The influence of neighboring group participation on the hydrolysis of 2-O-substituted methyl glucopyranosides

Article abstract of DOI:10.1021/ol202308y

Does neighboring group participation actually enhance the reactivity of the anomeric center when the participating group is inherently disarming? To investigate the influence of the neighboring group effect from a 2-O protective group on acidic glycoside hydrolysis, 10 methyl glucosides having different protective groups on O2 have been synthesized and a clear trend between anomeric configuration, participation of the protective group, and the rate of hydrolysis could be observed.

Full text of DOI:10.1021/ol202308y

ORGANIC
LETTERS
2011
Vol. 13, No. 22
5956–5959
The Influence of Neighboring Group
Participation on the Hydrolysis of
2-O-Substituted Methyl Glucopyranosides
Mads Heuckendorff, Christian M. Pedersen,* and Mikael Bols*
Universitetsparken 5, DK-2100 Copenhagen O, Denmark
bols@chem.ku.dk; cmp@chem.ku.dk
Received August 25, 2011
ABSTRACT
Does neighboring group participation actually enhance the reactivity of the anomeric center when the participating group is inherently disarming?
To investigate the influence of the neighboring group effect from a 2-O protective group on acidic glycoside hydrolysis, 10 methyl glucosides
having different protective groups on O2 have been synthesized and a clear trend between anomeric configuration, participation of the protective
group, and the rate of hydrolysis could be observed.
The influence of 2-O-protective groups in carbohydrates
on the stereochemical outcome of glycoside bond forma-
tion has been known for decades. Ester type protective
groups participate in the glycosylation reaction to form
2-O-acyl oxonium ions, which lead preferentially to the
trans coupling products, while ether type protective groups
cannot participate and result in low R/β selectivity. Yet
the precise rate increase obtained from such neighboring
group participation is much more qualitative. Obviously
some acceleration would be anticipated, or neighboring
group participation would not be a favorable path, but it
has neverbeen quantified. Incontrast Paulsen1 and Fraser-
Reid2 observed the generally enhanced reactivity of benzy-
lated vs acylated glycosyl donors and emphasized the im-
portance of the protective group on position 2 as the most
influential. It was therefore surprising when Demchenko,3
in contrast to the above, observed that perbenzylated
donors with a 2-O-benzoyl were significantly more reactive
than the fully benzylated counterpart. Similarly he noted
that the perbenzoylated 2-O-benzyl donor was less reactive
than the fully benzoylated donor. This was apparently a
case where neighboring group participation was enhancing
reactivity by overriding inductive effects. This was clarified
by Crich4 who demonstrated that anchimeric assistance
caused the 1,2 trans (β) 2-O-benzoyl to be more reactive
than the 1,2-trans-2-O-benzyl, while the 1,2 cis (R) 2-O-
benzoyl in fact was less reactive than the 1,2 cis (R) 2-O-
benzyl.
The significant rate accelerations observed in Demchenko’s
work prompted us to speculate if neighboring group parti-
citation could be used to facilitate glycoside hydrolysis as
well. Glycoside degradation methodologies are becoming
increasingly important with the growing need to efficiently
convert biomass. However, despite the many studies5 car-
ried out on hydrolysis of glycosides since the Fisher era only
very little quantitative information about anchimeric assis-
tance or neighboring group participation is available.
The inductive effect of a C2 substituent has of course been
studied and has a profound influence: The relative rate for
(1) Paulsen, H.; Richter, A.; Sinnwell, V.; Stenzel, W. Carbohydr.
Res. 1978, 64, 339–264.
(2) (a) Mootoo, D. R.; Konradson, P.; Udodong, U.; Fraser-Reid, B.
J. Am. Chem. Soc. 1988, 110, 5583. (b) Fraser-Reid, B.; Wu, Z.; Udodong,
U.; Ottoson, H. J. Org. Chem. 1990, 55, 6068. (c) Fraser-Reid, B.; Wu, Z.;
Andrews, C. W.; Skowrinski, E. J. Am. Chem. Soc. 1991, 113, 1434–1435.
(3) Kamat, M. N.; Demchenko, A. V. Org. Lett. 2005, 7, 3215–3218.
Premathilake, H. D.; Mydock, L. K.; Demchenko, A. V. J. Org. Chem.
2010, 75, 1095–1100.
(4) Crich, D.; Li, M. Org. Lett. 2007, 9, 4115–4118.
(5) Capon, B. Chem. Rev. 1969, 69, 407–498.
(6) Chemistry of the O-Glycosidic Bond: Formation & Cleavage;
Bochkov, A. F., Zaikov, G. E.; Pergamon Press: Oxford, 1979.
r
10.1021/ol202308y
Published on Web 10/18/2011
2011 American Chemical Society

according to the BellꢀEvansꢀPolanyi principle, lead to a
faster rate.6,7 The rate enhancement should be the same for
both anomers and depend on the ability of the neighboring
group to stabilize the oxacarbenium ion. Finally, if “back-
side attack” (anchimeric assistance) is at play, only the
β-anomer (1,2 trans) would be affected and hence be react-
ing significantly faster than the R-anomer (1,2 cis), wherethe
protective group on O2 cannot participate in a “back-side”
attack to push out the aglycon.8
In this work the goal was to see if and how acidic
glycoside hydrolysis was influenced by neighboring group
participation and therefore a series of methyl glucosides
were prepared with a 2-O-protective group with the ability
to perform anchimeric assistance. Solubility in water and
stability in acidic media were also necessary, and pivaloyl,
ethyl carbamoyl,9 and ethoxycarbonyl turned out to be
groups that met these requirements. The 2-O-mesyl and
2-O-methyl derivatives were selected for comparison as
two groups unable to perform neighboring group partici-
pation,10 but with widely different electronic effects.
Synthesis of model compounds was carried out using
selective protective group manipulations. Methyl R-D-glu-
copyranoside was 4,6-benzylidene protected to give the
known diol 1.11 Inspired by Jeanloz and Jeanloz,12 1 was
reacted withPivCltogivethe corresponding 2-O-protected
derivative in modest yield (24%) together with the 3-O-
derivatives (Scheme 1). Treatment of the diol with ethyli-
socyanate and pyridine only showed a minor conversion of
the starting material; this was overcome by using the CuCl
promoted reaction in DMF13 which gave the product in
32% yield. Debenzylidenation of the 2-O protected com-
pounds using Pd/C and hydrogen gave the model com-
pounds in quantitative yields (Scheme 3). Mesylation gave
an inseparable mixture of products, and to overcome
this the 3-OH was selectively benzylated using the method
developed by Hung and co-workers14 followed by mesyla-
tion or methylation and finally palladium-catalyzed hy-
drogenolysis to give the desired product 11 and 12 in an
overall good yield (66 and 92% respectively from the
mono-ol 4) (scheme 1). Reaction of 4 with ethyl chloro-
formate in dichloromethane and TMEDA as base gave 5,
which was hydrogenolyzed to give 8.15
Figure 1. Mechanism for acid-catalyzed hydrolysis of methyl
glucosides having different protective groups on O2. R being
H or alkyl; R0 a side chain on a carbonyl (ester, carbamate,
carbonate, etc.).
(8) Winstein, S.; Grunwald, E.; Buckles, R. E.; Hanson, C. J. Am.
Chem. Soc. 1948, 70, 816–821.
(9) Motokucho, S.; Sudo, A.; Endo, T. J. Polym. Sci., Part A: Polym.
the deoxy is 2500 and 2-chloro 0.042 as compared with the
parent 2-OH sugar.6 The general considerations are as
follows:6 First, electron-withdrawing groups on O2 reduce
the nucleophilicity of O1 and O5 and thereby slow the first
step (Figure 1a). Second, the intermediate conjugated acid
is destabilized by electron-withdrawing groups and hence
decomposes faster to the oxacarbenium ion, as compared
with the unprotected or alkylated equivalent. Under nor-
mal conditions the second step is the rate-determining step.
Third, the oxacarbenium ion may be resonance stabilized
by a participating group on O2 (Figure 1b); this can,
Chem. 2007, 45, 4459–4464.
(10) There is plenty of evidence that sulfonate is nonparticipating;
see: Srivastava, V. K.; Schuerch, C. Carbohydr. Res. 1980, 79, C13–C16.
Srivastava, V. K.; Schuerch, C. J. Org. Chem. 1981, 46, 1121–1126.
Crich, D.; Picione, J. Org. Lett. 2003, 5, 781–784. Crich, D.; Hutton,
T. K.; Banerjee, A.; Jayalath, P.; Picione, J. Tetrahedron: Assymmetry
2005, 16, 105–119. Abdel-Rahman, A. A. ꢀH.; Jonke, S.; Ashry, E. S. H.
E.; Schmidt, R. R. Angew. Chem., Int. Ed. 2002, 41, 2972–2974.
(11) Commercially avaliable.
(12) Jeanloz, R. W.; Jeanloz, D. A. J. Am. Chem. Soc. 1957, 79, 2579–
2583.
(13) Duggan, M. E; Imagire, J. S. Synthesis 1989, 131–132.
(14) Wang, C.; Lee, J.; Luo, S.; Fan, H.; Pai, C.; Yang, W.; Lu, L.;
Hung, S. Angew. Chem., Int. Ed. 2002, 41, 2360–2362.
(15) The methyl 2-O-methoxycarbonyl-R/β-D-glucopyranoside was
prepared as well, but methanol formed from hydrolysis of the methyl
carbonate complicated the rate measurements. See Supporting
Information.
€
(7) Bruckner, R. Reaktionsmechanismen: Organische Reaktionen,
Stereochemie, Moderne Synthesemethoden 3. Auflage; Elsevier GmbH:
M€unchen, 2004.
Org. Lett., Vol. 13, No. 22, 2011
5957

Scheme 1. Synthesis of 2-O-protected Methyl-R-D-glucosides
Scheme 2. Synthesis of 2-O-Protected Methyl β-D-Glucopyra-
nosides
glucose could be observed as an intermediate in the
hydrolysis of the pivalates. From the ethyl carbonates only
glucose could be observed. From 12 and 23, 2-O-mesyl
glucose was the final product, but the reaction was so
slow that no complete conversion of starting material was
observed after many days.
When comparing the anomers a clear trend is observed;
the β-anomer is 1.5 to3 times faster thanthe corresponding
R. Such difference in reactivity between the anomers is
well-known; itisusuallyinthe range 1.5ꢀ5 withβ being the
more reactive,6,18 and can to a large extent be explained by
the anomeric effect making the R-anomer more stable in
the ground state.
This difference in rate coefficient between the anomers is
essentially independent of the 2-O protective group, and
the small difference between the nonparticipating and
the participating groups suggests that “back-side attack”
anchimeric assistance plays a negligible role in the rate
enhancement of the 1,2-trans configurated methyl gluco-
sides. This is in line with the observations by Paulsen and
Meyberg,19 who observed that 1,2-trans acetates in cyclo-
hexane do not participate in a “back-side” anchimeric
assistance20 due to their equatorialꢀequatorial relation-
ship; it should be emphasized that this study was carried
out under water-free conditions.
Another remarkable observation is the small difference
between “arming” nonparticipating groups and the “dis-
arming” carbonyl-containing participating groups, where
the difference is within a factor of 5; far from the examples
of anchimeric assistance in the literature, where a factor of
The β-model compounds were prepared from the
commercially available methyl β-D-glucopyranoside,
which was 4,6-benzylidene protected followed by selective
3-O-benzylation via a tin-acetal intermediate to give the
mono-ol 1316 ready for further functionalization. Com-
pounds 15ꢀ19 were obtained using standard conditions
with an excess of reagents (Scheme 2). Excess ethylisocya-
nate, however, resulted in the side product 25, and there-
fore only 1 equiv of ethylisocyanate was used. Reaction of
the substrate with ethyl chloroformate in pyridine did not
go to completion. Instead, TMEDA was found superior
as base for the synthesis of 16 from the diol 14. Finally
palladium-mediated removal of the benzylidene and ben-
zyl groups gave the model compounds 20ꢀ24 in excellent
yield.
With all model compounds in hand hydrolysis rates
could be determined by following the conversion using
NMR spectroscopy. The reactions were carried out in D2O
containing DCl (2.0 M), and the development of the meth-
anol peak was followed in the initial part of the reaction
(see Supporting Information). By plotting the integral vs
time a rate was determined for the hydrolysis of each of the
model compounds (Table 1).
The hydrolysis products of complete hydrolysis were,
however, also determined.17 The products 2-O-methyl and
2-O-ethylcarbamoylglucosewereobtainedfrom10, 11, 22,
or 24, which were isolated in 30% and 22% yield, respec-
tively (see Supporting Information). From 8, 9, 20, and 21,
the end hydrolysis product was glucose, but 2-O-pivaloyl
(18) Ex. Lemieux in situ anomerization procedure: Lemieux, R. U.;
Hendriks, K. B.; Stick, R. V.; James, K. J. Am. Chem. Soc. 1975, 4056–
4062.
(19) Paulsen, H.; Meyborg, H. Tetrahedron Lett. 1972, 38, 3973–
3976.
(20) Winstein, S.; Buckles, R. E. J. Am. Chem. Soc. 1942, 64, 2780–
2787.
(16) Murphy, P. V.; O’Brien, J. L.; Smith, A. B., III. Carbohydr. Res.
2001, 334, 327–335.
(17) Defined as when starting material was essentially gone.
5958
Org. Lett., Vol. 13, No. 22, 2011

The acyloxonium ion that is generated from the carbamate
is resonance stabilized and therefore most stable. Pivaloyl
is second in stability because of the electron donation from
the tert-butyl substituent. The carbonate is less nucleophi-
lic and cannot stabilize a positive charge as well. This
results in a small deactivation due to the inductive effect
overriding neighboring group participation. The mesylates
are, asexpected, the least reactivetowardacidic hydrolysis,
due to the intense inductive destabilization of the inter-
mediate oxacarbenium ion. Methylated and nonfunctio-
nalized O-2 are of about the same reactivity, with the latter
being slightly more reactive due to the more electron-
donating capacity of H compared with a methyl group
(electronegativity of 2.176 vs 2.472 on the Pauling scale).21
No correlation between the rate coefficients and Hammett
parameters or field effects (F)22 (Table 1) was observed,
which shows that inductive effects are not the only factor
involved.23 To obtain an estimate of the degree of neigh-
boring group participation24 we calculated the presumed
rate of the acyl and carbamoyl derivatives if only field
effects played a role and compared it with the experimental
values.25 This showed that these derivatives hydrolyze
2ꢀ45 times more rapidly than predicted from the F value
(Table 1) and that the neighboring group effect has this
magnitude.
Table 1. Rate Constants and Relative Rates of Hydrolysis of
2-O-Substitued Glucopyranosides in 2 M DCl at 60 °C
From this study we conclude that participating groups
indeed enhance the hydrolysis rate of methyl glucosides,
but not due to a SN2 type back-side assisted push out of
methanol, but rather by stabilization of the intermediate
developing positivechargewhere the BellꢀEvansꢀPolanyi
principle is operating. The magnitude of this stabilization
is not large (2ꢀ45 times) and is mostly absorbed by the
inherent electron-withdrawing effect of the neighboring
group. The difference in rate between anomers is not
caused by neighboring group effects but mainly by the
greater stability of the R- over β-anomer due to the
anomeric effect.
a NB = neighboring group. Rate increase estimated as a result of
neighboring group particitation. This value was determined by using the
Hammett and Swain and Lupton equations (log(k/ko) = fF) to calculate
a rate solely based on induction (field) effects to see how much higher the
experimental value were.25
Acknowledgment. We thank FTP for supporting this
work.
2000 is a classical example of the rate enhancement caused
by neighboring group participation from an ester.8 The
relative rates are picturing the stability of the formed cyclic
intermediate, i.e. by the BellꢀEvansꢀPolanyi principle.
Supporting Information Available. Experimental de-
tails for preparation of new compounds, NMR spectra,
and kinetics plots. This material is available free of charge
via the Internet at http://pubs.acs.org.
(21) Inamoto, N.; Masuda, S. Chem. Lett. 1982, 1003–1006.
(22) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165–195.
(23) Similar conclusions were drawn from the work on C5 substitu-
ents: Timell, T. E.; Enterman, W.; Spencer, F.; Soltes, E. J. Can. J. Chem.
1965, 43, 2296–2305.
(24) Crich, D.; Hu, T.; Cai, F. J. Org. Chem. 2008, 73, 8942–8953.
(25) This was done by using the Hammett and the Swain and Lupton
equations: log k/ko = f F. The parameter f was obtained from the data of
the 2-OH and 2-O-mesylate and was ꢀ2.9 for the R-series and ꢀ2.7 for β.
3
Org. Lett., Vol. 13, No. 22, 2011
5959