Chelation-controlled regioselective endo cleavage and stereoselective C-1 alkylation of pentofuranosides

Article abstract of DOI:10.1039/a706166h

Combinations of Lewis acids and nucleophilic reagents trigger endo-opening of the furanoside ring of methyl furanosides 1, 10 and 13, resulting in the attachment of the nucleophilic group at C-1 of the carbohydrate. The stereoselectivity in the C-C bond-f

Full text of DOI:10.1039/a706166h

View Article Online / Journal Homepage / Table of Contents for this issue
Chelation-controlled regioselective endo cleavage and stereoselective
C-1 alkylation of pentofuranosides
Roger Olsson, Pontus Rundström and Torbjörn Frejd*
Organic Chemistry 1, Department of Chemistry, Lund University, PO Box 124, S-221 00 Lund,
Sweden
Combinations of Lewis acids and nucleophilic reagents trigger endo-opening of the furanoside ring of
methyl furanosides 1, 10 and 13, resulting in the attachment of the nucleophilic group at C-1 of the
carbohydrate. The stereoselectivity in the C᎐C bond-forming step is low for the 2-deoxyfuranosides but
very high (dr 1:99) for the furanosides carrying a methoxy group in the 2-position when a combination
of TiCl4 and Me2Zn is used. Different selectivities are obtained with Me2Zn as compared with Me3Al.
Reagents based on several organometallic reagents of Al, Si, Ti and Zn in combination with TiCl4 can
be used.
β--ribofuranoside with TMSCN and a catalytic amount of
Introduction
BF3ؒOEt2.33 Later, Martin et al. used the chelation control of
pentofuranosides and hexopyranosides for the stereoselective
allylation at C-1 by application of the Lewis acid–nucleophilic
reagent combination TiCl4/allyltrimethylsilane.34 In several
cases this allylative endo cleavage resulted in complete stereo-
control.
There is continuing interest in the regioselective cleavage of
acetals and subsequent stereoselective substitution of one of
the C᎐O bonds by means of Lewis acid–nucleophile systems.1–10
Glycosides, constituting a large group of mixed acetals carrying
several stereogenic centres, are recognized as potent starting
materials for the synthesis of a wide variety of complicated
natural products.11,12 In most carbohydrate examples the
exocyclic C᎐O bond has been substituted giving rise to new
O-glycosides13,14 or C-glycosides,15 both with a high degree of
stereocontrol (Scheme 1). A manifestation of the importance
Previously, we reported stereoselective C-alkylations via endo
C᎐O bond cleavage of benzyl pentopyranosides by reaction
with AlMe3 or organotitanium reagents.35,36 This methodology
was used in the synthesis of two γ-hydroxyisoleucine stereo-
isomers, (2R,3R,4R) and (2S,3R,4R), as well as their corre-
sponding γ-lactones.37 In these examples the organoaluminium
or organotitanium reagents acted as both Lewis acids and
nucleophile carriers. A limitation with these reagents was the
difficulty in varying the nucleophilic part. Therefore the com-
bination of a Lewis acid separate from the nucleophilic reagent
would be advantageous. In particular, octahedral organo-
titanium chelates as intermediates were recently shown to influ-
ence strongly the stereoselectivity of addition to chiral α-alkoxy
carbonyls.38,39
exo -
cleavage
O
O
Nu
Nu
+ ROA
Nu
O
OR
A
A
O
O
Nu
OR
endo -
cleavage
OR
We here report on the use of various easily accessible organo-
metallic reagents based on zinc,40,41 aluminium,42,43 silicon or
titanium,44–47 in combinations with TiCl4 in the chelation-
controlled regio- and stereo-selective C-alkylation of pento-
furanosides. Since this methodology increases the number of
transferable organyl groups to C-1, the general synthetic
strategy of using readily available inexpensive carbohydrate
derivatives48 as building blocks in natural product synthesis is
further expanded.
A = Brønsted or Lewis acid
Scheme 1
of the exo-cleavage is clearly seen in the tremendous success of
oligosaccharide synthesis in glycoscience.16–19
About 50 years ago Lindberg showed that glycosides, when
treated with Lewis acids, underwent anomerization via endo-
cleavage followed by ring closure.20 The endo-cleavage mode has
also been debated in relation to enzymic cleavage of carbo-
hydrates.21–24 The use of carbohydrates as stereochemically pure
starting materials for organic synthesis would benefit from
general methods taking advantage of the endo-cleavage mode.
Cases of non-stereoselective endo-cleavage reactions have been
Results and discussion
The simple α/β-furanosides 1, 10 and 13 were subjected to the
endo C᎐O bond-cleavage reactions and the results are shown in
Table 1. Addition of AlMe3 to β-1 (Scheme 2) resulted in a 2:1
dr of products of 2 and 3 in a low yield of 31% (Entry 1). While
25
reported using trimethylsilyl cyanide (TMSCN)/SnCl2 or
RCu 26,27 applied to simple tetrahydrofuran (THF) and tetra-
hydropyran (THP) ethers. More elaborate glycosides were
cleaved by the use of Me2BBr.28–31
OMe OMe
To our knowledge, the first example of alkylative endo cleav-
age was reported by Kawana et al.32 who clearly pointed to the
possibility of chelation of the reagent, in this case MeMgX,
to the O-4/O-5 positions of methyl furanosides. The stereo-
selectivities were not always high but in one case a diastereo-
isomeric ratio (dr) of 87:13 was obtained. Stereoselective
introduction of the cyano group was achieved via endo-cleavage
by Hashimoto and Hayakawa on treating benzyl tri-O-benzyl-
MeO
O
Lewis Acid
OMe
MeO
R
OH
MeO
2/3 R = Me
4/5 R = Et
6/7 R = Ph
8/9 R = CN
α- or β-1
Scheme 2
J. Chem. Soc., Perkin Trans. 1, 1998
785

View Article Online
Table 1 Results of the reaction of glycosides 1, 10 and 13 with TiCl4 and various organometallic reagents
Reagents
Reaction
time/h
dr/Isolated
Entry
Substrate
(mol equiv.)a
Delay/mina
Temp./ЊCb
Products
yield (%)c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
β-1
AlMe3 (3)
MeTiCl3 (1)
MeTiCl3 (2)
48
33
22
3
2
3
2
3
8
33
12
12
12
12
12
5
5
12
7
reflux
25
25
2/3
2:1/31
/0
β-1
2/3
β-1
2/3
3:1/52
1:1/49
2:1/37
2:1/47
3:2/27
3:2/55
3:2/12
3:2/9
2:1/65
3:2/40
1:1/85
1:1/70
3:2/60
1:3/65
1:99/61
2:1/67
1:99/60
2:1/71
1:99/56
β-1
TiCl4 (1)/ZnMe2 (3)
TiCl4 (1)/AlMe3 (3)
TiCl4 (1)/AlMe3 (3)
TiCl4 (1)/ZnEt2 (3)
TiCl4 (1)/ZnEt2 (3)
TiCl4 (1)/ZnEt2 (3)
TiCl3(OPri) (1)/ZnEt2 (3)
TiCl4 (1)/AlEt3 (3)
MeTiCl3 (1)/ZnEt2 (3)
TiCl4 (1)/ZnPh2 (3)
TiCl4 (1)/Me3SiCN (3)
TiCl4 (1)/ZnEt2 (3)
TiCl4 (1)/AlMe3 (3)
TiCl4 (1)/ZnMe2 (3)
TiCl4 (1)/AlMe3 (3)
TiCl4 (1)/ZnMe2 (3)
TiCl4 (1)/AlMe3 (3)
TiCl4 (1)/ZnMe2 (3)
60
60
10
60
10
1
10
10
10
30
10
30
60
60
45
35
45
35
Ϫ72
2/3
β-1
Ϫ72
Ϫ72
Ϫ72
Ϫ72
Ϫ72
25
Ϫ42
Ϫ42
Ϫ72
Ϫ42
Ϫ42
Ϫ42d
Ϫ42d
Ϫ42
Ϫ42
Ϫ42
Ϫ42
2/3
β-1
2/3
β-1
4/5
β-1
4/5
β-1
4/5
β-1
4/5
β-1
4/5
β-1
4/5
β-1
6/7
β-1
8/9
α-1
4/5
α/β-10
α/β-10
α-13
α-13
β-13
β-13
11/12
11/12
14/15
14/15
14/15
14/15
12
7
a
The Lewis acid to the left was added to the substrate, followed by a waiting period as indicated before addition of the reagent to the right. b All
reactions were started at Ϫ72 ЊC and the temperature of the cooling bath was thereafter slowly raised to the indicated value ( 3 ЊC). c The
diastereomeric ratios (drs) were determined by GLC analyses. d The temperature of the solution was allowed to reach 5 ЊC before quenching.
Cl
Cl
Cl
Thus, ZnEt2 turned out to work well as an ethyl donor although
the diastereoselectivity of products 4/5 was quite low (3:2). The
complexation time (1, 10 or 60 min) had no significant influence
on the selectivity, but the highest yield (55%) was achieved with
a 10 min delay (Entries 7–9). A slight increase in both yield and
selectivity was noticed when AlEt3 was used as the alkyl donor
instead of ZnEt2 (Entries 8 and 11). The weaker Lewis acid
TiCl3(OPri) (a higher temperature was necessary) together
with ZnEt2 gave at best 9% yield of products 4/5 (Entry 10). The
α-anomer of substrate 1 reacted with TiCl4/ZnEt2 to give the
same ratio of products 4/5 as the β-anomer in a reasonable yield
(Entry 15). For an interpretation see below.
Cl
Me
Cl
Cl
Cl
Me
Me
Cl
Cl
Cl
Ti
Ti
Ti
Me
Me
O
O
O
Me
O
O
Cl
O
Cl
O
O
O
O
O
O
A
B
C
Me
Me
Me
Hf = +46.7 kcal mol–1
Hf = 0 kcal mol–1
Hf = +24.6 kcal mol–1
Fig. 1 Ab initio-calculated energies of the three most reasonable che-
lates between TiCl4 and β-1. 1 cal = 4.184 J.
The combination of TiCl4 and ZnR2 or AlR3 resembles the
Ziegler–Natta catalyst, which is formed by the couple TiCl4/
AlEt3. Here the alkylated titanium species is the actual coordin-
ating catalyst while AlEt3 serves to initiate the reaction by alkyl
transfer to titanium.50 In order to find out whether a similar
alkyl-group transfer between the organyl donor and TiIV
precedes the substitution reaction of the endo C᎐O bond,
compound β-1 was first treated with one mole equivalent of
MeTiCl3 (instead of TiCl4) before the addition of ZnEt2. The
product formed (4/5) originated exclusively from attack of an
ethyl group, showing the same diastereoselectivity as the com-
bination TiCl4/ZnEt2 (Entry 12). If a ligand exchange precedes
the substitution reaction then both methyl and ethyl attack
at C-1 would be expected. Since this is not the case, ligand
exchange seems not to have any influence on this reaction. Once
the titanium species has formed the chelate with the carbo-
hydrate there may not be any possibility for alkyl transfer. The
delay between the addition of Lewis acid and the organyl donor
had an influence on the yields as seen in Entries 7–9.
Both a phenyl group and a cyano group could be introduced
at C-1 of substrate β-1 by TiCl4/ZnPh2 and TiCl4/Me3SiCN,
respectively, in good yields, showing that the organyl part can
be groups other than simple alkyls (Entries 13 and 14).
For compound β-1 as a substrate a slight tendency of
increased selectivity was noticed in the series ZnMe2 <
ZnEt2 < AlMe3 = AlEt3 < MeTiCl3 (see Entries 4, 8, 5, 11 and
3, in that order). Thus, the larger reagents gave the higher selec-
tivities. An increase of the steric hindrance in the substrate, e.g.
by introducing a substituent at C-2, would perhaps work in the
same direction. Indeed, when the α/β-anomeric mixture of the
permethylated ribofuranoside 10 was treated with TiCl4/AlMe3
a stoichiometric amount of MeTiCl3, accessible by Zn–Ti
exchange by mixing TiCl4 and ZnMe2 in a 2:1 ratio,45 gave no
reaction at all with substrate β-1 (Entry 2), use of two mole
equivalents led to a 52% isolated yield of products 2 and 3 in
a 3:1 dr (Entry 3). This indicated that the first mole equivalent
of Lewis agent formed a chelate probably involving the ring
oxygen (O-5), thereby activating the carbohydrate for alkyl
donation at C-1 by the second mole equivalent of the Lewis
acid serving only as an organyl donor. Different chelates may be
formed due to the presence of four coordinating oxygens in the
substrate. However, single-point ab initio calculations (3-21G)49
on preoptimized geometries (PM3) of TiCl4-chelates with sub-
strate β-1 suggested that the most favourable chelation was at
O-4/O-5 (complex B, Fig. 1). The other two reasonable chelates
(A and C) were of considerably higher energy. No chelating
effect at all was found at O-3/O-4.
This reasoning led us to first let the carbohydrate and the
chelating Lewis acid (TiCl4) form the complex (such as B) before
the nucleophile carrier was introduced. Thus, treating substrate
β-1 first with one mole equivalent of TiCl4 at Ϫ72 ЊC, and then,
after a 60 min complexation time, with the nucleophile carrier
ZnMe2 resulted in a 49% yield of products 2/3 but without any
selectivity (Entry 4). No reaction occurred in the absence of
TiCl4. Changing the organyl donor from ZnMe2 to AlMe3
(Entry 5) gave the same selectivity (2:1) as when only AlMe3
was used. The yields were only 31 and 37%, respectively, in
these cases, but shortening the complexation time from 60 min
to 10 min increased the yield to 47% (Entry 6).
It was now interesting to see whether it was possible to use
reagents carrying other nucleophiles than the methyl group.
786
J. Chem. Soc., Perkin Trans. 1, 1998

View Article Online
OMe OMe
Me
2Zn
MeO
O
OMe
Lewis Acid
OMe
MeO
MeO
MeO
Me
O
H
OH
OH
OH
OMe
MeO
MeO
C
H
+
O
H
TiCl4
–
OMe
11/12
MeO
10 α/β mixture
Me3Al
OMe
OMe
Fig. 2 The tentative intermediate from compound 13 is preferentially
attacked at opposite faces depending on the reagent
MeO
O
Lewis Acid
OMe
OMe
MeO
chain systems would not apply. Besides, both the zinc and the
aluminium reagents should be able to form chelates that would
lead to the same or similar selectivity in the chelating Felkin–
Ahn model, but this was not the case for substrate 13. The zinc
reagent attacks the intermediate originating from compound 13
from above, while the aluminium reagent attacks preferentially
from below the average ring plane as depicted in Fig. 2.
The relative configuration of product 14 was determined by
GLC/MS comparison of its permethylated derivative with
authentic compound 18, prepared in four steps from chloride
1655 as outlined in Scheme 4. Then the absolute configurations
OMe
14
α- or β-13
+
OMe
OMe
OMe
15
Scheme 3
(Scheme 3), a higher diastereoselectivity (11/12 1:3) was
reached (Entry 16) as compared with β-1 (2:1, Entries 5 and 6).
Surprisingly, changing the methyl donor to ZnMe2 resulted
in excellent selectivity, 1:99 (Entry 17). This is not what we
expected since we expected the aluminium reagent to be larger
than the zinc reagent. However, the 2-OMe group constitutes an
extra Lewis-acid-coordination site, making it very difficult to
draw any simple conclusions at present since there exists no
investigation of the complexation situation of carbohydrates
carrying multiple Lewis-acid-coordination sites. Also xylo-
furanoside 13 (Scheme 3) gave an excellent selectivity (14/15
1:99) with ZnMe2, while application of AlMe3 gave not only a
lower selectivity but also the reversed diastereoselectivity (2:1
as compared to 1:99, Entries 20 and 21, respectively).
Cl
OMe
OMe
O
ii,iii,i
RO
OMe
MeO
OMe OMe
RO
OR
18 (enantiomer of methylated 14)
16 R = H
17 R = Me
i
Scheme 4 Reagents and conditions: i, NaH, MeI, THF, room temp.; ii,
aq. H2SO4, 100 ЊC; iii, LiAlH4, THF, reflux
of compounds 14 and 15 follow from the known configurations
of the starting materials. The configuration at C-5 of the indi-
vidual isomers of the other diastereomeric pairs (2/3, 3/4, 6/7,
8/9 and 11/12) could not be determined since the NMR data
were not sufficiently informative due to overlapping peaks, and
the attempted chromatographic separations were unsuccessful
except for compounds 11 and 12. Other protecting groups than
methyl will be needed to solve these problems, e.g. by allowing
mild removal and the synthesis of cyclic derivatives for struc-
ture determination, which will be reported in due course.
With a derivative similar to compound 13 Martin et al.34
found that the β-anomer performed well in their reactions,
while the α-anomer was essentially unreactive. Such a difference
between the anomers was not observed in our reactions; in the
case of compound 13 the α and β anomers gave the same main
products and diastereomeric ratios with both AlMe3 and
ZnMe2 (Entries 18–21). Also, reactions performed separately
with the α and β anomers of substrate 1 resulted in the same
product mixtures of compounds 4 and 5 (cf. Entries 8 and 11).
Thus, the reaction with one anomer proceeds mainly by reten-
tion while the reaction with the other one gives inversion of the
stereocenter at C-1, a pattern that also was observed in our
earlier work with benzyl pentopyranosides.35 This indicates that
a common intermediate is formed before the transfer of the
organic group to C-1.
Computations performed to clarify the origin of the selec-
tivity in the reactions of pentopyranosides implied that a
hydrogen-bonded, seven-membered-ring intermediate was
formed prior to methyl transfer.51,52 In the furanoside cases
1, 10 and 13 such hydrogen-bonded structures will be six-
membered (Fig. 2) and would orient the oxocarbenium unit in
the average plane of the ring system. The nucleophile would
then have to attack either from below or from above this plane
depending on the nature of the nucleophilic reagent and the
substituents on the substrate. Similar hydrogen bonding was
recently postulated by Corey et al. to explain the enantio-
selectivity in reactions between chiral boron Lewis acids and
aldehydes.53,54
Conclusions
The endo-C᎐O bond was cleaved and a new organic group was
introduced at C-1 of furanosides by application of a Lewis acid
together with a nucleophilic reagent. It was earlier shown that
this reaction was feasible for pyranosides. A preliminary mech-
anistic model, which provides a plausible explanation for the
stereoselectivities observed, was suggested. We assume that the
reaction proceeds via chelation between O-5/O-4 of the furano-
side and the Lewis acids (MeTiCl3, TiCl4 and TiCl3OPri)
followed by transfer of the organic groups by using ZnMe2,
ZnPh2, ZnEt2, AlMe3, AlEt3 and MeTiCl3 as nucleophile
carriers. In the case of compound β-1 the cyano group was
transferred by using Me3SiCN and TiCl4. The different selec-
tivities of Me2Zn and Me3Al indicated that perhaps the more
useful reagent control could be further developed.
Experimental
Column chromatographic separations were performed by using
Merck SiO2 60A (0.035–0.070 nm) silica gel with ethyl acetate–
heptane (E/H) mixtures as eluents. TLC analyses were made on
Merck SiO2 60 F254 precoated glass plates and the spots were
visualized by charring with a solution of phosphomolybdic
acid (25 g), Ce(SO4)2ؒ4H2O (10 g) conc. H2SO4 (60 ml) in water
(940 ml). NMR spectra were recorded in CDCl3 at 21 ЊC on a
One may speculate about the origins of the selectivity differ-
ences of the reagents but the situation is complicated due to the
presence of multiple coordination sites and the use of excess of
the reagents. The cyclic hydrogen-bonded model (Fig. 2) would
be favoured over the open-chain alternative due to electrostatic
attraction. Thus, the Cram and Felkin–Ahn models of open-
J. Chem. Soc., Perkin Trans. 1, 1998
787

View Article Online
Bruker DRX 400 NMR spectrometer [(1H) 400 MHz, CHCl3
δ 7.27; and (13C) 100 MHz, CHCl3 δC 77.2]. J-Values are in
Hz. GLC analyses were performed with a DBwax column
(J&W Scientific) capillary column (30 m; 0.25 mm i.d., 0.25
µm stationary phase). Mps were measured on a Gallenkamp
melting point apparatus and are uncorrected.
57.4, 55.9 (2 C), 38.3, 35.8 and 19.2 (2 C); HRMS (CI-CH4):
C9H21O4 (M ϩ H); Found: m/z, 193.1447. Calc.: m/z, 193.1440.
(2R,3S,5R)- and (2R,3S,5S)-1,3,5-Trimethoxyheptan-2-ol 4
and 5 (entries 7–12, 15). The general procedure was applied
to compound 1 and the ethyl-group donors shown in Table 1.
Column chromatography (E/H 1:10 → 1:1) of the crude
product gave an inseparable mixture of products 4/5 as a syrup:
ν(film)/cmϪ1 3440; δH 3.87 and 3.72 (each 1 H, 2 m), 3.54 (8 H,
m), 3.41, 3.39, 3.38, 3.33, 3.32 and 3.32 (each 3 H, 6 s), 2.57
(each 1 H, 2 s), 1.76 (4 H, m), 1.55 (4 H, m) and 0.88 and 0.87
(6 H, 2 t, J 7.5); δC 79.0, 78.8, 78.4, 77.8, 73.5, 71.5, 71.3, 59.2,
59.1, 58.3, 57.2, 56.2, 56.1, 35.0, 32.2, 25.8, 25.4, 8.9 and 8.7;
HRMS (CI-CH4): C10H23O4 (M ϩ H); Found: m/z, 207.1600.
Calc.: m/z, 207.1596.
(2R,3S,5R)- and (2R,3S,5S)-1,3,5-Trimethoxy-5-phenyl-
pentan-2-ol 6 and 7 (entry 13). The general procedure was
applied to compound 1 and ZnPh2. Column chromatography
(E/H 1:10 → 1:1) of the crude product gave an inseparable
mixture of products 5/6 as a syrup: ν(film)/cmϪ1 3420 and 3100;
δH 7.32 (each 5 H, 2 m), 4.39 and 4.33 (each 1 H, 2 dd), 3.88 and
3.82 and 7.32 (each 1 H, 2 m), 3.50 (6 H, m), 3.47, 3.40, 3.39,
3.33, 3.21 and 3.23 (each 3 H, 6 s), 2.63 (each 1 H, 2 s) and 2.18,
1.88, 1.84 and 1.75 (each 1 H, 4 m); δC 142.6, 142.0, 128.7 (2 C),
128.6 (2 C), 127.9, 127.7, 126.8 (2 C), 126.6 (2 C), 80.3, 80.0,
79.0, 78.7, 74.0, 73.6, 71.8, 71.3, 59.4, 59.3, 58.8, 57.4, 56.7,
56.6, 40.21 and 37.3; HRMS (CI-CH4): C14H23O4 (M ϩ H);
Found: m/z, 255.1595. Calc.: m/z, 255.1597.
(2R,4S,5R)- and (2S,4S,5R)-5-Hydroxy-2,4,6-trimethoxy-
hexanonitrile 8 and 9 (entry 14). The general procedure was
applied to compound 1 and Me3SiCN. Column chrom-
atography (E/H 1:10 → 1:1) of the crude product gave an
inseparable mixture of products 8/9 as a syrup: ν(film)/cmϪ1
3460 and 2064; δH 4.32 and 4.28 (each 1 H, 2 dd), 3.86 and 3.85
(each 1 H, 2 m), 3.50 (6 H, m), 3.515, 3.505, 3.42, 3.41, 3.405
and 3.40 (each 3 H, 6 s), 2.50 and 2.40 (each 1 H, 2 d) and 2.17
and 2.05 (each 2 H, 2 m); δC 118.6, 118.3, 78.8, 77.0, 73.3, 73.2,
70.9, 70.8, 68.1, 67.2, 59.4 (2 C), 58.7, 58.3 (2 C), 58.1, 34.7 and
33.8; HRMS (CI-CH4): C9H18NO4 (M ϩ H); Found: m/z,
204.1236. Calc.: m/z, 204.1236.
IR, mass and optical rotation data were recorded with a
Nicolet Impact 410 Infrared Spectrometer, a JEOL JMS-SX
102 mass spectrometer and a Perkin-Elmer 241 Polarimeter,
respectively. All reactions were carried out in oven-dried flasks
or vials equipped with rubber septa and under an argon atmos-
phere. The organometallic reagents were transferred by dried,
argon-flushed syringes and cannulas. Heptane was distilled
from sodium. CH2Cl2 was distilled from CaH2 and stored over
molecular sieves. Ethyl acetate was distilled immediately before
use. TiCl4 and Ti(OPri)4 were purchased from Aldrich and
Janssen respectively, and diluted with dry CH2Cl2 to 2.0 
before use. ZnMe2 (2.0  in toluene; Merck), ZnEt2 (1.0  in
hexane; Merck), ZnPh2 (Alfa), AlMe3 (2.0  in hexane;
Aldrich), AlEt3 (1.0  in hexane; Merck) and Me3SiCN (Fluka
AG) were all used without further purification. MeTiCl3,
Cl3Ti(OPri)45 and the substrates 1056 and 1357 were prepared
according to the respective literature procedure.
Methyl 2-deoxy-3,5-di-O-methyl-â-D-erythro-pentofuranoside
and its á-D-anomer á/â-1
A solution of ‘2-deoxy--ribose’ (5.0 g, 37.6 mmol) in 0.05%
HCl/MeOH (190 ml) was stirred at room temp. for 30 min,
followed by addition of Ag2CO3 (1.0 g, 3.6 mmol). The result-
ing solution was stirred for 5 min and was then filtered, evapor-
ated and finally co-evaporated with THF (3 × 10 ml), to afford
6.4 g of an oily mixture of the anomeric methyl glycosides. This
mixture was dissolved in THF (85 ml) and the solution was
cooled to 0 ЊC. NaH (60%; 3.3 g, 83 mmol) and MeI (11 g,
5.2 ml, 83 mmol) were alternately added in portions as the
temperature was slowly raised to ambient. The resulting
slurry was stirred for 60 h, whereafter MeOH (4 ml) was
added and the suspension was poured into ice–water (100 ml).
CH2Cl2 (100 ml) was now added and the aqueous phase was
extracted with CH2Cl2 (3 × 35 ml). The combined extracts
were washed with water (2 × 100 ml), dried (MgSO4), and
concentrated in vacuo. Column chromatography (E/H
3:7 → 2:3 → 1:1) afforded the β- (1.6 g, 24%) and α-
anomer (2.4 g, 36%) of compound 1. Spectral data were as
reported in the literature.58
(2R,3R,4S,5R)- and (2R,3R,4S,5S)-1,3,4,5-Tetramethoxy-
hexan-2-ol 11 and 12 (entries 16 and 17). The general procedure
was applied to compound 10 and the methyl-group donors
shown in Table 1. Column chromatography (E/H 1:10 →
1:1) of the crude product gave compounds 11 and 12 as syrups
(H/E 1:2).
Compound 11: Rf 0.40; [α]2D0 ϩ24 (c 2.57, CHCl3); ν(film)/cmϪ1
3430; δH 3.89 (1 H, m), 3.55 (3 H, m), 3.47, 3.42, 3.41 and 3.35
(each 3 H, 4 s), 3.35 (2 H, m), 3.27 (1 H, d, J 3.5) and 1.21 (3 H,
d, J 6.3); δC 84.0, 80.7, 76.9, 74.0, 71.0, 59.8, 59.6, 59.3, 56.7
and 15.0; HRMS (CI-CH4): C10H23O5 (M ϩ H); Found: m/z,
223.1544. Calc.: m/z, 223.1546.
General procedure for the opening and C-1 alkylation of
substrates 1, 10 and 13
A solution of TiCl4 (1.0 mol equiv.; 2.0  in CH2Cl2) was added
to a solution of the substrate (0.40 mmol; 0.2  in CH2Cl2) at
Ϫ72 ЊC. After a certain time delay (see Table 1) the alkyl donor
was added dropwise and the solution was stirred for the time
and at the temperature indicated in Table 1. The progress of the
reaction was monitored by TLC (E/H 2:1). The reaction mix-
ture was quenched by slow addition into a vigorously stirred
cold water–ethyl acetate mixture (30:70). Stirring was con-
tinued for 1 h. The aqueous phase was extracted with ethyl
acetate, and the combined organic phases were washed with
water, dried (Na2SO4), and gently concentrated at reduced
pressure (12 mmHg; water bath ≈35 ЊC).
(2R,3S,5R)- and (2R,3S,5S)-1,3,5-Trimethoxyhexan-2-ol 2
and 3 (entries 1–6). The general procedure was applied to
compound 1 and the methyl-group donors shown in Table 1.
Column chromatography (E/H 1:10 → 1:1) of the crude
product gave an inseparable mixture of products 2/3 as a syrup:
ν(film)/cmϪ1 3420; δH 3.86 and 3.74 (each 1 H, 2 m), 3.50 (6 H,
m), 3.42, 3.40, 3.34, 3.33, 3.32 and 3.32 (each 3 H, 6 s), 2.59
(each 1 H, 2 s), 1.88 and 1.85 (each 1 H, 2 m), 1.67 and 1.58
(each 2 H, 2 m) and 1.17 and 1.16 (each 3 H, 2 d, J 6.3 and 6.4);
δC 79.0, 78.7, 73.8, 73.5, 73.4, 73.0, 71.6, 71.3, 59.2, 59.1, 58.4,
Compound 12: Rf 0.35: [α]2D0 ϩ13 (c 0.58, CHCl3); ν(film)/cmϪ1
3430; δH 4.02 (1 H, m), 3.56 (3 H, m), 3.54, 3.44, 3.42 and 3.38
(each 3 H, 4 s), 3.41 (1 H, m), 3.27 (1 H, dd, J 5.7 and 5.8), 3.05
(1 H, d, J 3.0) and 1.26 (3 H, d, J 6.4); δC 84.8, 81.6, 76.9,
74.0, 71.1, 61.3, 59.3, 59.2, 56.9 and 15.7; HRMS (CI-CH4):
C10H23O5 (M ϩ H); Found: m/z, 223.1545.
(2R,3S,4S,5R)- and (2R,3S,4S,5S)-1,3,4,5-Tetramethoxy-
hexan-2-ol 14 and 15 (entries 18–21). The general procedure was
applied to anomeric glycosides α-13 and β-13, respectively,
and the alkyl donors shown in Table 1. Column chrom-
atography (E/H 1:10 → 1:1) of the crude product gave
compound 15 or
a mixture of diastereomers 14/15 as
syrups.
Compound 15: [α]2D0 Ϫ9 (c 2.30, CHCl3); ν(film)/cmϪ1 3450; δH
3.88 (1 H, m), 3.60–3.44 (4 H, m), 3.53, 3.51, 3.38 and 3.32
(each 3 H, 4 s), 3.30 (1 H, m), 2.79 (1 H, d, J 5.2) and 1.24 (3 H,
d, J 6.3); δC 84.7, 81.1, 75.9, 74.1, 69.9, 61.0, 60.9, 59.2, 56.6
and 14.8; HRMS (CI-CH4): C10H23O5 (M ϩ H); Found: m/z,
223.1544.
788
J. Chem. Soc., Perkin Trans. 1, 1998

View Article Online
Mixture 14/15: δH 3.88 (2 H, m), 3.90–3.44 (8 H, m), 3.54,
3.53, 3.52, 3.51, 3.39, 3.38, 3.32 and 3.30 (each 3 H, 8 s), 3.30
(2 H, m), 2.81 (1 H, d, J 4.7), 2.79 (1 H, d, J 5.2), 1.24 (3 H, d,
J 6.3) and 1.23 (3 H, d, J 6.3); δC 84.7, 84.3, 81.2, 81.1, 77.5,
75.9, 74.1, 74.0, 70.5, 69.9, 61.2, 61.1, 60.9, 60.6, 59.3, 59.2,
56.6, 56.5, 14.9 and 14.8; HRMS (CI-CH4): C10H23O5 (M ϩ H);
Found: m/z, 223.1545.
extracts sequentially with water (3 × 5 ml), saturated aq.
NH4Cl and brine, followed by drying (Na2SO4), and removal
of the solvent under reduced pressure, to give a syrup (20 mg).
GLC-MS analysis showed a 2:1 mixture of the corresponding
methyl ethers of isomers 14/15. Gas chromatographic co-
injection on the non-chiral phase DBwax column with com-
pound 18 showed a perfect overlap of the first peak of
the diastereomeric mixture with that of compound 18, which
verified that the methyl ether of stereoisomer 14 was the
enantiomer of compound 18.
Methyl 6-chloro-6-deoxy-2,3,4-tri-O-methyl-á-D-glucopyrano-
side59 17
NaH (60%; 600 mg, 14.5 mmol) and MeI (2.1 g, 14.4 mmol)
were added to a solution of methyl 6-chloro-6-deoxy-α--
glucopyranoside55 16 (700 mg, 2.9 mmol) in THF (15 ml). The
resulting mixture was stirred at room temp. for 48 h under
argon, whereafter water (50 ml) was added. Work-up was as
follows: extraction with CH2Cl2 (3 × 50 ml), washing of the
collected organic extracts sequentially with water (3 × 20 ml),
saturated aq. NH4Cl and brine, followed by drying (Na2SO4),
and removal of the solvent under reduced pressure. The residue
was subjected to column chromatography (H/E 3:1) to give title
compound 17 (660 mg, 89%) as a solid; mp 64 ЊC (lit.,59 62–
64 ЊC); [α]2D0 ϩ163 (c 2.28, CHCl3) [lit.,59 162.2 (c 1, CHCl3)]; δH
4.83 (1 H, d, J 3.6), 3.74 (3 H, m), 3.61, 3.57, 3.50 and 3.41 (each
3 H, 4 s), 3.5 (1 H, m) and 3.17 (2 H, m); δC 97.7, 83.5, 81.9,
80.3, 69.9, 61.1, 60.9, 59.2, 55.5 and 44.8; HRMS (FABϩ):
C10H20ClO5 (M ϩ H); Found: m/z, 255.0999. Calc.: m/z,
255.0999. The discrepancies between our NMR data and those
reported in the literature59 are probably due to the different
solvents used; CDCl3 versus CDCl3–benzene (6:1).
Acknowledgements
We thank the Swedish Natural Science Research Council, the
Knut and Alice Wallenberg Foundation, and the Crafoord
Foundation for financial support. We also thank Dr Ulf
Wellmar for a sample of compound 1.
References
1 A. Alexakis and P. Mangeney, Tetrahedron: Asymmetry, 1990, 1,
477.
2 I. Mori, K. Ishihara, L. A. Flippin, K. Nozaki, Y. Yamamoto,
P. A. Bartlett and C. H. Heathcock, J. Org. Chem., 1990, 55,
6107.
3 S. E. Denmark and N. G. Almstead, J. Org. Chem., 1991, 56, 6458.
4 S. E. Denmark and N. G. Almstead, J. Am. Chem. Soc., 1991, 113,
8089.
5 T. Sammakia and R. S. Smith, J. Am. Chem. Soc., 1994, 116, 7915.
6 T. Sammakia and R. S. Smith, J. Org. Chem., 1992, 57, 2997.
7 T. Sammakia and R. S. Smith, J. Am. Chem. Soc., 1992, 114, 10 998.
8 K. Ishihara, N. Hanaki and H. Yamamoto, J. Am. Chem. Soc., 1993,
115, 10 695.
(2S,3R,4R,5R)-1,2,3,4,5-Pentamethoxyhexane 18
9 T.-M. Yuan, S.-M. Yeh, Y.-T. Hsieh and T.-Y. Luh, J. Org. Chem.,
1994, 59, 8192.
Compound 17 (500 mg, 2.0 mmol) was added to aq. H2SO4 (25
ml; 2 ). The resulting mixture was stirred at 100 ЊC for 60 h
and was then cooled to ambient temperature whereupon the
mixture was neutralized by addition of saturated aq. NaHCO3
and worked up as follows: extraction with EtOAc (3 × 50 ml),
washing of the collected organic extracts sequentially with
water (20 ml) and brine, followed by drying (Na2SO4), and
removal of the solvent under reduced pressure. The residue was
dissolved in THF (15 ml) and then LiAlH4 (300 mg, 8.0 mmol)
was added. The resulting mixture was stirred at reflux for 24 h
under argon and then was cooled to 0 ЊC. The excess of LiAlH4
was decomposed by careful addition of water (20 ml) and the
mixture was worked up as follows: extraction with EtOAc
(4 × 50 ml), washing of the collected organic extracts sequen-
tially with water (20 ml) and brine, followed by drying
(Na2SO4), and removal of the solvent under reduced pressure.
The residue was dissolved in THF (10 ml), whereafter NaH
(60%; 320 mg, 8.0 mmol) and MeI (1.1 g, 8.0 mmol) were
added. The resulting mixture was stirred at room temp. for 48 h
under argon, whereafter water (30 ml) was added. Work-up was
as follows: extraction with CH2Cl2 (3 × 40 ml), washing of the
collected organic extracts sequentially with water (3 × 10 ml),
saturated aq. NH4Cl and brine, followed by drying (Na2SO4),
and removal of the solvent under reduced pressure. The residue
was subjected to column chromatography (H/E 2:1) to give
title compound 18 (197 mg, 42%) as a syrup; [α]2D0 ϩ6 (c 0.40,
CHCl3); δH 3.61 (2 H, m), 3.50–3.27 (7 H, m) and 3.50, 3.48,
3.46, 3.30 and 3.28 (each 3 H, 5 s); δC 82.8, 81.0, 80.6, 77.6,
72.4, 60.8, 60.5, 59.3, 58.9, 56.4 and 15.0; HRMS (CI-
CH4): C11H25O5 (M ϩ H); Found: m/z, 237.1719. Calc.: m/z,
237.1702.
10 M. Kinugasa, T. Harada, T. Egusa, K. Fujita and A. Oku, Bull.
Chem. Soc. Jpn., 1996, 69, 3639.
11 S. Hanessian, Total Synthesis of Natural Products: the Chiron
Approach, Pergamon Press, Oxford, 1983.
12 A. Vasella, in Modern Synthetic Methods 1980, ed. R. Scheffold,
Salle and Sauerländer, Frankfurt am Main, 1980, vol. 2, p. 173.
13 A. F. Bochkov and G. E. Zaikov, Chemistry of the O-Glycosidic
Bond: Formation and Cleavage, Pergamon Press, Oxford, 1979.
14 P. M. Collins and R. J. Ferrier, Monosaccharides: Their Chemistry
and Their Roles in Natural Products, Wiley, Chichester, 1996.
15 D. E. Levy and C. Tang, The Chemistry of C-Glycosides, Elsevier
Science, Oxford, 1st edn., 1995, vol. 13.
16 Glycoscience. Synthesis of Oligosaccharides and Glycoconjugates,
ed. H. Driguez and J. Thiem, Springer-Verlag, Berlin, 1997, vol. 186.
17 Glycoscience. Synthesis of Substrate Analogs and Mimetics, ed.
H. Driguez and J. Thiem, Springer-Verlag, Berlin, 1997, vol. 187.
18 F. Barresi and O. Hindsgaul, Mod. Synth. Methods, 1995, 7, 283.
19 G. Magnusson, A. Y. Chernyak, J. Kihlberg and L. O. Kononov, in
Neoglycoconjugates: Preparation and Applications, ed. Y. C. Lee and
R. T. Lee, Academic Press, San Diego, 1994, p. 53.
20 B. Lindberg, Acta Chem. Scand., 1949, 3, 1153.
21 M. L. Sinnott, Bioorg. Chem., 1993, 21, 34.
22 J. L. Liras and E. V. Anslyn, J. Am. Chem. Soc., 1994, 116, 2645.
23 R. B. Gupta and R. W. Franck, J. Am. Chem. Soc., 1987, 109,
6554.
24 R. W. Franck, Bioorg. Chem., 1992, 20, 77.
25 K. Utimoto, Y. Wakabayashi, T. Horiie, M. Inoue, Y. Shishiyama,
M. Obayashi and H. Nozaki, Tetrahedron, 1983, 39, 967.
26 J. F. Normant, A. Alexakis and P. Mangeney, Tetrahedron, 1989, 45,
507.
27 A. Ghribi, A. Alexakis and J. F. Normant, Tetrahedron Lett., 1984,
25, 3075.
28 H. E. Morton and Y. Guindon, J. Org. Chem., 1985, 50, 5379.
29 Y. Guindon and P. C. Anderson, Tetrahedron Lett., 1987, 28,
2485.
30 Y. Guindon, M. A. Bernstein and P. C. Anderson, Tetrahedron Lett.,
1987, 28, 2225.
Structure determination of compound 14
31 Y. Guindon, P. C. Anderson, C. Yoakim, Y. Girard, S. Berthiaume
and H. E. Morton, Pure Appl. Chem., 1988, 60, 1705.
32 M. Kawana, H. Kuzuhara and S. Emoto, Bull. Chem. Soc. Jpn.,
1981, 54, 1492.
33 H. Hashimoto and M. Hayakawa, Chem. Lett., 1989, 1881.
34 O. R. Martin, S. P. Rao, T.-F. Yang and F. Fotia, Synlett, 1991, 702.
35 T. Inghardt and T. Frejd, J. Org. Chem., 1989, 54, 5539.
The 2:1 mixture of diastereomers 14/15 (20 mg, 0.1 mmol) was
dissolved in THF (2 ml), and then NaH (60%; 10 mg, 0.2 mmol)
and MeI (30 mg, 0.2 mmol) were added. The resulting mixture
was stirred at room temp. for 48 h under argon, whereafter
water (5 ml) was added. Work-up was as follows: extraction
with CH2Cl2 (3 × 5 ml), washing of the collected organic
J. Chem. Soc., Perkin Trans. 1, 1998
789

View Article Online
36 R. Olsson, U. Berg and T. Frejd, Tetrahedron, submitted.
37 T. Inghardt, T. Frejd and G. Svensson, Tetrahedron, 1991, 47,
6469.
38 M. T. Reetz, B. Raguse and T. Seitz, Tetrahedron, 1993, 49, 8561.
39 V. Jonas, G. Frenking and M. T. Reetz, Organometallics, 1995, 14,
5316.
50 P. Cossee, Trans. Faraday Soc., 1962, 58, 1226.
51 U. Berg, R. Olsson and T. Frejd, presented at the 5th European
Symposium on Organic Reactivity, Santiago de Compostela, Spain,
1995, Abstract C1B.
52 R. Olsson, U. Berg and T. Frejd, Tetrahedron Lett., 1997, 38, 5701.
53 E. J. Corey and J. J. Rohde, Tetrahedron Lett., 1997, 38, 37.
54 E. J. Corey, J. J. Rohde, A. Fischer and M. D. Azimioara,
Tetrahedron Lett., 1997, 38, 33.
40 P. Knochel and R. D. Singer, Chem. Rev., 1993, 93, 2117.
41 P. Knochel, Chemtracts, 1995, 8, 205.
42 K. Mauruoka and H. Yamamoto, Tetrahedron, 1988, 44, 5001.
43 H. Yamamoto, in Organometallics in Synthesis, ed. M. Schlosser,
Wiley, Chichester, 1994, vol. 7, p. 509.
44 B. Weidmann and D. Seebach, Angew. Chem., Int. Ed. Engl., 1983,
22, 31.
45 M. T. Reetz, in Organometallics in Synthesis: A Manual, ed.
M. Schlosser, Wiley, Chichester, 1994, vol. 3, p. 195.
46 M. T. Reetz, J. Westermann, R. Steinbach, B. Wenderoth, R. Peter,
R. Ostarek and S. Maus, Chem. Ber., 1985, 118, 1421.
47 M. T. Reetz, Top. Curr. Chem., 1982, 106, 1.
48 M. Bols, Carbohydrate Building Blocks, Wiley, New York, 1996.
49 SPARTAN version 4.0 Wavefunction, Inc., 18401 Von Karman
Ave., suite 370 Irvine, CA 92715, USA.
55 R. L. Whistler and A. K. M. Anisuzzaman, Methods Carbohydr.
Chem., 1980, 8, 227.
56 G. R. Barker, J. Chem. Soc., 1948, 2035.
57 E. E. Percival and R. Zobrist, J. Chem. Soc., 1952, 4306.
58 K. Krohn and G. Börner, J. Org. Chem., 1991, 56, 6038.
59 A. Klemer and G. Mersman, Carbohydr. Res., 1972, 22, 425.
Paper 7/06166H
Received 22nd August 1997
Accepted 22nd October 1997
790
J. Chem. Soc., Perkin Trans. 1, 1998