Direct glycosylation of unprotected and unactivated carbohydrates under mild conditions

Article abstract of DOI:10.1021/ol203329u

Ligand exchange acetalization of acetals in the presence of catalytic amounts of mandelic acid and titanium tert-butoxide is reported. This transformation is successfully extended to glycosylation of unprotected and unactivated pentoses. Even unreactive pentoses such as d-arabinose or d-lyxose can be transformed by this new methodology into corresponding isopropyl glycosides.

Full text of DOI:10.1021/ol203329u

ORGANIC
LETTERS
2012
Vol. 14, No. 3
792–795
Direct Glycosylation of Unprotected and
Unactivated Carbohydrates under Mild
Conditions
Matthias Pfaffe and Rainer Mahrwald*
Institute of Chemistry, Humboldt-University, Brook-Taylor Str. 2, 12489 Berlin,
Germany
rainer.mahrwld@rz.hu-berlin.de
Received December 13, 2011
ABSTRACT
Ligand exchange acetalization of acetals in the presence of catalytic amounts of mandelic acid and titanium tert-butoxide is reported. This
transformation is successfully extended to glycosylation of unprotected and unactivated pentoses. Even unreactive pentoses such as
D-arabinose or D-lyxose can be transformed by this new methodology into corresponding isopropyl glycosides.
Glycoconjugates or glycosides are very important ex-
amples of biomolecules. Due to different arranged and
configured hydroxyl groups of carbohydrates a general
applicable chemical synthesis to glycosides of all carbohy-
drates does not exist. Therefore the syntheses of glycosides
have been of great interest for a long time. In order to
achieve the required stereo-, chemo-, and regioselectivity
many different and complex methods of glycosidation
have been developed. Frequently, these synthetic maneu-
versare associatedwithanextensivehandlingof protecting
groups and additional activation of the anomeric car-
bon atom.1 In order to avoid these multistep approaches
to defined glycosides, several attempts have been made
to increase and to utilize reactivity differences between
the hemiacetal function and the remaining unprotected
hydroxyl groups of carbohydrates.2Also, methods have
been developed to deploy unprotected and unacti-
vated carbohydrates in direct glycosidation processes.
These transformations are associated with high reaction
(4) (a) Schmidt, R. R. Pure Appl. Chem. 1989, 61, 1257. (b) Schmidt,
R. R. In Modern Methods in Carbohydrate Synthesis; Khan, S. H., Neill,
R. A., Eds.; Harwood Academic Publishers: 1996. (c) Vauzeilles, B.; Dausse,
B.; Palmier, S.; Beau, J.-M. Tetrahedron Lett. 2001, 42, 7567.
(5) (a) Bishop, C. T.; Cooper, F. P. Can. J. Chem. 1963, 41, 2743.
(b) Fischer, E. Chem. Berichte 1893, 26, 2400. (c) Fischer, E. Chem.
Berichte 1895, 28, 1145. (d) Capon, B. Chem. Rev. 1969, 69. (e) Izumi,
M.; Fukase, K.; Kusumoto, S. Biosci. Biotechnol. Biochem. 2002, 66,
211. (f) Lee, R. T.; Lee, Y. C. Carbohydr. Res. 1974, 37, 193. (g) Wu, J.;
Serianni, A. S. Carbohydr. Res. 1991, 210, 51. (h) Wessel, H. P. Carbo-
hydr. Chem. 1988, 7, 263. (i) Mowery, D. F., Jr. J. Phys. Chem. 1974, 78,
1918. (k) Simon, E.; Cook, K.; Pritchard, M. R.; Stripe, W.; Bruch, M.;
Bendinskas, K. J. Chem. Educ. 2010, 87, 739. (l) Damez, C.; Bouquillon,
ꢀ
S.; Harakat, D.; Henin, F.; Muzart, J.; Pezron, I.; Komunjer, L.
Carbohydr. Res. 2007, 342, 154. (m) Evans, M. E.; Angyal, S. J.
Carbohydr. Res. 1972, 25, 43. (n) Angyal, S. J.; Bodkin, C. L.; Mills,
J. A.; Pojer, P. M. Aust. J. Chem. 1977, 30, 1259. (o) Angyal, S. J.;
Bodkin, C. L.; Parrish, F. W. Aust. J. Chem. 1975, 28, 1541.
(1) For current and comprehensive overviews in this field, see:
(a) Toshima, K.; Sasaki, K. In Comprehensive Glycoscience; Kamerling,
J. P., Ed.; Elsevier: 2007; Vol. 1, pp 261À310. (b) Brito-Arias, M. Synthesis
and Characterization of Glycosides; Springer: 2007. (c) Demchenko, A. V.
Handbook of Chemical Glycosylation; WILEY-VCH: Weinheim, 2008.
(2) (a) Hannesian, S.; Lou, B. Chem. Rev. 2000, 100, 4443.
(b) Bernardes, G. J. L.; Gamblin, D. P.; Davis, B. G. Angew. Chem.,
Int. Ed. 2006, 45, 4007.
ꢁ
(6) (a) Ferrieres, V.; Bertho, J.-N.; Plusquellec, D. Tetrahedron Lett.
ꢁ
1995, 36, 2749. (b) Bertho, J.-N.; Ferrieres, V.; Plusquellec, D. J. Chem.
Soc., Chem. Commun. 1995, 1391. (c) Velty, R.; Benvegnu, T.; Gelin, M.;
Privat, E.; Plusquellec, D. Carbohydr. Res. 1997, 299, 7. (d) Regeling, H.;
Zwanenburg, B.; Chittenden, G. J. F. Carbohydr. Res. 1998, 314, 267.
ꢁ
(3) (a) Gorityala, B. K.; Ma, J.; Pasunooti, K. K.; Cai, S.; Liu, X.-W.
ꢀ
(e) Ferrieres, V.; Benvegnu, T.; Lefeuvre, M.; Plusquellec, D.; Mack-
Green Chem. 2011, 13, 573. (b) Munoz, F. J.; Andre, S.; Gabius, H.-J.;
enzie, G.; Watson, M. J.; Haley, J. A.; Goodby, J. W.; Pindak, R.;
ꢁ
Sinisterra, J. V.; Hernaiz, M. J.; Linhardt, R. J. Green Chem. 2009, 11,
373. (c) Park, T.-J.; Weiwer, M.; Yuan, X.; Baytas, S. N.; Munoz, E. M.;
Murugesan, S.; Linhardt, R. J. Carbohydr. Res. 2007, 342, 614.
(d) Guchhait, G.; Misra, A. Q. K. Catal. Commun. 2011, 14, 52.
(e) Mamidyqala, S. K.; Finn, M. G. J. Org. Chem. 2009, 74, 8417.
Durbin, M. K. J. Chem. Soc., Perkin Trans. 2 1999, 951. (f) Ferrieres, V.;
Bertho, J.-N.; Plusquellec, D. Carbohydr. Res. 1998, 311, 25. (g) Blixt,
O.; Allin, K.; Pereira, L.; Datta, A.; Paulson, J. C. J. Am. Chem. Soc.
ꢀ
2002, 124. (h) Auge, J.; Sizun, G. Green Chem. 2009, 11, 1179. (i) Joniak,
D.; Polakova, M. J. Serb. Chem. Soc. 2001, 66, 81.
r
10.1021/ol203329u
Published on Web 01/20/2012
2012 American Chemical Society

temperatures3 and/or the deployment of strong bases,4
strong acids,5 or Lewis acids6 as catalysts. During our
ongoing studies on the application of ligand-exchange
mediated CÀC bond formation processes, we observed
substantial amounts of products derived from acetaliza-
tion. These reactions were performed in the presence of
titanium(IV) alkoxides and R-hydroxy acids.7 By optimiz-
ing our findings, we were able to achieve acetalization of
hemiacetals in the tetrahydrofuran and tetrahydropyran
series, under neutral reaction conditions (Scheme 1).
mentioned in Figure 1. The highest yields were obtained
by deployment of 2-deoxy-D-ribose (3a: 61%) and D-ribose
(4a: 56%). The lowest yields were detected for D-lyxose and
D-arabinose (6a; 7a < 10%). The yields can be increased
by longer reaction times. Isopropyl glycosides of D-xylose,
D-lyxose, and D-arabinose were isolated after 12 days at
room temperature, in approximately 30% yield (5a: 28%,
6a: 23%, 7a: 20%) (green line, Figure 1). Under these
conditions the exclusive formation of isopropyl furano-
sides is observed (kinetic control). Also, substantial amounts
of isopropyl pyranosides could not be detected after 12 days
at room temperature.
Scheme 1. Acetalization of Hemiacetals
Table 1. Yields and Anomeric Ratios of Isopropyl Furanosides
for D-Pentosesa
yields (%)
R/β ratio
entry
compd
2d
5d
12d
(%)
Only by deployment of titanium(IV) alkoxides and
R-hydroxy acids was acetalization of hemiacetals observed.
Corresponding isopropyl acetals 1 and 2 were isolated in
quantitative yields in both runs (n = 1 or 2) at room
temperature within 24 h.
1
2
3
4
5
3a
4a
5a
6a
7a
61
56
11
5
76
78
13
9
62
73
28
23
20
50/50
7/93
50/50
75/25
35/65
2
5
Encouraged by these results we tested this catalytic
system in the glycosylation of unprotected carbohydrates.
For this purpose, D-ribose was reacted with isopropanol
under the following reaction conditions. When titanium-
(IV) tert-butoxide (10 mol %) and D-mandelic acid
(50 mol %) were employed at room temperature the
corresponding isopropyl-D-riboside 4a was isolated in
good yield (Scheme 2).
a Reaction conditions: 50 mol % D-mandelic acid, 10 mol %
Ti(OtBu)4, iPrOH, rt. Deoxyribose: 4 mol % D-mandelic acid, 10 mol %
Ti(OtBu)4, iPrOH, rt.
These findings contrast with results obtained by the
Fischer- or other acid-mediated glycosidation methods.
Different ratios of pyranoid/furanoid glycosides were
detected under Fischer conditions depending on reaction
times.8 The following different anomeric ratios (R/β) were
detected for the isolated isopropyl furanosides under our
new conditions for D-pentoses (cf. Table 1).
Scheme 2. Direct Glycosylation of D-Ribose
This strong substrate selectivity is not consistent with
results obtained by Fischer glycosylation with MeOH.9
Both yield and anomeric ratios of glycosides formedbyour
method are dictated by the configuration of the hydroxy
groups of D-pentoses deployed.
In order to demonstrate the scope of this direct glyco-
sylation, several functionalized alcohols were reacted with
unprotected D-ribose under these conditions. Again, only
furanoid ribosides were observed after 2 days. A preference
Investigations revealed that primary as well as second-
ary alcohols can be successfully applied in the reactions
withD-ribose. Alsotert-butanol reactsunderthese reaction
conditions but in a lower yield (cf. Scheme 4). A ratio of 5:1
(D-mandelic acid/Ti(OtBu)4) and a ratio of 1:2 for 2-deoxy-
D-ribose was identified as optimal for glycosylation of
D-pentoses with isopropanol. Preformed titanium(IV)
alkoxide/D-mandelic acid complexes did not lead to glyco-
sylation products under these reaction conditions. Our
optimized reaction conditions were applied to glycosyla-
tion reactions with isopropanol for the D-pentoses
(8) Sanki, A. K.; Boucau, J.; Srivastava, P.; Adams, S. S.; Ronning,
D. R.; Sucheck, S. J. Bioorg. Med. Chem. 2008, 16, 5672.
(9) (a) Wu, J.; Serianni, A. S. Carbohydr. Res. 1991, 210, 51.
(b) Schulze, O.; Voss, J.; Adiwidjaja, G. Synlett 2001, 229.
(10) Lithium salts are known to promote and to direct glycosylation
processes. The mode of action is still under discussion. (a) Worm-
Leonhard, K.; Larsen, K.; Jensen, K. J. J. Carbohydr. Chem. 2007, 26,
349. (b) Vauzeilles, B.; Dausse, B.; Palmier, S.; Beau, J.-M. Tetrahedron
Lett. 2001, 42, 7567. (c) Schene, H.; Waldmann, H. Synthesis 1999, 1411.
(d) Lubineau, A.; Drouillat, B. J. Carbohydr. Chem. 1997, 16, 1179.
(e) Boehm, G.; Waldmann, H. Tetrahedron Lett. 1995, 36, 3843.
(f) Waldmann, H.; Boehm, G.; Schmid, U.; Roettele, H. Angew. Chem.,
Int. Ed. Engl. 1994, 33, 1994. (g) Mukaiyama, T.; Matsubara, K. Chem.
Lett. 1992, 21, 1041. (h) Uchiro, H.; Mukaiyama, T. Chem. Lett. 1996,
25, 271. (i) Mukaiyama, T.; Shimpuku, T.; Takashima, T.; Kobayashi,
S. Chem. Lett. 1989, 145. (k) Mukaiyama, T.; Takashima, T.; Katsur-
ada, M.; Aizawa, H. Chem. Lett. 1991, 533. (l) Mukaiyama, T.; Wariishi,
K.; Furuya, M.; Kobayashi, S. Chem. Lett. 1989, 1277.
(7) (a) Mahrwald, R.; Ziemer, B. Tetrahedron Lett. 2002, 43, 4459.
(b) Mahrwald, R. Org. Lett. 2000, 2, 4011. (c) Mahrwald, R. J. Prakt.
Chem. 1994, 336, 361.
Org. Lett., Vol. 14, No. 3, 2012
793

Figure 1. Yields of isopropyl glycosides for D-pentoses as a
function of time. Lines are drawn to guide the eye. Reaction
conditions: 50 mol % D-mandelic acid, 10 mol % Ti(OtBu)4,
iPrOH, rt. Deoxyribose: 4 mol % D-mandelic acid, 10 mol %
Ti(OtBu)4, iPrOH, rt.
Figure 2. Yields of isopropyl glycosides for D-pentoses as a
function of time in the presence of lithium bromideb and for
comparison without lithium bromide.a Lines are drawn to
guide the eye. bReaction conditions: 50 mol % D-mandelic acid,
10 mol % Ti(OtBu)4, 1 equiv of LiBr, iPrOH, rt. Deoxyribose:
4 mol % D-mandelic acid, 10 mol % Ti(OtBu)4, iPrOH, rt.
for the formation of β-configured ribosides was observed.
β-Configured anomers were solely detected in the diol
series (4dÀ4f, cf. Scheme 3). The results demonstrate that
under our new reaction conditions both the aglycon as well
as the unprotected carbohydrate have a great influence on
the anomeric ratio of glycosides formed.
Several attemptsweremade todecreasethelong reaction
times. To this end, different additives were tested. As a
result of this optimization, lithium salts, especially lithium
bromide, proved to be beneficial for these transforma-
tions.10 The results are depicted in Figure 2.
conditions (3a: 95%, 4a: 90%, 5a: 73%, 6a: 65%, 7a: 64%,
12 days, rt). Furanoid glycosides were the major products
after 2 days at room temperature (80À100%). Pyranoid
glycosides were detected after longer reaction times (cf.
Table 1 in the Supporting Information). To demonstrate
this property of lithium bromide, several alcohols were
reacted under the described reaction conditions for 2 days
(cf. Scheme 4). These results clearly emphasize the yield-
increasing feature of lithium bromide.
A marked improvement in yield was found between
Figures 1 and 2 for all D-pentoses used under the LiBr
Scheme 4. Comparison of Glycosylation
Scheme 3. Glycosidation of D-Ribose under Mild Reaction
Conditionsa
a In the presence of lithium bromide. bIn the absence of lithium
bromide .c>85% furanoid structure. dR/β ratio.
We next studied the effect of different solvents on the
glycosylation outcome. We were keen to reduce the
amount of aglycon and thus enable the application of solid
a R/β ratio.
794
Org. Lett., Vol. 14, No. 3, 2012

aglycons. Reactions were not observed in the presence of
DMSO or DMF.11
Scheme 6. Direct Glycosydation of D-Ribose and D-Xylose in
the Presence of Solvents and Lithium Bromide
Scheme 5. Direct Glycoslyation of D-Ribose in Different Solventsa
a R/β ratio.
conditions, the exclusive formation of furanoid glyco-
sides is observed. Also, the use of lithium bromide as an
additive strongly increases the yields of products ob-
served. When lithium bromide is employed in polar
solvents, thermodynamic control of glycosidation is
observed. As a result, pyranosides were almost exclu-
sively formed with high degrees of selectivity. Thus, an
optional access to furanoid or pyranoid glycosides is
provided by the choice of reaction conditions (solvents
and catalysts component). Further extension of these
findings to hexoses is underway.
The favored solvent for these reactions turned out to be
propylene carbonate. Comparable yields were obtained
when used with propylene carbonate and reduced amounts
of isopropanol (5 equiv, Scheme 5). Furanoid glycosides
were detected exclusively with similar anomeric ratios to
neat isopropanol.
To study the influence of lithium bromide under these
reaction conditions, the optimal solvents were used for
a survey of the glycosylation reactions of D-ribose and
D-xylose (cf. Scheme 6; Table 2 in the Supporting
Information). The formation of thermodynamically con-
trolled pyranoside was observed. These findings stand in
contrast to results observed in the absence of solvents (see
kinetic control, Scheme 4).12
Acknowledgment. The authors thank Bayer-Schering
Pharma AG, Bayer Services GmbH, BASF AG, and Sasol
GmbH for financial support. This work was supported by
Evonik Industries (M.P.).
In summary, we have developed a new method for
direct glycosylation of unprotected and unactivated
carbohydrates under mild reaction conditions. This
new transformation utilizes titanium(IV) tert-butoxide
and D-mandelic acid. By deployment of these reaction
Supporting Information Available. NMR data for all of
the synthesized compounds and full characterization of
novel compounds. This material is available free of
charge via the Internet at http://pubs.acs.org
(11) This is possibly due to the chelating effect of these solvents.
(12) For investigations in thermodynamically and kinetically con-
trolled glycosylation reactions, see: Bishop, C. T.; Coopper, F. P. Can. J.
Chem. 1962, 40, 224.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 3, 2012
795