.8 (m,
1H), 3.55-3.7 (m, 1H), 2.41 (br s, 1H), 2.05 (s) on 1.95-2.1
(m) [4H total], 1.6-1.75 (m, 1H), 1.44 (d, 3H, J ) 6.0 Hz); 13
C
NMR δ 171.2, 137.5, 129.5, 128.4, 126.6, 103.7, 84.2, 76.4, 58.4,
47.8, 32.9, 23.1, 18.3; HRMS (FAB) calcd for C15H22NO4 (M +
H+) 280.1549, found 280.1548.
Also obtained was 167 mg (14% yield) of amide 8a as the
1
more mobile chromatography fraction (pure by H NMR). An
analytical sample was recrystallized from CH2Cl2/hexanes: mp
163-165 °C; [R]23D +45.5° (c 1.45, CHCl3); IR (film) 3293, 1649
g, 24% yield; pure by 1H NMR) was 4b, isolated as an oil: [R]26
D
1
cm-1; H NMR δ 7.35-7.35 (m, 5H), 6.06 (br d, 1H, J ) 9.1
-15.3° (c 1.86, CHCl3); 1H NMR δ 7.35-7.55 (m, 5H), 6.07 (s,
1H), 4.63 (d, 1H, J ) 6.6 Hz), 4.25-4.45 (m, 3H), 2.85-3.2
(m, 2H), 1.45 (d, 3H, J ) 6.2 Hz); 13C NMR δ 156.8, 137.4,
129.2, 128.2, 126.2, 103.1, 78.5, 75.1, 68.6, 33.9, 17.5; HRMS
(FAB) calcd for C13H16NO3 (M + H+) 234.1130, found 234.1129.
Hz), 5.94 (s, 1H), 4.2-4.3 (m, 1H), 3.95-4.05 (m, 1H), 3.66 (d,
J ) 8.1 Hz) on 3.65-3.8 (m) [2H total], 3.45-3.6 (m, 1H), 3.34
(br s, 1H), 2.10 (s, 3H), 1.75-1.95 (m, 1H), 1.6-1.75 (m, 1H),
1.40 (d, 3H, J ) 6.0 Hz); 13C NMR δ 171.6, 138.0, 129.5, 128.5,
126.4, 104.1, 84.3, 75.9, 58.1, 45.2, 37.1, 23.2, 16.9; HRMS
(FAB) calcd for C15H22NO4 (M + H+) 280.1549, found 280.1546.
P r ep a r a tion of Ald eh yd e 6a . DMSO (0.14 mL, 2.0 mmol)
was added over 5 min to cold (-60 °C) oxalyl chloride [0.4 mL
of a 2 M CH2Cl2 solution; 0.81 mmol (COCl)2], and the resulting
solution was stirred for 5 min. A solution of 7a (192 mg, 0.69
mmol) in CH2Cl2 (3 mL) was added dropwise over 9 min.
Triethylamine (0.5 mL, 3.7 mmol) was added over 2 min, and
the reaction solution was stirred for 90 min at ambient
temperature. Aqueous 10% brine (5 mL) was added, and the
organic layer was separated. The aqueous layer was extracted
with CH2Cl2 (two 15-mL portions), and the combined organic
layers were worked up to afford 177 mg of crude solid product.
The less mobile isomer (5.09 g, 57% yield) was 4a , isolated
as a solid. Recrystallization from CH2Cl2/hexanes afforded
3.56 g of 4a : [R]26 +30.0° (c 1.16, CHCl3). An analytical
D
sample was obtained after two more recrystallizations: mp
82.5-83 °C; [R]26 +32.0° (c 1.41, CHCl3); 1H NMR δ 7.35-
D
7.55 (m, 5H), 5.94 (s, 1H), 4.53 (d, 1H, J ) 7.7 Hz), 4.40 (t, J
) 9.8 Hz) on 4.3-4.45 (m) [3H total], 3.13 (m, 2H), 1.50 (d,
3H, J ) 6.0 Hz); 13C NMR δ 157.0, 137.2, 129.5, 128.4, 126.6,
104.2, 77.8, 76.6, 68.7, 34.8, 17.2; HRMS (FAB) calcd for
C13H16NO3 (M + H+) 234.1130, found 234.1129. Anal. Calcd
for C13H15NO3: C, 66.94; H, 6.48; N, 6.00. Found: C, 66.81;
H, 6.38; N, 5.92.

3676 J . Org. Chem., Vol. 62, No. 11, 1997
Wade et al.
Recrystallization from CH2Cl2/hexanes gave 156 mg (82%
1H NMR δ 7.35-7.55 (m, 5H), 6.03 (s, 1H), 5.71 (br d, 1H, J )
8.4 Hz), 4.15-4.3 (m, 1H), 4.05 (quint, 1H, J ) 6.3 Hz), 3.81
(d, 1H, J ) 6.9 Hz), 3.65-3.75 (m, 1H), 3.4-3.55 (m, 1H), 2.55
(br s, 1H), 1.86 (s) on 1.8-2.0 (m) [4H total], 1.65-1.8 (m, 1H),
1.40 (d, 3H, J ) 6.3 Hz); 13C NMR (CDCl3) δ 171.6, 138.0,
129.4, 128.7, 125.9, 102.4, 85.6, 74.7, 58.1, 45.3, 36.9, 22.9, 18.0;
HRMS (FAB) calcd for C15H22NO4 (M + H+) 280.1549, found
280.1548.
1
yield; pure by H NMR) of 6a : mp 124-126 °C; [R]23 +16.2°
D
1
(c 1.73, CHCl3); IR (film) 3286, 1723, 1656 cm-1; H NMR δ
9.76 (d, 1H, J ) 2.0 Hz), 7.35-7.5 (m, 5H), 6.38 (br d, 1H, J )
9.3 Hz), 5.84 (s, 1H), 4.45-4.6 (m, 1H), 4.05-4.2 (m, 1H), 3.81
(t, 1H, J ) 7.5 Hz), 2.93 (ddd, 1H, J ) 2.0, 5.8, 17.7 Hz Hz),
2.83 (dd, 1H, J ) 4.6, 17.7 Hz), 1.99 (s, 3H), 1.42 (d, 3H, J )
6.0 Hz); 13C NMR δ 201.1, 169.8, 137.1, 129.5, 128.4, 126.6,
103.1, 82.7, 77.5, 46.8, 44.6, 23.2, 18.7; HRMS (FAB) calcd for
C15H20NO4 (M + H+) 278.1392, found 278.1395.
P r ep a r a tion of Ald eh yd e 6b. The procedure employed
for preparation of 6a was repeated on 7b (85 mg, 0.3 mmol)
1
P r ep a r a tion of 5a ,b fr om 6a . A solution of aldehyde 6a
(134 mg, 0.48 mmol) in 0.1 N methanolic HCl was stirred for
12 h at rt and was then concentrated at reduced pressure.
Acetic anhydride (3 mL) was added to the residue followed by
pyridine (0.2 mL), and the resultant was stirred for 3.5 h and
then concentrated. The residue was purified by preparative
TLC (CH2Cl2/MeOH, 90:10) to give 86 mg (73% yield) of 5 as
an anomeric mixture (5a /5b, 82:18 by 1H NMR): [R]23D -160.7°
(c 0.92, CHCl3). Repetitive flash chromatography (hexanes/
EtOAc/MeOH, 59:40:01) furnished the major anomer 5a : mp
to afford 72 mg (85% yield; pure by H NMR) of 6b: mp 124-
126 °C; [R]26D +15.0° (c 1.0, CHCl3); IR (film) 3278, 1717, 1644
cm-1 1H NMR δ 9.73 (s, 1H), 7.35-7.5 (m, 5H), 6.18 (br d,
;
1H, J ) 9.3 Hz), 5.92 (s, 1H), 4.45-4.6 (m, 1H), 4.20 (quint,
1H, J ) 6.2 Hz), 3.92 (t, 1H, J ) 6.2 Hz), 2.93 (ddd, 1H, J )
1, 6.2, 17.8 Hz), 2.73 (dd, 1H, J ) 4.3, 17.8 Hz), 1.98 (s, 3H),
1.38 (d, 3H, J ) 6.2 Hz); 13C NMR (CDCl3) δ 201.1, 169.8,
137.1, 129.4, 128.4, 126.4, 102.5, 83.7, 75.7, 47.0, 43.9, 23.2,
18.9; HRMS (FAB) calcd for C15H20NO4 (M + H+) 278.1392,
found 278.1385.
187-188 °C; [R]23 -210.2° (c 1.18, CHCl3); IR (film) 3316,
P r ep a r a tion of 5a ,b fr om 6b. The procedure employed
for 6a was repeated using 6b (68 mg, 0.24 mmol) to furnish
D
1
1738, 1651 cm-1; H NMR δ 5.34 (br d, 1H, J ) 7.5 Hz), 5.09
(d, 1H, J ) 2.6 Hz), 4.81 (br s, 1H), 4.5-4.65 (m, 1H), 4.05 (q,
1H, J ) 6.5 Hz), 3.35 (s, 3H), 2.19 (s, 3H), 1.94 (s, 3H), 1.7-
1.9 (m, 2H), 1.11 (d, 3H, J ) 6.5 Hz); 13C NMR δ 170.9, 169.7,
97.9, 71.3, 64.9, 54.8, 43.9, 30.5, 23.2, 20.9, 16.8; HRMS (FAB)
calcd for C11H20NO5 (M + H+) 246.1341, found 246.1332.
Also obtained from the flash chromatography was a fraction
consisting of a 5a ,b mixture: 5a /5b 17:83. The 1H NMR
spectrum of 5a was subtracted, and the remaining signals
matched the reported spectrum13 of pure 5b.
45 mg (75% yield) of 5 as an anomeric mixture (5a /5b, 83:17
1
by H NMR): [R]23 -146.6° (c 1.02, CHCl3).
D
P r ep a r a tion of Alk yn yl DHI 15. Butyllithium (1.2 mL
of a 2 M pentane solution; 2.4 mmol) was added dropwise over
15 min to a cold (dry ice) solution of alkyne 1413 (194 mg, 1.53
mmol) in THF (1 mL). The resulting solution was allowed to
warm to room temperature, and a solution of 3-nitro-4,5-
dihydroisoxazole (267 mg, 2.31 mmol) in THF (1 mL) was
added dropwise over 15 min. Water (2 mL) was added, and
organic products were extracted with CH2Cl2 (four 15-mL
portions). The combined extracts were worked up, and the
resulting residue was purified by preparative TLC (CH2Cl2/
MeOH, 95:5) to afford 227 mg (76% yield) of pure 15 as an oil:
F or m a tion of Ben zoa tes 10 a n d 11 fr om DHI Aceta l
4b. A 651 mg sample of 4b was stored for several weeks neat
in a stoppered flask. The partially air-oxidized sample was
then purified by preparative TLC (hexanes/EtOAc, 65:35) to
furnish 488 mg of 4b as a more mobile fraction and 64 mg of
an 85:15 mixture (1H NMR) of 10 and 11 as a less mobile
fraction. Attempts at separating the mixture of 10 and 11 by
preparative TLC or crystallization failed: [R]24D +28.3° (c 1.95,
CHCl3); IR (neat) 3458, 1716 cm-1; 1H NMR δ 8.0-8.1 (m, 2H),
7.5-7.65 (m, 1H), 7.4-7.5 (m, 2H), 5.76 (d, 1H of 11, J ) 4.1
Hz), 5.35-5.40 (m, 1H), 4.65 (d, 1H of 10, J ) 4.9 Hz), 4.3-
4.45 (m, 2H), 2.8-3.2 (m, 2H), 1.65 (br s, 1H), 1.45 (d, 3H of
10, J ) 6.5 Hz), 1.33 (d, 3H of 11, J ) 6.5 Hz); 13C NMR δ
166.0, 165.6*, 158.7, 133.5*, 133.2, 129.8*, 129.6, 128.4, 72.8*,
71.7, 70.6, 68.9, 68.5*, 67.8*, 35.9*, 34.6, 18.7*, 16.0 (*low
intensity signals attributed to 11); HRMS (FAB) calcd for
C13H15NO4 (M + H+) 250.1079, found 250.1081.
[R]23 +39.4° (c 1.7, MeOH); IR (film) 2233 cm-1 1H NMR
;
D
(CDCl3) δ 4.84 (t, 1H, J ) 6.2 Hz), 4.39 (t, 2H, J ) 10.4 Hz),
4.17 (dd, 1H, J ) 6.5, 8.2 Hz), 3.96 (dd, 1H, J ) 6.0, 8.2 Hz),
3.02 (t, 2H, J ) 10.4 Hz), 1.45 (s, 3H), 1.35 (s, 3H); 13C NMR
δ 142.3, 110.5, 94.6, 74.6, 69.7, 69.3, 65.4, 37.7, 25.9, 25.6;
LRMS (EI) m/ e 195 (M+). Anal. Calcd for C10H13NO3: C,
61.51; H, 6.73; N, 7.18. Found: C, 61.29; H, 6.49; N, 7.08.
P r ep a r a tion of Alk en yl DHI 17. A solution of alkyne 15
(0.61 g, 3.1 mmol) and quinoline (215 mg, 0.83 mmol) in
benzene (205 mL) was transferred to a Parr hydrogenator flask
containing a mixture of benzene (20 mL) and Lindlar catalyst
(60 mg) that had been preequilibrated (10 min) under hydro-
gen. The flask contents were subjected to vigorous shaking
under hydrogen (20 psi) for 3 h and were then filtered,
concentrated, and diluted with CH2Cl2 (100 mL). The result-
ing solution was worked up as described for 2b, affording 0.54
g of crude alkene (Z/ E, 90:10). An analytical sample (pure
by 1H NMR) of the Z-alkene was obtained by preparative TLC
(hexanes/EtOAc, 60:40): [R]23D +66.1° (c 0.95, MeOH); 1H NMR
δ 6.13 (d, 1H, J ) 11.8 Hz), 5.97 (dd, 1H, J ) 7.8, 11.8 Hz),
5.11 (q, 1H, J ) 7.0 Hz), 4.38 (t, 1H, J ) 9.8 Hz), 4.25 (dd, 1H,
J ) 6.5, 8.1 Hz), 3.63 (dd, 1H, J ) 7.1, 8.1 Hz), 2.95-3.2 (m,
2H), 1.45 (s, 3H), 1.39 (s, 3H); 13C NMR δ 154.4, 137.2, 119.1,
109.6, 73.2, 69.4, 69.0, 37.2, 26.6, 25.5; HRMS (FAB) calcd for
C10H26NO4 (M + H+) 198.1130, found 198.1123.
Red u ctive Clea va ge of DHI 4b. A solution of 4b (1.04 g,
4.46 mmol) in diethyl ether (100 mL) was added to LiBH4 (36
mL of a 2 M THF solution, 72 mmol of LiBH4), and the
resulting solution was refluxed for 12 d. More LiBH4 (two 15-
mL portions, 60 mmol of LiBH4) was added after the third day
and eighth day to replenish spent reducing agent. Repetition
of the procedure described for preparation of 4a gave 1.4 g of
crude boron-containing product. This was taken up in benzene
(100 mL), and ethanolamine (2 g) was added. The resulting
mixture was stirred vigorously for 8 d and was worked up as
described for 4a to give 0.93 g of boron-free crude product.
Flash chromatographic purification as described for 4a af-
forded 0.83 g of diastereomeric γ-amino alcohols (75:25 ratio)
as an oil. Acetylation followed by repetitive preparative TLC
furnished 241 mg (51% yield) of amide 7b as the less mobile
product (pure by 1H NMR). The analytical sample was
The crude (Z)-alkene (0.54 g, 2.76 mmol) and iodine (76 mg,
0.3 mmol) were dissolved in benzene (45 mL), and the solution
was stirred and irradiated with a sunlamp for 3 h. The
resulting solution was worked up as described for 2b. Pre-
parative TLC (hexanes/EtOAc 60:40) on the resulting residue
recrystallized from CH2Cl2/hexanes: mp 88-89 °C; [R]23
D
-39.8° (c 4.5, CHCl3); IR (film) 3300, 1650 cm-1; H NMR δ
provided 0.43 g (70% yield based on 14; pure by H NMR) of
1
1
7.35-7.5 (m, 5H), 6.37 (br d, 1H, J ) 9.0 Hz), 5.92 (s, 1H),
4.2-4.35 (m, 2H), 3.85-3.89 (dd, 1H, J ) 4.5, 5.7 Hz), 3.5-
3.75 (m, 2H), 1.97 (s) on 1.85-2.05 (m) [4H total], 1.45-1.65
(m, 1H), 1.36 (d, 3H, J ) 6.2 Hz); 13C NMR (CDCl3) δ 171.2,
137.1, 129.3, 128.3, 126.4, 102.3, 84.9, 74.6, 58.3, 47.8, 32.2,
22.8, 18.7; HRMS (FAB) calcd for C15H22NO4 (M + H+)
280.1549, found 280.1550.
(E)-alkenyl DHI 15: [R]23D +37.1° (c 1.3, MeOH); IR (film) 1650
cm-1; 1H NMR δ 6.70 (d, 1H, J ) 15.9 Hz), 5.93 (dd, 1H, J )
7.0, 15.9 Hz), 4.66 (q, 1H, J ) 7.2 Hz), 4.40 (t, 2H, J ) 10.1
Hz), 4.18 (dd, 1H, J ) 6.3, 8.2 Hz), 3.66, (dd, 1H, J ) 7.4, 8.2
Hz), 3.08 (t, 2H, J ) 10.1 Hz), 1.46 (s, 3H), 1.42 (s, 3H); 13C
NMR δ 156.4, 135.3, 121.9, 109.6, 75.8, 69.0, 68.9, 33.0, 26.5,
25.6. HRMS (FAB) calcd for C10H26NO4 (M + H+) 198.1130,
found 198.1132.
Also obtained from the more mobile chromatography frac-
1
tion was 56 mg (12% yield; pure by H NMR) of amide 8b as
P r ep a r a tion of Alk en yl DHI 16. A solution containing
17 (170 mg, 0.86 mmol) and pyridinium p-toluenesulfonate
an oil: [R]21D +43.1° (c 1.46, CHCl3); IR (film) 3290, 1654 cm-1
;

Synthesis of L-Daunosamine Derivatives
J . Org. Chem., Vol. 62, No. 11, 1997 3677
(265 mg, 1.05 mmol) in MeOH (29 mL) was refluxed for 2 d.
To the cooled solution were added NaHCO3 (1 g) and Na2SO4
(2 g). The resulting mixture was filtered and concentrated,
toluene being added with further concentration to entrain
residual pyridine. The residue was treated with Ac2O (0.6
mL), DMAP (1 crystal), pyridine (0.7 mL), and CH2Cl2 (1.5
mL). The resulting solution was stirred for 3 h and brine (5
mL) was added. The organic product was extracted into EtOAc
(six 30-mL portions), and the combined extracts were worked
up. Preparative TLC (acetone/ether 08:92) on the resulting
residue afforded 162 mg (78% yield; pure by 1H NMR) of 16
organic layers were worked up, and the crude product was
purified by preparative TLC (diethyl ether) to afford 21 mg
(82% yield) of tetraacetates 18 and 19 [98:2 mole ratio,
respectively, by analytical HPLC (hexanes/i-PrOH 90:10)]:
[R]23 -9.0° (c 1.1, CHCl3); IR (film) 1738 cm-1 1H NMR δ
;
D
5.78 (d, 1H, J ) 3.4 Hz), 5.51 (dd, 1H, J ) 3.4, 8.2 Hz), 5.2-
5.3 (m, 1H), 4.1-4.5 (m, 4H), 2.9-3.1 (m, 2H), 2.14 (s, 3H),
2.12 (s 3H), 2.09 (s) and 2.08 (s) [6H total];13C NMR δ 170.5,
169.7, 169.6, 154.7, 69.0, 68.6, 68.3, 67.0, 61.6, 35.3, 20.7, 20.6,
20.6, 20.5; HRMS (FAB) calcd for C15H22NO9 (M + H+)
360.1295, found 360.1296.
as an oil: [R]23 +36.6° (c 1.1, CHCl3); IR (film) 1736, 1651
D
AD Rea ction of Alk en yl DHI 16 Usin g (DHQ)2-P HAL.
A mixture of (DHQ)2-PHAL (65 mg, 0.08 mmol), methane-
sulfonamide (17 mg, 0.18 mmol), K3Fe(CN)6 (143 mg, 0.43
mmol), K2CO3 (67 mg, 0.48 mmol), K2OsO4‚2H2O (4 mg, 0.01
mmol), water (0.6 mL), and t-BuOH (0.6 mL) was stirred for
10 min at rt. A solution of the (E)-alkene 16 (31 mg, 0.13
mmol) in t-BuOH (0.6 mL) was added, and the procedure used
for the preceding (DHQD)2-PHAL reaction was repeated.
Preparative TLC (diethyl ether) of the crude product furnished
38.6 mg (85% yield) of mainly tetraacetate 19 [18/19 5:95 by
analytical HPLC (hexanes/i-PrOH, 90:10)]: [R]23D +40.0° (c 1.4,
cm-1; H NMR δ 6.68 (d, 1H, J ) 16.1 Hz), 5.92 (dd, 1H, J )
1
5.9, 16.1 Hz), 5.6-5.7 (m, 1H), 4.40 (t, 2H, J ) 10.2 Hz), 4.31
(dd, 1H, J ) 3.9, 11.8 Hz), 4.14, (dd, 1H, J ) 6.5, 11.8 Hz),
3.07 (t, 2H, J ) 10.2 Hz), 2.13 (s, 3H), 2.08 (s, 3H); 13C NMR
δ 170.3, 169.6, 156.3, 131.6, 122.8, 70.7, 69.3, 64.2, 33.6, 21.0,
20.8; HRMS (FAB) calcd for C11H16NO5 (M + H+) 242.1028,
found 242.1027.
AD Rea ction of Alk en yl DHI 16 Usin g (DHQD)2-
P HAL. A mixture consisting of (DHQD)2-PHAL (37 mg,
0.045 mmol), methanesulfonamide (10.5 mg, 0.11 mmol),
K3Fe(CN)6 (80 mg, 0.24 mmol), K2CO3 (39 mg, 0.28 mmol),
K2OsO4‚2H2O (3.8 mg, 0.01 mmol), water (0.6 mL), and
t-BuOH (0.6 mL) was stirred for 10 min at rt. To the mixture
was added a solution of 16 (16.9 mg, 0.07 mmol) in t-BuOH
(0.6 mL). Stirring was continued for 21 h, and anhydrous
Na2SO3 (1 g) was added with additional stirring for 1 h. The
mixture was concentrated in vacuo, and the solids were
triturated for 30 min (MeOH/acetone 20:80; 200 mL). The
resulting slurry was filtered through silica gel (1 g), and the
filtrate was concentrated. Traces of water were removed by
azeotropic distillation with two portions of absolute EtOH
followed by drying in vacuo. The resulting crude diol was then
taken up in CH2Cl2 (1 mL) and the solution cooled (0-5 °C).
Pyridine (0.1 mL), Ac2O (0.1 mL), and DMAP (1 crystal) were
added, and stirring at 0-5 °C was continued for 2 h. Saturated
aqueous brine (5 mL) was then added, and the mixture was
extracted with EtOAc (six 30-mL portions). The combined
1
CHCl3); IR (film) 1740 cm-1; H NMR δ 5.76 (d, 1H, J ) 6.4
Hz), 5.56 (dd, 1H, J ) 4.5, 6.4 Hz), 5.25-5.35 (m, 1H), 4.3-
4.45 (m, 3H), 4.03 (dd, 1H, J ) 6.1, 11.9 Hz), 2.9-3.1 (m, 2H),
2.13 (s), 2.12 (s), 2.11 (s) [9H total], 2.06 (s, 3H); 13C NMR δ
170.4, 169.9, 169.7, 169.5, 154.2, 69.5, 69.0, 68.9, 67.4, 61.7,
35.3, 20.7, 20.6; HRMS (FAB, NaBr) calcd for C15H21NO9Na
(M + Na+) 382.1114, found 382.1115.
Su p p or tin g In for m a tion Ava ila ble: An ORTEP drawing
1
of 4a and H NMR spectra of all new products (21 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
J O962293D