Free-radical mediated synthesis of enantiomerically pure, highly functionalized inositols from carbohydrates

Article abstract of DOI:10.1016/S0008-6215(01)00116-1

We report the synthesis, free-radical cyclization of precursors 1,2,7-trideoxy-7-iodo-3,4:5,6-di-O-isopropylidene-D-gluco-hept-1-enitol (1), methyl 7-O-acetyl-6-O-benzyl-8-bromo-2,3,8-trideoxy-4,5-O-isopropylidene-D-gluco- oct-2-enonate (2) and 5-O-acetyl-4-O-benzyl-6-bromo-6-deoxy-2,3-O-isopropylidene-D-glucose-O- benzyloxime (3), readily prepared from D-glucose, and some selected transformations of the carbocycles obtained from these intermediates. In compound 1 we have installed a terminal double bond and an iodide as radical acceptor and leaving group, respectively. Compounds 2 and 3 are ε-bromo aldehydes substituted with α,β-unsaturated ester and oxime ether functions as radical traps, respectively. The tributyltin hydride mediated ring closure of these radical precursors have afforded a series of interesting, diverse and highly functionalized carbocycles which can be considered useful building blocks for the synthesis of branched-chain cyclitols, aminocyclitols and aminoconduritols. In these processes, a good chemical yield and high stereoselectivity has been found in the newly formed stereocenters. Particularly interesting has been the finding that the stereochemical outcome of the free-radical cyclization is independent of the ratio of isomers (E or Z) in oxime ether 3. These results show the power and the state of art of this strategy for the stereocontrolled synthesis of enantiomerically pure inositols from carbohydrates.

Full text of DOI:10.1016/S0008-6215(01)00116-1

www.elsevier.nl/locate/carres
Carbohydrate Research 332 (2001) 341–349
Free-radical mediated synthesis of enantiomerically pure,
highly functionalized inositols from carbohydrates
Jose´ Marco-Contelles,* Carmen Pozuelo, Elsa de Opazo
Laboratorio de Radicales Libres, Instituto de Qu´ımica Organica General (CSIC), C/ Juan de la Cier6a 3,
E-28006 Madrid, Spain
Received 19 February 2001; accepted 20 April 2001
Abstract
We report the synthesis, free-radical cyclization of precursors 1,2,7-trideoxy-7-iodo-3,4:5,6-di-O-isopropylidene-
gluco-hept-1-enitol (1), methyl 7-O-acetyl-6-O-benzyl-8-bromo-2,3,8-trideoxy-4,5-O-isopropylidene- -gluco-oct-2-
enonate (2) and 5-O-acetyl-4-O-benzyl-6-bromo-6-deoxy-2,3-O-isopropylidene- -glucose-O-benzyloxime (3), readily
prepared from -glucose, and some selected transformations of the carbocycles obtained from these intermediates. In
D-
D
D
D
compound 1 we have installed a terminal double bond and an iodide as radical acceptor and leaving group,
respectively. Compounds 2 and 3 are o-bromo aldehydes substituted with a,b-unsaturated ester and oxime ether
functions as radical traps, respectively. The tributyltin hydride mediated ring closure of these radical precursors have
afforded a series of interesting, diverse and highly functionalized carbocycles which can be considered useful building
blocks for the synthesis of branched-chain cyclitols, aminocyclitols and aminoconduritols. In these processes, a good
chemical yield and high stereoselectivity has been found in the newly formed stereocenters. Particularly interesting
has been the finding that the stereochemical outcome of the free-radical cyclization is independent of the ratio of
isomers (E or Z) in oxime ether 3. These results show the power and the state of art of this strategy for the
stereocontrolled synthesis of enantiomerically pure inositols from carbohydrates. © 2001 Elsevier Science Ltd. All
rights reserved.
Keywords: Carbocyclization; Free-radicals; Tributyltin hydride; Inositols; Carbohydrates
1. Introduction
free-radical cyclization of the radical precur-
sors 1–3 (Schemes 1 and 2), and some selected
subsequent transformations of the carbocycles
obtained. In compound 1 (Scheme 1), we have
installed a terminal double bond and an io-
dide, as radical acceptor and leaving group,
respectively. Compounds 2 and 3 (Scheme 2)
are o-bromo aldehydes substituted with a,b-
unsaturated ester and oxime ether functions,
respectively. These radical precursors should
afford a series of interesting, diverse and
highly functionalized inositols, useful building
blocks for the synthesis of branched-chain
During the last 13 years we have developed
a series of new, free-radical based methodolo-
gies1–4 for the synthesis of enantiomerically
pure, highly functionalized carbocycles (‘inosi-
tols’) from carbohydrates.5,6
In this paper we describe some of our recent
developments in the free-radical cyclization of
acyclic, conveniently functionalized radical
precursors.5 We report here the synthesis and
* Corresponding author. Tel.: +34-91-5622900; fax: +34-
91-5644853.
E-mail address: iqoc21@iqog.csic.es (J. Marco-Contelles).
cyclitols,
aminocyclitols
and
amino-
conduritols.7
0008-6215/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(01)00116-1

342
J. Marco-Contelles et al. / Carbohydrate Research 332 (2001) 341–349
1
2. Results and discussion
isolation we could obtain, and analyze by H
NMR spectroscopy, pure isomer E-3 [l 7.46
(d, J1,2 6.4 Hz, 1 H, H-1)], a mixture of
isomers E-3 and Z-3, in a 1:1 ratio, and pure
isomer Z-3 [l 6.94 (d, J1,2 5.0 Hz, 1 H, H-1)].
This allowed us for the first time in our free-
radical projects11 to study the influence of the
geometry at the double bond of the oxime
ether in the yield and in the stereochemical
outcome of the 6-exo-trig12 free-radical
cyclization.
Synthesis of the radical precursors 1–3.—
The synthesis of 1 has been achieved from
alcohol 4 (Scheme 1). After aldehyde depro-
tection as described,8 to give aldehyde 5, fol-
lowed by Wittig reaction, final iodination of
alcohol 6, under Garegg’s conditions,9 af-
forded iodide 1.
The synthesis of the radical precursor 2 and
3 has been performed from a common inter-
mediate, aldehyde 11, under standard condi-
tions, as shown in Scheme 2. This aldehyde
has been synthesized from 7, a very well
known building block prepared by us several
years ago.10 This synthetic sequence provided
the desired compound after selective acid hy-
drolysis of the 1,2-O-isopropylidene group at
C-5 and C-6 to give 8 (70% yield), bromina-
tion at C-6 yielding halosugar 9 (61%), acety-
lation of the hydroxyl at C-5 and final
deprotection of the masked aldehyde at C-1
giving aldehyde 11 via compound 10. These
transformations afforded the desired interme-
diates in good yield and multigram quantities
(see Section 3). All new compounds showed
good analytical and spectroscopic data.
Wittig reaction on aldehyde 11 afforded the
expected a,b-unsaturated ester 2, isolated as
an inseparable mixture of E (H-2: 6.66 ppm,
J1,2 15.6, J2,3 6.0 Hz) and Z (H-2: 6.19 ppm,
J1,2 11.5, J2,3 8.6 Hz) isomers in 72% combined
yield, which were processed together in the
next free-radical cyclization.
The synthesis of oxime ether 3 was achieved
under the standard conditions in 74% yield
from aldehyde 11 (Scheme 2). During the
6-exo-Trig free-radical cyclization of precur-
sors 1–3.—Free-radical cyclization of precur-
sor 1, under standard conditions mediated by
tributyltin hydride and AIBN, provided car-
bocycle 12(C5R), unidentified product a and
unidentified product b (Scheme 1), in 40%
overall yield and in a 2:0.1:0.4 ratio (deter-
mined by coupled GLC–MS analysis in the
crude reaction mixture), respectively. After
column chromatography, we were able to iso-
late major resulting product, carbocycle
12(C5R), in 86% diastereomerically pure
form. After extensive spectroscopic analysis
we could analyze the relevant vicinal coupling
constant around carbon C-5 (J4,5 10.0 Hz);
this is a typical data for axial(H-4)–axial(H-5)
protons, which allowed us to assign as R the
absolute configuration at C-5. In summary,
compound 12(C5R) was the expected 6-exo-
trig cyclization derivative. We have hypothe-
sized that the other two detected minor
compounds resulting in the carbocyclization
reaction, products a and b, possibly corre-
sponded to the other two possible isomeric
cyclization products: the epimer at C-5,
12(C5S), and the 7-endo product, 13. This
Scheme 1.

J. Marco-Contelles et al. / Carbohydrate Research 332 (2001) 341–349
343
Scheme 2.
result of a free-radical cyclization not only is
important the presence of the isopropylidene
protecting groups for defining the course of
the radical cyclization, but the relationship
between the stereogenic sequence of the asym-
metric carbons and the location of the carbon
centered radical and the radical trap. In com-
pound 1 the primary radical is vicinal to a
syn-isopropylidene group, while in Redlich’s
example, in compound 14,14 the primary radi-
cal is vicinal to an anti-isopropylidene group.
These results moved us to test the free-radi-
cal cyclization of precursor 2 (Scheme 2). It
was of interest to see whether a simple iso-
propylidene group, of anti configuration and
located vicinal to the carbon that becomes the
radical acceptor, could efficiently direct the
stereochemical outcome of the process. The
free-radical cyclization of this precursor, un-
der the usual conditions, afforded only 16
(Scheme 2) in 63% yield, diastereomerically
pure at newly formed stereocenter. The basic
was based on the comparative analysis of the
mass spectra corresponding to these products,
obtained in the coupled GLC–MS experi-
ment. These two products showed almost
identical mass spectrum [similar to 12(C5R)
mass spectrum], but unfortunately, no un-
equivocal proof could be obtained about their
real identity. Independently of the exact na-
ture of these secondary products, the present
carbocyclization has been highly stereo- and
regioselective, the major isomer being the 6-
exo-trig product and the C5R isomer,
[12(C5R)].
The stereochemical outcome of this process
is quite similar and follows the expected
trends observed in our previous results in re-
lated substrates during the 6-exo-trig free-rad-
ical cyclization reactions.11 According to
Beckwith,13 major isomer results from the
free-radical cyclization of the radical species in
a chair-like conformation with most of the
substituents in preferred pseudo-equatorial
orientation.
The formation of the exclusive 6-exo-trig
product in this case was rather surprising in
view of the Redlich’s results on the cyclization
of radical precursor 14 to give carbocycle 15
(Scheme 3).14 This is indeed an unique 7-endo-
trig ring case closure in a sugar derivative.
Apparently, this result proves that for the final
Scheme 3.

344
J. Marco-Contelles et al. / Carbohydrate Research 332 (2001) 341–349
showed strong effects between H-2 and H-1/
H-4, between H-5 and H-3, and between H-4
and H-6b. In summary, in spite of the moder-
ate chemical yield for the cyclization, this ring
closure has been always completely stereose-
lective giving the expected b-substituted iso-
mer at C-5, in agreement with the cyclization
of precursor 2 (see above). We have also
observed that the stereoselectivity of the cy-
clization is independent of the E/Z ratio in the
precursor, obtaining always the same pure
major isomer. This is a fact that Bartlett and
colleagues could demonstrate for the 5-exo-
trig free-radical cyclization of precursors
derived from sugars having oxime ethers as
radical traps,15 but a conclusion that we could
not demonstrate until now in our similar and
pioneered 6-exo-trig free-radical cyclizations
on sugar templates.
hydrolysis of 16 gave alcohol 17 in 75% yield.
The stereochemistry at the newly formed
stereocenter in 16 was easily established by
1
analysis of the H NMR spectrum and selec-
1
tive H–1H decoupling experiments. Doing
that, and as expected, for a chair-like confor-
mation, we analyzed proton H-4 at 3.07 ppm,
with typical axial–axial vicinal coupling con-
stants (J3,4 9.0, J4,5a 10.8 Hz), which allowed
us to establish as S the absolute configuration
at C-5. Similar data and identical conclusions
were also obtained from the analysis of 17. In
summary, the stereochemical outcome of this
cyclization was the same which we have ob-
served in other previous studies,5 and clearly
indicates that the presence of only one iso-
propylidene group at C-2 and C-3 is enough
for a total control of the stereochemistry dur-
ing the ring closure. An additional bonus, due
to the type of the radical acceptor, was the
absence of endo cyclization products.
In summary, these results prove that the
free-radical cyclization strategy is one of the
most efficient and convenient methods for the
synthesis of differently substituted and func-
tionalized inositols.
The free-radical cyclization of a CꢀN dou-
ble bond-containing radical acceptor (the ox-
ime ether 3) (Scheme 2) was very interesting.
Cyclization of the mixture of oximes E-3+Z-
3, in a 1:1 ratio, using the slow addition of
reagents technique (see Section 3), afforded 18
in poor yield (20%) accompanied with other
more complex products whose structure could
not be determined. When the cyclization was
performed with pure precursor (E)-3 in the
standard conditions, but at 0.01 M in toluene,
and using the ‘one-pot’ addition technique,
only 18 was obtained but in a poor yield
(27%). Finally, for precursor (Z)-3 in the stan-
dard conditions, but at 0.05 M in toluene, and
using again the ‘one-pot’ addition technique,
only 18 was obtained in a good yield (73%).
This compound, after basic hydrolysis, af-
forded alcohol 19 in 85% yield (Scheme 2).
Compounds 18 and 19 showed analytical and
spectroscopic data in good agreement with
3. Experimental
General methods.—Reactions were moni-
tored by TLC using precoated silica gel alu-
minum plates containing
a
fluorescent
indicator (E. Merck, 5539). Detection was
done by UV (254 nm) followed by charring
with sulfuric–AcOH spray, 1% aq KMnO4
solution or 0.5% phosphomolybdic acid in
95% EtOH. Anhydrous Na2SO4 was used to
dry organic solutions during work-ups and the
removal of solvents was carried out under
vacuum with a rotary evaporator. Flash-
column chromatography was performed using
silica gel 60 (230–400 mesh, E. Merck) and
hexane–EtOAc mixtures as eluent unless oth-
1
1
these structures. Careful analysis of the H
erwise stated. H spectra were recorded with a
NMR spectrum of derivative 18 along with
Varian VXR-300(400)S spectrometers, using
tetramethylsilane as an internal standard and
13C NMR spectra were recorded with a
Bruker WP-200-SY. Values with (*) can be
interchanged. The coupled GLC–MS experi-
ment was run in a column. Compounds 4, 5
and 7 were prepared according to published
procedures.8,10
1
selective H–1H decoupling experiments al-
lowed us to assign the chemical shifts and
determine relevant coupling constants for
configurational determination (J5,6b 11.0 Hz).
This value clearly showed that the configura-
tion at the newly formed stereocenter was R.
In addition, selective NOE experiments

J. Marco-Contelles et al. / Carbohydrate Research 332 (2001) 341–349
345
mL), under Ar and at reflux, was treated with
triphenylphosphine (284 mg, 1.1 mmol, 2.4
equiv), imidazole (74 mg, 1.1 mmol, 2.4 equiv)
and iodine (172 mg, 0.7 mmol, 1.5 equiv).
After 30 min the mixture was cooled, and the
mixture was washed with a 10% aq solution of
NaHCO3 and brine. The organic phase was
separated, dried, the solvent was evaporated,
and the residue submitted to chromatography
(99:1 hexane–EtOAc) affording 1 (115 mg,
70%): oil; [h]2D5 −9° (c 0.57, CHCl3); IR (film)
w 2985, 2934, 1455, 1380, 1216, 1061 cm−1; 1H
NMR (300 MHz, CDCl3): l 5.82 (ddd, J1,2
17.4, J1%,2 10.1, J2,3 7.3 Hz, 1 H, H-2), 5.43
(dm, 1 H, H-1), 5.31 (dm, 1 H, H-1%), 4.51 (q,
J6,7=J5,6=J6,7% 7.0 Hz, 1 H, H-6), 4.39 (J3,4
8.5 Hz, 1 H, H-3), 4.08 (dd, J4,5 1.7 Hz, 1 H,
H-5), 3.77 (dd, 1 H, H-4), 3.38 (dd, J7,7% 2.5
Hz, 2 H, 2 H-7), 1.52, 1.43 (6 H), 1.38 [3 s,
2×OC(CH3)2, 12 H]; 13C NMR (75 MHz,
CDCl3): l 134.6 (C-2), 119.9 (C-1), 109.8,
109.3 [2 C, 2×OC(CH3)2O], 79.1 (C-3), 78.4
(C-4), 77.9 (C-6), 74.2 (C-5), 27.2, 26.9, 26.7,
25.5 [4 C, 2×OC(CH3)2O], 2.5 (C7); EIMS
General method for free-radical cycliza-
tion.—To a solution of the precursor in tolu-
ene (0.02 M), previously deoxygenated by
bubbling Ar into the solution, under Ar, at
reflux, a solution of AIBN (0.5 equiv) and
Bu3SnH (2 equiv) was slowly added via sy-
ringe pump in the indicated time. The mixture
was heated until complete reaction, the sol-
vent was removed, the residue dissolved in a
mixture of 1:1 ethyl ether–15% aq solution of
KF and stirred vigorously overnight. Then,
the organic layer was separated, dried, filtered,
evaporated and the residue was submitted to
chromatography (eluting with hexane–EtOAc
mixtures) to give the product.
1,2-Dideoxy-3,4:5,6-di-O-isopropylidene- -
D
gluco-hept-1-enitol (6).—To a suspension of
hemiacetal 5 (1.15 g, 4.41 mmol) and methylt-
riphenylphosphonium bromide (4.91 g, 13.7
mmol, 3.1 equiv) in dry THF (7 mL), cooled
at −20 °C, under Ar, n-BuLi (5.5 mL, 8.8
mmol, 1.6 M in hexane) was slowly added.
The mixture was warmed at rt, stirred for 2 h
and quenched with an aq satd solution of
ammonium chloride. Then, the mixture was
diluted with CH2Cl2, washed with an aq satd
solution of NaHCO3, brine, dried, filtered and
evaporated. The residue was submitted to
flash chromatography (7:3 hexane–EtOAc) to
give 6 (960 mg, 84%): oil; [h]2D5 −12° (c 0.23,
CHCl3); IR (film) w 3460 (OH), 2986, 2936
+
’
+
(70 eV): m/z 368 ([M ], 1), 353 ([M −15],
12), 241 (13), 183 (23), 155 (10), 127 (52), 108
(21), 98 (100). Anal. Calcd for C13H21IO4: C,
42.40; H, 5.75. Found: C, 42.36; H, 5.81.
Carbocyclization of radical precursor 1.—
Compound 1 (88.6 mg, 0.24 mmol) was sub-
mitted to the procedure in Section 3.2 (time
for the slow addition: 9 h, and 24 for addi-
tional reaction time) to give Fraction (A) [15
mg of a mixture of 12(C5R) (tr=16.84 min, in
the GLC–MS experiment) and product a
(tr=19.27 min, in the GLC–MS experiment),
in a 93:7 ratio, according to GLC] and Frac-
tion (B) [6 mg of a mixture of 12(C5R),
product b (tr=17.91 min, in the GLC–MS
experiment), and product a, in a 51:10:39 ra-
tio, respectively]. Total yield: 21 mg (40%).
12(C5R) (86% diastereomerically pure)
1
1380, 1216, 1050 cm−1; H NMR (300 MHz,
CDCl3): l 5.79 (ddd, J1,2 17.3, J1%,2 10.1, J2,3
7.6 Hz, 1 H, H-2), 5.41 (dm, 1 H, H-1), 5.28
(dm, 1 H, H-1%), 4.45 (dd, J3,4 8.6 Hz, 1 H,
H-3), 4.23 (dt, J6,7 5.2, J5,6 6.7 Hz, 1 H, H-6),
4.05 (dd, J4,5 1.7 Hz, 1 H, H-5), 3.76 (t, J7,OH
5.2 Hz, 2 H, 2 H-7), 3.68 (dd, 1 H, H-4), 2.84
(t, 1 H, OH), 1.50, 1.42 (6 H), 1.35 [3 s,
2×OC(CH3)2, 12 H]; 13C NMR (75 MHz,
CDCl3): l 134.4 (C-2), 120.1 (C-1), 109.8,
108.8 [2 C, 2×OC(CH3)2O], 79.3 (C-3), 78.2
(C-6), 77.2 (C-5), 73.3 (C-3), 61.3 (C-7), 27.1,
26.8, 26.4, 25.5 [4 C, 2×OC(CH3)2O]; EIMS
(70 eV): m/z 243 ([M+−15], 12), 141 (5), 127
(21), 113 (10), 98 (36), 69 (52), 43 (100). Anal.
Calcd for C13H22O5: C, 60.45; H, 8.58. Found:
C, 60.30; H, 8.81.
[1 (1,2,4/3,5)-1,2:3,4-di-O-isopropylidene-5-
L
C-methylcyclohexan-1,2,3,4-tetrol]: oil; [h]D25
−44° (c 0.52, CHCl3); IR (film) w 2984, 2927,
1
1456, 1381, 1232, 1055 cm−1; H NMR (300
MHz, CDCl3): l 4.29 (td, J1,2 =J1,6b 4.9, J1,6a
1.7 Hz, 1 H, H-1), 4.13 (dd, J2,3 9.0 Hz, 1 H,
H-2), 3.52 (t, J3,4 9.2 Hz, 1 H, H-3), 2.94 (t,
J4,5 10.0 Hz, 1 H, H-4), 2.26 (ddd, J6a,6b 15.5,
J5,6a 4.7 Hz, 1 H, H-6a), 1.75 (m, 2 H, H-5,
1,2,7-Trideoxy-7-iodo-3,4:5,6-di-O-isopropy-
lidene- -gluco-hept-1-enitol (1).—A solution
D
of 6 (116 mg, 0.45 mmol) in dry toluene (5.4

346
J. Marco-Contelles et al. / Carbohydrate Research 332 (2001) 341–349
(24), 301 (51), 287 (33), 250 (26), 171 (34), 91
(100). Anal. Calcd for C20H31BrO4S2: C, 50.10;
H, 6.52; S, 13.37. Found: C, 50.22; H, 6.31; S,
13.45.
H-6b), 1.51, 1.42, 1.41, 1.35 [4 s, 2×
OC(CH3)2O], 1.05 (d, J 6.5 Hz, 3 H, CH3); 13C
NMR (75 MHz, CDCl3): l 110.1, 108.9 [2 C,
2×OC(CH3)2O], 81.8 (C-3), 81.0 (C-4), 78.6
(C-2), 74.5 (C-1), 35.3 (C-6), 28.6, 26.9, 26.2 (2
C) [4 C, 2×OC(CH3)2O], 17.8 (CH3); EIMS
(70 eV): m/z 227 ([M+−15], 100), 169 (39),
127 (39), 109 (61), 81 (92). Anal. Calcd for
C13H22O4: C, 64.44; H, 9.15. Found: C, 64.37;
H, 9.31.
5-O-Acetyl-4-O-benzyl-6-bromo-6-deoxy-
2,3-O-isopropylidene-
D
-glucose
diethyl
dithioacetal (10).—Compound 9 (1.0 g, 2.1
mmol) was treated with Ac2O (9.5 mL) and
pyridine (9.4 mL) at rt overnight. The solvent
was removed and the residue was submitted to
chromatography (17:3 hexane–EtOAc) to give
bromide 10 (590 mg, 51%): oil; [h]2D5 −7° (c
0.55, CHCl3); IR (film) w 3020, 2980, 1720,
4-O-Benzyl-2,3-O-isopropylidene- -glucose
D
diethyl dithioacetal (8).—Compound 7 (9.6 g,
21 mmol) was treated with 7:3 AcOH–water
(50 mL) at rt overnight. The solvent was
removed and the residue was submitted to
chromatography (4:1 hexane–EtOAc) to give
diol 8 (6.5 g, 70%): oil; [h]2D5 −178° (c 0.11,
CHCl3); IR (film) w 3600–3100, 3030, 3010,
1
1465, 1340, 1250–1180, 1100–1025 cm−1; H
NMR (300 MHz, CDCl3): l 7.35–7.26 (m, 5
H, OCH2C6H5), 5.23 (ddd, J5,6 7.2, J5,4 2.8,
J5,6% 4.4 Hz, 1 H, H-5), 4.76 (q, J 11.4 Hz, 2 H,
OCH2C6H5), 4.33 (dd, J2,3 7.7, J1,2 5.7 Hz, 1
H, H-2), 4.23 (dd, J4,3 7.3 Hz, 1 H, H-4),
3.97–3.93 (m, 2 H, H-6%, H-3), 3.83 (d, 1 H,
H-1), 3.67 (dd, J6,6% 11.5 Hz, 1 H, H-6), 2.76–
2.71 (m, 4 H, 2×SCH2CH3), 2.13 (s, 3 H,
OCOCH3), 1.43, 1.38 [2 s, 6 H, OC(CH3)2O],
1.25 (t, J 7.5 Hz, 6 H, 2×SCH2CH3); EIMS
(70 eV): m/z 459–461 (2, 2), 4512–414 (4, 4),
385–387 (4, 4), 327–329 (7, 7), 135 (51), 91
(100). Anal. Calcd for C22H33BrO5S2: C, 50.67;
H, 6.38; S, 12.29. Found: C, 50.44; H, 6.31; S,
12.17.
1
2920, 1425, 1345, 1225, 1050 cm−1; H NMR
(300 MHz, CDCl3): l 7.36–7.34 (m, 5 H,
OCH2C6H5), 4.70 (q, J 11.3 Hz, 2 H,
OCH2C6H5), 4.42–4.37 (m, 2 H), 4.01–3.97
(m, 1 H), 3.88 (d, J1,2 4.6 Hz, 1 H, H-1),
3.81–3.77 (m, 3 H), 3.10–3.00 (br s, 1 H,
OH), 2.80–2.60 (m, 4 H, 2×SCH2CH3),
2.65–2.58 (br s, 1 H, OH), 1.45, 1.43 [2 s,
OC(CH3)2O], 1.23 (t, J 7.5 Hz, 6 H, 2×
SCH2CH3); EIMS (70 eV): m/z 308 (16), 297
(8), 223 (12), 136 (20), 135 (37), 91 (100).
Anal. Calcd for C20H32O5S2: C, 57.66; H, 7.74;
S, 15.39. Found: C, 57.47; H, 7.61; S, 15.22.
4-O-Benzyl-6-bromo-6-deoxy-2,3-O-iso-
Methyl
7-O-acetyl-6-O-benzyl-8-bromo-
-gluco-
2,3,8-trideoxy-4,5-O-isopropylidene-
D
oct-2-enonate (2).—Compound 10 (590 mg,
1.13 mmol) was dissolved in 18:1 acetone–wa-
ter. Then, mercuric oxide (490 mg, 2.26 mmol)
and mercuric chloride (614 mg, 2.26 mmol)
were added. The mixture was refluxed
overnight. The flask was cooled, the solvent
evaporated, and the residue was dissolved in
CH2Cl2, washed with a 5% aq solution of KI,
brine, dried, filtered, and evaporated. The
crude was submitted to chromatography (3:1
hexane–EtOAc) to give aldehyde 11 (298 mg,
63%) which was immediately used. To an
aliquot of this aldehyde (228.3 mg, 0.55
mmol), dissolved in dry toluene (10 mL),
methoxycarbonyltriphenylphosphorane (273
mg, 0.82 mmol, 1.5 equiv) was added, and the
mixture was stirred for 24 h at rt. The solvent
was removed and the residue was submitted to
chromatography (7:3 hexane–EtOAc) to give
2 (186.6 mg, 72%), as an inseparable mixture
propylidene-
D
-glucose
diethyl
dithioacetal
(9).—Compound 8 (1.5 g, 3.6 mmol) was
dissolved in dry THF (45 mL) and treated
with triphenylphosphine (1.88 g, 7.2 mmol, 2
equiv) and carbon tetrabromide (3.5 g, 10.8
mmol) at rt for 45 min. The solvent was
removed and the residue was submitted to
chromatography (3:2 hexane–EtOAc) to give
bromide 9 (1.0 g, 61%): oil; [h]2D5 −16° (c
0.69, CHCl3); IR (film) w 3560–3200, 3025,
1
2970, 1465, 1390, 1220, 1130 cm−1; H NMR
(300 MHz, CDCl3): l 7.37–7.35 (m, 5 H,
OCH2C6H5), 4.69 (q, J 11.3 Hz, 2 H,
OCH2C6H5), 4.41 (d, J 3.3 Hz, 2 H), 4.12 (q,
J 6.0 Hz, 1 H, H-5), 3.03 (d, J 6.0 Hz, 1 H,
OH), 2.77–2.61 (m, 4 H, 2×SCH2CH3), 1.46,
1.43 [2 s, OC(CH3)2O], 1.25 (t, J 7.3 Hz, 6 H,
2×SCH2CH3); EIMS (70 eV): m/z 479–481
([M++1], 2), 451 (4), 419 (28), 385 (28), 325

J. Marco-Contelles et al. / Carbohydrate Research 332 (2001) 341–349
347
1 (1,2,4/3,5)-2-O-Benzyl-3,4-O-isopropyli-
L
of E and Z isomer in a 1.4:1 ratio: oil; IR
(CHCl3) w 2986, 1726, 1438, 1373, 1071 cm−1
dene-5-C-methoxycarbonylmethylcyclohexan-
1,2,3,4-tetrol (17).—Compound 16 (33 mg,
0.084 mmol) was treated with a mixture of
1:4:5 triethylamine, water and MeOH (1 mL)
at rt for 2 days. The solvent was removed and
the crude was submitted to chromatography
(7:3 hexane–EtOAc) to give 17 (22 mg, 75%):
oil; [h]2D5 −9° (c 0.14, CHCl3); IR (film) w
3454, 2930, 1739, 1436, 1372, 1237, 1096
;
1H NMR (300 MHz, CDCl3): l 7.44–7.28 (m,
5 H, OCH2C6H5), 6.66 (dd, J2,3 15.6, J3,4 6.0
Hz, 1 H, H-3(E)), 6.19 (dd, J2,3 11.5, J3,4 8.6
Hz, 1 H, H-3(Z)), 5.98 (dd, J2,4 1.0 Hz, 1 H,
H-2(Z)), 5.90 (dd, J2,4 1.3 Hz, 1 H, H-2(E)),
5.70 (ddd, J4,5 15.8 Hz, 1 H, H-4(Z)), 5.28–
5.14 (m, 2 H), 4.84 (q, J 11.1 Hz, 2 H,
(Z)-OCH2C6H5), 4.71 (q, J 11.5 Hz, 2 H,
(Z)-OCH2C6H5), 4.45 (ddd, J4,5 14.3 Hz, 1 H,
H-4(E)), 3.98–3.60 (m, 3 H), 3.73 (s, 3 H,
(E)-CO2CH3), 3.68 (s, 3 H, (Z)-CO2CH3), 2.10
(s, 3 H, (E)-OCOCH3), 2.09 (s, 3 H, (Z)-
OCOCH3), 1.40, 1.38, 1.37 (3 s, OC(CH3)2);
EIMS (70 eV): m/z 455–457 ([M+−15], 8/8),
247 (25), 167 27), 127 (54), 91 (100). Anal.
Calcd for C21H27BrO7: C, 53.51; H, 5.77.
Found: C, 53.55; H, 5.63.
1
cm−1; H NMR (300 MHz, CDCl3): l 7.32–
7.21 (m, 5 H, OCH2C6H5), 4.77 (q, J 12.0 Hz,
2 H, OCH2C6H5), 4.09 (m, Wh/2 1.4 Hz, 1 H,
H-1), 3.89 (t, J3,4=J2,3 9.5 Hz, 1 H, H-3), 3.67
(s, 3 H, ꢀCO2CH3), 3.54 (dd, J2,1 3.4 Hz, 1 H,
H-2), 3.08 (dd, J4,5a 10.7 Hz, 1 H, H-4), 2.66
(dd, J7,7% 15.0, J7,5a 4.4 Hz, 1 H, H-7), 2.60–
2.12 (m, 5 H, OH, H-5a, H-7%, 2 H-6), 1.43,
1.40 [2 s, OC(CH3)2O]; EIMS (70 eV): m/z 168
(10), 140 (9), 107 (9), 91 (100). Anal. Calcd for
C19H26O6: C, 65.13; H, 7.48. Found: C, 65.29;
H, 7.55.
Carbocyclization of radical precursor 2.—
Compound 2 (146.3 mg, 0.3 mmol) was sub-
mitted to the procedure in Section 3.2 by
adding quickly HSnBu3 (161 mL, 0.6 mmol, 2
equiv) and AIBN (24.6 mg, 0.5 equiv) in one
single operation, to give product 16 (77 mg,
63%) after chromatography (4:1 hexane–
5-O-Acetyl-4-O-benzyl-6-bromo-6-deoxy-
2,3-O-isopropylidene- -glucose O-benzyloxime
D
(3).—Aldehyde 11 (443 mg, 1.06 mmol) was
dissolved in CH2Cl2 (19 mL) and treated with
O-benzylhydroxylamine hydrochloride (192
mg, 1.1 mmol, 1.1 equiv), pyridine (91 mg, 1.1
mmol, 1.1 equiv) and water (ten drops). The
mixture was stirrred at rt for 24 h. The mix-
ture was diluted with more CH2Cl2, washed
with brine, dried, filtered and the solvent was
removed. The residue was submitted to chro-
matography (19:1 hexane–EtOAc) to give
pure compound (E)-3 (120 mg), (E)+(Z)-3
(229 mg, in a 1:1 ratio) and pure (Z)-3 (156
mg). Total: 405 mg (74% yield). (E)-3: oil; IR
(film) w 3064, 3032, 2987, 1745, 1455, 1372,
EtOAc). Compound 16 [1 (1,2,4/3,5)-1-O-ace-
L
tyl-2-O-benzyl-3,4-O-isopropylidene-5-C-meth
oxycarbonylmethylcyclohexan - 1,2,3,4 - tetrol]:
oil; IR (film) w 2986, 1738, 1437, 1372, 1241,
1
1092 cm−1; H NMR (300 MHz, CDCl3): l
7.32–7.21 (m, 5 H, OCH2C6H5), 5.46 (q,
J1,2=J1,6a=J1,6b 3.0 Hz, 1 H, H-1), 4.75 (q, J
12.5 Hz, 2 H, OCH2C6H5), 3.83 (t, J3,4=J2,3
9.5 Hz, 1 H, H-3), 3.64 (s, 3 H, CO2CH3), 3.56
(dd, 1 H, H-2), 3.07 (dd, J4,5a 10.8 Hz, 1 H,
H-4), 2.65 (dd, J7,7% 15.2, J7,5a 4.0 Hz, 1 H,
H-7), 2.50–2.00 (m, 4 H, H-5a, H-7%, 2 H-6),
2.08 (s, 3 H, OCOCH3), 1.43, 1.40 [2 s,
OC(CH3)2O]; 13C NMR (75 MHz, CDCl3): l
172.1, 170.2 (OCOCH3, CO2CH3), 137.9–
127.5 (OCH2C6H5), 110.3 [OC(CH3)2O], 80.3
(C-3)*, 78.5 (C-2), 77.8 (C-4)*, 71.0
(OCH2C6H5), 68.4 (C-1), 51.6 (CO2CH3), 36.7
(C7)*, 33.6 (C-6)*, 32.3 (C-5), 26.9, 26.8 [2 C,
OC(CH3)2O], 21.1 (OCOCH3); EIMS (70 eV):
m/z 392 (2), 377 ([M+−15], 6), 334 (6), 303
(17), 226 (19), 168 (46), 91 (100). Anal. Calcd
for C21H28O7: C, 64.27; H, 7.19. Found: C,
64.37; H, 7.11.
1
1242, 1083, 1025 cm−1; H NMR (300 MHz,
CDCl3): l 7.46 (d, J1,2 6.4 Hz, 1 H, H-1),
7.37–7.31 (m, 10 H, 2×OCH2C6H5), 5.15–
5.10 (m, 1 H, H-5), 5.12 (s, 2 H, OCH2C6H5),
4.67 (q, J 11.3 Hz, 2 H, OCH2C6H5), 4.58 (dd,
J2,3 8.1, J1,2 6.3 Hz, 1 H, H-2), 4.11 (dd, J4,3
3.6 Hz, 1 H, H-3), 3.87 (dd, J5,6 2.8, J6,6% 11.3
Hz, 1 H, H-6), 3.73 (dd, J4,5 4.8 Hz, 1 H, H-4),
3.70 (dd, J5,6% 6.5 Hz, 1 H, H-6%), 2.08 (s, 3 H,
OCOCH3), 1.41, 1.38 [2 s, 6 H, OC(CH3)2O];
13C NMR (75 MHz, CDCl3): l 169.7
(OCOCH3), 147.6 (C-1), 137.8-127.9 (2×
OCH2C6H5), 110.0 [OC(CH3)2O], 79.0, 76.5,
76.2–74.8 (2×OCH2C6H5), 73.9, 73.4, 32.0

348
J. Marco-Contelles et al. / Carbohydrate Research 332 (2001) 341–349
2×OCH2C6H5),
5.88
(br
s,
1
H,
(C-6),
26.5–26.4
[OC(CH3)2O],
20.8
NHOCH2C6H5), 5.55 (dt, J1,2 =J1,6a 3.2, J1,6b
3.0 Hz, 1 H, H-1), 4.69 (q, J 13.2 Hz, 2 H,
OCH2C6H5), 4.68 (s, 2 H, OCH2C6H5), 3.88
(dd, J3,4 10.1, J2,3 10.3 Hz, 1 H, H-3), 3.61 (dd,
1 H, H-2), 3.41–3.33 (m, 2 H, H-4, H-5), 2.20
(dt, J5,6a 3.2, J6a,6b 14.7 Hz, 1 H, H-6a), 2.09
(s, 3 H, OCOCH3), 1.60 (ddd, J5,6b 11.0 Hz, 1
H, H-6b), 1.47, 1.45 [2 s, 6 H, OC(CH3)2O];
13C NMR (75 MHz, CDCl3): l 170.1
(OCOCH3), 137.8–127.5 (2×OCH2C6H5),
110.9 [OC(CH3)2O], 77.69 (C-3), 77.64 (C-2),
77.2 (C-4), 77.1–71.1 (2×OCH2C6H5), 68.5
(C-1), 57.1 (C-5), 32.0 (C-6), 26.9, 26.8
[OC(CH3)2O], 21.1 (OCOCH3); EIMS (70
(OCOCH3); EIMS (70 eV): m/z 506–504
([M+−15], 2, 1), 464–462 (1, 1), 296 (13), 234
(10), 181 (9), 91 (100). Anal. Calcd for
C25H30BrNO6: C, 57.70; H, 5.81; N, 2.69.
Found: C, 57.47; H, 5.61; N, 2.77. (Z)-3: oil;
IR (film) w 3064, 3032, 2987, 1745, 1455, 1372,
1242, 1083 cm−1
CDCl3):
;
1H NMR (300 MHz,
7.38–7.18 (m, 10 H, 2×
l
OCH2C6H5), 6.94 (d, J1,2 5.0 Hz, 1 H, H-1),
5.20–5.13 (m, 2 H, H-5, H-2), 5.16–5.09 (2 d,
J 11.4 Hz, 2 H, OCH2C6H5), 4.44 (q, J 11.5
Hz, 2 H, OCH2C6H5), 4.09 (dd, J3,4 1.8, J3,2
7.3 Hz, 1 H, H-3), 3.92 (m, 1 H, H-4), 3.91
(dd, J5,6 7.1, J6,6% 11.6 Hz, 1 H, H-6), 3.61 (dd,
J5,6% 6.5 Hz, 1 H, H-6%), 2.12 (s, 3 H,
OCOCH3), 1.39 [s, 6 H, OC(CH3)2O]; 13C
NMR (75 MHz, CDCl3): l 169.6 (OCOCH3),
151.6 (C-1), 137.7–127.1 (2×OCH2C6H5),
110.7 [OC(CH3)2O], 79.9, 77.1, 76.9–74.6
(2×OCH2C6H5), 73.9, 71.4, 32.4 (C-6), 26.3
[2 C, OC(CH3)2O], 20.9 (OCOCH3); EIMS
(70 eV): m/z 464–462 (1, 1), 428–430 (1, 1),
370–372 (2, 2), 354 (4), 296 (13), 234 (9), 181
(8), 91 (100). Anal. Calcd for C25H30BrNO6:
C, 57.70; H, 5.81; N, 2.69. Found: C, 57.65;
H, 5.90; N, 2.53.
+
’
+
eV): m/z 441 ([M ], 12), 440 ([M −1], 34),
382 (25), 181 (15), 105 (111), 91 (100). Anal.
Calcd for C25H31NO6: C, 68.01; H, 7.08; N,
3.17. Found: C, 67.96; H, 7.18; N, 3.43.
[1 (1,2,4/3,5) - 2 - O - Benzyl - 5 - benzyloxy-
L
amino-3,4-O-isopropylidenecyclohexan-1,2,3,4-
tetrol] (19).—Acetate 18 (96.8 mg, 0.22 mmol)
was treated with a with a mixture of 1:4:5
triethylamine, water and MeOH (1 mL) at rt
for 2 days. The solvent was removed and the
crude was submitted to chromatography (7:3
hexane–EtOAc) to give 19 (75 mg, 85% yield):
oil; [h]2D5 2° (c 0.35, CHCl3); IR (film) w 3477,
3030, 2929, 1496, 1454, 1371, 1232, 1091, 1027
Carbocyclization of radical precursor 3
Carbocyclization of radical precursor (E)-
3+(Z)-3 (1:1). This mixture (159 mg, 0.3
mmol) was treated according Section 3.2 (slow
addition in 6 h, and additional heating at
100 °C for 10 h) to give 18 (26 mg, 20%).
Carbocyclization of radical precursor pure
(E)-3. This compound (105 mg, 0.2 mmol)
was treated according to Section 3.2 (0.01 M
in toluene, ‘one-pot’ addition, and additional
heating at 100 °C for 2 h) to give (18) (24 mg,
27%).
1
cm−1; H NMR (300 MHz, CDCl3): l 7.38–
7.27 (m, 10 H, 2×OCH2C6H5), 5.80 (br s, 1
H, NHOCH2C6H5), 4.76 (q, J 11.9 Hz, 2 H,
OCH2C6H5), 4.70 (s, 2 H, OCH2C6H5), 4.12
(m, 1 H, H-1), 3.93 (t, J3,4=J2,3 9.3 Hz, 1 H,
H-3), 3.57 (dd, J2,1 3.3 Hz, 1 H, H-2), 3.41 (dt,
J5,4=J5,6b 10.7, J5,6a 3.9 Hz, 1 H, H-5), 3.37
(dd, 1 H, H-4), 2.25 (dt, J1,6a 3.9, J6a,6b 14.4
Hz, 1 H, H-6a), 1.55–1.48 (m, 1 H, H-6b),
1.45 [s, 6 H, OC(CH3)2O]; EIMS (70 eV): m/z
+
’
Carbocyclization of radical precursor pure
(Z)-3. This compound (134 mg, 0.26 mmol)
was treated according to Section 3.2 (0.05 M
in toluene, ‘one-pot’ addition, and additional
heating at 100 °C for 5 h) to give 18 (86 mg,
350 ([M ], 8), 188 (5), 91 (100). Anal. Calcd
for C23H29NO5: C, 69.15; H, 7.32; N, 3.51.
Found: C, 69.06; H, 7.27; N, 3.33.
73%). [1
L
(1,2,4/3,5)-1-O-Acetyl-2-O-benzyl-5-
Acknowledgements
benzyloxyamino - 3,4 - O - isopropylidenecyclo-
hexan-1,2,3,4-tetrol] (18): oil; [h]2D5 −23° (c
1.05, CHCl3); IR (film) w 3031, 2985, 1739,
This work was supported by CICYT, CAM
and COST Action No. D12 (European
Union). E.O. is a fellow of the Consejer´ıa de
Educacio´n y Cultura (CAM).
1
1497, 1454, 1372, 1238, 1093 cm−1; H NMR
(300 MHz, CDCl3): l 7.35–7.31 (m, 10 H,

J. Marco-Contelles et al. / Carbohydrate Research 332 (2001) 341–349
349
8. Wong, M. J. H.; Gray, G. R. Carbohydr. Res. 1980, 80,
87–98.
References
9. Garegg, P. J.; Samuelson, B. J. Chem. Soc., Perkin Trans.
1 1980, 2866–2869.
10. Marco-Contelles, J.; Sa´nchez, B. J. Org. Chem. 1993, 58,
4293–4297.
11. Marco-Contelles, J.; Pozuelo, C.; Jimeno, M. L.; Mar-
t´ınez, L.; Mart´ınez-Grau, A. J. Org. Chem. 1992, 57,
2625–2631.
12. Baldwin, J.E. J. Chem. Soc., Chem. Commun. 1976, 734–
736.
1. Giese, B. Radicals in Organic Synthesis: Formation of
Carbon–Carbon Bonds; Pergamon: Oxford, 1986.
2. Giese, B.; Kopping, B.; Go¨bel, T.; Dickhaut, J.; Thoma,
G.; Kulicke, K. J.; Trach, F. Org. React. 1996, 48,
301–856.
3. Curran, D. P. Synthesis 1988, 417–439, 489–513.
4. Motherwell, W. B.; Crich, D. Free Radical Chain Reac-
tions in Organic Synthesis; Academic: London, 1991.
5. Mart´ınez-Grau, A.; Marco-Contelles, J. Chem. Soc. Re6.
1998, 27, 155–162.
13. Beckwith, A. L. J.; Schiesser, C. H. Tetrahedron 1985, 41,
3925–3941.
6. Marco-Contelles, J.; Alhambra, C.; Mart´ınez-Grau, A.
Synlett 1998, 693–699; Corrigendum: Synlett 1999, 376.
7. Balci, M.; Su¨tbeyaz, Y.; Sec¸en, H. Tetrahedron 1990, 46,
3715–3742.
14. Redlich, H.; Sudau, W.; Szardenings, A. K.; Vollerthum,
R. Carbohydr. Res. 1992, 226, 57–78.
15. Bartlet, P. A.; McLaren, K. L.; Ting, P. C. J. Am. Chem.
Soc. 1988, 110, 1633–1634.
.