Carbohydrate chiral-pool approach to four enantiomerically pure 2-naphthylmethyl 3-hydroxy-2-methylbutanoates

Article abstract of DOI:10.1016/j.tet.2008.03.088

d-Glucose, l-xylose, and d- and l-arabinose were sources of chirality to obtain four enantiomerically pure 3-hydroxy-2-methylbutanoic acids, which were reacted with 2-naphthyldiazomethane to furnish their fluorescent 2-naphthylmethyl esters.

Full text of DOI:10.1016/j.tet.2008.03.088

Tetrahedron 64 (2008) 5551–5562
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Carbohydrate chiral-pool approach to four enantiomerically pure
2-naphthylmethyl 3-hydroxy-2-methylbutanoates
,
b
,
,
Bogdan Doboszewski a y, Piet Herdewijn a
*
a Laboratory of Medicinal Chemistry, Rega Institute for Medical Research, Katholieke Universiteit Leuven, Minderbroedersstraat 10, 3000 Leuven, Belgium
b Departamento de Quimica, UFRPE, 52171-900 Recife, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
D-Glucose, L-xylose, and D- and L-arabinose were sources of chirality to obtain four enantiomerically pure
Received 28 February 2008
Received in revised form 26 March 2008
Accepted 26 March 2008
3-hydroxy-2-methylbutanoic acids, which were reacted with 2-naphthyldiazomethane to furnish their
fluorescent 2-naphthylmethyl esters.
Ó 2008 Elsevier Ltd. All rights reserved.
Available online 29 March 2008
Keywords:
Chiral pool
Fluorescent tag
Monosaccharides
Stereoselective
Stereospecific
1. Introduction
stereogenic carbon atoms in 1 can be obtained, which would be an
initial step to solve the still lacking full stereochemical constitution
Batumin 1, 2E,10Z,12E-20-(3-aminocarboxy)-2-methyl-1-oxo-
butylamino-7-methylene-17-oxo-19-oxy-3,5,15-trimethyleicosa-2,
10,12-trienoic acid, is an antibiotic produced by the strains of
Pseudomonas batumici (Fig. 1).1 It shows very good activity partic-
ularly against Gram-negative bacteria, e.g., Staphylococcus aureus,2,3
which pose a serious health problem due to their resistance against
antibiotics in current use. Batumin’s basic structural elucidation
was performed, except for establishing the configurations of the
five stereogenic carbon centers. Compound which has the same
gross molecular composition dubbed as kalimantacin A, was iso-
lated from a fermentation broth of Alcaligenes sp. YL-02632S. Kali-
mantacin A has the same set of configurations of the C]C bonds
but its absolute configurations remain unknown.4–6 It is not clear at
this time if batumin and kalimantacin A are the same compounds
or if they differ at the configurations of the stereogenic centers. In
principle, compound 1 can be degraded in a basic medium to re-
lease 3-hydroxy-2-methylbutanoic acid 2 (Fig. 1). It should be
possible to establish absolute configurations of 2 by chiral HPLC by
comparison of its mobility with the four reference 3-hydroxy-2-
methylbutanoic acids, which have the known configurations. In this
way, the configurations of the left fragment containing two
of 1.
Reported here is a carbohydrate-based chiral-pool approach to
obtain all four enantiomerically pure (within accuracy of their 1H
and 13C NMR spectra recorded at 500 MHz and 125 MHz, re-
spectively) 3-hydroxy-2-methylbutanoic acids 20, 33, 48, and 56 in
a form of their fluorescent 2-naphthylmethyl esters. Preliminary
results of this work were published.7
The syntheses of chiral 2 with variable enantiomeric purities
were realized before via aldol condensations8–16 (in some cases
enzymatic resolution step was involved8,9), enantioselective re-
ductions of carbonyl group,17,18 or by acetoxymercuration followed
by chiral resolution.19
In this work, we opted for the application of monosaccharides to
obtain the enantiomerically pure targets 20, 33, 48, and 56 in order
to avoid a necessity to perform chiral separations, which can be
difficult. Additionally, the starting sugars have well-established
absolute configurations, which assure that the targets have pre-
dictable stereochemistry.
2. Results and discussion
The basic idea of transformation of the starting sugars 1,2;5,
6-di-O-isopropylidene-
thylsilyl-1,2-O-isopropylidene-
diphenylsilyl-1,2-O-isopropylidene-
D
-gluco-furanose 3, 5-O-tert-butyldime-
-xylo-furanose 25, 5-O-tert-butyl-
-arabino-furanose 35, and its
* Corresponding author. Tel.: þ32 16 337387; fax: þ32 16 337340.
E-mail address: piet.herdewijn@rega.kuleuven.be (P. Herdewijn).
On leave from the Departamento de Quimica.
L
y
D
0040-4020/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.03.088

5552
B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
CH3
O
O
CH2 CH3 CH3
E
CO2H
H2N
O
N
H
E
Z
CH3
OH
O
CH3
1 Batumin
CH3
CO2H
HO
CH3
2
Figure 1. Structure of batumin and of 3-hydroxy-2-methylbutanoic acid, a possible degradation product of batumin.
L
-enantiomer 49 into the targets 20, 33, 48, and 56, respectively, is
process),23 and an unexpected monoisopropylidenated compound
8. Cleavage of 5,6-acetonide was most probably a result of the
presence of small quantity of PdCl2 in the batch of the catalyst. In
a hydrogen atmosphere, PdCl2 was reduced to liberate HCl, which
cleaved the more reactive 5,6-acetonide. The same explanation was
given to rationalize the alleged hydrogenolysis of the tert-butyldi-
methylsilyl ethers, which eventually turned out to be a simple
shown in Figure 2. In all four cases, the stereocenters at the atoms
C4 in the substrates were preserved and the bulky 1,2-O-iso-
propylidene moieties were used to install predictably the C3 methyl
groups in the intermediates 7, 28, 39, and 54 via oxidation, Wittig
methylenation, and hydrogenation. In the case of
nose, this sequence was 100% stereoselective, whereas for
cose 3 and -xylose 25 the selectivity was ca. 90%, however, a simple
crystallization furnished pure stereoisomers at the C3 at a later
stage (see below). The lateral groups attached to C4 in the sub-
strates were transformed into the methyl groups and the C2 atoms
functioned as the latent carboxyl groups.
The synthesis of the (2R,3R) compound 20 is shown in Scheme 1.
Commercial 1,2;5,6-di-O-isopropylidene-D-gluco-furanose 3 was
oxidized to 4 using CrO3–Py–Ac2O system by analogy to the other
oxidations20,21 and subjected to Wittig methylenation to give 5.
Catalytic hydrogenation over Pd–C rather than RaNi22 furnished
a separable mixture of 6 and 7 in a proportion ca. 1:10 by 1H NMR
(roughly the same proportion of 6 to 7 was obtained using much
more synthetically demanding thiol-catalyzed free-radical redox
D- and
L
-arabi-
D-glu-
L
hydrolysis.24,25 The
chromatography and subjected to ‘dehomologation’26,27 ((a) H5IO6–
EtOAc and (b) NaBH4–EtOH) to give 3-deoxy-3-C-methyl- -ribo
D-allo stereoisomer 7 was isolated by flash
D
epimer 10 via the aldehyde 9. Conventional tosylation furnished
a highly crystalline 11, which served as a substrate for ‘deoxy-
genative substitution’ (–CH2OTs/–CH3) using LiBH(Et)3 to get
another methyl group (11/13). Since chromatographic separation
of 6 and 7 was inconvenient, the ‘dehomologation’ step was per-
formed on the mixture 6 and 7 obtained after hydrogenation of 5
over Adam’s catalyst (PtO2). The mixture of both epimeric com-
pounds 12 thus obtained was subjected to tosylation and crystal-
lization from EtOAc. This simple crystallization step furnished
a pure
D
-ribo compound 11 in 54% yield counting on 6 and 7. The
O
O
O
O
Me
(R)
Me
OH
CO2R
O
O
O
OH
OH
H
H
O
O
(R)
Me
Me
7 D-allo
Me OH
O
O
3 D-gluco
20 2R,3R
15 D-ribo
100%
+ ~10% D-gluco
H
(S)
O
O
O
Me
OH
CO2R
O
O
Me
OH
OH
Me
O
O
t
Me
BuMe2SiO
HO
(S)
Me
HO
H
31 L-ribo
100%
33 2S,3S
28 L-ribo
+ ~10% L-xylo
25 L-xylo
Me
(R)
O
O
Me
t
O
BuPh2SiO
HO
OH
CO2R
O
Me
O
Me
OH
OH
O
O
H
Me
(S)
HO
H
35 D-arabino
44 D-lyxo
39 D-lyxo
48 2S,3R
100%
H
O
OH
O
O
OH
OH
CO2R
(S)
Me
t
BuPh2SiO
HO
O
O
Me
Me
H
(R)
Me
56 2R,3S
O
O
Me
OH
54 L-lyxo
55 L-lyxo
49 L-arabino
100%
Figure 2. Utilization of chirality present in D-glucose, L-xylose, and D- and L-arabinose to obtain the targets 20, 33, 48, and 56 with predictable configurations.

B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
5553
O
O
O
O
Ph3P+CH3(Br-)
BuLi, THF
O
O
PtO2
CrO3, Py
3
6 + 7
6:7=1:10 95%
EtOH, H2
Ac2O, CH2Cl2
84%
O
O
-78 °
rt
O
H2C
O
59%
O
5
4
O
O
O
O
HO
HO
O
O
Me
O
Pd/C cont. with PdCl2
H2, EtOH
5
+
+
O
O
O
O
Me
Me
O
O
7 more polar
6 less polar
minor
8 33%
major
6+7 53%
OHC
O
HO
O
TsO
O
H5IO6
NaBH4
O
EtOH
TsCl, Py
7
EtOAc
Me
O
CH2Cl2, DMAP
75%
O
O
Me
O
O
Me
11
9
10
61% for two steps
pure ribo epimer
by selective cryst.
54% for three
1.H5IO6
EtOAc
HO
O
TsCl, Py
6/7
CH2Cl2, DMAP steps
O
2.NaBH4
EtOH
Me
O
12
Me
Me
Me
O
O
O
LiBH(Et)3, THF
H2SO4, H2O
MeOH
OH
11
O
O
Me
BEt
Me
Me
O
O
OH
13
15
14
tentative structure
73% from 11
Me
Me
H
Me
H
NaClO2
OCHO
CO2R
OH
OCHO
CHO
35%H2O2
NaIO4, H2O
H
+
CO2R
H
H
15
H
NaH2PO4
CH3CN
EtOH
Me
Me
Me
16
R=H
17
18 R=H
23,CH2Cl2
R=
20
42%
37%
19
NH4OH, MeOH, 89%
-
CHO
CH=NNH2
+
CH=N=N
.
H2NNH2 H2O
EtOH
yellow HgO, CH2Cl2
cat. KOH
2292%
21
23
Scheme 1. Obtention of 2-naphthylmethyl (3R)-hydroxy-(2R)-methylbutanoate 20 from D-glucose.
configuration (3R) in the tosylate 11 was confirmed by X-ray
analysis.28 Tosylate 11 was then treated with LiBH(Et)3, a donor of
a very soft ‘supernucleophilic’ hydrogen anion, to give 3,5-dideoxy-
evaporations of the extracts with MeOH furnished 3,5-dideoxy 3-C-
methyl- -ribo-furanose 15 isolated in 73% counting on 11. (Two
D
alternative procedures to remove the isopropylidene function in
13: I2–MeOH31 and FeCl3$6H2O–CH2Cl2,32 were inefficient.) In this
way, both methyl groups were properly installed and the carboxyl
group was ready to be formed. The vicinal diol system in 15 was
cleaved with NaIO4 to give the transient aldehyde 16, which was
immediately subjected to oxidation using NaClO2–H2O2–NaH2PO4
system.33 TLC at this stage showed two products: a major for-
myloxy acid 17 and a minor more polar hydroxy acid 18, visualized
as yellow spots on a blue background upon revelation with an acid–
base indicator bromocresol green.34,35 Both 17 and 18 were not
isolated, but instead they were esterified with freshly prepared
2-naphthyldiazomethane 23 obtained from the hydrazone 2236
by KOH catalyzed HgO oxidation by analogy to the other
3-C-methyl-1,2-O-isopropylidene-
Graf described a different route to 13 starting from 3,5-dideoxy-1,2-
O-isopropylidene-3C-(E,Z)-methylthiomethylene-a- -erythro-pen-
tofuranose by RaNi hydrogenation/desulfurization followed by
preparative gas chromatography to separate it from the C3 -xylo
D-ribo-furanose 13. Tronchet and
D
D
epimer.29 The isolation of 13 was skipped and acid hydrolysis of the
1,2-acetonide function was performed by the addition of aq H2SO4
directly to the reaction mixture still containing the boron com-
pounds to give the diol 15 presumably via a cyclic borate 14 (see
Section 3). Borates derived from vicinal diols are known to be quite
stable, e.g., they survive on HBr–AcOH or PBr3 treatment but can be
easily cleaved by methanolysis.30 Consequently, repeated co-

5554
B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
arylhydrazones.37,38 The esterifications with 23 can be referred to
as titrations: loss of the red-orange color of a CH2Cl2 solution of 23
was nearly instantaneous and was accompanied by evolution of
nitrogen. Addition of 23 was stopped when a reddish color in-
dicating the presence of the excess of 23 persisted. The procedure
routinely took ca. 1–2 min. TLC of these mixtures showed a strong
fluorescent spot of the formate 19 (major) and a more polar 2-
naphthylmethyl (3R)-hydroxy-(2R)-methylbutanoate 20 (minor),
accompanied by UV absorbing compounds having Rfs values similar
to those of 19. The presence of these impurities was not surprising
since it is known that in situ generated aryl diazomethanes are in
general not more than 85% pure,39 and also the transient naph-
thylmethyl carbenes can decompose to form by-products.
It should be pointed out that tagging of the organic acids with
fluorescent 1-napthylmethyl or 1-napthylethyl groups was applied
to facilitate their detections during HPLC analysis.40 Both 19 and the
target 20 could be isolated at this stage, and the formate 19 could be
easily transformed to 20 by ammonolysis using aq NH4OH in
MeOH. However, it was much more convenient to perform this
deprotection on a crude esterification mixture and to isolate 20 in
a cumulative yield of ca. 70% for four steps counting on the diol 15.
Comparison of the 1H and 13C NMR spectra of the (2R,3R) target 20
with those of the (2S,3R) epimer 48 (see below) showed that no
inversion of configuration at the C2 atom via enolization took place
during this sequence and that 20 is a pure enantiomer within the
limits of the 500 MHz (1H) and 125 MHz (13C) measurements.
2-naphthylmethyl (3S)-hydroxy-(2S)-methylbutanoate 33 proceeded
exactly as described above for the -series. Again, both (2S,3S) 33 and
(2R,3R) 20 differ only in the sign of their optical rotations.
The other two targets, which have (2S,3R) and (2R,3S) configu-
ration, viz. 48 and 56, respectively, were obtained in a stereospecific
D
way using
Selective tert-butyldiphenylsilylation of
isopropylidenation was performed by analogy to the known pro-
cess described for
-arabinose20,21 to furnish tert-butyldiphe-
nylsilyl-1,2-O-isopropylidene-
-arabino-furanose 35.43 Oxidation
D
-arabinose (Scheme 3) and
L-arabinose (Scheme 4).
D
-arabinose followed by
L
D
at the position C3 followed by Wittig methylenation furnished
the olefin 36. Tebbe’s reagent (m-chloro-m-methylene[bis-
(cyclopentadienyl)titanium]-dimethylaluminum) could also be
used,21 but its price is rather high. Two-step Peterson olefination
((1) Me3SiCH2Li and (2) NaH) was an alternative,21 but a single step
Wittig reaction seemed to be the simplest, and permitted the
obtention of 36 in 72% yield for two steps. Desilylation furnished 37,
which was subjected to hydrogenation over Adam’s catalyst to
furnish two products. The less polar 38 had a double bond migrated
to the 3,4-position and was isolated in 13% yield, whereas the
necessary
3-deoxy-1,2-O-isopropylidene-3-C-methyl-D-lyxofur-
anose 39 was obtained in 71% yield. Formation of 38 undoubtedly
reflects a steric difficulty to install four syn oriented substituents
inside a lyxofuranosyl ring. At the same time, it is known that
tetrasubstituted olefins are quite resistant toward hydrogenation,44
so persistence of 38 under such conditions is not surprising. The
alcohol 39 was converted to its 5-O-tosylate 40, which, however,
turned out to be unreactive toward substitution using LiBH(Et)3.
This inertness again reflects a steric congestion in the lyxo ring in
40. The triflate 41, however, reacted with LiBH(Et)3 to give 42 even
though decomposition was visible on TLC. Further steps to obtain
The enantiomeric (2S,3S) product 33 was obtained from
as shown in Scheme 2. 1,2-O-Isopropylidene- -xylo-furanose 24
was prepared in a one-pot reaction by analogy to the -enantio-
mer.41 Selective protection of the primary OH group by silylation
(/25) by analogy to its
-enantiomer,27,42 followed by CrO3–
L-xylose
L
D
D
Py–Ac2O oxidation and Wittig methylenation as described before
(3/4/5) furnished the olefin 26. Conventional desilylation and
hydrogenation over Adam’s catalyst furnished inseparable mixture
of the major (w90%) L-ribo epimer 28 and the L-xylo epimer 29,
formed in the same proportion as compounds allo 7 and gluco 6.
Tosylation of this mixture and subsequent crystallization furnished
pure C3 L-ribo epimer 30, which is an enantiomer of D-compound 11
described above. Obviously, the NMR data of 30 match with
those of 11 and both compounds have the opposite optical
rotations. Transformation of 30 to the diol 31 and further to
the 3,5-dideoxy-3-C-methyl-
D
-lyxofuranose 44 proceeded as de-
-ribo (30/31) series. It is
scribed above for the -ribo (11/15) and
D
L
evident that a combination of the best known leaving group (–OTf)
and a supernucleophilic properties of the LiBH(Et)3 partly offsets
the steric hindrance in 41, even though a yield of the isolated 44
was 27%, much less than in the case of the less congested com-
pounds 11 and 30. It should be pointed out that application of the
triflate 45, which is less congested than 41 due to lack of the C3
methyl group, did not improve the yield of 44 (data not shown).
Further transformation of 44 proceeded following the procedure
O
O
O
HO
O
HO
O
1.acetone,H2SO4
2.Na2CO3, H2O
t
BuMe2SiCl
OH
O
O
TBDMSO
HO
imidazole,
DMF, 75%
one pot, 75%
HO
HO
25
OH
L-xylose
1.CrO3, Py, Ac2O
24
2.Ph3P+CH3(Br-)
t
TBDMS: BuMe2Si-
BuLi, 70% two steps
O
O
O
O
O
O
O
Me
O
O
PtO2
H2
+
HO
O
O
O
Bu4NF
HO
TBDMSO
HO
~100%
Me
27
26
29 L-xylo, minor
~100%, inseparable, 28:29=~10:1
28 L-ribo, major
TsCl, Py
O
H
1.NaIO4
OH
CO2R
O
(S)
OH
OH
O
1.LiBH(Et)3
2.NaClO2, H2O2
Me
Me
Me
O
Me
(S)
2.aqH2SO4
NaH2PO4
3.23
4.NH4OH, MeOH
TsO
Me
31, 64%
H
30, pure L-ribo by cryst.
69%
33 2S,3S
58% for four steps
R= 2-naphthylmethyl
OMe
O
Me
OH
Me
32
Scheme 2. Obtention of 2-naphthylmethyl (3S)-hydroxy-(2S)-methylbutanoate 33 from L-xylose.

B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
5555
t
O
HO
BuPh2SiO
t
O
O
t
BuPh2SiCl
BuPh2SiO
acetone
H+
OH
OH
O
OH
imidazole
O
HO
OH
D-arabinose
HO
34
HO
35
1.CrO3, Ac2O, Py
2.Ph3P+CH3(Br-)
BuLi
HO
O
O
O
O
O
t
HO
H3C
HO
O
O
BuPh2SiO
O
O
O
O
O
Bu4NF
+
PtO2, H2
H3C
37
36
39 71%
38 13%
72% two steps
92%
TsCl, Py, CH2Cl2
or Tf2O, Py, CH2Cl2
RO
H3C
O
O
O
O
H3C
H3C
O
H3C
H3C
B-Et
O
O
H
+
LiBH(Et)3
O
O
40 R= -Ts 67%
41 R= -Tf
42
43
tentative structure
MeOH
Me
Me
Me
O
OH
OCHO
OH
Me
OH
1.NaIO4
CO2H
CO2H
H
H
+
Me
2.NaClO2, 35%H2O2
NaH2PO4
Me
44
H
H
27% from 39 via 41
46
47
1. 23
2. NH4OH, MeOH
TfO
O
O
Me
(R)
OH
O
CO2CH2
H
Me
(S)
45
H
48 54%-65% from 44
Scheme 3. Obtention of 2-naphthylmethyl (3R)-hydroxy-(2S)-methylbutanoate 47 from D-arabinose.
used above (15/20 and 31/33) via the intermediates 46 and 47,
which were deformylated without separation to furnish the third
target 2-naphthylmethyl (3R)-hydroxy-(2S)-methylbutanoate 48.
The last target 56, which has (2R,3S) configuration was obtained
applied to
nylsilyl-1,2-O-isopropylidene-
olefin 50 with two noticeable differences. Firstly, attempted hy-
drogenation of 50 over Pd–C catalyst resulted in migration of the
carbon–carbon double bond in a much higher extend than in the
D
-arabinose, starting with the known tert-butyldiphe-
-arabino-furanose 4920,21 via the
L
from
L-arabinose (Scheme 4) following the same directions as those
t
OSi BuPh2
O
O
OH
Pd/C
1.CrO3, Py, Ac2O
t
BuPh2SiO
t
O
O
BuPh2SiO
O
2.Ph3P+CH3(Br-)
H2
O
O
BuLi
O
50
H3C
51
49
O
46%
56% for two steps
Bu4NF
O
O
+
PtO2*
O
HO
HO
O
O
H2
O
O
t
Me
O
BuPh2SiO
53 95%
54 83%
O
Me
52 36%
1.Tf2O, Py
2.LiBH(Et)3
3.H+
H
O
OH
OH
(S)
Me
CO2CH2
H
Me
Me
(R)
Me
56 2R,3S
63% from 55
OH
55 27%
O
O
TsO
*Different batch of PtO2 was used; no migration
of C=C bond, cf. 37
O
Scheme 4. Obtention of 2-naphthylmethyl (3S)-hydroxy-(2R)-methylbutanoate 56 from L-arabinose.

5556
B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
case of the desilylated
was obtained in 46% yield, whereas the necessary
D
-compound 37, and the unwanted olefin 51
-lyxo-3-C-
became warm again. Nearly all CrO3 solubilized. After 30 min, solid
1,2;5,6-di-O-isopropylidene- -gluco-furanose 3 (6.55 g, 25.2 mmol)
L
D
methyl product 52 was obtained in 36% yield. This was presumably
a result of a bulk of the tert-butyldiphenylsilyl group. Also, palla-
dium promotes migrations of the C]C bonds to a greater extent
than platinum.45 Secondly, hydrogenation of the desilylated in-
termediate 53 over Adam’s catalyst (a different batch of PtO2 was
used than the one for hydrogenation of 37) resulted in the absence
of the transposition of the C]C bond and permitted isolation of the
was added immediately followed by Ac2O (9.8 ml, 100 mmol). The
mixture became warm again. TLC in 2:1 hexane–EtOAc after 45 min
showed an elongated spot of the ulose 4, which is slightly more
polar than 3. Most of CH2Cl2 was evaporated. Addition of 100 ml of
1:1 toluene–EtOAc mixture precipitated most of the chromium
compounds. The decanted solution and the washings of the
precipitated black solid tar were pushed through
a
necessary C3-Me-
ferences probably reflect variations of the properties of surfaces of
platinum catalyst.
It should be pointed out that a tosylate 57 was unreactive
toward LiBH(Et)3 just like the more congested 40.
Further transformation of the alcohol 54 to 3,5-dideoxy-3-
L
-lyxo product 54 in 83% yield. These subtle dif-
60–200 mesh silica gel column prepared in 1:2 toluene–EtOAc
using an overpressure. Elution with 1:2 toluene–EtOAc, evapora-
tion, co-evaporation with xylenes, and final drying furnished 4
(5.44 g, 84%), which was used for a Wittig reaction without
characterization.
A
magnetically stirred suspension of methyltriphenyl-
C-methyl-
L
-lyxofuranose 55 and its subsequent conversion to 2-
phosphonium bromide (98%, 10.7 g, 30 mmol) in THF (100 ml)
under nitrogen with EtOH–dry ice external cooling was treated
with 2.5 M n-BuLi in hexane (11.2 ml, 28 mmol) dropwise. The
mixture turned yellow. After the addition of n-BuLi, cooling was
stopped for 30 min and re-applied again. A solution of 4 in 60 ml of
THF was added dropwise for ca. 10 min. Cooling bath was removed.
After 1.5 h, TLC showed that all 4 reacted forming a less polar 5 with
Rf 0.36 in 9:1 hexane–EtOAc. The solids were filtered on a sintered
glass. Evaporation and flash chromatography in 9:1 hexane–EtOAc
gave 5 (3.2 g, 59% counted on 4) as a colorless oil. [a]2D4 þ109.4 (c 2.6,
CHCl3).
naphthylmethyl (3S)-hydroxy-(2R)-methylbutanoate 56 proceeded
exactly as described for their enantiomeric equivalents.
In summary, all four enantiomerically pure 3-hydroxy-2-meth-
ylbutanoic acids as fluorescent 2-naphthylmethyl esters were
synthesized starting from easily available carbohydrates.
The application of the targets 20, 33, 48, and 56 to establish
absolute configurations of the degradation product 2 after its
conversion to 2-naphthylmethyl ester will be published in due
course.
3. Experimental
3.1. General
1H (CDCl3, 300 MHz): 5.81 (d, J12¼4.0 Hz, 1H, H1), 5.50 (dd,
J¼1.3, 1.8 Hz, 1H, ]CH2), 5.45 (dd, J¼0.8, 2.2 Hz, 1H, ]CH2), 4.89
(apparent dd, J¼1.0 Hz, J21¼4.0 Hz, 1H, H2), 4.67–4.64 (m, 1H),
4.09–4.03 (m, 2H), 3.97–3.90 (m, 1H), 1.52, 1.44, 1.37, and 1.36 (four
s, 3H each, isopropylidene Me). 13C (75 MHz): 146.8, 113.4, 112.5,
109.7, 104.5, 82.1, 79.2, 77.3, 66.7, 27.3, 27.0, 26.5, 25.4.
All glasswares were dried at 104 ꢀC. Evaporations were per-
formed on a Bu¨chi evaporator coupled to a membrane pump.
Vacuums were broken using a balloon filled with nitrogen in the
case of moisture-sensitive products. Moisture-sensitive liquids
were transferred using dry syringes under atmosphere of nitrogen
or argon. THF was freshly distilled from sodium–benzophenone.
CH2Cl2 was dried by distillation from P2O5. Anhydrous acetone was
prepared by shaking with P2O5 for ca. 20 min, filtration, and dis-
tillation. Dry DMF containing molecular sieves was obtained from
Aldrich. NMR spectra were recorded on a Bruker 300 MHz and
Bruker Avance II 500 MHz spectrometers using CDCl3–TMS unless
otherwise stated. Electron impact exact mass measurements were
performed on the MS5OTS Kratos instrument at 50 eV, whereas the
electrospray measurements on the APEX-Qe Bruker instrument.
Optical rotations were taken on a Perkin–Elmer 341 automatic
polarimeter in CHCl3 (distilled from P2O5, i.e., ethanol-free) using
a 1-dm tube for Na line at ca. 24 ꢀC. Flash chromatography was
performed using a 230–400 mesh silica gel from ICN or Acros. TLC
plates (Fluka) with a fluorescent indicator were used for analysis of
the reaction mixtures. The chromatograms were visualized using
10% H2SO4 in MeOH unless otherwise stated and charring at 120–
130 ꢀC. ‘Bromocresol green reagent’ used for detection of the car-
boxylic acids means 0.4% solution of bromocresol green in EtOH
and 0.1 N aq NaOH added just to change a color to blue.34,35 ‘CrO3–
H2SO4 system’ means 2% CrO3 in 10% H2SO4. ‘Xylene’ means
a mixture of isomers. Pyridine was stored over pellets of KOH. Dry
ethyl acetate was prepared by distillation. Early distillate contain-
ing azeotropic mixture with water was rejected and the late frac-
tion was used. Na2SO4 was used to dry the extracts.
3.3. 3-Deoxy-1,2;5,6-di-O-isopropylidene-3-C-methyl-
furanose 6, 3-deoxy-1,2;5,6-di-O-isopropylidene-3-C-methyl-
-allo-furanose 7, and 3-deoxy-1,2-O-isopropylidene-3-
D-gluco-
D
C-allo/gluco-furanose 8
3.3.1. Hydrogenation of 5 using Pd–C contaminated with PdCl2
To a solution of 5 (3.2 g, 12.5 mmol) in EtOH (20 ml, 96%) was
added Pd–C (0.31 g 10%) under a blanket of argon. Hydrogenation
was performed in a Parr apparatus at the initial pressure of 35 psi.
After 7 h of shaking, TLC showed the presence of the minor 6,
slightly less polar than the major 7 (Rf of a mid-point was 0.40 in 9:1
hexane–EtOAc, run twice), and 8 at a start. Rf of 8 was 0.44 in 20:1
CH2Cl2–MeOH. The mixture was passed through a bed of Celite
(attention: used Pd catalyst is pyrophoric and should not be left
dry) and the solvent was evaporated. Flash chromatography in 20:3
hexane–EtOAc gave 0.07 g of 6, 0.02 g of a mixture of 6 and 7, and
0.72 g of 7. Total yield of 6 and 7 was 53%. Further elution with 20:1
CH2Cl2–MeOH gave 8 (0.89 g, 33%).
3.3.2. Hydrogenation of 5 using PtO2
To a solution of 5 (7.39 g, 28.9 mmol) in 20 ml of 96% EtOH was
added PtO2$H2O (0.33 g). Hydrogenation in a Parr apparatus at the
initial pressure of 40 psi for 7 h followed by filtration through Celite
(attention: used Pt is pyrophoric and should not be left dry) and
evaporation furnished 7.11 g, 95% of a mixture 6 and 7. This mixture
was used in a dehomologation step without chromatography.
Compound 6. 1H (300 MHz): 5.78 (d, J12¼3.5 Hz, 1H, H1), 4.36 (d,
J21¼3.5 Hz, 1H, H1), 4.13 (dd, J¼5.4, 7.9 Hz, 1H), 4.09–3.99 (m, 2H),
3.94 (dd, J¼5.4, 7.9 Hz, 1H), 2.45 (dq, J34¼4.0 Hz, JH3–Me¼7.4 Hz, 1H,
H3),1.52,1.41,1.35,1.31(four s, 3H each, isopropylidene Me), 0.96 (d,
JMe–H3¼7.5 Hz, 3H, C3–Me). 13C (75 MHz): 111.2, 109.2, 104.8, 86.5,
80.7, 73.5, 68.4, 40.6, 26.8, 26.7, 26.1, 25.3, 11.1.
3.2. 3-Deoxy-1,2;5,6-di-O-isopropylidene-3-C-methylene-
D-ribo-hexofuranose 5
To a magnetically stirred mixture of CrO3 (10.0 g, 100 mmol) in
100 ml of dry CH2Cl2 was added pyridine (16.2 ml, 200 mmol)
under nitrogen. The mixture immediately turned dark brown and

B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
5557
Compound 7. 1H (300 MHz): 5.76 (d, J12¼3.6 Hz, 1H, H1), 4.55 (t,
J¼4.2 Hz, 1H, H2), 4.11–4.02 (m, 2H), 3.97–3.89 (m, 1H), 3.73 (ddd,
an oil pump to furnish epimeric alcohols 12. Dry CH2Cl2 was added
followed by pyridine (7.5 ml), cat. quantity of DMAP, and TsCl
(10.1 g, 53.2 mmol). After an overnight reaction, TLC showed a sin-
gle spot of the products (Rf 0.26 in 4:1 hexane–EtOAc) and a fast
moving spot of TsCl. Extraction was performed as described above.
TsCl still present was removed by flash chromatography in hexane–
EtOAc, gradient 4:1/3:1, and the epimeric tosylates were crys-
tallized from EtOAc at rt overnight to furnish a pure ribo epimer 11
(5.22 g, 54%) together with 2.55 g of the mixture of the ribo/xylo
tosylates.
J¼3.6, 6.9, 13.6 Hz, 1H), 1.95 (ddq, J32¼4.7 Hz, J34¼9.7 Hz, JH3–Me
¼
6.8 Hz, 1H, H3), 1.51, 1.43, 1.35, 133 (four s, 3H each, isopropylidene
Me), 1.19 (d, JMe–H3¼6.8 Hz, 3H, C3–Me). 13C (75 MHz): 111.6, 109.5,
104.9, 83.5, 82.7, 77.6, 67.2, 42.4, 26.8, 26.5, 26.3, 25.2, 10.1.
Compound 8. 1H (300 MHz), major epimer: 5.77 (d, J12¼3.6 Hz,
1H, H1), 4.56 (t, J¼4.2 Hz, 1H, H2), 3.87 (dd, J¼4.2, 10.1 Hz, 1H),
3.81–3.71 (unresolved, 2H), 2.71 and 2.50 (two br s, exchangeable,
–OH), 2.04 (ddq, J32¼4.6 Hz, JH3–Me¼6.6 Hz, J34¼10.7 Hz, 1H, H3),
1.52 and 1.33 (two s, 3H each, isopropylidene Me), 1.15 (d, JMe–H3
¼
Mp 120–123 ꢀC (EtOAc), [a]D24 þ22.1 (c 6, CHCl3). 1H (300 MHz):
7.80 (d, J¼8.2 Hz, 2H, H aromatic), 7.34 (d, J¼8.2 Hz, 2H, H aro-
matic), 5.68 (d, J12¼3.5 Hz, 1H, H1), 4.52 (t, J¼4.1 Hz, 1H, H2), 4.23
6.8 Hz, 3H, C3 Me). 13C (75 MHz): 11.7, 104.6, 83.8, 83.4, 72.6, 63.3,
40.3, 26.7, 26.3, 10.0. HRMS (electrospray) calcd for C10H18O5þNaþ:
241.10521, found: 241.10432.
´
0
0
00
00
(dd, J5 4¼2.5 Hz, J5 5 ¼11.1 Hz, 1H, H5), 4.07 (dd, J5 4¼3.9 Hz,
00
00
0
0
00
J5 5 ¼11.1 Hz, 1H, H5 ), 3.86 (dt, J45 ¼J45 ¼3.2 Hz, J43¼10.3 Hz, 1H,
H4), 2.44 (s, 3H, Me), 2.02 (ddq, J32¼4.5 Hz, J3-Me¼7.5 Hz, J34¼10.0H,
1H, H3), 1.46 and 1.31 (two s, 3H each, isopropylidene Me), 1.03 (d,
JMe–H3¼6.8 Hz, 3H, C3–Me). 13C (75 MHz): 144.9, 132.8, 129.8,
128.0, 111.7, 104.8, 82.5, 79.9, 68.5, 39.2, 29.7, 26.3, 21.6, 9.0. HRMS
3.4. 3-Deoxy-1,2-O-isopropylidene-3-C-methyl-D-ribo-
furanose 10
To a magnetically stirred solution of the allo epimer 7 (0.72 g,
2.8 mmol) in dry EtOAc (15 ml) was added H5IO6 (0.76 g, 3.4 mmol).
The mixture turned opaque within several seconds. Stirring was
continued for 1 h 45 min. TLC in 18:3 hexane–EtOAc showed that all
substrate with Rf 0.40 reacted to form the aldehyde 9 with Rf 0.18.
The mixture was filtered through a sintered glass and nearly all
solvent was evaporated. Small quantity of solid material appeared.
Et2O was added and the mixture was passed through a bed of Celite.
The volatiles were evaporated. The residue was solubilized in
technical grade EtOH (20 ml) and NaBH4 (0.07 g, 1.9 mmol) was
added in one portion while maintaining magnetic stirring. After
2 h, TLC showed a new more polar alcohol 10 with Rf 0.51 in 20:0.5
CH2Cl2–MeOH. Evaporation and flash chromatography in 20:0.4
CH2Cl2–MeOH gave 10 (0.32 g, 61%).
(electrospray) calcd for
343.12031.
C
16H22O6SþHþ: 343.12099, found:
3.6. 3,5-Dideoxy-3-C-methyl-D-ribo-furanose 15
To a solution of the tosylate 11 (2.9 g, 8.4 mmol) in THF (20 ml)
under nitrogen was injected 1 M LiBH(Et)3 (15 ml) at rt and the
mixture was left overnight. TLC showed that all 11 with Rf 0.26
disappeared to form a faint, weakly charring spot of 13 with Rf 0.65
in 4:1 hexane–EtOAc. H2SO4 (1 M, 30 ml) was cautiously added
(1.5 M H2SO4 could also be used). Two layers were formed. Small
volume of technical grade THF (w10 ml) was added to achieve
homogeneity. Slightly opaque solution was magnetically stirred
overnight. TLC showed that the product having a putative structure
14 was less polar than the acetonide 13. The mixture was trans-
ferred to a separatory funnel charged with CH2Cl2–water and ex-
haustive extraction (3ꢁ) was performed. The organic phase was
washed once with water. Combined water phases were extracted
with EtOAc (3ꢁ) and the combined organic phases were washed
with water. To the CH2Cl2 phase was added methanol (200 ml) and
the volatiles were evaporated. To the residue, MeOH was added
again (300 ml) and was subsequently evaporated. This was re-
peated two more times. The EtOAc layer was treated in the same
manner. TLC of the combined residues showed that 14 was no
longer present. The newly formed diol 15 showed Rf 0.38 in 20:1.3
CH2Cl2–MeOH. Flash chromatography in this system furnished 15,
0.82 g as an oil in 73% cumulative yield counted on a tosylate 11.
Diol 15 chars more intensely while using 2% CrO3 in 10% aq H2SO4
than using 10% H2SO4 in MeOH. In another run yield of 15 was 54%.
[a]2D4 þ27.9 (c 5.7, CHCl3, after 30 min). 1H (500 MHz): 5.383 (d,
J12¼3.8 Hz, H1), 5.191 (s, H1), 5.11 (br s, exchangeable, –OH), 4.038
(t, J¼4.5 Hz, H2), 3.943 (d, J23¼4.3 Hz, H2), 3.891 (dq, J4-Me¼6.0 Hz,
J43¼10.2 Hz, H4), 3.859 (dq, J4-Me¼6.1 Hz, J43¼10.3 Hz), 3.70 and
3.28 (two br s, exchangeable, –OH), 2.006 (m of 14 lines,
J3-Me¼6.9 Hz, J34¼w9.6 Hz, J32¼4.3 Hz, H3), 1.705 (m of 14 lines,
J3-Me¼6.6 Hz, J34¼9.6 Hz, J32¼w5.5 Hz, H3), 1.277 (d, JMe–H4¼6.2 Hz,
terminal Me), 1.183 (d, JMe–H4¼6.1 Hz), 1.001 (d, JMe–H2¼7.0 Hz, C2–
Me), 0.995 (d, JMe–H2¼6.9 Hz, C2–Me). 13C (75 MHz): 102.29, 97.13,
80.91, 78.81, 78.51, 73.58, 44.84, 42.28, 20.68, 18.82, 9.73, 9.16.
HRMS (EI) calcd for C6H12O3–OH: 115.07589, found: 115.07717.
1H (300 MHz): 5.80 (d, J12¼3.6 Hz, 1H, H1), 4.58(t, J¼4.1 Hz, 1H,
H2), 3.92–3.81 (three groups of multiplets0,0 2H), 3.55 (ddd,
00
00
00
0
0
J5 4¼3.6 Hz, J5 -OH¼7.4 Hz, J5 5 ¼12.0 Hz, 1H, H5 ), 2.29 (dd, JOH-5
¼
00
4.9 Hz, JOH-5 ¼7.8 Hz, 1H, exchangeable, –OH), 2.07 (ddq,
J32¼4.7 Hz, J3-Me¼6.8 Hz, J34¼10.2 Hz, 1H, H3), 1.51 and 1.34 (two s,
3H each, isopropylidene Me), 1.07 (d, JMe-3¼6.8 Hz, 3H, C3-Me). 13
C
(75 MHz): 111.5, 104.8, 83.1, 83.0, 61.3, 38.0, 26.6, 26.2, 9.1. HRMS
(electrospray) calcd for C9H16O4þNaþ: 211.09410, found: 211.09376.
3.5. 3-Deoxy-1,2-isopropylidene-3C-methyl-5-O-tosyl-
ribo-furanose 11
D-
3.5.1. From D-ribo alcohol 10
Compound 10 (0.32 g, 1.7 mmol) was conventionally tosylated
using TsCl (0.65 g, 3.4 mmol) in CH2Cl2 (20 ml), pyridine (1 ml), and
cat. DMAP overnight. Few drops of water were added to destroy the
excess of TsCl for 1 h. Partition between CH2Cl2 and dil HCl, washing
the organic phase with water, drying, evaporation, and crystalli-
zation from EtOAc gave 11 (0.45 g, 75%) with Rf 0.26 in 4:1 hexane–
EtOAc.
3.5.2. From a mixture of 6 and 7 obtained by hydrogenation of 5
over Adam’s catalyst via epimeric alcohols 12
To a cold (ice bath) magnetically stirred solution of 6 and 7
(7.3 g, 28.3 mmol) in dry EtOAc was added H5IO6 (8.4 g, 36.8 mmol)
in one portion under the atmosphere of nitrogen. The cooling bath
was removed. After 2.5 h, the mixture was worked-up as described
for 9 to furnish yellowish oil. This oil was solubilized in 80 ml of 96%
EtOH, cooled in ice bath, and treated with NaBH4 (1.0 g, 28.3 mmol)
added portionwise while maintaining magnetic stirring; 30 min
later dil acetic acid was added to neutrality (indicator paper) and
the volatiles were evaporated. The residue was solubilized in EtOAc
and the resulting opaque solution was washed with water. The
organic phase was dried, filtered, evaporated, and finally dried on
3.7. 2-Naphthylmethyl (3R)-formyloxy-(2R)-methylbutanoate
19 and 2-napthylmethyl (3R)-hydroxy-(2R)-methylbutanoate 20
3.7.1. Procedure A
To
a magnetically stirred solution of the diol 15 (011 g,
0.83 mmol) in EtOH (8 ml) was added NaIO4 (0.21 g, 1 mmol),

5558
B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
solubilized in minimum volume of distd H2O (w2 ml). The
mixture became opaque immediately. After 1.5 h, TLC showed
a high migrating very weakly charring spot (Rf 0.87 in 20:1
CH2Cl2–MeOH) and a small amount of unreacted 15. Few milli-
grams of solid NaIO4 were added to complete the diol cleavage.
The mixture was filtered through sintered glass and evaporated
at ca. 20 ꢀC. Addition of CH3CN (8 ml) precipitated some more
white solid. Filtration and evaporation were repeated. To the
residue solubilized in CH3CN (25 ml) was added NaClO2 (80%
pure, 0.26 g, 2.21 mmol) in minimum volume of distd water,
followed by 0.2 ml of 35% H2O2 and NaH2PO4$2H2O (0.066 g,
0.42 mmol) in minimum volume of distd water. The additions
were done in this order. The mixture was stirred for 2 h. TLC run
in 20:1 CH2Cl2 and revealed in bromocresol green reagent
showed two yellow elongated spots on a blue background: the
upper more intense one belonging to the O-formylated acid 17
and the more polar spot belonging to the hydroxy acid 18. The
reaction mixture was filtered through Celite and the volatiles
were evaporated and dried on an oil pump. The residual oil was
solubilized in CH2Cl2 (20 ml) and a freshly prepared solution of
2-naphthyldiazomethane 23 was added using a Pasteur pipette.
Addition of 23 was continued until reddish color persisted and
evolution of nitrogen ceased. TLC at this point showed the
presence of strongly UV absorbing spots of 19 with Rf 0.65 and
20 with Rf 0.30 in 4:1 hexane–EtOAc. Several other UV absorbing
spots were present presumably arising from the decomposition
of 23. Upon spraying with a CrO3–H2SO4 system and heating at
120–130 ꢀC both 19 and 20 form reddish-brown spots. Evapora-
tion of the solvent and flash chromatography using a gradient of
EtOAc in hexane, 1:19/1:3, furnished 19 (0.10 g, 42%) and 20
(0.081 g, 37%).
3.8. 2-Naphthylmethyl hydrazone 22
To a magnetically stirred solution of 2-naphthylcarbaldehyde 21
(15.0 g, 96 mmol) in techn. EtOH (70 ml) was added hydrazine
hydrate (100%, 9.6 g, 9.3 ml, 188 mmol) via a syringe for w5 min
with external cooling in ice bath. After the addition was complete,
cooling bath was removed and semi-solid slurry was left overnight
at rt. The semi-solid mixture was cooled, filtered on a sintered glass,
and washed with an ice-cold 1:1 EtOH–Et2O and finally with ice-
cold Et2O. Final drying on an oil pump gave the crude hydrazone 22
(15.0 g, 92%). Mp 148–151 ꢀC (EtOH), lit.36 149–150 ꢀC.
1H (300 MHz, CDCl3þa drop of DMSO-d6): 7.88–7.76 (m, 6H),
7.51–7.41 (m, 2H), 5.9 (br s, exchangeable, –NH2). 13C (75 MHz):
141.3, 132.7, 132.6, 125.68, 125.52, 125.46, 122.2. HRMS (electro-
spray) calcd for C11H10N2þHþ: 171.09167, found: 172.09165.
3.9. 2-Naphthylmethyldiazomethane 23
A mixture of the hydrazone 22 (1.08 g, 6.4 mmol) in CH2Cl2
(40 ml), yellow HgO (1.96 g, 9.0 mmol), and a small crushed pellet
of KOH was magnetically stirred for 3 h. The initial bright yellow-
orange color of the heterogeneous mixture gradually faded and
a gray deposit of mercury appeared. Stirring was stopped and after
sedimentation of the solids the orange-red solution of 23 was
pipetted for the esterification step. This solution was always freshly
prepared. Although we have never experienced any problem with
this preparation, it should be remembered that any diazo com-
pound is potentially unstable.
3.10. 1,2-O-Isopropylidene-L-xylo-furanose 24
L-Xylose (20 g, 134 mmol) in 520 ml of dry acetone and 20 ml of
3.7.2. Procedure B: 2-naphthylmethyl (3R)-hydroxy-(2R)-methyl-
butanoate 20 from 2-naphthylmethyl (3R)-formyloxy-(2R)-methyl-
butanoate 19
To a solution of the formate 19 in MeOH (6 ml) was added one
drop of concd NH4OH. After 1 h, TLC showed a complete conversion
of 19 to a more polar 20 (system as above). Evaporation and flash
chromatography furnished 20 (0.080 g, 89%).
concd H2SO4 were stirred magnetically for 50 min. The flask was
cooled in ice-water bath and a solution of 26 g, 246 mmol of anhyd
Na2CO3 in 225 ml of water was gradually added to keep the internal
temperature below 20 ꢀC and to avoid excessive frothing; 2.5 h after
the end of addition, solid anhyd Na2CO3 (14 g,132 mmol) was added
for 15 min to neutralize all H2SO4. Stirring was continued for 10 min
more. TLC showed a spot with Rf 0.45 (CH2Cl2–MeOH). The solids
were filtered and acetone was evaporated. The residual oil was
purified by flash chromatography in CH2Cl2–MeOH, 30:1/20:1, to
furnish 24 (19.0 g, 75%) as a syrup, which solidified upon storage in
a refrigerator. An alternative one-pot preparation described for
3.7.3. Procedure C: 2-naphthylmethyl (3R)-hydroxy-(2R)-methyl-
butanoate 20 from a mixture of 19 and 20 without chromatographic
separation as in procedure A
Isolation of the formate 19 can be skipped and a crude mixture
after the esterification with 23 can be subjected to ammonolysis.
Thus, evaporation of CH2Cl2, solubilization in MeOH (10 ml), and
addition of few drops of concd NH4OH transformed 19 to 20 for
w45 min. Evaporation of the solvent and flash chromatography
using a gradient of EtOAc in hexane, 1:4/1:3, furnished 20 in
w70% yield counting on the diol 15.
L
-xylose is much longer and was considered less convenient.46
[a]2D4 þ18.2 (c 3.4, CHCl3), commercial
D-form from the Aldrich
has [a]2D4 ꢂ19.2 (c 1, H2O). 1H (300 MHz): 5.99 (d, J12¼3.6 Hz, 1H,
H1), 4.53 (d, J21¼3.6 Hz, 1H, H2), 4.33 (br s, 1H), 4.19–4.02 (m, 4H,
one proton exchangeable), 1.49 and 1.33 (two s, 3H each, iso-
propylidene Me). 13C (75 MHz): 111.8, 104.8, 85.6, 78.6, 76.9, 61.1,
26.7, 26.1. HRMS (electrospray) calcd for C8H14O5þNaþ: 213.07736,
found: 213.07320.
Compound 19. 1H (300 MHz): 7.90 (s, 1H, CHO), 7.84–7.81 (m, 4H,
H aromatic), 7.50–7.43 (m, 3H, H aromatic), 5.29 (s, 2H, –OCH2–),
5.27 (quintette, J¼6.6 Hz, 1H, H3), 2.80 (quintette, J¼7.2 Hz, 1H, H2),
1.26 (d, J¼6.4 Hz, 3H, terminal Me), 1.21 (d, J¼7.1 Hz, 3H, C2–Me).
13C (75 MHz): 173.3, 160.2, 133.2, 133.1, 128.4, 128.0, 127.7, 127.5,
126.4, 126.3, 125.9, 71.5, 66.6, 44.6, 17.0, 12.8. HRMS (electrospray)
calcd for C17H18O4þNaþ: 309.11029, found: 309.10968; calcd for
3.11. 5-O-tert-Butyldimethylsilyl-1,2-O-isopropylidene-
xylo-furanose 25
L-
To a solution of the diol 24 18.5 g, 97.4 mmol in dry DMF
(Aldrich, sold with molecular sieves) 100 ml under a blanket of
nitrogen, was added imidazole 16.5 g, 240 mmol and tert-butyldi-
methylsilyl chloride 14.8 g, 110 mmol. After an overnight reaction,
TLC showed one spot with Rf 0.45 in 4:1 hexane–EtOAc. Conven-
tional extraction and flash chromatography in hexane–EtOAc,
17:3/4:1, furnished 22 g, 75% of 25.
C
17H18O4þKþ: 325.08420, found: 325.08357.
Compound 20. [a]2D4 ꢂ24.1 (c 5, CHCl3). 1H (500 MHz): 7.85–7.82
(4H, H aromatic), 7.51–7.44 (3H, H aromatic), 5.32 (s, 2H, –OCH2–),
3.92 (quintette of broadened lines, J¼6.3 Hz, 1H, H3), 2.64 (br s,
exchangeable, –OH), 2.54 (quintette, J¼7.2 Hz, 1H, H2), 1.222 (d,
J¼6.4 Hz, 3H, terminal Me), 1.216 (d, J¼7.2 Hz, 3H, C2–Me). 13C
(75 MHz): 175.7, 133.14, 133.09, 128.4, 128.0, 127.7, 127.3, 126.32,
126.29, 125.7, 69.4, 66.5, 47.1, 20.7, 14.1. HRMS (electrospray) calcd
for C16H18O3þNaþ: 281.11483, found: 281.11465.
[a]2D4 þ11.0 (c 5.4, CHCl3). 1H (300 MHz): 5.96 (d, J12¼3.6 Hz, 1H,
H1), 4.51 (d, J21¼3.6 Hz, 1H, H2), 4.38 (d, J¼2.7 Hz, 1H), 3.37 (ap-
parent t, J¼2.5 Hz, 2H, one H exchangeable), 4.15–4.08 (unresolved,
2H, H5), 1.48 and 1.32 (two s, 3H each, isopropylidene Me), 0.89 (s,

B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
5559
t
9H, Bu), 0.11 (s, 6H, Me2Si). 13C (75 MHz): 111.5, 105.0, 85.6, 78.1,
performed in a Parr apparatus for 2.5 h. Filtration of Pt (attention:
pyrophoric) on the bed of Celite, evaporation, and drying furnished
3.6 g, approximately quantitative yield, of both epimeric 3-C-
methyl products 28 and 29 in w10:1 proportion, which are in-
separable under the flash chromatography conditions.
77.1, 62.4, 26.8, 26.1, 25.7, 18.1, ꢂ5,5, ꢂ5.7. HRMS (electrospray)
calcd for C14H28O5SiþHþ: 305.17786, found: 305.17726.
3.12. 5-O-tert-Butyldimethylsilyl-3-deoxy-1,2-O-
isopropylidene-3-C-methylene-
L
-erythro-pentofuranose 26
13C (75 MHz): 111.6, 111.5, 104.9, 104.5, 86.8, 83.1, 80.2,
62.1, 40.5, 26.7, 26.3, 26.1, 10.9. HRMS (electrospray): calcd for
C9H16O4þNaþ¼211.09410, found: 211.09380.
A complex of CrO3$Py was prepared by the addition of pyridine
(16.9 ml, 209 mmol) to CrO3 (10.4 g, 104 mmol) in dry CH2Cl2
(200 ml) under nitrogen while maintaining stirring for 30 min. To
this dark brown solution was added 25 (10.0 g, 32.9 mmol) in 40 ml
of dry CH2Cl2 immediately followed by Ac2O (10.2 ml, 108 mmol).
The mixture became warm. TLC (hexane–EtOAc 17:3) showed
a conversion of 25 into a slightly less polar product. After 10 min,
the excess of chromium compounds was precipitated by the addi-
tion of 240 ml of 1:1 EtOAc and toluene mixture. Black solids were
washed twice with 1:1 EtOAc–toluene. The original supernatant
and the washings were applied on top of a silica gel (70–200 mesh)
column prepared in 2:1 EtOAc–toluene. Slight overpressure was
used to push the solution through the gel. Elution with 2:1 EtOAc–
toluene was performed. The fractions containing the product were
pooled together, evaporated, and co-evaporated with xylenes. Final
drying was performed on an oil pump.
3.15. 3-Deoxy-1,2-O-isopropylidene-3-C-methyl-5-O-tosyl-
ribo-furanose 30
L-
Tosylation of the 28 and 29 mixture and crystallization of the
major C-3-methyl -ribo epimer proceeded as described for the
enantiomeric compound 11 derived from -glucose. Compound 30
was obtained in 69% yield together with 26% of the mixture 3-C-
methyl-5-O-tosyl -ribo/xylo compounds, inseparable under the
L
D
L
conditions of flash chromatography.
Mp 119–122 ꢀC (EtOAc), [a]D24 ꢂ21.4 (c 7, CHCl3). For 1H and 13C
NMR data see enantiomeric compound 11. HRMS (electrospray)
calcd for C16H22O6SþHþ: 343.12097, found: 343.12054.
3.16. 3,5-Dideoxy-3C-methyl-L-ribo-furanose 31
To a suspension of Ph3PþMe(Brꢂ) (98%, 17.5 g, 48 mmol) in
600 ml of THF cooled with EtOH–dry ice was injected 2.5 M BuLi in
hexane (18.2 ml, 45.5 mmol) with constant magnetic stirring. After
the addition, cooling bath was removed for 50 min and re-applied
again. A solution of the abovementioned oxidation product in
100 ml of THF was added via a syringe and the mixture was left to
reach rt overnight. TLC showed a complete conversion of the sub-
strate with Rf 0.44 into a strongly charring spot of the olefin 25 with
Rf 0.80 in 17:3 hexane–EtOAc. The solids were filtered on a sintered
glass and THF was evaporated. The resulting yellow oil was solubi-
lized in CH2Cl2 and washed with water. The extract was dried and
evaporated, and the residue was purified by flash chromatography in
19:1 hexane–EtOAc to furnish 26 (6.9 g, 70%) as an oil for two steps.
[a]2D4 ꢂ127.6 (c 3.9, CHCl3). 1H (300 MHz): 5.85 (d, J12¼4.1 Hz, 1H,
H1), 5.42 (dd, J¼1.2, 2.2 Hz, 1H), 5.27 (t, J¼1.7 Hz, 1H), 4.91–4.87 (m,
Compound 30 was converted to 31 in 64% yield following the
procedure described for the transformation of its D-ribo enantiomer
11 to obtain 15.
[a]2D4 ꢂ26.0 (c 1.8, CHCl3). For NMR data see compound 15. HRMS
(EI) calcd for C6H12O3ꢂOH: 115.07589, found: 115.07611.
3.17. O-Methyl 3,5-dideoxy-3-C-methyl-a-L-ribo-furanoside 32
An unexpected glycosylation took place during conversion of 30
to 31 in one case, evidently due to the presence of acid in the crude
CH2Cl2 and EtOAc extracts. The residue after evaporation of both
solvents was solubilized in MeOH and left overnight at rt, and
evaporation was reassumed next day to destroy any borates pres-
ent. TLC (CH2Cl2–MeOH 20:1.3) showed the presence of two com-
pounds: 32 with Rf 0.62 and 31 with Rf 0.41. Flash chromatography
in the same system furnished 32 (0.43 g, 28%) and 31 (0.30 g 22%).
This unwanted glycosylation can be avoided by performing all
evaporations and chromatography in 1 day.
0
0
0 00
1H), 4.78–4.74 (m, 1H), 3.76 (dd, J5 4¼4.1 Hz, J5 5 ¼10.6 Hz, 1H, H5 ),
00
00
00 0
3.68 (dd, J5 4¼3.8 Hz, J5 5 ¼10.6 Hz, 1H, H5 ), 1.50 and 1.39 (two s,
3H each, isopropylidene Me), 0.88 (s, 9H, tBu), 0.06 and 0.05 (two s,
6H, Me2Si). 13C (75 MHz): 147.6, 112.5, 111.5, 104.9, 81.99, 81.96,
80.8, 65.6, 27.5, 27.3, 25.8, 18.2, ꢂ5.4, ꢂ5.5. HRMS (electrospray)
calcd for C15H28O4SiþHþ: 301.18295, found: 301.18239.
1H (500 MHz): 4.74 (s, 1H, H1), 3.97 (d, J23¼3.9 Hz, 1H, H1), 3.90
(dq, J4-Me¼6.1 Hz, J43¼9.3 Hz, 1H, H4), 3.35 (s, 3H, –OMe), 2.51 (br s,
exchangeable, –OH), 1.98 (ddq, J34¼9.3 Hz, J23¼3.8 Hz, J3-Me¼7.0 Hz,
1H, H3), 1.28 (d, JMe-4¼6.1 Hz, 3H, C4-Me), 1.04 (d, JMe-4¼7.0 Hz, 3H,
C3–Me). 13C (75 MHz): 108.9, 80.9, 80.7, 54.3, 42.7, 20.9, 9.3. HRMS
(EI) calcd for C7H17O3ꢂCH3: 115.07590, found: 115.07692.
3.13. 3-Deoxy-1,2-O-isopropylidene-3-C-methylene-L-erythro-
pentofuranose 27
Thesilane26(6.2 g, 20.7 mmol)intechnicalgradeTHF(30 ml)was
deprotected with 1 M Bu4NF–THF (22 ml) overnight. Rf of the product
is 0.43 in 1:1 hexane–EtOAc. Evaporation of THF and flash chroma-
tography in the same system furnished 3.8 g, quantitative yield, of 27.
[a]2D4 ꢂ183.7 (c 2.8, CHCl3). 1H (300 MHz): 5.87 (d, J12¼4.0 Hz, 1H,
H1), 5.48 (dd, J¼0.9, 2.3 Hz,1H), 5.19 (dd, J¼1.2, 2.1 Hz,1H), 4.93–4.91
3.18. 2-Naphthylmethyl (3S)-hydroxy-(2S)-methyl-
butanoate 33
Transformation of the diol 31 to the target 33 was performed as
described for the enantiomeric compound 15/20 in 58% cumu-
lative yield.
0
0
00
(m, 1H), 4.85–4.80 (m, 1H), 3.88 (d, J5 5 ¼11.9 Hz, 1H, H5 ), 3.67 (dd,
J5 4¼4.4 Hz, J5 5 ¼12.0 Hz, 1H, H5 ), 2.16 (br s, exchangeable, 1H,
–OH), 1.52 and 1.39 (two s, 3H each, isopropylidene Me). 13C
(75 MHz): 145.5, 112.6, 112.2, 104.4, 82.0, 80.0, 63.4, 21.4, 27.1. HRMS
(electrospray) calcd for C9H14O4þNaþ: 209.07845, found: 209.07818.
[a]2D4 þ24.0 (c 6.8, CHCl3). For NMR data see compound 20.
HRMS (electrospray) calcd for C16H18O3þKþ: 297.08874, found:
297.08869.
00
00
00 0
3.19. 5-O-tert-Butyldiphenylsilyl-D-arabino-furanose 34
3.14. 3-Deoxy-1,2-O-isopropylidene-3-C-methyl-L-ribo-
furanose 28 and 3-deoxy-1,2-O-isopropylidene-3-C-methyl-
xylo-furanose 29
L
-
D
-Arabinose (30 g, 0.2 mol) was solubilized in dry DMF (300 ml)
with warming to ca. 100 ꢀC. When the temperature of this solution
dropped to ca. 50 ꢀC, imidazole (27.2 g, 0.4 mol) and tert-butyldi-
phenylsilyl chloride (55 g, 52 ml, 0.2 mol) were added under ni-
trogen, and the whole was stirred at 55–60 ꢀC for 2.5 h. Extraction
Hydrogenation of the olefin 27 (3.6 g, 19.3 mmol) in 96% EtOH
(22 ml) and PtO2$H2O (0.102 g) at the initial pressure of 45 psi was

5560
B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
(1 M HCl–CH2Cl2), washing of the organic phase with water, drying,
and evaporation furnished a yellowish oil. TLC (hexane–EtOAc, 2:3)
showed the product with Rf 0.4 and impurities moving with the
solvent front. Flash chromatography using a gradient of EtOAc in
hexane, 2:3/1:1/3:2, gave 34 (49 g, 63%) as a colorless oil.
1H (300 MHz): 5.43 (br s, H1a, integration of both signals was ca.
3:1), 5.30 (dd, J12¼3.6 Hz, J1-OH¼7.5 Hz, H1b, integration of both
signals was ca. 3:1). 13C (75 MHz): 135.64, 135.60, 131.86, 131.64,
130.24, 130.20, 130.13, 128.01, 127.96, 103.5, 97.0, 87.3, 83.1, 78.8,
78.2, 77.9, 76.7, 64.5, 64.1,26.8, 26.7, 19.2, 19.0. HRMS (electrospray)
calcd for C21H28O5SiþNaþ: 411.15984, found: 411.15892.
13C (75 MHz): 145.7, 135.7, 134.8, 133.4, 133.3, 129.7, 129.6, 127.70,
127.67, 114.0,113.1,105.4, 83.1, 81.4, 67.2, 27.3, 26.8, 26.5,10.2. HRMS
(electrospray) calcd for
447.19620.
C
25H32O4SiþNaþ: 447.19677, found:
3.22. 3-Deoxy-1,2-O-isopropylidene-3-C-methylene-
D-threo-furanose 37
Olefin 36 (7.9 g, 18.6 mmol) in technical grade THF (50 ml) was
treated with 1 M Bu4NF (20 ml, 20 mmol) for 2 h. TLC showed
a new spot with Rf 0.31 in 3:2 hexane–EtOAc. Evaporation of the
solvent and flash chromatography in 55:45 hexane–EtOAc gave 37
(4.5 g, 92%) as oil.
3.20. 5-O-tert-Butyldiphenylsilyl-1,2-O-isopropylidene-
D
-arabino-furanose 35
A convenient shortcut to get 37 is to skip a chromatography after
the Wittig reaction and to add Bu4NF directly to the THF solution of
the crude extract. Thus, starting from 16.2 g of 35, 5.1 g of the al-
cohol 37 was obtained in a cumulative yield of 72% for three steps.
[a]2D4 ꢂ28.4 (c 6.2, CHCl3). 1H (300 MHz): 5.85 (d, J12¼3.9 Hz, 1H,
H1), 5.49 (dd, J¼1.0, 2.1 Hz,1H), 5.27 (t, J¼1.5 Hz,1H), 4.90 (dt, J¼1.0,
1.0, 3.9 Hz, 1H, H2), 4.65 (ddd, J¼2.0, 4.1, 6.3 Hz, 1H, H4), 3.82 (dd,
Compound 34 (32 g, 82.5 mmol) in dry acetone (360 ml), concd
H2SO4 (1.6 ml), and CuSO4 (dried at ca. 150 ꢀC for 4 h, 36 g) were
stirred under nitrogen for 7 h. The solids were filtered on a sintered
glass. The filtrate was neutralized using cold concd NH4OH and the
solids were filtered again. TLC showed a spot of 35 with Rf 0.39 (in
hexane–EtOAc 4:1) and a minor less polar compound. Evaporation
and flash chromatography in hexane–EtOAc, 5:1/4:1, gave 35
(25 g, 71%) as a syrup. For the next oxidation step, a crude olefi-
nation product could be used.
J5 4¼7.1 Hz, J5 5 ¼11.7 Hz, 1H, H50), 3.70 (dd, J5 4¼4.2 Hz,
0
0
00
00
00
00
0
J5 5 ¼11.7 Hz, 1H, H5 ), 2.70 (br s, exchangeable, –OH), 1.56 and 1.36
(two s, 3H each, isopropylidene –Me). 13C (75 MHz): 145.0, 113.7,
113.4, 105.2, 83.7, 81.2, 65.8, 27.2, 26.5. HRMS (electrospray) calcd
for C9H14O4þNaþ: 209.07900, found: 209.07799.
[a]2D4 þ5.6 (c 3.2, CHCl3);
L
-enantiomer:20 [a]D24 ꢂ5 (c 1.2, CHCl3).
1H (300 MHz): 7.68–7.65 (m, 4H, H aromatic), 7.44–7.35 (m, 6H, H
aromatic), 5.87 (d, J12¼4.0 Hz, 1H, H1), 4.53 (d, J21¼4.0 Hz, 1H, H2),
4.42 (t, J¼3.1 Hz, 1H, H3), 4.06 (dt, J¼2.4, 6.6, 6.6 Hz, 1H), 3.87–3.77
(m, 2H, H5), 1.32 and 1.28 (two s, 3H each, isopropylidene Me), 1.06
(s, 9H, tBu). 13C (75 MHz): 135.61, 135.58, 133.21, 133.17, 129.81,
129.79, 127.8, 112.53, 105.6, 87.5, 87.1, 76.3, 63.7, 26.9, 26.8, 26.1,
19.2. HRMS (electrospray) calcd for C24H32O5SiþNaþ: 451.19168,
found: 451.19078.
3.23. 4-Hydroxymethyl-3,4-ene-1,2-O-isopropylidene-3-
C-methyl-L-glycero-pentofuranose 38 and 3-deoxy-1,2-O-
isopropylidene-3-C-methyl-D-lyxo-furanose 39
Olefin 37 (2.41 g, 12.9 mmol) in abs EtOH (20 ml) and 0.1777 g of
PtO2 were hydrogenated in a Parr apparatus at the initial pressure
of 38 psi for 2.5 h. TLC showed the olefin 38 with Rf 0.58, the main
compound 39 with Rf 0.36 (hexane–EtOAc 1:1), and two more polar
unidentified compounds present in small concentration. Com-
pound 39 is slightly more polar than the substrate 37. Filtration
through Celite (attention: dry Pt is pyrophoric), evaporation of the
solvent, and flash chromatography in 1:1 hexane–EtOAc gave 38
(0.31 g, 13%) and 39 (1.73 g, 71%). Both 38 and 39 are oils.
Compound 38. [a]2D4 þ4.5 (c 4.8, CHCl3). 1H (300 MHz): 5.96 (d,
J12¼5.3 Hz, 1H, H1), 5.11 (d of unresolved quintettes, four J¼0.9 Hz,
J21¼5.3 Hz, 1H, H1), 4.17 (s, 2H, H5), 2.259 (br s, exchangeable, 1H,
–OH), 1.76, 1.45 and 1.43 (three s, 3H each, isopropylidene –Me, C3–
Me). 13C (75 MHz): 151.0,111.9,108.0,103.8, 87.2, 55.7, 28.0, 27.9, 9.1.
HRMS (electrospray) calcd for C9H14O4þNaþ: 209.07900, found:
209.07840.
3.21. 3-Deoxy-5-O-tert-butyldiphenylsilyl-1,2-O-isopro-
pylidene-3-C-methylene-D-threo-furanose 36
To a stirred suspension of CrO3 (10.1 g, 100 mmol) in dry CH2Cl2
(200 ml) under nitrogen was added pyridine (16.2 ml, 200 mmol).
After 25 min, a solution of 35 (11.0 g, 25.7 mmol) in dry CH2Cl2
(120 ml) was added followed immediately by Ac2O (9.5 ml,
100 mmol). Stirring was continued for 17 min. The reaction was
quenched by the addition of 200 ml of 1:1 EtOAc–toluene. The dark
supernatant was applied on top of a 60–200 mesh silica gel column
prepared in 2:1 EtOAc–toluene. Black residue after decantation was
washed twice with 1:1 EtOAc–toluene and the washings were also
applied on the column. Elution with 2:1 EtOAc–toluene, evapora-
tion of the volatiles, co-evaporation with xylene, and drying on an
oil pump gave colorless oil (Rf 0.32 in 6:1 hexane–EtOAc), which
was solubilized in dry THF and added dropwise to a Wittig reagent
prepared from Ph3PþCH3(Brꢂ) (98%, 14.3 g, 40 mmol) suspended in
350 ml of dry THF and 2.5 M BuLi in hexane (15.2 ml, 38 mmol) as
described for the synthesis of 5 and 26. The cooling bath was re-
moved. After 4 h, TLC showed that the substrate with Rf 0.32
reacted to form strongly charring new product with Rf 0.58 in 6:1
hexane–EtOAc. Filtration through Celite and evaporation furnished
yellow-brown oil, which was solubilized in CH2Cl2 and washed
with water. The organic phase was dried and evaporated. The res-
idue was subjected to flash chromatography in 15:1 hexane–EtOAc
to yield 36 as oil (7.9 g, 72%) for two steps.
Compound 39. [a]2D4 þ44.4 (c 1.4, CHCl3). 1H (300 MHz): 5.84 (d,
J12¼3.9 Hz, 1H, H1), 4.58 (dd, J21¼3.9 Hz, J23¼5.3 Hz, 1H, H1), 4.23
00
0
(ddd, J45 ¼4.0 Hz, J43¼8.3 Hz, J045 ¼9.8 Hz, 1H, H4), 3.94 (dd,
0
0
00
00
J5 4¼10.0 Hz, J5 5 ¼11.4 Hz, 1H, H5 ), 3.52 (dt, J5 -OH¼3.9 Hz, J¼10.6,
10.6 Hz, 1H, H5) [after decoupling of the doublet at d 2.24: 3.96 (dd,
0
0
00
00
00 0
J5 4¼9.8 Hz, J5 5 ¼11.5 Hz), 3.53 (dd, J5 4¼4.0 Hz, J5 5 ¼11.5 Hz)],
2.46 (m of 14 lines, J32¼5.4 Hz, J3-Me¼7.4 Hz, J34¼8.3 Hz, 1H, H3),
2.24 (d, J¼9.1 Hz, exchangeable, 1H, –OH), 1.56 and 1.31 (two s, 3H
each, isopropylidene –Me), 1.10 (d, JMe-3¼7.4 Hz, 3H, C3–Me).
13C (75 MHz): 112.1, 105.9, 84.2, 82.3, 62.6, 38.7, 26.5, 25.5, 7.7.
HRMS (electrospray) calcd for C9H16O4þNaþ: 211.09465, found:
211.09403.
3.24. 3-Deoxy-1,2-O-ispopropylidene-3-C-methyl-5-O-
tosyl-D-lyxo-furanose 40
[a]2D4 þ3.2 (c 7.2, CHCl3). 1H (300 MHz): 7.71–7.66 (m, 4H, H ar-
omatic), 7.41–7.33 (m, 6H, H aromatic), 5.80 (d, J12¼3.9 Hz, 1H, H1),
5.45 (dd, J¼0.8, 2.0 Hz,1H), 5.32 (t, J¼1.4 Hz,1H), 4.85 (dt, J¼0.9, 0.9,
Alcohol 39 (0.090 g, 0.48 mmol) in dry CH2Cl2 (10 ml), pyridine
(1 ml), cat. DMAP, and TsCl (0.45 g, 2.4 mmol) were stored under
nitrogen for 24 h. TLC (hexane–EtOAc 1:1) showed a complete
conversion to the less polar 40. Extraction (CH2Cl2–dil HCl),
0
3.9 Hz, 1H), 4.60 (tt, J¼1.8, 1.8, 6.7, 6.7 Hz, 1H), 3.89 (dd, J5 4¼6.4 Hz,
0
00
0
00
00
00 0
J5 5 ¼10.0 Hz, 1H, H5 ), 3.76 (dd, J5 4¼6.9 Hz, J5 5 ¼10.0 Hz, 1H, H5 ),
1.34 and 1.31 (two s, 3H each, isopropylidene –Me), 1.06 (s, 9H, tBu).

B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
5561
washing of the organic phase with water, drying, evaporation, and
flash chromatography in 3:2 hexane–EtOAc gave the product
(0.11 g, 67%) as a syrup.
H2O and 35% H2O2 (0.3 ml) in this order. The resulting yellow ho-
mogenous mixture was stirred for 30 min. TLC at this stage showed
the main spot of 46 and much less intense spot of the more polar 47.
Revelation was done using a bromocresol green reagent: both 45
and 46 formed yellow spots on a blue background. The mixture was
passed through a bed of Celite, evaporated, and solubilized in distd
CH2Cl2. Addition of freshly prepared 2-naphthyldiazomethane 23
was continued until evolution of nitrogen subsided and a reddish
color persisted. The solvent was evaporated. The residue was sol-
ubilized in MeOH (15 ml) and seven drops of concd NH4OH were
added. After 40 min, TLC (hexane–EtOAc 3:1) showed an UV ab-
sorbing spot of 48 with Rf 0.34 accompanied by several much less
polar spots. Spraying with a CrO3–H2SO4 system and heating re-
veals 48 as a reddish-brown spot. Evaporation of the solvent and
flash chromatography in 3:1 hexane–EtOAc furnished 48 as a glassy
material, 0.17 g, 54% counting on 44.
1H (300 MHz): 7.81(d, J¼8.2 Hz, 2H, H aromatic), 7.34 (d,
J¼8.2 Hz, 2H, H aromatic), 5.74 (d, J12¼3.7 Hz, 1H, H1), 4.49 (t,
J¼4.3 Hz, 1H, H2), 4.30–4.19 (unresolved, 3H), 2.34 (s, 4H, H3 and
–PhMe), 1.30 and 1.24 (two s, 3H each, isopropylidene –Me), 1.08 (d,
JMe-3¼7.3 Hz, C3-Me). 13C (75 MHz): 144.7, 132.7, 129.7, 128.0, 111.8,
105.9, 81.9, 80.3, 70.0, 39.2, 26.2, 25.2, 21.5, 8.1. HRMS (electrospray)
calcd for C16H22O6SþNaþ: 365.10349, found: 365.10276.
3.25. 3,5-Dideoxy-3-C-methyl-D-lyxo-furanose 44
To a cold (ice bath) solution of the alcohol 39 (1.3 g, 6.9 mmol) in
dry CH2Cl2 (30 ml) and pyridine (2 ml) with magnetic stirring was
added dropwise trifluoromethanesulfonic anhydride (2.4 g, 1.4 ml,
8.5 mmol). After 25 min, extraction was performed (CH2Cl2–dil
HCl). The organic phase was washed with ice water, dried, filtered
through sintered glass, evaporated, and dried on an oil pump to
furnish red-orange thick oil. The vacuums were broken using ni-
trogen in a balloon. All the work-up should be performed rapidly
(ca. 30 min) to avoid decomposition of the triflate 41. Dry THF
(30 ml) was added and the flask was cooled in an ice bath. To the
stirred solution was injected 1 M LiBH(Et)3 in THF (10 ml). The
initial red-orange color turned yellow immediately. The homoge-
nous mixture was left overnight at rt. Small quantity of unreacted
41 was still present and more LiBH(Et)3 (3.0 ml) was added and the
reaction was continued for 24 h more. TLC showed that the sub-
strate 41 with Rf 0.22 all reacted forming a faint brown-yellow spot
of the substitution product 42 with Rf 0.38 (in hexane–EtOAc 10:1)
and a black spot at the application point. The reaction mixture was
cooled in ice bath and 4 ml of 3 M H2SO4 was cautiously added
(frothing) followed by 10 ml of 1 M H2SO4. Cooling bath was re-
moved and hydrolysis was continued overnight. TLC (hexane–
EtOAc 10:1) showed the presence of a new compound, presumably
43, which was less polar than 42. Exhaustive extraction was per-
formed using CH2Cl2 (3ꢁ) and EtOAc (3ꢁ) as described for 14 and
15. The extracts were evaporated and co-evaporated several times
with MeOH to break the putative borate 43 and to release the diol
44. Flash chromatography of the residual yellowish oil in 20:1.3
CH2Cl2–MeOH furnished 44 (0.25 g, 27%) as oil, counting on the
alcohol 39.
[a]2D4 þ0.5 (c 6, CHCl3). 1H (500 MHz, after D2O exchange): 7.84–
7.81 (m, 4H, H aromatic), 7.50–7.43 (m, 3H, H aromatic), 5.30 (s, 2H,
–OCH2–), 4.09 (dq, J32¼4.0 Hz, J3-Me¼6.3 Hz, 1H, H3), 2.62 (br s, re-
sidual OH), 2.57 (dq, J23¼3.9 Hz, J2-Me¼7.2 Hz,1H, H2),1.22 (d, JMe-2
¼
7.3 Hz, 3H, C2–Me), 1.17 (d, JMe-3¼6.4 Hz, 3H, terminal Me). 13C
(125 MHz): 175.70, 133.10, 133.07, 128.43, 127.93, 127.67, 127.35,
126.32, 126.30, 125.70, 67.95, 66.50, 45.57, 19.80, 10.98. HRMS
(electrospray) calcd for
281.11477.
C
16H18O3þNaþ: 281.11483, found:
3.27. 5-O-tert-Butyldiphenylsilyl-3-deoxy-1,2-O-isopro-
pylidene-3-C-methylene- -threo-pentofuranose 50
L
Compound 5021 was obtained from 4920,21 following the di-
rections for the -enantiomer 35 in 56% yield for two steps, rather
D
than using Tebbe’s reagent or Peterson’s olefination as described
before.21
[a]D ꢂ3.3 (c 6, CHCl3). For NMR data see 36.
3.28. 5-O-tert-Butyldiphenylsilyl-3,4-ene-1,2-O-
isopropylidene-3-C-methyl-
O-tert-butyldiphenylsilyl-3-deoxy-1,2-O-isopropylidene-3-C-
methyl- -lyxo-furanose 52
D-glycero-pentofuranose 51 and 5-
L
Olefin 50 (0.41 g, 1.37 mmol) in EtOH (25 ml) and 10% Pd–C
(0.05 g) were hydrogenated at the initial pressure of 40 psi in a Parr
apparatus for 5 h. TLC showed traces of unreacted substrate and
a less polar compound 51 with Rf 0.46 and the product 52 with Rf
0.19 (hexane–EtOAc 15:1). Filtration through Celite, evaporation,
and flash chromatography in a gradient of EtOAc in hexane,
1:15/2:15, gave 51 (oil, 0.19 g, 46%) and 52 (oil, 0.15 g, 36%).
Compound 51. [a]2D4 þ2.7 (c 2.3, CHCl3). 1H (300 MHz): 7.71–7.69
(m, 4H, H aromatic), 7.41–7.36 (m, 6H, H aromatic), 5.92 (d,
J12¼5.3 Hz, 1H, H1), 5.01 (d, J12¼5.3 Hz, 1H, H2), 4.19 (s, 2H, H5),
1.61 (s, 3H) and 1.45 (s, 6H): isopropylidene Me and C3–Me, 1.05 (s,
[a]2D4 þ32.3 (c 1.6, CHCl3, after 30 min). 1H (500 MHz, after D2O
exchange): 5.28 (d, J12¼4.6 Hz, H1), 5.27 (d, J12¼1.6 Hz, H1), 4.42
(quintette, J4-Me¼6.6 Hz, J43¼6.6 Hz, H4), 4.12 (quintette, J4-Me
¼
6.5 Hz, J43¼6.5 Hz, H4), 4.10 (dd, J21¼5.1 Hz, J23¼6.4 Hz, H2), 4.08
(dd, J21¼1.5 Hz, J23¼5.4 Hz, H2), 2.46 (m of 10 lines, J32¼5.5 Hz,
J34¼7.3 Hz, J3-Me¼7.3 Hz, H3), 2.28 (m of 6 lines, J¼7.0 Hz, H3), 1.22
(d, JMe–H¼6.6 Hz, terminal –Me), 1.17 (d, JMe–H¼6.6 Hz, terminal
–Me), 1.02 (d, JMe–H¼7.3 Hz, C2-Me), 0.98 (d, JMe–H¼7.3 Hz, C2–Me).
13C (75 MHz): 101.6, 96.5, 79.1, 77.1, 76.5, 73.5, 39.0, 38.2, 17.9, 17.6,
8.2, 7.5. HRMS (EI) calcd for C6H12O3ꢂCH3: 117.05517, found:
117.05518.
t
9H, Bu). 13C (75 MHz): 151.1, 135.70, 135.67, 133.21, 133.16, 129.8,
127.8, 111.7, 108.0, 103.9, 87.4, 57.5, 28.0, 26.8, 19.3, 9.2. HRMS
(electrospray) calcd for
447.19566.
C
25H32O4SiþNaþ: 447.19622, found:
3.26. 2-Naphthylmethyl (3R)-hydroxy-(2S)-methylbutanoate 48
Compound 52. [a]2D4 ꢂ32.8 (c 3.8, CHCl3). 1H (300 MHz): 7.69–
7.66 (m, 4H, H aromatic), 7.43–7.33 (m, 6H, H aromatic), 5.75 (d,
J12¼3.9 Hz, 1H, H1), 4.49 (t, J12¼J23¼4.5 Hz, 1H, H2), 4.24 (dt,
To a magnetically stirred solution of the diol 44 (0.16 g,
1.6 mmol) in EtOH (20 ml) was added NaIO4 (0.38 g, 1.8 mmol) in
4 ml of distd H2O. The mixture turned opaque immediately. After
2 h, TLC (CH2Cl2–MeOH 20:1.3) showed that all substrate has
reacted. The solids were filtered on a sintered glass and EtOH was
evaporated. To the residue was added CH3CN (20 ml). Small
quantity of the precipitated solids was again removed by filtration.
To the resulting clear solution was added NaClO2 (80%, 0.38 g,
3.2 mmol) solubilized in a minimum volume of distd H2O followed
by NaH2PO4$2H2O (0.076 g, 0.49 mmol) in minimum volume of
00
0
0
J45 ¼5.6 Hz, J43¼7.5 Hz, J45 ¼7.6 Hz, 1H, H4), 3.97 (dd, J5 4¼7.6 Hz,
0
00
0
00
00
00 0
J5 5 ¼10.2 Hz, 1H, H5 ), 3.77 (dd, J5 4¼5.3 Hz, J5 5 ¼10.3 Hz, 1H, H5 ),
2.42 (ddq, J32¼4.8 Hz, J34¼7.4 Hz, J3-Me¼7.3 Hz, 1H, H3), 1.27 and
1.24 (two s, 3H each, isopropylidene Me), 1.19 (d, JMe-3¼7.3 Hz, 3H,
C3-Me), 1.05 (s, 9H, tBu). 13C (75 MHz): 135.58, 135.54, 133.6, 133.4,
129.54, 129.52, 127.6, 111.6, 105.8, 84.2, 82.5, 64.3, 39.3, 26.8, 26.4,
25.5, 19.1, 8.6. HRMS (electrospray) calcd for C25H34O4SiþNaþ:
449.21187, found: 449.21146.

5562
B. Doboszewski, P. Herdewijn / Tetrahedron 64 (2008) 5551–5562
3.29. 3-Deoxy-1,2-O-isopropylidene-3-C-methylene-
threo-pentofuranose 53
L-
Dubyankova and Mr. Luc Baudemprez (Rega Institute) are thanked
for the 500 MHz spectra.
Compound 53 was obtained from 50 as described for its
antiomer 37 in 95% yield.
D-en-
References and notes
[a]2D4 þ28.8 (c 5, CHCl3). For NMR data see 37. HRMS (electro-
1. Smirnov, V. V.; Churkina, L. N.; Perepnikhatka, V. I.; Mukvich, N. S.; Garagulya,
A. D.; Kiprianova, E. A.; Kravets, A. N.; Dovzhenko, S. A. Appl. Biochem. Microbiol.
2000, 36, 46 and references cited therein.
2. Smirnov, V. V.; Kiprianova, E. A.; Gvozdyak, O. R.; Garagulya, A. D.; Churkina,
L. N.; Proskuryakova, N. B.; Kharchenko, L. A. Zh. Mikrobiol. Epidemiol. Im-
munobiol. 1999, 5, 77.
spray) calcd for C9H14O4þNaþ: 209.07845, found: 209.07787.
3.30. 3-Deoxy-1,2-O-isopropylidene-3-C-methyl-L-lyxo-
furanose 54
3. Witte, W.; Cuny, C. Mollmann, U. Chemother. J. 1997, 6, 48; Chem. Abstr. 1997,
127, 63039.
Compound 53 was subjected to hydrogenation over Adam’s
catalyst as described for its -enantiomer 37 to furnish 54 in 83%
4. Kamikiri, K.; Myazaki, S.; Nagai, K.; Suzuki, Y.; Yamaguchi, Y.; Shibazaki, M.;
Washisaki, S.; Tokunaga, T.; Butsudo, R.; Ratona, M. R. Patent JP 06319574A,
1994.
5. Kamigiri, K.; Suzuki, Y.; Shibazaki, M.; Morioka, M.; Suzuki, K.-I.; Tokunaga, T.;
Seitawan, B.; Rantiatmodjo, R. M. J. Antibiot. 1996, 49, 136.
6. Tokunaga, Y.; Kamigiri, K.; Orita, M.; Nishikawa, T.; Shimizu, M.; Kaniwa, H.
J. Antibiot. 1996, 49, 140.
7. Doboszewski, B.; Herdewijn, P. Tetrahedron Lett. 2008, 49, 1331.
8. Neri, C.; Williams, J. M. J. Adv. Synth. Catal. 2003, 345, 835.
9. Neri, C.; Williams, J. M. J. Tetrahedron Lett. 2002, 43, 4257.
10. Roush, W. R.; Bannister, T. D.; Wendt, M. D.; Jablonowaski, J. A.; Scheidt, K. A.
J. Org. Chem. 2002, 67, 4275.
D
yield. A different batch of Adam’s catalyst was used than in the case
of 37 and no migration of the C]C bond to the C3–C4 position was
seen.
[a]2D4 ꢂ44.4 (c 2.5, CHCl3). For NMR data see 39. HRMS (EI) calcd
for C9H16O4ꢂMe: 173.08137, found: 173.08187.
3.31. 3,5-Dideoxy-3-C-methyl-L-lyxo-furanose 55
11. Vicario, J. L.; Badia, D.; Dominguez, E.; Rodriguez, M.; Carrillo, J. J. Org. Chem.
2000, 65, 3754.
12. Harris, R. C.; Cutter, A. L.; Weisman, K. J.; Hanefeld, U.; Timoney, M. C.; Staunton,
J. J. Chem. Res. S. 1998, 283.
Compound 55 was obtained as described for its D-enantiomer 44
in 27% cumulative yield.
[a]2D4 ꢂ38.1 (c 3.2, CHCl3, after 30 min). For NMR data see 44.
HRMS (EI) calcd for C6H12O3ꢂMe: 117.05517, found: 117.05431.
13. Lu¨tzen, A.; Ko¨ll, P. Tetrahedron: Asymmetry 1997, 8, 1193.
14. Raimundo, B. C.; Heathcock, C. H. Synlett 1995, 1213.
15. Hoffmann, R. W.; Dresely, S. Chem. Ber. 1989, 122, 903.
16. Davis, S. G.; Dordor-Hedgecock, J. M.; Warner, P. Tetrahedron Lett. 1985, 26, 2125.
17. Tai, A.; Morimoto, N.; Yoshikawa, M.; Uehara, K.; Sugimura, T.; Kikukawa, T.
Agric. Biol. Chem. 1990, 54, 1753.
18. Tai, A.; Imaida, M. Bull. Chem. Soc. Jpn. 1978, 51, 1114.
19. Maskens, K.; Polgar, N. J. Chem. Soc., Perkin Trans. 1 1973, 109.
20. Dahlman, O.; Garegg, P. J.; Mayer, H.; Schramek, S. Acta Chem. Scand. B 1986,
40, 15.
3.32. 2-Naphthylmethyl (3S)-hydroxy-(2R)-methyl-
butanoate 56
Compound 56 was prepared from 55 as described for its enan-
tiomer 48 in 63% cumulative yield.
[a]2D4 ꢂ0.2 (c 3, CHCl3). For NMR data see 48. HRMS (electro-
21. Doboszewski, B.; Herdewijn, P. Tetrahedron 1996, 52, 1651.
22. Martin, O. R.; Nabinger, R. C.; Ali, Y.; Vyas, D. M.; Szarek, W. A. Carbohydr. Res.
1983, 121, 302.
spray) calcd for C16H18O3þHþ: 259.13286, found: 259.13281.
23. Dang, H.-S.; Roberts, B. P. J. Chem. Soc., Perkin Trans. 1 2002, 1161.
24. Ikawa, T.; Hattori, K.; Sajiki, H.; Hirota, K. Tetrahedron 2004, 60, 6901.
25. Ikawa, T.; Sajiki, H.; Hirota, K. Tetrahedron 2004, 60, 6189.
26. Xie, X.; Berges, D. A.; Robins, M. J. J. Org. Chem. 1996, 61, 5178.
27. Robins, M. J.; Doboszewski, B.; Timoshchuk, V. A.; Peterson, M. A. J. Org. Chem.
2000, 65, 2939.
28. De Armas, H. N.; Doboszewski, B.; Herdewijn, P.; Blaton, N. Acta Crystallogr.
2007, E63, 2678.
29. Tronchet, J. M. J.; Graf, R. Helv. Chim. Acta 1972, 55, 1141.
30. Dahlhoff, W. V.; Ko¨ster, R. Heterocycles 1982, 18, 421.
31. Szarek, W. A.; Zamojski, A.; Tiwari, K. N.; Ison, E. R. Tetrahedron Lett. 1986, 27,
3827.
3.33. 3-Deoxy-1,2-O-ispropylidene-3-C-methylene-5-O-
tosyl-L-threo-pentofuranose 57
Compound 53 (0.22 g, 1.2 mmol) in dry CH2Cl2 (10 ml), pyridine
(1 ml), cat. DMAP, and TsCl (0.5 g, 2.6 mmol) were left for 48 h. TLC
showed a new compound with Rf 0.46 in 3:1 hexane–EtOAc. Ex-
traction (CH2Cl2–dil HCl), drying, evaporation, and flash chroma-
tography using a gradient of EtOAc in hexane, 1:4/1:3, furnished
0.29 g, 72% of glassy 57.
32. Sen, S. E.; Roach, S. L.; Boggs, J. K.; Ewing, G. J.; Magrath, J. J. Org. Chem. 1997, 62,
6684.
1H (300 MHz): 7.81–7.78 (m, 2H), 7.35 (br s,1H, H aromatic), 7.33
(br s, 1H, H aromatic), 5.77 (d, J12¼3.7 Hz, 1H, H1), 5.50 (dd, J¼0.8,
1.9 Hz, 1H, ]CH2), 5.30 (t, J¼1.3 Hz, 1H, ]CH2), 4.81 (dt, J¼0.9,
0.9 Hz, J21¼3.7 Hz, 1H, H2), 4.68 (q of triplets, J¼1.8, 1.8 Hz,
33. Dalcanale, E.; Montanari, F. J. Org. Chem. 1986, 51, 567.
34. Krebs, K. G.; Heusser, D.; Wimmer, H. Thin-Layer Chromatography: A Laboratory
Handbook; Stahl, E., Ed.; Springer: Berlin, 1969; p 854.
35. Touchstone, J. C.; Dobbins, M. F. Practice of Thin Layer Chromatogaphy, 2nd ed;;
Wiley Interscience: New York, NY, 1983; p 163.
36. Asis, S. E.; Bruno, A. M.; Martinez, A. R.; Sevilla, M. W.; Gaozza, C. H.; Romano,
A. M.; Coussio, J. D.; Ciccia, G. H. Il Farmaco 1990, 54, 517.
37. Miller, J. B. J. Org. Chem. 1959, 25, 560.
00
0
0
0 00
J45 ¼5.4 Hz, J45 ¼7.3 Hz, 1H, H4), 4.19 (dd, J5 4¼7.4 Hz, J5 5 ¼10.1 Hz,
1H, H50), 4.10 (dd, J5 4¼5.3 Hz, J5 5 ¼10.1 Hz, 1H, H5 ), 2.44 (s, 3H,
Ph–Me), 1.34 and 1.30 (two s, 3H each, isopropylidene Me). 13C
(75 MHz): 144.8, 143.6, 132.7, 129.8, 128.0, 115.4, 113.4, 105.6, 80.9,
79.5, 71.8, 27.0, 16.2, 21.6. HRMS (electrospray) calcd for
00
00
00 0
38. Gutsche, C. D.; Jason, E. F. J. Am. Chem. Soc. 1956, 78, 1184.
39. Smets, G.; Boutemburg, A. J. Polym. Sci., Part A1 1970, 8, 3251.
40. Matthes, D. P.; Purdy, W. C. Anal. Chim. Acta 1979, 109, 61.
C
16H20O6SþNaþ: 363.08730, found: 363.08764.
´
´
41. Moravcova, J.; Capkova, J.; Stanek, J. Carbohydr. Res. 1994, 263, 61.
42. Yoshimura, Y.; Sano, T.; Matsuda, A.; Ueda, T. Chem. Pharm. Bull. 1988, 36,
162.
Acknowledgements
43. Martin, O. R.; Rao, S. P.; El-Shenawy, H. A.; Kurz, K. G.; Cutler, A. B. J. Org. Chem.
1988, 53, 3287.
44. Smith, M. B.; March, J. March’s Advanced Organic Chemistry, 6th ed.; Wiley-
Interscience: New Jersey, NJ, 2007; p 1053.
45. Keith, J. Comprehensive Organic Functional Group Transformations; Katritzky,
A. R., Meth-Cohn, O., Rees, C. W., Eds.; Pergamon: Oxford, 1995; p 71.
46. Streicher, H.; Meisch, J.; Bohner, C. Tetrahedron 2001, 57, 8851.
Mr. Bert Demarsin (Department of Chemistry, KULeuven) is
acknowledged for the electron impact high resolution mass mea-
surements, and Mrs. Stephanie Vandenwaeyenbergh and Prof. Jef
Rozenski (Rega Institute) for the electrospray HRMS. Dr. Natalya