Zinc-mediated fragmentation of methyl 6-deoxy-6-iodo-hexopyranosides

Article abstract of DOI:10.1007/s007060200021

An improved procedure was developed for the zinc-mediated fragmentation of protected and unprotected methyl 6-deoxy-6-iodo-hexopyranosides. The method employs sonication of the iodoglycoside with zinc dust in a THF/H₂O mixture.

Full text of DOI:10.1007/s007060200021

Monatshefte fuÈr Chemie 133, 467±472 ꢀ2002)
Zinc-Mediated Fragmentation of Methyl
6-Deoxy-6-iodo-hexopyranosides
Ã
Philip R. Skaanderup, Lene Hyldtoft, and Robert Madsen
Department of Chemistry, Technical University of Denmark, DK-2800 Lyngby, Denmark
Summary. An improved procedure was developed for the zinc-mediated fragmentation of protected
and unprotected methyl 6-deoxy-6-iodo-hexopyranosides. The method employs sonication of the
iodoglycoside with zinc dust in a THF=H2O mixture.
Keywords. Fragmentation; Glycosides; Halogens; Sonication; Zinc.
Introduction
The zinc-mediated reductive fragmentation of methyl 6-deoxy-6-halo-hexopyrano-
sides has been described for the ®rst time by Bernet and Vasella in 1979 [1]. The
5,6-dideoxy-hex-5-enoses thus generated are useful chiral synthons and have found
many applications in carbohydrate chemistry. Most notable is the synthesis of ®ve-
membered carbocycles by 1,3-dipolar cycloadditions [1,2] or radical cyclizations
[3]. In addition, the fragmentation has also been used for the synthesis of azasugars
[4] and complex natural products [5]. The corresponding methyl 5-deoxy-5-halo-
pentofuranosides also undergo this fragmentation [6].
However, in spite of the many applications, the reductive fragmentation of !-
haloglycosides continues to cause signi®cant experimental dif®culties. The original
procedure calls for re¯ux with acid-washed zinc powder in aqueous alcohols [1].
This is still the most widely used protocol, but the yields are often moderate due to
the instability of 5,6-dideoxy-hex-5-enoses at re¯ux and the formation of several
byproducts including methyl 6-deoxy-hexopyranosides ꢀC-6 reduction), 2,5,6-
trideoxy-hex-5-enoses ꢀC-2 deoxygenation), and 5,6-dideoxy-hex-5-enose dialkyl
acetals ꢀacetal formation) [1,3]. Recently, the addition of vitamin B12 has been
shown to give some improvement, allowing the fragmentation to be carried out at
room temperature [7]. Using a better leaving group at the anomericcenter, like the
p-methoxyphenyl group, also gives higher yields [8], but this is not a convenient
method for general use. Another procedure for the fragmentation employs the more
reactive zinc-silver graphite in THF at room temperature [9]. Yields are here
generally higher, and less byproducts are formed than when re¯uxing with zinc
powder in aqueous alcohols.
Ã
Corresponding author. E-mail: rm@kemi.dtu.dk

468
P. R. Skaanderup et al.
We have recently developed a domino reaction where methyl !-haloglycosides
are fragmented with zincand the resulting aldehydes are subsequently alkylated in
a Barbier reaction [10]. Zinc serves a dual function in this reaction. First, it
mediates the fragmentation of the haloglycoside; secondly, it activates an alkyl
halide for the Barbier alkylation. During this work we had to ®nd an ef®cient and
high-yielding protocol for the fragmentation of methyl !-haloglycosides. Herein,
we report a full account on these studies which resulted in the development of an
improved fragmentation procedure.
Results and Discussion
We ®rst studied the fragmentation with benzyl-protected methyl glucopyranosides
1 and 2 ꢀScheme 1). Re¯uxing bromide 1 with acid-washed zinc powder in
aqueous 2-propanol gave enal 3 in about 65% yield in accordance with previous
observations [1]. The major byproducts formed were 6-deoxy glucoside 4 together
with a small amount of acetal 5. In order to avoid these byproducts, more reactive
forms of zincwere investigated. Unfortunately, bromide 1 and iodide 2 did not
react with Rieke zinc, zinc graphite, or zinc-silver graphite in THF at room
temperature or at re¯ux. Presumably, this lack of reactivity is due to the steric bulk
of 1 and 2, as sterically demanding protecting groups have previously been shown
to inhibit the fragmentation [9].
Addition of an alcohol has been shown to enhance the rate of the fragmentation
with these reactive forms of zinc [9b]. Indeed, treatment of iodide 2 with Rieke zinc
in a THF=methanol mixture at re¯ux afforded complete fragmentation.
Unfortunately, the product 3 had epimerized at C-2, maybe due to traces of
potassium in the Rieke zinc. In order to avoid possible epimerizations, acetic acid
was then added instead of methanol. However, this led to reduction of the iodide to
afford deoxyglycoside 4. Virtually complete reduction to 4 was also observed upon
re¯uxing 2 with zinc powder and acetic acid in aqueous 2-propanol. Bromide 1
gave similar results; however, it is less reactive than iodide 2, and the latter was
therefore chosen for general use.
As it appeared important to have an alcohol and=or an acid present during the
fragmentation, several experiments were then conducted with a Lewis acid.
Treatment of 2 with Rieke zincand TMSCl in THF=2-propanol caused almost
complete conversion to acetal 5. Interestingly, no reduction of the iodide was
observed under these conditions. To prevent acetal formation the reaction was
repeated without 2-propanol. Gratifyingly, this now gave very clean formation of
the desired enal 3. The fragmentation was most conveniently carried out with zinc
Scheme 1

Fragmentation of Halohexopyranosides
469
Table 1. Zinc-mediated fragmentation of methyl 6-deoxy-6-iodo-hexopyranosides; conditions: A: Zn,
TMSCl, THF, sonication, 40ꢀC, 4±6 h; B: Zn, THF:H2O ˆ 9:1, sonication, 40ꢀC, 1 h; C: Zn, THF:
H2O ˆ 4:1, sonication, 40ꢀC, 1 h
Entry
Iodoglycoside
Conditions
Product
Yield=%
1
2
2
2
A
B
3
3
85
94
3
4
A
B
86
89
5
6
A
B
85
93
7
8
9
C
C
C
94
94
95
powder, as Rieke zincdid not provide better results and is more tedious to prepare.
Re¯uxing 2 with zincpowder and TMSCl worked well when starting with less than
1 g of 2. On larger scale, however, the reaction proved less ef®cient, presumably
due to precipitation of zinc salts onto the metallic zinc surface. Therefore, more
reproducible conditions were sought and sonication was investigated. To the best of
our knowledge this fragmentation has never been achieved under sonication
conditions prior to our ®rst report in 1999 [10a]. It turned out that sonication at
40ꢀC caused clean and reproducible conversion into enal 3 ꢀTable 1, entry 1). Even
on a 10 g scale this procedure proved very reliable. A more ®nely grinded zinc
ꢀ<10 mm) was used for these sonication experiments. Under re¯ux conditions this
type of zincwas less ef®cient due to aggregation, and a more coarse powder was

470
P. R. Skaanderup et al.
used for these experiments. The general applicability of the sonication protocol is
further illustrated by the fragmentation of protected mannoside 6 and galactoside 7
ꢀTable 1, entries 3 and 5).
Good yields were obtained under these conditions, but it still bothered us that
some material was always lost. This was presumably due to slight decomposition
of the enals 3, 11, and 12, as no other products were isolated from these reactions.
The problem seemed to be the acidic conditions and the 4±6 h of reaction time.
TMSCl was therefore replaced by H2O hoping that a continuous wash of the zinc
surface during the reaction would improve the reactivity. This turned out to work
very well. The fragmentation was ®nished in less than 1 h giving slightly better
yields of enals 3, 11, and 12 ꢀTable 1, entries 2, 4, and 6).
The fragmentation in a THF=H2O mixture also proved well suited for
unprotected iodoglycosides 8±10 giving rise to furanoses 13±15 ꢀTable 1, entries
7±9). The very high yields of all six fragmentation products 3 and 11±15 are
noteworthy and constitute a testament to the ef®ciency of these reaction conditions.
When unprotected iodoglucoside 8 was fragmented in anhydrous THF containing
TMSCl the product was not furanose 13, but instead the corresponding methyl
furanoside of 13 was formed. Apparently, a Fischer glycosylation with the
liberated methanol takes place under these conditions.
Zinc appears to be the metal of choice for the reductive fragmentation of !-
haloglycosides. Magnesium has been shown to give Wurtz-type homocoupling,
presumably by a radical mechanism [11]. In our previously developed domino
reaction we were able to replace zinc with indium [10b]. However, indium is a less
reactive metal. In fact, when we sonicated unprotected glucoside 8 with indium in a
THF=H2O mixture only a sluggish reaction was observed. Adding TMSCl to the
mixture gave full conversion of 8 in 12 h to give furanose 13 in 93% yield.
Protected glucoside 2 could not be fragmented with indium, and only decom-
position was observed on prolonged treatment. No fragmentation was also the
result when 8 was sonicated with tin, antimony, or bismuth.
In conclusion, we have developed an improved procedure for the zinc-mediated
fragmentation of methyl !-iodoglycosides. The fragmentation is most ef®ciently
carried out with zinc dust under sonication in a THF=H2O mixture. For protected
iodoglycosides a procedure under anhydrous conditions has also been developed
employing TMSCl. The latter method is useful when further synthetictransforma-
tions do not allow for the presence of H2O [10b].
Experimental
Zincwas activated and dried immediately before use [10b]. Zincpowder ꢀFluka 96453) was used for
fragmentations at re¯ux, whereas zinc dust ꢀAldrich 20.998-8) was used under sonication. All
sonications were performed in a Branson 1210 ultrasonic bath. Thin-layer chromatography ꢀTLC)
was performed on aluminum plates precoated with silica gel ꢀMerck 1.05554). Compounds were
visualized by heating after dipping in a solution of 2.5 g CeꢀSO4)2 and 6.25 g ꢀNH4)6Mo7O24 in
250 cm3 10% aqueous H2SO4. Flash chromatography was performed using silica gel 60. Reverse
phase ¯ash chromatography was carried out with silica gel 60-50 C18 ꢀMacherey-Nagel). NMR
spectra were recorded on a Varian Unity Inova 500 spectrometer. Microanalyses were conducted by
the Department of Chemistry at the University of Copenhagen ꢀdata were in accordance with
calculated values).

Fragmentation of Halohexopyranosides
471
General procedure for the fragmentation of benzyl
protected iodoglycosides "conditions A and B)
To a solution of 500 mg ꢀ0.87 mmol) methyl 2,3,4-tri-O-benzyl-6-deoxy-6-iodo-ꢀ-D-glycopyrano-
side in 20 cm3 THF, 569 mg ꢀ8.7 mmol) pre-activated zinc and 0.11 cm3 ꢀ0.87 mmol) TMSC1
ꢀcondition A) or 2.2 cm3 H2O ꢀcondition B) were added. The resulting suspension was sonicated
at 40ꢀC until TLC showed full conversion ꢀA: 4±6 h, B: 1 h). Then, 30 cm3 diethyl ether and
10 cm3 H2O were added. The resulting biphasicsystem was ®ltered, and the layers were
3
separated. The organicphase was washed with 10 cm H2O, 10 cm3 brine, dried ꢀK2CO3), ®ltered,
and concentrated to give a bright yellow syrup which was puri®ed by ¯ash chromatography
ꢀhexane:ethyl acetate ˆ 5:1).
2,3,4-Tri-O-benzyl-5,6-dideoxy-D-xylo-hex-5-enose ꢀ3; C27H28O4)
Colorless syrup; spectroscopic data in accordance with literature [7].
2,3,4-Tri-O-benzyl-5,6-dideoxy-D-lyxo-hex-5-enose ꢀ11; C27H28O4)
Colorless syrup; spectroscopic data in accordance with literature [7].
2,3,4-Tri-O-benzyl-5,6-dideoxy-L-arabino-hex-5-enose ꢀ12; C27H28O4)
1
Colorless syrup; Rf ˆ 0.45 ꢀhexane:ethyl acetate ˆ 5:1); H NMR ꢀCDCl3, 500 MHz): ꢁ ˆ 9.62 ꢀd,
J ˆ 1.7 Hz, 1 H), 7.40±7.20 ꢀm, 15H), 5.90 ꢀddd, J ˆ 17.5, 10.2, 7.7 Hz, 1 H), 5.44 ꢀdd, J ˆ 17.5,
1.7 Hz, 1 H), 5.42 ꢀdd, J ˆ 10.2, 1.7 Hz, 1 H), 4.65 ꢀd, J ˆ 12.0 Hz, 1 H), 4.61 ꢀd, J ˆ 11.1 Hz, 1 H),
4.56 ꢀd, J ˆ 11.5 Hz, 1 H), 4.55 ꢀd, J ˆ 11.5 Hz, 1 H), 4.50 ꢀd, J ˆ 11.1 Hz, 1 H), 4.20 ꢀd, J ˆ 11.5 Hz,
1 H), 4.11±4.08 ꢀm, 2H), 3.89 ꢀdd, J ˆ 7.7, 3.8 Hz, 1 H) ppm; 13C NMR ꢀCDCl3, 125 MHz):
ꢁ ˆ 202.43, 137.84, 137.50, 137.18, 135.47, 128.35, 128.23, 128.16, 128.07, 127.96, 127.73, 127.70,
127.52, 120.16, 83.92, 81.20, 79.23, 74.24, 73.39, 70.06 ppm.
General procedure for fragmentation of unprotected iodoglycosides "condition C)
To a solution of 200 mg ꢀ0.66 mmol) methyl 6-deoxy-6-iodo-ꢀ-D-glycopyranoside in 12 cm3 THF
and 3 cm3 H2O, 430 mg ꢀ6.6 mmol) pre-activated zinc were added, and the resulting suspension
was sonicated at 40ꢀC for 1 h. The reaction mixture was then ®ltered and concentrated to
approximately 5 cm3. The resulting solution was puri®ed by reverse phase ¯ash chromatography
eluting with H2O.
5,6-Dideoxy-D-xylo-hex-5-enofuranose ꢀ13; C6H10O4)
Colorless syrup; Rf ˆ 0.21 ꢀCHCl3:MeOH ˆ 9:1); spectroscopic data in accordance with literature
[12].
5,6-Dideoxy-D-lyxo-hex-5-enofuranose ꢀ14; C6H10O4)
Colorless syrup; Rf ˆ 0.24 ꢀCHCl3:MeOH ˆ 9:1); 4:1 mixture of ꢀ:ꢂ anomers; 1H NMR ꢀD2O,
500 MHz): ꢀ-anomer: ꢁ ˆ 5.86 ꢀddd, J ˆ 17.5, 10.5, 7.0 Hz, 1 H), 5.37±5.29 ꢀm, 2H), 5.25 ꢀd,
J ˆ 4.5 Hz, 1 H), 4.19±4.13 ꢀm, 2H), 4.09 ꢀt, J ˆ 4.5 Hz, 1 H) ppm; ꢂ-anomer: ꢁ ˆ 5.23 ꢀd, J ˆ 5.0 Hz,
1 H) ppm; 13C NMR ꢀD2O, 125 MHz): ꢀ-anomer: ꢁ ˆ 133.32, 120.18, 101.34, 82.03, 78.25,
73.50 ppm, ꢂ-anomer: ꢁ ˆ 134.18, 119.90, 96.12, 82.21, 72.45, 72.13 ppm.

472
P. R. Skaanderup et al.: Fragmentation of Halohexopyranosides
5,6-Dideoxy-L-arabino-hex-5-enofuranose ꢀ15: C6H10O4)
Colorless syrup; Rf ˆ 0.18 ꢀCHCl3:MeOH ˆ 9:1); 1:1 mixture of ꢀ:ꢂ anomers; 1H NMR ꢀD2O,
500 MHz, both anomers): ꢁ ˆ 5.91±5.80 ꢀm, 2H), 5.38±5.23 ꢀm, 5H), 5.20 ꢀd, J ˆ 3.0 Hz, 1 H) 4.38±
4.35 ꢀm, 1 H), 4.07±3.91 ꢀm, 4H), 3.84±3.82 ꢀm, 1 H) ppm; 13C NMR ꢀD2O, 125 MHz, both
anomers): ꢁ ˆ 136.92, 135.49, 120.18, 119.63, 101.44, 95.45, 84.23, 82.57, 82.18, 80.03, 78.20,
76.38 ppm.
Acknowledgments
We thank the Danish Natural Science Research Council for ®nancial support.
References
[1] Bernet B, Vasella A ꢀ1979) Helv Chim Acta 62: 1990
[2] a) Ferrier RJ, Furneaux RH, Prasit P, Tyler PC, Brown KL, Gainsford GJ, Diehl JW ꢀ1983)
J Chem SocPerkin Trans 1, 1621; b) Nakata M, Akazawa S, Kitamura S, Tatsuta K ꢀ1991)
Tetrahedron Lett 32: 5363; c) Drans®eld PJ, Moutel S, Shipman M, Sik V ꢀ1999) J Chem Soc
Perkin Trans 1, 3349
ꢀ
ꢀ
[3] Desire J, Prandi J ꢀ2000) Eur J Org Chem 3075
[4] Bernotas RC, Papandreou G, Urbach J, Ganem B ꢀ1990) Tetrahedron Lett 31: 3393
[5] a) Villalobos A, Danishefsky SJ ꢀ1990) J Org Chem 55: 2776; b) Furstner A, Baumgartner J
ꢀ1993) Tetrahedron 49: 8541
[6] Kobori Y, Myles DC, Whitesides GM ꢀ1992) J Org Chem 57: 5899

[7] Kleban M, Kautz U, Greul J, Hilgers P, Kugler R, Dong H-Q, Jager V ꢀ2000) Synthesis 1027
[8] Ovaa H, Codee JDC, Lastdrager B, Overkleeft HS, van der Marel GA, van Boom JH ꢀ1999)
ꢀ
Tetrahedron Lett 40: 5063
[9] a) Furstner A, Weidmann H ꢀ1990) J Org Chem 55: 1363; b) Furstner A, Jumbam D, TeslicJ,
Weidmann H ꢀ1991) J Org Chem 56: 2213; c) Furstner A, Baumgartner J, Jumbam DN ꢀ1993)
J Chem SocPerkin Trans 1, 131
[10] a) Hyldtoft L, Poulsen CS, Madsen R ꢀ1999) Chem Commun 2101; b) Hyldtoft L, Madsen R
ꢀ2000) J Am Chem Soc 122: 8444

[11] a) Stout EI, Doane WM, Trinkus VC ꢀ1976) Carbohydr Res 50: 282; b) Furstner A, Weidmann H
ꢀ1989) J Org Chem 54: 2307
[12] de Raadt A, Ebner M, Ekhart CW, Fechter M, Lechner A, Strobl M, Stutz AE ꢀ1994) Catalysis

Today 22: 549
Received September 9, 2001. Accepted October 30, 2001