Mechanism of Hydrolysis of Isopropenyl Glucopyranosides. Spectroscopic Evidence Concerning the Greater Reactivity of the α Anomer

Article abstract of DOI:10.1021/jo9721249

The mechanism of hydrolysis of isopropenyl α- and β-glucopyranosides (1 and 2, respectively) has been studied by temperature-dependent reaction kinetics, stereochemical analysis of products by ¹H NMR, solvent ¹⁸O- and ²H-labeling experiments, kinetic solvent deuterium isotope effects, and α-values for general acid catalysis. Compounds 1 and 2 are the first vinyl acetals to be shown to undergo hydrolysis exclusively by cleavage of the vinyl ether and not the acetal C-O bond. While both glucopyranosides undergo irreversible, rate-limiting C-protonation, 1 hydrolyzes approximately four times faster than 2 and has an enthalpy of activation that is 5.8 kcal mol^-1 lower than that of 2, suggesting that 1 is kinetically more basic than 2. Spectroscopic evidence indicates that conjugation of the glycosidic oxygen with the alkenyl double bond is greater in 1 than in 2, probably because of steric or conformational constraints.

Full text of DOI:10.1021/jo9721249

J . Org. Chem. 1998, 63, 833-840
833
Mech a n ism of Hyd r olysis of Isop r op en yl Glu cop yr a n osid es.
Sp ectr oscop ic Evid en ce Con cer n in g th e Gr ea ter Rea ctivity of th e
r An om er †
H. Keith Chenault* and Laura F. Chafin
Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716
Received November 19, 1997
The mechanism of hydrolysis of isopropenyl R- and â-glucopyranosides (1 and 2, respectively) has
been studied by temperature-dependent reaction kinetics, stereochemical analysis of products by
2
1H NMR, solvent 18O- and H-labeling experiments, kinetic solvent deuterium isotope effects, and
R-values for general acid catalysis. Compounds 1 and 2 are the first vinyl acetals to be shown to
undergo hydrolysis exclusively by cleavage of the vinyl ether and not the acetal C-O bond. While
both glucopyranosides undergo irreversible, rate-limiting C-protonation, 1 hydrolyzes approximately
four times faster than 2 and has an enthalpy of activation that is 5.8 kcal mol-1 lower than that
of 2, suggesting that 1 is kinetically more basic than 2. Spectroscopic evidence indicates that
conjugation of the glycosidic oxygen with the alkenyl double bond is greater in 1 than in 2, probably
because of steric or conformational constraints.
In tr od u ction
As part of an investigation of isopropenyl glycosides
as potential substrates for enzyme-catalyzed1 and elec-
trophile-mediated2 transglycosylation reactions, we have
examined the mechanism and rates of hydrolysis of
isopropenyl R-D-glucopyranoside (1) and its â anomer (2).3
hydrogen atom abstraction,10 and substitution by Grig-
nard reagents.11 In the case of hydrolysis, the differences
in rates of reaction are often relatively small. Alkyl
â-glycosides hydrolyze 1.3-3.2 times faster than their
corresponding R anomers, and aryl R-glycopyranosides
hydrolyze 4.1-8.7 times faster than their corresponding
â-glycopyranosides.12 Considerable controversy sur-
rounds the mechanistic explanations for these small rate
differences.13-17 It is unclear whether the primary influ-
ence of nOfσ*CO orbital interactions is on ground states
or transition states of the reactions. Additionally, the
influence of the anomeric effect on the equilibrium
populations of various protonated and twist-boat inter-
mediates adds another layer of uncertainty to the prob-
lem.9,18
Before our work, it was not apparent whether hydroly-
sis of these vinyl glycosides would proceed with cleavage
of the glycosidic (acetal) or vinyl ether C-O bond.
Propenyl acetals, such as 3, hydrolyze with exclusive
cleavage of the acetal C-O bond.4 On the other hand,
hydrolysis of 2-alkoxy-2,3-dihydropyrans, such as 4,
involves competing cleavage of both acetal and vinyl ether
moieties.5,6 The rate of acetal hydrolysis in 4 may be
attenuated by intramolecular collapse of the oxocarbe-
nium ion with the enol leaving group.
Another point to be studied was whether the anomeric
configuration of 1 or 2 influences the rate or mechanism
of hydrolysis. The R and â anomers of alkyl glycopyra-
nosides and related compounds are known to differ in
the rates at which they undergo hydrolysis,7 oxidation,8,9
(7) Feather, M. S.; Harris, J . F. J . Org. Chem. 1965, 30, 153-157.
(8) (a) Isbell, H. S. J . Res. Nat. Bureau Standards 1962, 66, 233-
239. (b) Angyal, S. J .; J ames, K. Aust. J . Chem. 1970, 23, 1209-1221.
(c) Deslongchamps, P.; Moreau, C. Can. J . Chem. 1971, 49, 2465-2467.
(d) Deslongchamps, P.; Moreau, C.; Fre´hel, D.; Atlani, P. Can. J . Chem.
1972, 50, 3402-3404.
(9) Deslongchamps, P. Pure Appl. Chem. 1993, 65, 1161-1178.
(10) Remy, G.; Cottier, L.; Descotes, G. Can. J . Chem. 1980, 58,
2660-2665.
(11) Eliel, E. L.; Nader, F. W. J . Am. Chem. Soc. 1970, 92, 584-
† This paper is dedicated to Professor Melvyn D. Schiavelli on the
occasion of his 55th birthday.
(1) Castro, A. Ph.D. Dissertation, University of Delaware, 1995,
Chapter 1.
590.
(12) Capon, B. Chem. Rev. (Washington, D.C.) 1969, 69, 407-498.
(13) The Anomeric Effect and Associated Stereoelectronic Effects;
Thatcher, G. R. J ., Ed.; ACS Symposium Series 539; American
Chemical Society: Washington, DC, 1993.
(14) Deslongchamps, P. Stereoelectronic Effects in Organic Chem-
istry; Pergamon: New York, 1983.
(2) (a) Chenault, H. K.; Castro, A. Tetrahedron Lett. 1994, 35, 9145-
9148. (b) Chenault, H. K.; Castro, A.; Chafin, L. F.; Yang, J . J . Org.
Chem. 1996, 61, 5024-5031.
(3) Chenault, H. K.; Chafin, L. F. J . Org. Chem. 1994, 59, 6167-
(15) Kirby, A. J . The Anomeric Effect and Related Stereoelectronic
Effects at Oxygen; Springer-Verlag: New York, 1983.
(16) (a) Sinnott, M. L. Adv. Phys. Org. Chem. 1988, 24, 113-204.
(b) Sinnott, M. L. Chem. Rev. (Washington, D.C.) 1990, 90, 1171-1202.
(17) Ratcliffe, A. J .; Mootoo, D. R.; Andrews, C. W.; Fraser-Reid, B.
J . Am. Chem. Soc. 1989, 111, 7661-7662.
6169.
(4) Chiang, Y.; Chwang, W. K.; Kresge, A. J .; Yin, Y. J . Am. Chem.
Soc. 1989, 111, 7185-7190.
(5) Kankaanpera¨, A. Acta Chem. Scand. 1969, 23, 1465-1470.
(6) Burt, R. A.; Chiang, Y.; Kresge, A. J . Can. J . Chem. 1980, 58,
2199-2202.
(18) Deslongchamps, P. In ref 13, pp 26-54.
S0022-3263(97)02124-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/22/1998

834 J . Org. Chem., Vol. 63, No. 3, 1998
Chenault and Chafin
∆H‡ ) 18.39 ( 0.27 kcal mol-1 and ∆S‡ ) -10.0 ( 0.9
eu. Thus, the greater reactivity of 1 is due to a signifi-
cantly lower enthalpy of activation that is only partially
offset by a more negative entropy of activation.
Vin yl Eth er C-O Bon d Clea va ge. Chemoselective
hydrolysis of the vinyl ether of 1 and 2 was first
1
suggested by the observation made by H NMR that 1
hydrolyzes exclusively to give R-glucopyranose as the
initially formed product while 2 hydrolyzes exclusively
to give â-glucopyranose as the initially formed product.3
This result would be hard to explain if hydrolysis were
proceeding by cleavage of the glycosidic (acetal) C-O
bond. Reaction of 1 and 2 in methanol-d4 containing 1
mM DCl resulted in the addition of methanol-d4 across
the alkenyl ether double bond to form glucosyl methyl
acetonides 5 and 6, respectively, as products and not
methyl glucopyranosides. Again, this result is consistent
with a mechanism of hydrolysis involving cleavage of the
vinyl ether C-O bond and not cleavage of the glycosidic
C-O bond.
F igu r e 1. Eyring plots for the hydrolysis of 1 (b) and 2 (9)
in 1.00 mM HCl, I ) 0.100 M, at 20-40 °C. Data points are
the averages of three determinations.
We report here that, at pH 3.0 and 25 °C, both 1 and
2 hydrolyze exclusively by irreversible, rate-limiting
protonation of the â-carbon of the enol ether followed by
alkenyl ether C-O bond cleavage, not glycosidic C-O
bond cleavage. Compounds 1 and 2 are the first alkyl
vinyl acetals reported to hydrolyze exclusively by cleav-
age of the vinyl ether C-O bond. Furthermore, com-
pounds 1 and 2 exhibit a kinetic anomeric effect, with 1
hydrolyzing approximately four times faster than 2. UV,
1H and 13C NMR, and IR spectroscopies indicate that, in
the ground state, conjugation of the vinyl ether of 1 is
greater than that of 2, probably as a result of steric or
conformational constraints.
To prove unequivocally which C-O bond is cleaved
during hydrolysis, 1 and 2 were hydrolyzed in [18O]water
(97.9%), and aliquots were analyzed by mass spectros-
copy. GC-MS showed complete incorporation of one atom
of solvent-derived 18O into the acetone formed while
electrospray-MS indicated no incorporation of 18O into the
glucose formed. On the basis of the relative magnitudes
of the M+ peak for acetone and the M + 2 peak for
glucose, less than 0.5% of the hydrolysis observed in-
volves cleavage of the acetal moiety.
Resu lts a n d Discu ssion
Kin etics. Isopropenyl glucopyranosides 1 and 2 were
prepared as previously reported.3 Hydrolyses were per-
formed in dilute (0.5-4.0 mM) hydrochloric or perchloric
acid or in dilute (1-25 mM) solutions of carboxylic acid
buffers, pH 3-4. The ionic strength (I) of solutions was
typically maintained at 0.100 M by addition of NaCl.
Although initial reactions were monitored by enzymatic
assay for the release of glucose,3 the rates of later
reactions were measured more conveniently by observing
the decrease in absorbance at 215 nm due to the cleavage
of the vinyl ether. Rates of hydrolysis were first order
with respect to the glucopyranoside for at least 4 half-
lives and were found to be first-order with respect to
hydronium ion by varying the concentration of acid. At
25 °C, [H+] ) 1.00 mM and I ) 0.002 M, rate constants
for the hydrolysis of 1 and 2 were (5.11 ( 0.07) × 10-3
and (1.50 ( 0.06) × 10-3 s-1, respectively. At I ) 0.100
M, the corresponding rate constants for 1 and 2 were
Ir r ever sibility of C-P r oton a tion . When the elec-
trophilic addition of methanol-d4 to 1 and 2 was observed
by 1H NMR as described above, no deuterium was
incorporated into the starting material and only one atom
of deuterium was observed to be incorporated into 5 and
6.3 Similarly, when 1 and 2 were hydrolyzed in D2O,
analysis by 1H NMR and GC-MS indicated no exchange
of deuterium for vinyl protons in the starting material
and incorporation of only one atom of deuterium into the
acetone produced. These results indicate that protona-
tion of the â-carbon of the enol ether of 1 and 2 is
irreversible. On the basis of the magnitude of the M +
2 peak for the acetone produced, less than 0.5% of the
hydrolysis proceeds with incorporation of two atoms of
deuterium into product.
(3.64 ( 0.04) × 10-3 and (0.960 ( 0.009) × 10-3 s-1
,
respectively. These rate constants are 10-100 times
greater than those for the hydrolysis of alkyl and aryl
glucosides at 60 °C and pH 0.0.12
Kin etic Solven t Deu ter iu m Isotop e Effects. When
the hydrolyses of 1 and 2 in buffered D2O were observed
by 1H NMR, the rates appeared to be more than six times
slower than those measured under essentially identical
conditions in H2O. Careful spectrophotometric measure-
ment of the rates of hydrolysis of 1 and 2 at 25 °C in
H2O and D2O solutions of 10 mM chloroacetate buffer,
at a nominal pH of 3.0, confirmed this finding. Apparent
solvent deuterium isotope effects for 1 and 2 were 7.66
( 0.25 and 7.56 ( 0.47, respectively. Taking into account
Depending on the reaction conditions, 1 hydrolyzes
approximately four times faster than 2. Although this
difference in reactivity corresponds to a relatively small
change in free energy of activation, temperature-depend-
ent kinetics reveal a notable difference between the
enthalpies of activation for the hydrolyses of 1 and 2
(Figure 1). At 25 °C, [H+] ) 1.00 mM and I ) 0.100 M,
the activation parameters for the hydrolysis of 1 are
∆H‡ ) 12.56 ( 0.21 kcal mol-1 and ∆S‡ ) -26.8 ( 0.7
eu. The corresponding activation parameters for 2 are

Hydrolysis of Isopropenyl Glucopyranosides
J . Org. Chem., Vol. 63, No. 3, 1998 835
F igu r e 2. Observed rate constants for the hydrolysis of 1 (b)
and 2 (9) in chloroacetate buffer, nominal pH 3.0, at 25 °C.
F igu r e 3. Observed rate constants for the hydrolysis of 1 (O)
and 2 (0) in carboxylic acid buffers (1-10 mM, I ) 0.100 M,
25 °C) plotted as a function of the actual concentration of
hydronium ion.
Ta ble 1. Solven t Isotop e Effects on th e Hyd r olysis of 1
a n d 2 a t 25 °C a n d I ) 0.002 M
pronounced, but the overall contribution of buffer to
catalysis was so small that the rates of hydrolysis were
determined largely by catalysis by the hydronium ion.
Rates of hydrolysis of 1 and 2 at low concentrations (1-
10 mM) of buffer exhibited essentially no dependence on
the concentration of buffer. In fact, when buffer failure
was taken into account, a plot of the observed rate
constants versus actual concentration of hydronium ion
gave a straight line, regardless of the concentrations or
identities of the buffers used (Figure 3). From this plot,
1
2
-1
kH , M s-1
5.11 ( 0.07
1.70 ( 0.03
3.01 ( 0.07
1.50 ( 0.06
0.480 ( 0.008
3.13 ( 0.14
+
-1
kD , M s-1
+
+
+
kH /kD
the differences in lyonium ion concentration19 due to the
suppression of ionization of chloroacetic acid in D2O20 and
the fact that lyonium ion was the dominant catalytic
species present (see below), the solvent isotope effects on
catalysis by lyonium ion, kH /kD , were calculated to be
3.06 ( 0.10 and 3.02 ( 0.19 for 1 and 2, respectively.
Measurement of the rates of hydrolysis of 1 and 2 in
dilute solutions of perchloric acid in H2O and D2O allowed
+
+
+
the values of kH for 1 and 2 at I ) 0.100 M were
determined to be 3.67 ( 0.04 and 0.967 ( 0.008 M-1 s-1
,
respectively. These rate constants are identical, within
experimental error, with those determined by varying the
concentration of mineral acid (see Kinetics above).
Another method of determining the R coefficient for
general acid catalysis involves measuring the effective
solvent isotope effects for a reaction run in mixtures of
H2O and D2O.21,24,25 This solvent isotope effect method
has been found to give values of R that are essentially
identical with those determined by normal buffer dilution
techniques.21,23,25 The agreement between the two meth-
ods is not surprising since both methods provide informa-
tion about the extent to which a proton in flight has been
transferred in the transition state. For a proton being
transferred from the lyonium ion to a substrate contain-
ing no other exchangeable hydrogen atomssor in which
the fractionation factors for all other exchangeable
hydrogens remain constant during the activation of the
substrate26sR can be determined from eq 1, in which ø
represents the mole fraction of exchangeable deuterium
contained in the solvent, kø/kH represents the kinetic
isotope effect in a solvent of deuterium content ø, and λ
represents the fractionation factor for the lyonium ion,
which, in this case, is 0.69.27 The dependence of kø/kH
on ø is particularly sensitive to the magnitude of R for
+
+
the catalytic rate constants, kH and kD , to be determined
individually (Table 1). From these rate constants, the
+
+
values of kH /kD for the hydrolysis of 1 and 2 were
calculated to be 3.01 ( 0.07 and 3.13 ( 0.14, respectively.
These normal solvent deuterium isotope effects indicate
rate-limiting C-protonation, and their magnitude is typi-
cal for the hydrolysis of vinyl ethers.5,21,22 By contrast,
+
+
an inverse solvent isotope effect (kH /kD < 1) would have
been expected for a process involving preequilibrium
proton transfer, as is typical for acetal hydrolysis.
r Coefficien ts for Gen er a l Acid Ca ta lysis. Brøn-
sted R coefficients for the hydrolyses of 1 and 2 were
investigated by examining the effects of varying concen-
trations of carboxylic acid buffers (chloroacetic, 3-chlo-
rolactic, formic, glycolic, and propionic) on the hydrolyses
of 1 and 2. At moderate concentrations of buffer,
however, standard buffer dilution plots exhibited nega-
tive curvature that exceeded that due to normal buffer
failure (Figure 2). This type of curvature has been
observed by others and has been attributed to self-
association of buffer components in solution.23 At lower
concentrations of buffer, the observed curvature was less
(19) The concentration of lyonium ion was set by the nominal buffer
ratio, ([ClCH2CO2H]/[ClCH2CO2-])nom ) 0.741, and the effective dis-
sociation of chloroacetic acid in H2O and D2O.
+
+
reactions in which kH /kD > 2. Hydrolysis of 1 and 2 in
(20) The pKa of chloroacetic acid in D2O is greater than that in H2O
by 0.49: (a) McDougall, A. O.; Long, F. A. J . Phys. Chem. 1962, 66,
429-433. (b) Bell, R. P., Kuhn, A. T. Trans. Faraday Soc. 1963, 59,
1789-1793.
(23) (a) Gold, V.; Waterman, D. C. A. J . Chem. Soc. B 1968, 839-
849. (b) Gold, V.; Waterman, D. C. A. J . Chem. Soc. B 1968, 849-855.
(24) Kresge, A. J . Pure Appl. Chem. 1964, 8, 243-258.
(25) Kresge, A. J .; Chiang, Y. J . Chem. Soc. B 1967, 58-61.
(26) It is reasonable to assume that the fractionation factors for all
hydroxyl groups of the glucosyl residue remain constant at 1.0 during
the activation of the substrate (ref 27).
(27) (a) Schowen, R. L. In Isotope Effects on Enzyme-Catalyzed
Reactions; Cleland, W. W., O’Leary, M. H., Northrop, D. B., Eds.;
University Park Press: Baltimore, 1977; pp 64-99. (b) Schowen, K.
B. J . In Transition States of Biochemical Processes; Gandour, R. D.,
Schowen, R. L., Eds.; Plenum: New York, 1978; pp 225-283.
(21) Salomaa, P.; Kankaanpera¨, A.; Lajunen, M. Acta Chem. Scand.
1966, 20, 1790-1801.
(22) (a) Kresge, A. J .; Chwang, W. K. J . Am. Chem. Soc. 1978, 100,
1249-1253. (b) Kresge, A. J .; Leibovitch, M. Can. J . Chem. 1990, 68,
2129-2130. (c) Kresge, A. J .; Leibovitch, M. J . Am. Chem. Soc. 1992,
114, 3099-3102. (d) J ones, J .; Kresge, A. J . Can. J . Chem. 1993, 71,
38-41. (e) Chiang, Y.; Eliason, R.; Guo, G. H.-X.; Kresge, A. J . Can. J .
Chem. 1994, 72, 1632-1636.

836 J . Org. Chem., Vol. 63, No. 3, 1998
Chenault and Chafin
C(1)-O(5) bond would leave the glycosidic oxygen of 1
more available for conjugation with its attached isopro-
penyl group than is the glycosidic oxygen of 2.
Indeed, UV spectra indicate greater conjugation in the
isopropenyl ether of 1 than in the isopropenyl ether of 2.
Although both compounds exhibit absorbance maxima for
the isopropenyl ether at 193 nm, the molar absorptivities
of 1 and 2 are 4963 and 6473 M-1 cm-1, respectively. By
comparison, the molar absorptivities for 2-methyl-5,6-
dihydropyran (7) and 9-oxabicyclo[3.3.1]non-1-ene (8) at
λmax ) 190 nm are 5700 and 6200 M-1 cm-1, respec-
tively.29 Because of conformational restrictions in 8, the
lone-pair electrons on oxygen are clearly less conjugated
with the carbon-carbon double bond than are those in
7. The less conjugated vinyl ether has the greater molar
absorptivity.
1H NMR shows a clear correlation between the reactiv-
ity and spectroscopic behavior of 1 and 2 and those of a
series of simple alkyl vinyl ethers. While the difference
in 1H NMR chemical shift (∆δH) between the terminal
methylidene protons of 1 is greater than that of 2, the
geminal coupling constant (2J HH) between the same
protons is less negative in 1 than in 2 (Table 2).
Similarly, the terminal vinyl protons of a series of simple
alkyl vinyl ethers, ROCHdCH2, in which R is tert-butyl,
isopropyl, and methyl, display a trend toward decreasing
difference in chemical shift but increasingly negative
geminal coupling constant.31 The same series of vinyl
ethers decreases in its reactivity toward acid-catalyzed
hydrolysis (Table 2),30 as well as cationic olefin polym-
erization.32,36
Comparison of 2-chloroethyl vinyl ether with the other
simple alkyl vinyl ethers in Table 2 suggests that tert-
butyl vinyl ether is the most reactive of the simple alkyl
vinyl ethers because its alkyl group is most electron
releasing and allows the greatest resonance between
oxygen and the carbon-carbon double bond. This inter-
pretation is supported by kinetic and spectroscopic data
for aryl vinyl ethers, which also show a correlation
between decreasing rate of acid-catalyzed hydrolysis,
decreasing difference in methylidene proton chemical
shifts, and increasingly negative geminal coupling con-
stant between methylidene protons (Table 2, Figure 5).
There are, however, some significant differences be-
tween the spectroscopic data for the simple alkyl vinyl
ethers and those for the aryl vinyl ethers. The decreasing
difference in chemical shift between the methylidene
protons of alkyl vinyl ethers is due, in general, to an
F igu r e 4. Relationship between the apparent solvent isotope
effect, kø/kH, and the mole fraction deuterium content of the
solvent, ø, for the hydrolysis of (a) 1 and (b) 2. The solid lines
represent the values of R determined by nonlinear regression
while the dashed lines represent R ( 0.1.
mixtures of H2O and D2O containing 1 mM perchloric
acid provided kø/kH. Nonlinear regression of eq 1 to the
curved relationship between kø/kH and ø depicted in
Figure 4 gave R ) 0.627 ( 0.013 for 1 and R ) 0.646 (
0.031 for 2. Thus, within experimental error, the values
of R for 1 and 2 are the same.
kø (1 - ø + øλ1-R)2(1 - ø + øλ2R+1kD/kH)
)
(1)
(1 - ø + øλ)3
kH
Sp ectr oscop ic Da ta . Because the mechanism of
hydrolysis of 1 and 2 involves rate-limiting C-protonation,
the difference in reactivity between 1 and 2 implies a
difference in kinetic basicity between the two compounds.
Although solvent deuterium isotope effects and R coef-
ficients for general acid catalysis indicate little basis for
the difference in kinetic basicity and reactivity, spectro-
scopic methods were expected to be considerably more
sensitive probes of the ground-state electronic structures
of the isopropenyl ether moieties of 1 and 2. It was
expected that the more reactive compound, 1, would show
evidence of greater conjugation in the isopropenyl ether
and hence greater electron density on the â carbon of the
double bond than 2. The endocyclic nOfσ*CO interaction
that is present in 1 but absent in 213-15 should lengthen
the glycosidic C-O bond and shorten the endocyclic
C(1)-O(5) bond,28 thereby raising the energy of the C(1)-
O(5) σ* orbital. The consequent reduction in the exocyclic
nfσ* interaction between the glycosidic oxygen and the
(29) Quinn, C. B.; Wiseman, J . R. J . Am. Chem. Soc. 1973, 95, 1342-
1343.
(30) Ledwith, A.; Woods, H. J . J . Chem. Soc. B 1966, 753-757.
(31) Feeney, J .; Ledwith, A.; H., S. L. J . Chem. Soc. 1962, 2021-
2023.
(32) Masuda, T. J . Polym. Sci., Polym. Chem. Ed. 1973, 11, 2713-
(28) (a) Altona, C.; Romers, C. Acta Crystallogr. 1963, 16, 1225-
1232. (b) Paulsen, H.; Luger, P.; Heiker, F. R. In Anomeric Effect:
Origin and Consequences; Szarek, W. A., Horton, D., Eds.; ACS
Symposium Series 87; American Chemical Society: Washington, DC,
1979; Chapter 5, pp 63-79. (c) Briggs, A. J .; Glenn, R.; J ones, P. G.;
Kirby, A. J .; Ramaswamy, P. J . Am. Chem. Soc. 1984, 106, 6200-
6206. (d) J ones, P. G.; Kirby, A. J . J . Chem. Soc., Chem. Commun.
1986, 444-445.
2717.
(33) Potts, W. J .; Nyquist, R. A. Spectrochim. Acta 1959, 679-708.
(34) McClelland, R. A. Can. J . Chem. 1977, 55, 548-551.
(35) Afonin, A. V.; Vashchenko, A. V.; Fujiwara, H. Bull. Chem. Soc.
J pn. 1996, 69, 933-945. We thank Professor Fujiwara for kindly
providing us with the Supporting Information for this paper.
(36) Yuki, H.; Hatada, K.; Takeshita, M. J . Polym. Sci., Part A-1
1969, 7, 667-678.

Hydrolysis of Isopropenyl Glucopyranosides
J . Org. Chem., Vol. 63, No. 3, 1998 837
Ta ble 2. Cor r ela tion betw een Ra tes of Hyd r olysis a n d Sp ectr oscop ic Da ta for th e Meth ylid en e Gr ou p of Vin yl Eth er s
vinyl ether
krel
δH
δH
∆δH
2J HH, Hz
δC
δC
â
1J CH, Hz
ν
CH(wag), cm-1
cis
trans
R
1
2
100a
29a
4.35
4.29
4.22
4.23
0.13
0.06
-1.6
-2.2
157.7
158.3
87.3
87.0
156.9, 160.7
157.6, 161.2
825
812
tBuOCH)CH2
100b
39b
4.20c
4.15c
4.00c
4.05c
3.85c
3.88c
3.85c
4.17c
0.35
0.27
0.15
0.12
-0.1c
-1.2c
-2.2c
-2.7c
146.7d
150.9d
153.0d
91.2d
88.1d
86.1d
824e
iPrOCH)CH2
810, 823e
813e
CH3OCH)CH2
ClCH2CH2OCH)CH2
2.5b
p-CH3O(C6H4)OCH)CH2
p-CH3(C6H4)OCH)CH2
C6H5OCH)CH2
p-Cl(C6H4)OCH)CH2
p-O2N(C6H4)OCH)CH2
100f
76f
4.60g
4.67g
4.69g
4.72g
4.97g
4.28g
4.33g
4.35g
4.41g
4.68g
0.32
0.34
0.34
0.31
0.29
-1.5g
-1.6g
-1.5g
-1.8g
-2.0g
149.6g
148.8g
148.4g
148.0g
146.1g
93.6g
94.3g
95.1g
95.8g
99.1g
157.6, 161.6g
157.7, 161.6g
157.9, 161.8g
158.3, 162.1g
159.3, 162.8g
48f
15f
1.1f
Relative kH for hydrolysis in water at 25 °C. Relative kH for hydrolysis in 95% aqueous acetone at 25 °C (ref 30). c Reference 31.
a
b
+
+
Calculated from chemical shifts reported in ref 32, assuming that δC(CS2) ) 193.7. e Reference 33. f Relative rate of hydrolysis in 1.00
N HCl at 25 °C (ref 34). Reference 35.
d
g
the aryl group becomes more electron deficient and the
vinyl ether becomes less reactive. The general decrease
in chemical shift difference between methylidene protons
results from a greater downfield shift of the signal for
the proton trans to the aryloxy group (δHtrans). Although
Hammett σ values and the rates of hydrolysis of both
para- and meta-substituted aryl vinyl ethers correlate
1
well, Hammett plots of the H NMR data for the same
aryl vinyl ethers show considerable scatter (Figure 5).
2
Meta substituents, in particular, influence ∆δH and J HH
more strongly than para substituents. In addition,
phenyl and p-methylphenyl vinyl ethers in the case of
∆δH or phenyl and p-methoxyphenyl vinyl ethers in the
case of 2J HH lie significantly off the line defined by other
para-substituted phenyl vinyl ethers. It should be noted
2
that the trends in ∆δH and J HH exhibited by aryl vinyl
ethers are opposite from those predicted by the assump-
tion that increased electron donation from the ethereal
oxygen leads to greater sp2 f sp3 rehybridization of the
methylidene carbon.30 Such rehybridization would lead
to decreased ∆δH (as the protons become more magneti-
2
37
cally equivalent) and to more negative J HH
.
Although 13C NMR has been used previously to exam-
ine vinyl ethers32,35,38 and is considered to be a sensitive
probe of hybridization and electron density associated
with individual carbon atoms,39 1 and 2 exhibit only small
differences between the chemical shifts (δCR and δCâ) and
one-bond 13C-1H coupling constants (1J CH) of their iso-
propenyl ether moieties (Table 2). Still, the differences
in 13C NMR chemical shift between 1 and 2 correlate with
those of the alkyl vinyl ethers in Table 2. As the vinyl
ether becomes more reactive, δCR shifts upfield and δCâ
shifts downfield. For aryl vinyl ethers, however, the
trend is exactly opposite, with δCR shifting downfield and
δCâ shifting upfield as the reactivity of the vinyl ether
increases (Table 2, Figure 6). Nevertheless, the trend
in one-bond 13C-1H coupling constants for the meth-
ylidene groups of 1 and 2 agrees with that of the aryl
vinyl ethers. As the reactivity of the vinyl ether in-
creases, both values of 1J CH decrease (Table 2, Figure 6).
Although the correlations between reactivity and 13C
F igu r e 5. Hammett correlations for (a) the rates of acid-
catalyzed hydrolysis, (b) the differences in 1H NMR chemical
shift (∆δH ) δHcis - δHtrans) between terminal methylidene
protons, and (c) the geminal coupling constants (2J HH) between
terminal methylidene protons of substituted aryl vinyl ethers.
Aryl substituents are para (circles) or meta (squares). Tri-
angles represent para-substituted or unsubstituted aryl vinyl
ethers that generated anomolous data. Solid lines are the best
least-squares fits to all data. Dashed lines are the best least-
squares fits to only the data represented by circles. Data taken
from ref 35.
(37) Pople, J . A.; Bothner-By, A. A. J . Chem. Phys. 1965, 42, 1339-
1349.
(38) Reynolds, W. F.; McClelland, R. A. Can. J . Chem. 1977, 55,
536-540.
(39) (a) Hansen, P. E. Prog. Nucl. Magn. Reson. Spectrosc. 1981,
14, 175-296. (b) Webb, J . G. K.; Yung, D. K. Can. J . Chem. 1983, 61,
488-493. (c) Tolman, C. A.; English, A. D.; Manzer, L. E. Inorg. Chem.
1975, 14, 2353-2356.
upfield shift of the signal for the proton cis to the alkoxy
group (δHcis). With the aryl vinyl ethers, however, NMR
signals for both methylidene protons shift downfield as

838 J . Org. Chem., Vol. 63, No. 3, 1998
Chenault and Chafin
catalysis indicate a relatively late transition state with
significant transfer of an acidic proton to the â-carbon of
the vinyl ether.
Because protonation of the alkenyl ether is involved
in the rate-limiting step of hydrolysis, the greater reac-
tivity of 1 indicates that 1 is kinetically more basic than
2. This assertion is supported by the fact that the
enthalpy of activation for the hydrolysis of 1 is signifi-
cantly lower than that for the hydrolysis of 2. The lower
activation barrier of 1 is partially offset by a more
negative entropy of activation. Presumably, the more
sterically encumbered isopropenyl group of 1 requires
greater increase in order upon protonation and nucleo-
philic attack than does that of 2.
The cause for the greater kinetic basicity of 1 cannot
be steric accessibility of the enol ether or dipole-dipole
interactions within the substrate since both of these
should make 2 more reactive than 1. The difference in
UV molar absorptivities of 1 and 2 clearly indicates a
difference in the extent of conjugation of the glycosidic
oxygens with their attached alkenyl double bonds. How-
ever, comparison of 1 and 2 with 7 and 8 suggests that
the difference in conjugation is due to a conformational
difference, not a difference in the electronic states of the
glycosidic oxygens. 1H NMR and 13C NMR spectroscopies
show parallels between the reactivities and spectroscopic
properties of isopropenyl glucopyranosides and those of
simple alkyl vinyl ethers and substituted aryl vinyl
ethers. These similarities indicate that electron donation
from the glycosidic oxygen into the alkenyl ether double
bond by resonance is greater in 1 than in 2. Trends in
13C NMR chemical shift data, however, suggest that the
difference in reactivity between 1 and 2 originates from
steric constraints. Correlations between the reactivity
and 13C NMR chemical shifts of 1 and 2 are identical with
those of simple alkyl vinyl ethers but opposite of those
of substituted aryl vinyl ethers (Table 2). While the
difference between aryl vinyl ethers is certainly elec-
tronic, the major difference between the simple alkyl
vinyl ethers is the size of the alkyl group and any change
in ground-state conformational population that might
attend changes in steric strain. Even the differences in
1H NMR chemical shift between methylidene protons
show greater similarity between the isopropenyl gluco-
sides and alkyl vinyl ethers than between the isopropenyl
glucosides and aryl vinyl ethers. Similarities in the IR
between the methylidene wagging frequencies of 1 and
2 and those of tert-butyl and methyl vinyl ethers also
indicate that the increased reactivity of 1 is due to an
effect of sterics on the ground-state conformation or
stability. Thus, while various spectroscopic evidence
indicates a difference in the degree of conjugation within
the enol ether moieties of 1 and 2, it appears that the
increased reactivity of 1 is due more to steric and
conformational strain in the ground state than to a
difference in the electronic states of the glycosidic oxygen
atoms of 1 and 2 resulting from nOfσ*CO interactions.
F igu r e 6. Hammett correlations for (a) the 13C NMR chemical
shifts of the R vinyl carbon atoms, (b) the 13C NMR chemical
shifts of the â vinyl carbon atoms, (c) the one-bond 13C-1H
coupling constants between the â vinyl carbon atoms and the
protons cis to the aryloxy groups (1J CHcis), and (d) the one-bond
13C-1H coupling constants between the â vinyl carbon atoms
and the protons trans to the aryloxy groups (1J CHtrans) of
substituted aryl vinyl ethers. See Figure 5 for an explanation
of the symbols used. Data taken from ref 35.
NMR data are better than those involving 1H NMR data,
for 13C NMR chemical shifts, it is again phenyl and
p-methylphenyl vinyl ethers that deviate from the trends
established by the other aryl vinyl ethers. In the
Hammett plot of 1J CHtrans, p-amino- and p-(dimethylami-
no)phenyl vinyl ethers are the notable outliers.
In the IR, the methylidene protons of 1 and 2 exhibit
the same trend in their strong, out-of-plane wagging
frequencies as do the methylidene protons of the simple
alkyl vinyl ethers in Table 2. The more reactive vinyl
ethers display higher methylidene wagging frequencies.
In the case of the simple alkyl vinyl ethers, the different
wagging frequencies have been interpreted as represent-
ing different conformers about the alkyl33 or vinyl40 C-O
bond. Methyl vinyl ether is believed to exist in the s-cis
conformation while tert-butyl vinyl ether is believed to
exist in a gauche conformation. Isopropyl vinyl ether
displays two methylidene wagging frequencies because
it exists as an approximately 1:1 mixture of two possible
conformers.
Con clu sion
All experimental evidence indicates that the hydrolyses
of 1 and 2 proceed by irreversible, rate-limiting C-
protonation, followed by alkenyl ether C-O bond cleav-
age, not glycosidic C-O bond cleavage. Compounds 1
and 2 are the first vinyl acetals demonstrated to hydro-
lyze exclusively by cleavage of the vinyl ether C-O bond.
The magnitudes of both the solvent deuterium kinetic
isotope effects and the R coefficients for general acid
Exp er im en ta l Section
Gen er a l P r oced u r es. Anhydrous methanol was used as
purchased from Aldrich. Enzymes from Sigma and Biozyme
and other reagents were used as purchased. Synthetic reac-
tions were performed under an inert atmosphere of dry argon,
using dry glassware and standard syringe/septa techniques.
Analytical TLC was performed on plates (Whatman or Anal-
tech) containing a 250 µm-thick layer of 60 Å silica gel.
(40) Ledwith, A.; Woods, H. J . J . Chem. Soc. B 1970, 310-314.

Hydrolysis of Isopropenyl Glucopyranosides
J . Org. Chem., Vol. 63, No. 3, 1998 839
Compounds were visualized by dipping plates into p-anisal-
dehyde stain and heating. Preparative liquid chromatography
was performed on 230-400 mesh 60 Å silica gel (Aldrich).
Hyd r olysis of 1 a n d 2 in H218O. Isopropenyl glucopyra-
noside (1.0 mg) was dissolved in 100 µL of 10 mM chloroacetate
buffer, pH 3.0, prepared in 97.9% H218O. Hydrolysis of each
glucopyranoside was monitored by GC-MS and electrospray-
MS. Aliquots (10 µL) of each reaction were removed periodi-
cally, added to a microcentrifuge tube, and extracted with
CHCl3 (10 µL). Portions of the CHCl3 solutions were removed
by syringe and used to monitor the formation of acetone by
GC-MS. Aliquots (10 µL) of the reactions were also removed
to monitor the formation of glucose by electrospray-MS.
Solven t Deu ter iu m In cor p or a tion d u r in g Hyd r olysis
of 1 a n d 2. Isopropenyl glucopyranoside (4.0 mg) was dis-
solved in 650 µL of 10 mM chloroacetate buffer in D2O, pH
1H NMR spectra were obtained at a nominal resonance
frequency of 250 MHz, using D2O as solvent and sodium
3-(trimethylsilyl)propionate-2,2,3,3-d4 (TSP) as the internal
reference (δ 0.00 ppm). Liquid secondary ion mass spectros-
copy (LSIMS) was performed using Cs+ (20 eV) as the ionizing
beam and glycerol/NaI as the matrix. Elemental analysis was
performed by Robertson Microlit Laboratories, Inc.
Kinetic experiments were maintained at constant temper-
ature by a digital Peltier controller. Disappearance of the vinyl
ether functional group was observed spectroscopically by the
decrease in absorbance at λ ) 215 nm. Apparent rate constants
were calculated by nonlinear regression of UV/vis data, using
Enzfitter version 1.05 software.
1
3.0, and hydrolysis was monitored by H NMR. Aliquots (10
µL) of each reaction were removed periodically and extracted
with CHCl3 (10 µL). Portions of the CHCl3 solutions were
analyzed by GC-MS to monitor the formation of acetone.
Hyd r on iu m Ion Ca ta lytic Coefficien ts a n d Solven t
Isotop e Effects. Rates of hydrolysis of 1 and 2 were
measured at 25 °C in solutions of 0.50, 1.00, 2.00, or 4.00 mM
Isop r op en yl r-D-Glu cop yr a n osid e (1). To 1.00 g (1.80
mmol) of isopropenyl 2,3,4,6-tetra-O-pivaloyl-R-D-glucopyra-
noside3 in 5 mL of methanol was added 0.360 mL of a 20%
solution of sodium methoxide in methanol. The reaction was
stirred overnight at room temperature. Dowex-50W×8 resin
(H+, 0.300 g) was added to the solution, and the solution was
filtered quickly. The solvent was evaporated in vacuo to afford
a white solid. Purification by flash chromatography using
chloroform/ethyl acetate/methanol (5:2:1) as the eluent yielded
0.325 g (82%) of 1: mp 157-159 °C; Rf 0.84 (80% aqueous
acetonitrile containing 9.5 mM tetrabutylammonium hydrox-
+
+
perchloric acid made in H2O or D2O. Values of kH and kD
were calculated by linear regression of the averages of
triplicate determinations of the apparent rate constant at each
concentration of perchloric acid.
Apparent solvent isotope effects in isotopically mixed sol-
vents were determined by mixing the appropriate amounts of
1.00 mM perchloric acid in H2O and 1.00 mM perchloric acid
in D2O and measuring the rates of hydrolysis of 1 and 2 at 25
°C. Rates of hydrolysis at each ratio of H2O and D2O were
determined in quadruplicate and averaged.
Bu ffer Dilu tion Stu d ies. Rates of hydrolysis of 1 and 2
were determined in a series of carboxylic acid buffer solutions
of constant buffer ratio but varying buffer concentration.
Chloroacetic acid, 3-chlorolactic acid, formic acid, glycolic acid,
and propionic acid were used to make buffers. Buffers were
constructed by mixing known quantities of each free acid with
a standardized solution of sodium hydroxide. The “nominal
pH” of each buffer was calculated from the proportions of free
acid and sodium hydroxide mixed, using the Henderson-
Hasselbalch equation. Each stock solution of buffer was
diluted to various concentrations so as to maintain a constant
nominal ratio of [HA]/[A-]. At low concentrations of buffer,
however, this ratio declined due to dissociation of the free acid
into A- + H+ (buffer failure).41 All buffers were maintained
at a constant ionic strength of 0.100 M by including sodium
chloride in the stock solutions and dilutions, taking buffer
failure into account in determining the degree of ionization of
each carboxylic acid. Chloroacetate buffer was made fresh
daily since it hydrolyzes upon sitting. All other buffers were
used within 3-4 days of being made.
1
ide); H NMR δ 5.45 (d, J ) 3.7 Hz, 1H), 4.35 (d, J ) 1.7 Hz,
1H), 4.22 (dq, J ) 0.8, 1.7 Hz, 1H), 3.85-3.61 (m, 5H), 3.46 (t,
J ) 9.4 Hz, 1H), 1.88 (s, 3H); 13C NMR δ 157.7, 95.2, 87.3,
73.4, 72.6, 71.3, 69.7, 60.7, 19.7; IR (KBr) 3456 (s), 3247 (s),
2926 (m), 1654 (w), 1637 (m), 1274 (s), 1145 (m), 1112 (s), 1086
(m), 1053 (s), 1034 (s), 825 (s) cm-1; LSIMS m/z 243 (M + Na+),
163, 145, 127, 109. Anal. Calcd for C9H16O6: C, 49.09; H, 7.32.
Found: C, 49.15; H, 7.54.
Isop r op en yl â-D-Glu cop yr a n osid e (2). To 1.00 g (2.59
mmol) of isopropenyl 2,3,4,6-tetra-O-acetyl-â-D-glucopyrano-
side3 in 10 mL of methanol was added 0.300 mL of a 20%
solution of sodium methoxide in methanol. After 8 h, the
reaction was quenched with 0.100 g of Dowex-50W×8 resin,
and the solution was filtered quickly. The solvent was
evaporated in vacuo. Purification of the white solid by flash
chromatography using chloroform/ethyl acetate/methanol (5:
2:1) as the eluent yielded 0.541 g (95%) of 2: mp 137-138 °C;
Rf 0.78 (80% aqueous acetonitrile containing 9.5 mM tetrabu-
1
tylammonium hydroxide); H NMR (D2O) δ 4.94 (d, J ) 7.8
Hz, 1H), 4.29 (d, J ) 2.2 Hz, 1H), 4.23 (dq, J ) 0.8, 2.2 Hz,
1H), 3.92 (dd, J ) 2.1, 12.4 Hz, 1H), 3.72 (dd, J ) 5.6, 12.4
Hz, 1H), 3.61-3.54 (m, 2H), 3.47-3.39 (m, 2H), 1.88 (d, J )
0.54 Hz, 3H); 13C NMR δ 158.3, 98.5, 87.0, 76.4, 75.9, 73.0,
69.8, 61.0, 19.7; IR (KBr) 3446 (s), 2927 (m), 1654 (m), 1637
(m), 1560 (w), 1508 (w), 1374 (m), 1271 (s), 1106 (s), 1073 (s),
1019 (s), 812 (s) cm-1; LSIMS m/z 243 (M + Na+), 163, 145,
127, 109. Anal. Calcd for C9H16O6: C, 49.09; H, 7.32.
Found: C, 48.92; H, 7.50.
The observed rate constants for hydrolysis of 1 and 2 were
adjusted for the decrease in hydronium ion concentration at
low concentrations of buffer due to buffer failure, using the
method of Kresge (eq 2).41 The observed rate constants, kobs
,
kadj ) kobs + ([H+]max - [H+]act)kH
(2)
+
1
Hyd r olysis of 1 a n d 2 Mon itor ed by H NMR. A 10 mM
sodium salicylate buffer, pH 3.0, in D2O was prepared by
dissolving 69.0 mg (0.500 mmol) of salicylic acid in 40 mL of
D2O, adjusting the pH to 3.0 with 1 M NaOD in D2O, and
adding D2O to bring the final volume to 50 mL. In an NMR
tube, 600 µL of salicylate buffer was added to 50 µL of 364
mM isopropenyl glucopyranoside in D2O. Hydrolysis of the
glucopyranoside was monitored by acquisition of 1H NMR
spectra and observing the disappearance of the anomeric
proton of 1 or 2 and the appearance of the anomeric proton of
R- or â-glucopyranose, respectively.
Meth a n olysis of 1 a n d 2. A 25 mM solution of DCl in
methanol-d4 was prepared by adding 3.0 µL of benzoyl chloride
to 977 µL of methanol-d4. In an NMR tube, 4.0 mg of
isopropenyl glucopyranoside was dissolved in 625 µL of
methanol-d4, and an 1H NMR spectrum was obtained. After
the addition of 25 µL of 25 mM DCl in methanol-d4 to the
glucopyranoside solution, subsequent 1H NMR spectra were
acquired.
[H+]act ) (-(γH + γA-)[A-]
+ Ka +
nom
((γ + γ -)[A-]nom + Ka)2 + 4γH+γ - [HA]nomKa)/
x
H
A
A
2γH+γ (3)
-
A
[HA]act ) [HA]nom - [H+]act
(4)
(5)
kadj ) kH+[H+] + kHA[HA]act
act
measured at various concentrations of a given buffer were
adjusted to the hydronium ion concentration of the most
concentrated buffer ([H+]max), using hydronium ion concentra-
(41) Kresge, A. J .; Chen, H. L.; Chiang, Y.; Murrill, E.; Payne, M.
A.; Sagatys, D. S. J . Am. Chem. Soc. 1971, 93, 413-423.

840 J . Org. Chem., Vol. 63, No. 3, 1998
Chenault and Chafin
tions ([H+]act) calculated from acid dissociation constants (eq
and electrospray-MS experiments. We thank J ie Yang
for acquiring the proton-coupled 13C NMR spectra of 1
and 2.
+
3) and the hydronium ion catalytic coefficients (kH ) reported
+
-
in Table 1. Activity coefficients (γH and γA ) of 0.83 and 0.78
were used for hydronium and carboxylate ions, respectively,
at an ionic strength 0.100 M.42 The actual concentrations of
carboxylic acid present in solution were calculated by adjusting
for depletion due to dissociation (eq 4). The adjusted rate
constants and actual concentrations of general acid were then
fit to eq 5 by linear regression.
Su p p or tin g In for m a tion Ava ila ble: Partial assignments
of 1H and 13C NMR spectra of 1 and 2, as well as tables of
rate data versus temperature, concentration of hydronium ion,
solvent deuterium content, and concentration of buffer (3
pages). This material is contained in libraries on microfiche,
immediately follows this article in the microfilm version of the
journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
Ack n ow led gm en t. We thank Dr. Gordon Nicol for
providing LSIMS data and for help with the GC-MS
(42) Kielland, J . J . Am. Chem. Soc. 1937, 59, 1675-1678.
J O9721249