Deoxygenation at the C3 position of d- and l-arabinofuranose: Stereospecific access to enantiomeric cordycepose derivatives

Article abstract of DOI:10.1016/j.carres.2013.07.017

Efficient synthesis of 3-deoxy-1,2-O-isopropylidene-β-d- and β-l-threo-pentofuranose (1,2-O-isopropylidene-β-d- and β-l-cordycepose) was accomplished starting from d- and l-arabinofuranose derivatives, respectively, by the action of LiBH(Et)₃ on corresponding intermediate 3-O-lyxofuranosyl trifluoromethanesulfonates.

Full text of DOI:10.1016/j.carres.2013.07.017

Carbohydrate Research 380 (2013) 143–148
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Note
Deoxygenation at the C3 position of D- and L-arabinofuranose:
stereospecific access to enantiomeric cordycepose derivatives
Fábio da Paixão Soares a, Maria Joselice e Silva b, Bogdan Doboszewski a,
⇑
a Departamento de Química, Universidade Federal Rural de Pernambuco, 52171-900 Recife, PE, Brazil
b Departamento de Farmácia, Universidade Federal do Rio Grande do Norte, 59010-180 Natal, RN, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 April 2013
Received in revised form 17 July 2013
Accepted 31 July 2013
Available online 9 August 2013
Efficient synthesis of 3-deoxy-1,2-O-isopropylidene-b-
b-D- and b-L-cordycepose) was accomplished starting from D- and L-arabinofuranose derivatives, respectively,
by the action of LiBH(Et)3 on corresponding intermediate 3-O-lyxofuranosyl trifluoromethanesulfonates.
Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved.
D
- and b-
L
-threo-pentofuranose (1,2-O-isopropylidene-
Keywords:
Arabinose
Chiral pool
Cordycepose
Deoxygenation
Stereoselective synthesis
Chiral pool synthesis is a convenient approach to obtain the
target compounds stereoselectively and with predetermined con-
figuration. Besides aminoacids1,2 and terpenes,3,4 carbohydrates5–
file of the Bignoniaceae. The initial stage of such project is to
establish a convenient and stereospecific access to the derivatives
of 3-deoxy-
respectively, which is the objective of the present communication.
Attention was directed to the known derivatives of -arabinose
D- and L-arabinose (D- and L-cordycepose) 6 and 20,
9
have been widely explored for this purpose since: (i) their abso-
lute configurations are known, (ii) many of them are available and
(iii) methodologies of their functionalizations are well established.
Of particular importance are the sugars available in both enantio-
meric forms because a synthesis developed using one enantiomer
can be repeated with another enantiomer, permitting a uniform
access to both enantiomeric targets using the same set of condi-
tions. Considering commercial access and affordable prices, three
sugars can be considered to be easy starting materials for chiral
D
311,12 and its enantiomer L-17,11,13 available in two steps ((i)
tertBuPh2SiCl, imidazole; (ii) acetone, H+; an alternative which calls
for arabinodithioethyl acetal is much less convenient14,15 due to
use of badly smelling ethanethiol and necessity to use environ-
mentally unacceptable mercury compounds). Both 3 and 17 were
previously used as the starting material to synthesize degradation
products of the antibiotic Batumin/Kalimantacin A11 active again
Gram-negative bacteria, to obtain new type of pentofuranosyl
nucleoside analogs for antiviral/antisense research,16,17 to synthe-
size potential inhibitors of biosynthesis of 3-deoxy-mannooctulo-
sonic acid (KDO),18 an essential constituents of the
lipopolysaccharides (LPS) of Gram-negative bacteria, and to syn-
thesize components of the LPS from Coxiella burnetti.13 Our ap-
proaches to obtain the target D-6 are shown in Scheme 1.
pool stereoselective synthesis:
and ascorbic and isoascorbic acid. Reported here is application of
both arabinoses to obtain 3-deoxy-1,2-O-isopropylidene-b-
threo-pentofuranose (1,2-O-isopropylidene-b- -cordycepose)
and its -enantiomer 20 as general substrates for further synthetic
work to obtain chiral b-hydroxyaldehydes A (Fig. 1 shows the
application of -arabinose for this purpose), which can be elabo-
D- and L-arabinose, D- and L-xylose,
D
-
6
D
L
D
rated to obtain either 1,3-diols B or the alcohols C, among others.
The compounds having the structure B and C are constituents of
the waxes isolated from plants belonging to the Bignoniaceae
family, and in many cases their absolute configurations are
unknown.10 Consequently, stereospecific synthesis of the targets
B and C will contribute to the knowledge of the phytochemical pro-
Apparently simple Barton deoxygenation19,20 at the position C3
in 3 via phenoxythiocarbonate 4 using Bu3SnH/AIBN (newer
versions of this reaction include (Me3Si)3SiH/AIBN,18 Bu3SnH/
B(Et)3/air,21,22 B(Me)3/H2O,23 and (Bu4N)2S2O8/HCO2Na24
) fur-
nished the product 5, but the process was unexpectedly low-yield-
ing (38%) contrary to similar cases18,25 and was accompanied by
free-radical de-esterification to form the substrate 3 recovered in
31% yield (Scheme 2). Reversion to the starting alcohol during
free-radical deoxygenation is a known phenomenon and has been
⇑
Corresponding author. Tel.: +55 81 33206382; fax: +55 81 33206374.
E-mail address: bdoboszewski@hotmail.com (B. Doboszewski).
0008-6215/$ - see front matter Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.carres.2013.07.017

144
F. da Paixão Soares et al. / Carbohydrate Research 380 (2013) 143–148
Figure 1. Selected targets A, B, and C available from the general substrate 6.
Scheme 2. Free-radical de-esterification of phenoxycarbonate
deoxygenation.
4 during Barton
reported before.26 Due to low yield in this approach we used a
modified Wolff–Kishner reaction via tosylhydrazone 10
(Scheme 1). Such type of deoxygenation was successfully applied
to get 3-deoxy-
-xylose27 but low yield (20%) was reported in an-
other case.28 Likewise, problems were encountered during a syn-
thesis of
tropane alkaloid ferrugine29 which shows some
a
The next subsequent approach was a Barton type deoxygen-
ation using a cyclic 3,5-lyxo thiocarbonate 14 which should fur-
nished 6 via the more stable secondary radical. A precursor of 14
L
a
potential in treatment of Alzheimer’s disease, and DIBAL30 fur-
nished the highest yield out from several other reagents.29 Com-
pound 10 was obtained by the reaction of the tosylhydrazide 931
with the ulose 8 obtained by CrO3/Py/Ac2O13 oxidation of 3. Both
geometrical isomers were formed, which have different Rf values.
The less polar tosylhydrazone 10 of unknown configuration was
selectively crystallized from the mixture (31%). The mother liquor
consisted of approximately 1:1 mixture of E/Z isomers (51%). Both
fractions were separately reacted with excess of DIBAL to furnish 5
via 11 again in low yield (ꢀ23%) in both cases.
was the tert-butyldiphenylsilyl-1,2-O-isopropylidene-b-D-lyxof-
uranose 12, obtained with a total selectivity by NaBH4 reduction
of the above-mentioned ulose 8. This selectivity can be a conse-
quence of either conformational property of 8 and/or the presence
of two bulky substituents flanking the carbonyl group. For this
reason the hydride ion attacked only the more exposed re side of
the C3 atom of 8. Desilylation to get 13 and treatment with
thiocarbonyldiimidazole followed by Bu3SnH/AIBN indeed fur-
nished the expected 3-deoxy compound 6, but again in low yield
(22%). Some reversion to the diol 13 took place (31%). Similar
Scheme 1. Approaches to obtain 3-deoxy-1,2-O-isopropylidene-b-D-threo-pentofuranose 6 from D-arabinose.

F. da Paixão Soares et al. / Carbohydrate Research 380 (2013) 143–148
145
difficulties to use cyclic thiocarbonates were encountered during
free-radical deoxygenations of C-arabinofuranosyl derivatives,32,33
even though the procedure functioned satisfactorily in the case of
xylofuranosyl substrate.24 Possible side reactions during these pro-
cesses may include free-radical rearrangement of the cyclic thio-
carbonate to thiolcarbonate34 or a reduction to form methylene
acetal.35
Scheme 3. Synthesis of 3-deoxy-1,2-O-isopropylidene-b-
and its 5-O-tosylate 21 from -arabinose.
L-threo-pentofuranose 20
L
These difficulties prompted us to change an approach and to use
1.2. 5-O-tert-Butyldiphenylsilyl-a,b-D-arabinofuranose (2) and
nucleophilic substitution using
a hydride ion coming from
LiBH(Et)3 (SuperhydrideÒ), and a trifluoromethanesulfonate 15 de-
rived from the lyxo compound 12. This ‘substitutive deoxygen-
ation’11 functioned very well at room temperature and furnished
3-deoxy product 5 in 77% yield for two steps (from 12 via 15). Still
its
L
-enantiomer
A slurry of D-arabinose 1 (10 g, 67 mmol) in DMF (150 mL, dried
by azeotropic removal of water using benzene) was warmed using
an oil bath at ca. 90–100 °C with magnetic stirring until a homog-
enous solution was obtained, whereupon oil bath was removed.
When the temperature of the solution dropped to 55 °C, imidazole
(9.1 g, 134 mmol) was added followed by dropwise addition (syr-
inge) of tert-butylchlorodiphenylsilane (18.3 g, 17.3 mL, 67 mmol)
during 10 min. All these operations were performed under a blan-
ket of nitrogen. After 3 h at ca. 55 °C, TLC showed the main com-
pound having Rf 0.40 (hexane–EtOAc 2:3 or CH2Cl2–MeOH
20:0.7) together with impurities which moved with the solvent
front. Extraction (CH2Cl2–dil HCl), washing of the organic phase
with aq NaHCO3, drying (MgSO4), filtration, evaporation of the vol-
atiles, and co-evaporation with xylenes, furnished yellowish oil.
This material was chromatographically purified (gradient of CH2-
higher yield was obtained for the L-compound 19 (82%). It is visible
that a combination of very soft hydride ion together with the best
known leaving group constitutes an effective procedure to get
deoxygenated derivatives 5 and 19 via nucleophilic substitution.
It should be pointed out however that the arabino triflate 16 did
not react under these conditions.
Conventional desilylation of 5 using Bu4NF or NH4F36 to get 6
followed by tosylation furnished highly crystalline compound 7,
whose X-ray structure37 shows a pentofuranosyl ring in two dis-
tinctive conformations 4E and 4T3, even though its 1H NMR spec-
trum shows only one set of the averaged signals, evidently
resulting from rapid equilibration in the NMR time-scale. It is gen-
erally argued to what extent the solid state conformation is related
to the solution conformation38 and substantial difference between
both conformations was published.39
Cl2–MeOH 20:0.5 ? 20:1 or
a
gradient of hexane–EtOAc
3:2 ? 2:3) to furnish the product 2 (16.5 g, 64%); ½a D26
ꢁ
+22.0 (c
2.4, CHCl3); lit.13 for the
portion of the
NMR signals (CDCl3):
exchange with D2O. 13C NMR (CDCl3):
L
-enantiomer: ½a D26
ꢂ22 (c 1, CHCl3). Pro-
ꢁ
The above-mentioned procedure of ‘substitutive deoxygenation’
a
/b anomers was ca. 5:2 by integration of the 1H
was applied to the
L
-arabinofuranose 1711–13 via
L
-lyxofuranose 18
a: d 5.41, s; b: d 5.29 d, J = 3.8 Hz, both after
and furnished 3-deoxy-1,2-O-isopropylidene -b -
L
-threo-pentofura-
a
: 103.54; b: 96.95.
nose 20 and its 5-O-tosylate 21 in comparative yields (Scheme 3).
Application of the alcohols 6 and 20, together with their corre-
sponding tosylates 7 and 21, respectively, in the stereoselective
synthesis will be reported in due course.
The enantiomeric 5-O-tert-BuPh2Si-a,b-L-arabinofuranose was
obtained using the same procedure and in similar yield; ½a D26
ꢁ
ꢂ22.8 (c 2.2, CHCl3); lit.13
½
a 2D6
ꢁ
ꢂ22 (c 1, CHCl3).
In conclusion, efficient and stereospecific synthesis of 3-deoxy-
1.3. 5-O-tert-Butyldiphenylsilyl-1,2-isopropylidene-b-D-
arabinofuranose (3)
1,2-O-isopropylidene-b-
was devised starting from
corresponding - and -lyxofuranosyl 3-O-trifluormethanesulfo-
D
- and
L
-threo-pentofuranoses 6 and 20
D
- and
L-arabinose, respectively, via their
D
L
Compound 2 (15.4 g, 39.7 mmol) in acetone (200 mL, dried by
shaking of techn. grade acetone with ca. 5% wt./vol. of P2O5 for
15 min, filtration, and distillation), concd H2SO4 (0.8 mL), and pul-
verized CuSO4 (15 g, dried at 140–150 °C for 6 h) were magneti-
cally stirred under a blanket of nitrogen during 8 h. TLC (hexane–
EtOAc 3:1) showed the product 3 (Rf 0.32). The solids were filtered
off using a sintered glass funnel and washed with acetone. Concd
NH4OH (3 mL) was added, and precipitated (NH4)2SO4 was re-
moved by filtration. The volatiles were evaporated and the residue
was passed through a silica gel column (gradient of hexane–EtOAc
4:1 ? 3:1) to furnish 3 as colorless viscous oil (12.5 g, 74%). The
nates. This procedure complements the existing methods of synthe-
sis of differently substituted 3-deoxyarabinoses (cordyceposes)
which were obtained by deoxygenation using 3-O-imidazoylthio-
carbonate/(Me3Si)3SiH/AIBN18 or 3-O-methylxanthate/Bu3SnH/
AIBN,25 starting from 1,2:5,6-di-O-isopropylidene-
a-D-glucofura-
nose,40–42 3-deoxy- -mannose,43 -xylose,44 from the lyxofurano-
D D
syl epoxides,45–47 or finally obtained via a total synthesis using
bromoacetic acid48 or (S)-1,2-O-isopropylidene glycerol.49
1. Experimental
yield was slightly higher than published before;11,13
2.3 CHCl3); lit.11
+5.6 (c 3.2, CHCl3).
½
a 2D6
+5.3 (c
ꢁ
1.1. General methods
½ ꢁ
a 2D6
Column chromatography was performed using silica gel G
70–230 mesh, and TLC analysis on aluminum plates precoated
with 60 F254, both from Merck. NMR spectra were recorded on
the Varian or Bruker spectrometers operating at 300 and 75 MHz
for 1H and 13C NMR, respectively. Electrospray exact mass mea-
surements of the compounds judged to be ca. 99% pure by the 1H
NMR were performed using Shimadzu or Bruker instruments in po-
sitive mode. Optical rotations were measured on Krüss automatic
polarimeter using 1 dm tube at ca. 26 °C. Nitrogen dried by passage
through ‘blue silica gel’ provided inert atmosphere for reactions
involving moisture-sensitive compounds. Evaporations were per-
formed below 40 °C. TLC chromatograms were revealed using
10% H2SO4 in MeOH, 2% CrO3 in 10% aq H2SO4 or using ‘Hanessian
stain’, followed by heating at ca. 110 °C.
1.4. 5-O-tert-Butyldiphenylsilyl-1,2-isopropylidene-b-L-
arabinofuranose (17)
The enantiomer 17 was prepared as described for 3 in similar
yield. ½aLꢁ2D6 ꢂ5.2 (c 2.1, CHCl3); lit.13
½
a 2D6
ꢁ
ꢂ5.0 (c 1.2, CHCl3). The
NMR data of both 3 and 17 match the previously published
ones.11,13
1.5. 5-O-tert-Butyldiphenylsilyl-3-deoxy-1,2-O-isopropylidene-
b-D-threo-pentofuranose (5)
To a solution of
D-arabino compound 3 (2.0 g, 4.7 mmol) in dry
CH2Cl2 (10 mL) and pyridine (0.8 mL), was injected chlorophenoxy-

146
F. da Paixão Soares et al. / Carbohydrate Research 380 (2013) 143–148
thioformate (0.8 mL, 1.1 g, 6 mmol). After overnight reaction the
substrate 3 was converted to the less polar 4 (Rf 0.44, hexane–
EtOAc 15:1). Extractive work-up (partition between CH2Cl2 and
dil. HCl followed by washing of the organic phase with aq NaHCO3,
drying, filtration, and evaporation of the solvent) furnished yellow
oil which was purified by a short bed chromatography using
hexane–EtOAc 15:1. The product 4 (without characterization)
was solubilized in dry degassed toluene (50 mL), warmed to ca.
100 °C, and treated dropwise with a solution of Bu3SnH (1.9 g,
1.8 mL, 6.3 mmol) and AIBN (0.10 g, 0.6 mmol) in toluene
(10 mL). After overnight warming, addition of the same amount
of Bu3SnH and AIBN was repeated. After a total of 48 h at ca.
100 °C, TLC (hexane–EtOAc 6:1) showed small amount of unre-
acted substrate 4 (Rf 0.64), the product 5 (Rf 0.53), unidentified
UV absorbing compound (Rf 0.44), and the alcohol 3 (Rf 0.17). Tol-
uene was evaporated. Chromatography in hexane–EtOAc 10:1 fur-
nished 0.73 g (38%) of the deoxygenated compound 5 as oil. Elution
with hexane–EtOAc (5:1) permitted recovery of the alcohol 3
4.69 (t, J = 4.6 Hz, 1H), 4.36 (ddd, J = 2.0 Hz, J = 6.5 Hz, J = 6.5 Hz,
J = 8.4 Hz, 1H), 4.19 (dd, J = 6.9 Hz, J = 9.7 Hz, 1H), 4.11 (dd,
J = 6.6 Hz, J = 9.7 Hz, 1H), 2.44 (s, 3H), 2.17 (ddd, J = 5.7 Hz,
J = 8.5 Hz, J = 14.4 Hz, 1H), 2.04 (dd, J = 1.6 Hz, J = 14.5 Hz, 1H),
1.33 and 1.25 (two s, 3H each). 13C NMR (CDCl3): 145.0, 133.0,
130.0, 128.2, 112.3, 107.0, 80.3, 77.9, 71.3, 33.7, 26.7, 25.7, 21.7. Ex-
act mass cacld for C15H20O6S [M+Na]+: 351.0873, found: 351.0873.
1.8. 5-O-tert-Butyldiphenylsilyl-1,2-O-isopropylidene-b-D-threo-
pentofuranosyl-3-ulose (8)
To a suspension of CrO3 (10 g, 100 mmol) in dry CH2Cl2
(150 mL) was added dry pyridine, (16.2 mL, 15.9 g, 201 mmol) in
one portion. The mixture turned dark brown and slight warming
took place. This mixture was magnetically stirred for 25 min. A
solution of 3 (10.8 g, 25.2 mmol) in CH2Cl2 (90 mL) was added,
immediately followed by acetic anhydride (9.8 mL, 10.6 g,
104 mmol) using a syringe. The dark brown mixture was stirred
for 15 min. TLC (hexane–EtOAc 4:1) showed a conversion of 3 (Rf
0.24) to the ulose 8 (Rf 0.45). A mixture of EtOAc–toluene 1:1
(200 mL) was added, which resulted in precipitation of black solid.
The supernatant was decanted and solid material was washed
twice with the same mixture of solvents. Combined solutions were
applied on top of a short bed of silica gel column prepared in
EtOAc–toluene 2:1. Elution with EtOAc–toluene 2:1 followed by
evaporation and repeated co-evaporation with xylenes furnished
8 as yellowish oil, which was used immediately without character-
ization to get either a tosylhydrazone 10 or a lyxose 12.
(0.63 g, 31%). ½a D26
ꢁ
+11.5 (c 7.8, CHCl3). 1H NMR (CDCl3): 7.68–
7.66 and 7.43–7.33 (10H, H aromatic), 5.78 (d, J1,2 = 4.0 Hz, 1H,
0
00
H-1), 4.71 (ddd, J2;3 = 1.1 Hz, J2,1 = 4.0 Hz, J2,3 = 7.1 Hz, 1H, H-2),
0
0
00
4.28 (m of 12 lines, J4,3 = 2.9 Hz, J4,5 = 5.8 Hz, J4,3 = 7.8 Hz,
00
0
0
00
J4,5 = 7.8 Hz, 1H, H-4), 3.84 (dd, J5 ,4 = 5.9 Hz, J5 ,5 = 9.8 Hz, 1H, H-
50), 3.78 (dd, J5 ,4 = 7.7 Hz, J5 ,5 = 9.8 Hz, 1H, H-500), 2.26 (ddd,
00
00
0
J3 ,2 = 1.1 Hz, J3 ,4 = 2.6 Hz, J3 ,3 = 14.1 Hz, 1H, H-30), 2.13 (ddd,
0
0
0
00
J3 ,2 = 6.0 Hz, J3 ,4 = 8.0 Hz, J3 ,3 = 14.2 Hz, 1H, H-300), 1.32 and 1.27
(two s, 3H each), 1.06 (s, 9H). 13C NMR (CDCl3): 135.6, 135.6,
133.5, 133.5, 129.6, 127.7, 112.1, 106.7, 81.36, 80.7, 65.9, 33.6,
27.0, 26.9, 25.9, 19.2. Exact mass calcd for C24H32O4Si [M+Na]+:
435.1962, found: 435.1973.
00
00
00
0
1.9. p-Toluenesulfonylhydrazide (9)
1.6. 3-Deoxy-1,2-O-isopropylidene-b-
D
-threo-pentofuranose (6)
A solution of p-toluenesulfonyl chloride (4.7 g, 25 mmol) in
benzene or toluene (100 mL) and 50% aq hydrazine (5 g, 50 mmol)
was magnetically stirred until solidification (ca. 10 min), and the
mixture was left for 3 h. The colorless solid material was filtered,
washed with ice-cold water, and crystallized from hot water to
yield 9 (3.1 g, 67%); mp 109–111 °C; lit.31 mp 112 °C. 1H NMR
(DMSO-d6): 8.31 (s, 1H, exchangeable), 7.70 (d, J = 8.1 Hz, 2H),
7.40 (d, J = 8.0 Hz, 2H), 4.06 (bs, 2H, exchangeable), 2.39 (s, 3H).
13C NMR (DMOS-d6): 142.9. 135.2, 129.5, 127.7, 21.0.
Compound 5 (0.83 g, 2 mmol) in THF (15 mL) was desilylated
overnight using 1 M Bu4NF solution (4 mL). TLC (hexane–EtOAc
5:7) showed the product 6 (Rf 0.32) together with less and more
polar impurities. Evaporation of the solvent and chromatography
in hexane–EtOAc 2:3 furnished 6 (0.14 g, 40%). ½a D26
ꢁ
+32.0 (c 3.5,
CHCl3); lit. for the
L
-enantiomer: ½a D26
ꢁ
ꢂ32.6 (c 1, CHCl3),50
½ ꢁ
a 2D6
ꢂ30.86 (c 1 CHCl3).51 1H NMR (CDCl3): 5.82 (d, J1,2 = 3.9 Hz, 1H,
00
0
H-1), 4.76 (ddd, J2,3 = 1.2 Hz, J2,1 = 4.0 Hz, J2,3 = 6.2 Hz, 1H, H-2),
0
00
0
4.33 (dddd, J4,3 = 3.1 Hz, J4,5 = 4.2 Hz, J4,5 = 7.6 Hz, J4,3 = 8.6 Hz,
1.10. 5-O-tert-Butyldiphenylsilyl-3-deoxy-1,2-O-isopropylidene
0
0
0
00
1H, H-4), 3.83 (ddd, J5 ,OH = 2.8 Hz, J5 ,4 = 8.0 Hz, J5 ,5 = 11.0 Hz, 1H,
-b-
(10)
D
-threo-pentofuranosyl-3-(N-p-toluenesulfonyl) hydrazone
H-50), 3.61 (ddd, J5 ,4 = 4.2 Hz, J5 ,OH = 7.6 Hz, J5 ,5 = 11.6 Hz, 1H,
00
00
00
0
00
0
00
H-5 ), 2.34 (dd, exchangeable, JOH,5 = 3.8 Hz, JOH,5 = 11.6 Hz, 1H,
OH), 2.21 (ddd, J3,2 = 6.3 Hz, J3 ,4 = 8.6 Hz, J3 ,3 = 14.4 Hz, 1H, H-30),
1.99 (dddd, J = ca 0.5 Hz, unidentified long range coupling,
The ulose 8 was prepared and isolated as described above under
0
0
00
1.8 using 3 (3.7 g), CrO3 (3.4 g), pyridine (5.5 mL), and Ac2O
(3.4 mL) and proportional volumes of CH2Cl2. To this compound
was added a solution of p-toluenesulfonylhydrazide 9 (1.3 g,
7 mmol) in toluene (45 mL) and techn EtOH (8 mL). The resulting
homogenous solution was left overnight. TLC (hexane–EtOAc 3:1)
showed that all 8 (Rf 0.47) reacted to form two overlapping spots
Rf 0.42 and 0.39 of both geometrical isomers of 10; tosylhydrazide
did not migrate in this solvent system. Ethyl acetate was added and
the resulting solution was washed with brine, dried, filtered, and
the volatiles were evaporated. The residue spontaneously solidi-
fied. Ethyl ether was added and colorless crystalline material was
filtered and dried to yield 1.57 g, 30.6% of the less polar isomer
of unknown configuration. The mother liquor which consisted of
00
00
00
0
00
J3 ,2 = 1.1 Hz, J3 ,4 = 3.0 Hz, J3 ,3 = 14.2 Hz, 1H, H-3 ), 1.56 and
1.32(two s, 3H each). 13C NMR (CDCl3): 112.4, 106.4, 81.7, 80.7,
65.1, 33.2, 27.0, 25.9. Exact mass calcd for C8H14O4 [M+Na]+:
197.0784, found: 197.0780.
1.7. 3-Deoxy-1,2-O-isopropylidene-5-O-p-toluenesulfonyl-b-
threo-pentofuranose (7)
D-
Compound 6 (0.17 g, 1 mmol) in CH2Cl2 (20 mL) and Et3N
(1 mL) was reacted with p-toluenesulfonyl chloride (0.28 g,
1.5 mmol) overnight. TLC showed a new compound having Rf
0.36 (hexane–EtOAc 2:1). Extraction (CH2Cl2-dil. NaHCO3), wash-
ing the organic phase with water, drying, evaporation of the sol-
vent and chromatography in hexane–EtOAc (3:2) furnished 7
(0.33 g, quantitative yield) which spontaneously crystallized. Mp
a
mixture of both compounds in roughly equal proportion
weighted 2.65 g, 51.5%. Both yields are counted on 3. The less polar
isomer: mp 167–169 °C (dec.); ½a D26
ꢁ
+25.0 (c 2.5, DMSO). 1H NMR
73–75 °C (Et2O–hexane), lit.42 data for the
L
-enantiomer: mp
(CDCl3 and 1 drop of DMSO-d6): 9.31 (s, 1H, exchangeable), 7.53–
7.09 (H aromatic, 14H), 5.76 (d, J = 3.9 Hz, 1H), 4.89–4.84 (unre-
solved, 2H), 3.94 (dd, J = 5.0 Hz, J = 9.3 Hz, 1H), 3.84 (dd,
J = 9.5 Hz, J = 10.4 Hz, 1H), 2.34 (s, 3H), 1.25 and 1.12 (two s, 3H
76–77 °C (EtOH); ½a D26
ꢁ
+42.8 (c 1.2, CHCl3); lit.42 data for the
L-
enantiomer: ½a 2D6
ꢁ
ꢂ48.8 (c 0.7, CHCl3). 1H NMR (CDCl3): 7.80 (d,
J = 8.3 Hz, 2H), 7.34 (d, J = 8.3 Hz, 2H), 5.76 (d, J = 3.7 Hz, 1H),

F. da Paixão Soares et al. / Carbohydrate Research 380 (2013) 143–148
147
each), 1.07 (s, 9H). 13C NMR (CDCl3 and 1 drop of DMSO-d6): 156.3,
143.9, 135.4, 135.3, 132.2, 131.2, 130.2, 130.2, 129.5, 129.4, 128.5,
128.1, 128.1, 128.0, 113.7, 130.8, 79.5, 76.6, 27.1, 26.6, 26.0, 21.6,
19.2. Exact mass calcd for C31H38N2O6SSi [M+H]+: 595.2292, found:
595.2304.
stirred at 29 °C. After 1 h the deprotection was completed and the
product showed Rf 0.47 (CH2Cl2–MeOH 4:1). The solvent was evap-
orated and the residual slurry was applied on top of the silica gel
column. Elution with CH2Cl2–MeOH (4:1) furnished 13 (0.22 g,
84%); mp 96-97 °C (MeOH); ½a D26
ꢁ
+21.4 (c 1, MeOH); ½a D26
ꢁ
+19.2 (c
1, DMSO); lit.52 for the
-enantiomer: mp 109–111 °C (EtOAc);
L
1.11. Synthesis of compound (5) via Wolff–Kishner reaction
using hydrazone (10)
½
a 2D6
ꢁ
ꢂ20.5 (c 1.55, DMSO). 1H NMR (DMSO-d6; the signals were
identified using COSY spectrum): 5.59 (d, J = 3.0 Hz, 1H, H-1)),
4.97 (d, J = 4.2 Hz, 1H, exchangeable, 3-OH), 4.40 (apparent t,
J = 3.4 Hz, 1H, H-2), 4.27 (t, J = 4.2 Hz, 1H, exchangeable, 5-OH),
4.15 (apparent q, J = 4.5 Hz, 1H [after D2O exchange: dd,
J = 4.2 Hz, J = 5.4 Hz], H-3), 3.86 (dt, J = 3.3 Hz, J = 5.5 Hz,
J = 5.5 Hz, 1H, H-4), 3.63 (dd, J = 3.6 Hz, J = 9.0 Hz, 1H, H-50), 3.58
The crystalline, less polar compound 10, 0.597 g (1 mmol) was
solubilized in dry CH2Cl2 (25 mL) under a blanket of nitrogen and
the resulting solution was chilled in ice-bath. 1 M DIBAL in hexane
(3 mL, 3 mmol) was injected while maintaining magnetic stirring.
The solution immediately became yellow. Cooling bath was re-
moved. TLC showed that the substrate was no longer present after
30 min. Rf of the product 5 was 0.70 (hexane–EtOAc 3:1). Several
more and less polar compounds were also present. MeOH (1 mL)
was added and extraction was performed (CH2Cl2–brine). The gel
of aluminum compounds appeared. The whole mixture was passed
through sintered glass. The organic phase was separated, dried, fil-
tered, and evaporated. Chromatography in hexane–EtOAc 20:3 fur-
nished 5 (0.095 g, 23%). The same yield was obtained when the
above mentioned mixture of both geometrical isomers 10 was
used. Thus, using 2.65 g (4.5 mmol) of 10 and 13.5 mL (13.5 mmol)
of 1 M DIBAL in hexane, 0.42 g (23%) of 5 was obtained.
(dd, J = 4.5 Hz, J = 9.0 Hz, 1H, H-5 ). 13C NMR (DMSO-d6, identifica-
00
tion established using HSQC spectrum): 111.98, CMe2; 103.96, C-1;
82.7, C-4; 79.4, C-2; 70.2, C-3; 60.8, C-5; 26.5 and 26.0, CMe2. Exact
mass calcd for C8H14O5 [M+Na]+: 213.0743, found: 213.0731.
1.15. Synthesis of 3-deoxy-1,2-O-isopropylidene-b-D-threo-
pentofuranose (6) using cyclic thiocarbonate (14)
To the solution of the D-lyxo diol 13 (0.37 g, 1.9 mmol) in dry
DMF (15 mL) was added thiocarbonyldiimidazole (0.54 g,
2.4 mmol, 90% pure) and the resulting bright yellow solution was
kept for 18 h at room temperature. TLC showed the absence of
the substrate 13 (Rf 0.48 in CH2Cl2–MeOH 4:1) and formation of
less polar compound 14 migrating near solvent front. Extractive
work-up using CH2Cl2–brine–dil. HCl, washing the organic phase
with aq NaHCO3, drying (MgSO4), filtration, and evaporation of
the solvent furnished yellow syrup, which was dried on oil pump
and used for the next step without further purification nor charac-
terization. Toluene (40 mL) was added and the resulting solution
was purged with N2 for 10 min., and warmed to ca. 100 °C (oil
bath). Bu3SnH (1.9 mL, 2.06 g, 7.1 mmol) and AIBN (0.11 g,
0.7 mmol) in toluene (20 mL) were added dropwise during
15 min. while maintaining magnetic stirring. After 8 h TLC showed
a spot of the deoxygenated product 6, starting diol 13 and uniden-
tified byproducts. Toluene was evaporated and the residue was
purified by chromatography to furnish 6 (0.074 g, 22%; elution
with hexane–EtOAc 10:15) and recovered diol 13 (0.11 g, 31%; elu-
tion with CH2Cl2–MeOH 10:1). Both yields were calculated for two
steps.
1.12. 5-O-tert-Butyldiphenylsilyl-1,2-O-isopropylidene-b-D-
lyxofuranose (12)
To an ice cold solution of the ulose 8 prepared from 3 (10.8 g,
25.2 mmol) as described under 1.8, in techn EtOH (80 mL) was
added NaBH4 (2.6 g, 69 mmol) in small portion over 10 min while
maintaining magnetic stirring. Cooling bath was removed after
completion of the addition and the reduction was continued for a
total of 1 h. TLC showed the following pattern of mobilities (hex-
ane–EtOAc 4:1): arabinose 3: Rf 0.24, ulose 8: Rf 0.45, lyxose 12:
Rf 0.34. The whole reaction mixture was transferred to a separatory
funnel charged with CH2Cl2, brine, and NH4Cl (to destroy any ex-
cess of NaBH4), and extraction was performed. Organic phase
was dried, filtered, and the solvent was evaporated. The product
was purified by chromatography in hexane–EtOAc 16:5 to furnish
12 (8.6 g, 80%) for two steps (from 3 via 8), as a syrup which solid-
ified after ca. 1 year in a refrigerator. Mp 48 °C; ½a D26
ꢁ
+7.5 (c 3.2,
CHCl3). 1H NMR (CDCl3): 7.71–7.67 and 7.44–7.34 (H aromatic,
10H), 5.71 (d, J = 4.1 Hz, 1H), 4.60 (dd, J = 4.1 Hz, J = 5.8 Hz, 1H),
4.33 (apparent quartet, J = 5.9 Hz; after D2O exchange: t,
J = 5.8 Hz, 1H), 4.19–4.10 (unresolved, 2H), 3.92–3.83 (m, 1H),
1.42 and 1.35 (two s, 3H each), 1.06 (s, 9H). 13C NMR (CDCl3):
135.7, 135.6, 133.2, 133.0, 129.8, 127.7, 114.1, 104.9, 81.3, 79.8,
70.9, 63.2, 26.8, 26.8, 26.5, 19.2. Exact mass calcd for C24H32O5Si
[M+Na]+: 451.1911, found: 451.1911.
1.16. Synthesis of 5-O-tert-butyldiphenylsilyl-3-deoxy-1,2-O-
isopropylidene-b-
D-threo-pentofuranose (5) via D-lyxofuranosyl-
3-O-trifluoromethanesulfonate (15)
To the magnetically stirred cold solution (ice-water bath) of 12
(5.4 g, 12.6 mmol) in CH2Cl2 (40 mL) and pyridine (5 mL, 4.9 g,
62 mmol) was injected trifluoromethanesulfonyl anhydride
(2.7 mL, 4.5 g, 16 mmol) during 5 min. under nitrogen atmo-
sphere. After 50 min TLC showed that all 12 reacted to form a less
polar 15 (Rf 0.57 in hexane–EtOAc 5:1). Extraction (CH2Cl2-ice
cold dil. HCl), washing with ice cold water, drying, filtration,
evaporation, and brief drying on oil pump, furnished yellow-red
syrup. Freshly dist. THF (60 mL) was added, and 1 M LiBH(Et)3
in hexane (16 mL, 16 mmol) was injected while maintaining mag-
netic stirring under nitrogen atmosphere. Slight warming took
place. Next day TLC showed that the triflate 15 and the product
5 have the same Rf, but both spots displayed different colors dur-
ing revelation with either 2% CrO3 in aq 10% H2SO4 or 10% H2SO4
in MeOH. Methanol (1 mL) was added and the volatiles were
evaporated. The residue was partitioned between CH2Cl2 and
water, and organic phase was evaporated. The residue was puri-
1.13. 5-O-tert-Butyldiphenylsilyl-1,2-O-isopropylidene-b-L-
lyxofuranose (18)
The same procedure as described above (1.12) was repeated
using -arabinose 17 to furnish 5-O-tert-butyldiphenylsilyl-1,2-O-
L
isopropylidene-b-L-lyxofuranose 18 in approximately the same
yield. Mp 48 °C; ½a D26
ꢂ7.8 (c 3.5, CHCl3). Exact mass found:
ꢁ
451.1899. For the NMR data: see 12.
1.14. 1,2-O-Isopropylidene-b-D-lyxofuranose (13)
To a solution of 5-O-tert-butyldiphenylsilyl-1,2-O-isopropyli-
dene-b- -lyxofuranose 12 (0.6 g, 1.4 mmol) in MeOH (15 mL) was
added NH4F36 (0.3 g, 8.1 mmol) and the mixture was magnetically
D
fied by chromatography (hexane–EtOAc 10:1) to furnish
(4.01 g, 77%) for two steps.
5

148
F. da Paixão Soares et al. / Carbohydrate Research 380 (2013) 143–148
10. Lima, C. S. D. A. Ph.D. Thesis, Universidade Federal da Paraiba, João Pessoa, PB,
Brazil, 2003.
11. Doboszewski, B.; Herdewijn, P. Tetrahedron 2008, 64, 5551–5562.
12. Martin, O. R.; Rao, S. P.; El-Shenawy, H. A.; Kurz, K. G.; Cutler, A. B. J. Org. Chem.
1988, 53, 3287–3292.
13. Dahlman, O.; Garegg, P. J.; Mayer, H.; Schramek, S. Acta Chem. Scand. B 1986, 40,
15–20.
14. Vázquez-Tato, M. P.; Seijas, J. A.; Fleet, G. W. J.; Mathews, C. J.; Hemnings, P. R.;
Brown, D. Tetrahedron 1995, 51, 959–974.
15. Chevallier, O. P.; Migaud, M. E. Beilstein J. Org. Chem. 2006, 2.
16. Herdewijn, P.; De Winter, H.; Doboszewski, B.; Verheggen, I.; Augustyns, K.;
Hendrix, C.; Saison-Behmoaras, T.; De Ranter, C.; Van Aerschot, A. ACS Symp.
Ser. 1994, 580, 80–99.
1.17. 5-O-tert-Butyldiphenylsilyl-3-deoxy-1,2-O-
isopropylidene-b- -threo-pentofuranose (19)
L
The procedure presented under 1.16 was used to obtain 5-O-tert-
butyldiphenylsilyl-3-deoxy-1,2-O-isopropylidene-b- -threo-pen-
L
tofuranose 19 starting from
L
-lyxose 18 in 82% yield.
Compound 19: ½a 2D6
ꢁ
ꢂ11.0 (c 6, CHCl3). Exact mass calcd for
C
24H32O4Si [M+Na]+: 435.1962, found: 435.1970. For the NMR
data: see enantiomeric compound 5.
17. Doboszewski, B.; Herdewijn, P. Tetrahedron 1996, 52, 1651–1668.
18. Airoldi, C.; Merlo, S.; Nicotra, F. J. Carbohydr. Chem. 2010, 29, 30–38.
19. Studer, A.; Amrein, S. Synthesis 2002, 835–849.
20. Barton, D. H. R.; McCombie, S. W. J. Chem. Soc., Perkin Trans, I 1975, 1574–1585.
21. Barton, D. H. R.; Jang, D. O. Tetrahedron Lett. 1990, 31, 3991–3994.
22. Nozaki, K.; Oshima, K.; Utimoto, K. Tetrahedron Lett. 1988, 29, 6125–6126.
23. Spiegel, D. A.; Wiberg, K. B.; Schachere, M. R.; Medeiros, M. R.; Wood, J. L. J. Am.
Chem. Soc. 2005, 127, 12513–12515.
1.18. 3-Deoxy-1,2-isopropylidene-b-
L
-threo-pentofuranose (20)
Desilylation of 19 using Bu4NF as described under 1.6 for the
D
-enantiomer 6, furnished 20 in 77% yields. ½a D26
ꢂ33.0 (c 3.4,
ꢁ
CHCl3); lit. ½a 2D6
ꢁ
ꢂ32.6 (c 1, CHCl3),50
½
a 2D6
ꢁ
ꢂ30.86 (c 1 CHCl3).51 Ex-
act mass calcd for C8H14O4 [M+Na]+: 197.0784, found: 197.0779.
24. Park, H. S.; Lee, H. Y.; Kim, Y. H. Org. Lett. 2005, 7, 3187–3190.
25. Gurjar, M. K.; Bera, S. Org. Lett. 2002, 4, 3569–3570.
26. Acton, E. M.; Goerner, R. N.; Ryan, K. J.; Henry, D. W. J. Med. Chem. 1979, 22,
518–525.
For the NMR data: see enantiomeric compound 6.
1.19. 1,2-O-isopropylidene-5-O-p-toluenesulfonyl-b-L-threo-
27. Xiang, Y.; Cavalcanti, S.; Chu, C. K.; Schinazi, R. F.; Pai, S. B.; Shu, Y.-L.; Scheng,
Y.-C. Bioorg. Med. Chem. Lett. 1995, 5, 877–880.
pentofuranose (21)
28. Nair, V.; Sinhababu, A. K. J. Org. Chem. 1978, 43, 5013–5017.
29. Sienkiewicz, M.; Wilkaniec, U.; Lazny, R. Tetrahedron Lett. 2009, 50, 7196–7198.
30. Lightner, D. A.; Paquette, L. A.; Chayangkoon, P.; Lin, H.-S.; Peterson, J. R. J. Org.
Chem. 1988, 53, 1973–1979.
31. Albert, A.; Royer, R. J. Chem. Soc. 1949, 1148–1151.
32. Doboszewski, B. Nucleosides Nucleotides 1997, 16, 1049–1052.
33. Doboszewski, B. Nucleosides, Nucleotides Nucleic Acids 2009, 28, 875–901.
34. Tsuda, T.; Sato, Y.; Kanemitsu, K.; Hosoi, S.; Shibayama, K.; Nakao, K.; Ishikawa,
Y. Chem. Pharm. Bull. 1996, 44, 1465–1475.
This compound was obtained as described under 1.7 for its
D
-enantiomer 7. ½a D26
ꢂ41.2 (c 1.5, CHCl3); mp 73–75 °C (Et2O–hex-
ꢁ
ane), lit.42 mp 76–77 °C (EtOH); ½a 2D6
ꢂ48.8 (c 0.7, CHCl3). Exact
ꢁ
mass calcd for C15H20O6S [M+Na]+: 351.0873, found: 351.0870.
For the NMR data: see enantiomeric compound 7.
Acknowledgments
35. De Angelis, F.; Marzi, M.; Minetti, P.; Misiti, D.; Muck, S. J. Org. Chem. 1997, 62,
4159–4161.
36. Zhang, W.; Robins, M. J. Tetrahedron Lett. 1992, 33, 1177–1180.
37. Doboszewski, B.; Silva, M. J. e; Nazarenko, A.; Nemykin, V. N. Acta Crystallogr., E
2012, E68, 1109–1110.
38. Raap, J.; van Boom, J. H.; van Lieshout, H. C.; Haasnoot, C. A. G. J. Am. Chem. Soc.
1988, 110, 2736–2743.
Financial support of this project was provided by the Post-Grad-
uation Program in Chemistry of this University and Instituto Nac-
ional de Ciência e Tecnologia para Inovação Farmacêutica.
39. Haasnoot, C. A. G.; Liskamp, R. M. J.; van Dael, P. A. W.; Noordik, J. H.;
Ottenheim, H. C. J. J. Am. Chem. Soc. 1983, 105, 5406–5414.
40. Brown, D. M.; Jones, G. H. J. Chem. Soc. (C) 1967, 249–252.
41. Walkup, R. D.; Cunningham, R. T. Org. Prep. Proced. Int. 1988, 20, 129–134.
42. Srivastava, V. K.; Lerner, L. M. J. Med. Chem. 1979, 22, 24–28.
43. Prokop, J.; Murray, D. H. J. Pharm. Sci. 1968, 57, 1697–1703.
44. Lerner, L. M. Carbohydr. Res. 1984, 132, 168–172.
References
1. Williams, R. M.; Burnett, C. M. ACS Symp. Ser. 2009, 1009, 420–442.
2. Del Valle, J. R.; Wang, X.; Ersmark, K.; Hanessian, S. ACS Symp. Ser. 2009, 1009,
443–461.
3. Liu, W. Terpenes: The Expansion of the Chiral Pool. In Handbook of Chiral
Chemicals; Ager, D., Ed., 2nd ed.; CRC Press: Boca Raton, 2005; pp 59–74.
4. Ho, T.-L. Enantioselective Synthesis of Natural Products from Chiral Terpenes; John
Wiley & Sons: New York, 1992.
45. El Khadem, H. S.; Audichaya, T. D.; Withee, M. J. Carbohydr. Res. 1974, 33, 329–
337.
46. Lemieux, R. U.; Watanabe, K. A.; Pavia, A. A. Can. J. Chem. 1969, 47, 4413–4426.
47. Casini, G.; Goodman, L. J. Am. Chem. Soc. 1964, 86, 1427–1432.
48. Solladie, G.; Hutt, J. Tetrahedron Lett. 1987, 28, 797–800.
49. Schreiber, S. L.; Goulet, M. T. Tetrahedron Lett. 1987, 28, 1043–1046.
50. Yadav, J. S.; Gupta, M. K.; Prathap, I. Synthesis 2007, 1343–1348.
51. Mereyala, H. B.; Gadikota, R. R. Tetrahedron: Asymmetry 2000, 11, 743–751.
52. Kuzuhara, H.; Terayama, H.; Ohrui, H.; Emoto, S. Carbohydr. Res. 1971, 20, 165–
169.
5. Inch, T. D. Tetrahedron 1984, 40, 3161–3213.
6. Vasella, A. Chiral Building Blocks in Enantiomer Synthesis from Sugars. In
Modern Synthetic Methods; Scheffold, R., Ed.; Salle and Sauerlander: Frankfurt/
Aarau, 1980; pp 173–267.
7. Hanessian, S. Total Synthesis of Natural Products: The Chiron Approach; Pergamon
Press: Oxford, 1983.
8. Hollingsworth, R. I.; Wang, G. Chem. Rev. 2000, 100, 4267–4282.
9. Knauer, S.; Kranke, B.; Krause, L.; Kunz, H. Curr. Org. Chem. 2004, 8, 1739–1761.