Enthalpic nature of the CH/π interaction involved in the recognition of carbohydrates by aromatic compounds, confirmed by a novel interplay of NMR, calorimetry, and theoretical calculations

Article abstract of DOI:10.1021/ja903950t

Specific interactions between molecules, including those produced by a given solute, and the surrounding solvent are essential to drive molecular recognition processes. A simple molecule such as benzene is capable of recognizing and differentiating among very similar entities, such as methyl 2,3,4,6-tetra-O-methyl-α-D-galactopyranoside (α-Me₅Gal), methyl 2,3,4,6-tetra-O-methyl-β-D-galactopyranoside (β-Me ₅Gal), 1,2,3,4,6-penta-O-acetyl-β-D-galactopyranose (β-Ac₅Gal), and methyl 2,3,4,6-tetra-O-methyl-α-D- mannopyranoside (α-Me₅Man). In order to determine if these complexes are formed, the interaction energy between benzene and the different carbohydrates was determined, using Calvet microcalorimetry, as the enthalpy of solvation. These enthalpy values were -89.0 ± 2.0, -88.7 ± 5.5, -132.5 ± 6.2, and -78.8 ± 3.9 kJ mol^-1 for the four complexes, respectively. Characterization of the different complexes was completed by establishing the molecular region where the interaction takes place using NMR. It was determined that β-Me₅Gal is stabilized by the CH/π interaction produced by the nonpolar region of the carbohydrate on the α face. In contrast, α-Me₅Man is not specifically solvated by benzene and does not present any stacking interaction. Although α-Me₅Gal has a geometry similar to that of its epimer, the obtained NMR data seem to indicate that the axial methoxy group at the anomeric position increases the distance of the benzene molecules from the pyranose ring. Substitution of the methoxy groups by acetate moieties, as in β-Ac ₅Gal, precludes the approach of benzene to produce the CH/π interaction. In fact, the elevated stabilization energy of β-Ac ₅Gal is probably due to the interaction between benzene and the methyl groups of the acetyls. Therefore, methoxy and acetyl substituents have different effects on the protons of the pyranose ring.

Full text of DOI:10.1021/ja903950t

Published on Web 11/23/2009
Enthalpic Nature of the CH/π Interaction Involved in the
Recognition of Carbohydrates by Aromatic Compounds,
Confirmed by a Novel Interplay of NMR, Calorimetry, and
Theoretical Calculations
Karla Ram´ırez-Gualito,*,⊥ Rosa Alonso-R´ıos,⊥ Beatriz Quiroz-Garc´ıa,⊥
Aaro´n Rojas-Aguilar,‡ Dolores D´ıaz,§ Jesu´s Jime´nez-Barbero,§ and Gabriel Cuevas⊥
Instituto de Qu´ımica, UniVersidad Nacional Auto´noma de Me´xico, Apdo. Postal 70213, 04510
Coyoaca´n, Circuito Exterior, Me´xico D.F., México, Departamento de Qu´ımica, Centro de
InVestigacio´n y de Estudios AVanzados, Instituto Polite´cnico Nacional, Apdo. Postal 14-740,
07000 Me´xico D.F., México, and Centro de InVestigaciones Biolo´gicas, CSIC, Ramiro de
Maeztu 9, 28040 Madrid, Spain
Received May 19, 2009; E-mail: karamirezg@prodigy.net.mx
Abstract: Specific interactions between molecules, including those produced by a given solute, and the
surrounding solvent are essential to drive molecular recognition processes. A simple molecule such as
benzene is capable of recognizing and differentiating among very similar entities, such as methyl 2,3,4,6-
tetra-O-methyl-R-D-galactopyranoside (R-Me5Gal), methyl 2,3,4,6-tetra-O-methyl-ꢀ-D-galactopyranoside (ꢀ-
Me5Gal), 1,2,3,4,6-penta-O-acetyl-ꢀ-D-galactopyranose (ꢀ-Ac5Gal), and methyl 2,3,4,6-tetra-O-methyl-R-
D-mannopyranoside (R-Me5Man). In order to determine if these complexes are formed, the interaction energy
between benzene and the different carbohydrates was determined, using Calvet microcalorimetry, as the
enthalpy of solvation. These enthalpy values were -89.0 ( 2.0, -88.7 ( 5.5, -132.5 ( 6.2, and -78.8 (
3.9 kJ mol-1 for the four complexes, respectively. Characterization of the different complexes was completed
by establishing the molecular region where the interaction takes place using NMR. It was determined that
ꢀ-Me5Gal is stabilized by the CH/π interaction produced by the nonpolar region of the carbohydrate on the
R face. In contrast, R-Me5Man is not specifically solvated by benzene and does not present any stacking
interaction. Although R-Me5Gal has a geometry similar to that of its epimer, the obtained NMR data seem
to indicate that the axial methoxy group at the anomeric position increases the distance of the benzene
molecules from the pyranose ring. Substitution of the methoxy groups by acetate moieties, as in ꢀ-Ac5Gal,
precludes the approach of benzene to produce the CH/π interaction. In fact, the elevated stabilization energy
of ꢀ-Ac5Gal is probably due to the interaction between benzene and the methyl groups of the acetyls.
Therefore, methoxy and acetyl substituents have different effects on the protons of the pyranose ring.
proteins that recognize carbohydrates such as galectins9 and
mediate the interaction between carbohydrates and the aromatic
Introduction
Interactions between carbohydrates and proteins mediate
many fundamental and complex phenomena, such as allergic
reactions, embryogenesis, tissue development, fertilization,
metastasis, bacterial cell wall recognition, and protein hydration
and stability.1-7 It has been demonstrated that these interactions
are facilitated by the formation of hydrogen bonds.8 However,
CH/π stacking is very common in the recognition sites of
residues of the side chains of Trp, Tyr, Phe, and His present in
the receptors. CH/π interactions have their origin in dispersion
forces, which have an impact on the enthalpic term of the free
energy.10 Depending on the chemical nature of the interacting
groups, the CH/π interaction may contribute more stabilizing
energy than a hydrogen bond; therefore, it may become difficult
to evaluate the actual energy contribution of each type of
interaction when they are competing. Because of the importance
of CH/π interactions and the intrinsic difficulty in their proper
description, multiple efforts have been made to characterize
them. When the C-H group is highly acidic, as is the case for
the C-H group of chloroform, the CH/π interaction is intense.
However, when the hydrogen group donor is not very acidic,
the participation of several C-H bonds in the proper orientation
⊥ Universidad Nacional Auto´noma de Me´xico.
‡ Centro de Investigación y de Estudios Avanzados.
§ Centro de Investigaciones Biolo´gicas, CSIC.
(1) Kitov, P. I.; Bundle, D. R. J. Am. Chem. Soc. 2003, 125, 16271–
16284.
(2) Thobhani, S.; Ember, B.; Siriwardena, A.; Boons, G.-J. J. Am. Chem.
Soc. 2003, 125, 7154–7155.
(3) Bertozzi, C. R.; Kiessling, L. L. Science 2001, 291, 2357–2364.
(4) Simanek, E. E.; McGarvey, G. J.; Jablonowski, J. A.; Wong, C. H.
Chem. ReV. 1998, 98, 833–862.
(8) Scheiner, S. Molecular Interactions: From Van der Waals to Strongly
Bound Complexes; Wiley: New York, 1997.
(9) Lis, H.; Sharon, N. Chem. ReV. 1998, 98, 637–674.
(10) Anslyn, E. V.; Dougherty, D. A. Modern Physical Organic Chemistry;
University Science Books: Sausalito, CA, 2005.
(5) Wies, W. L.; Drickamer, K. Annu. ReV. 1996, 65, 441–473.
(6) Vyas, N. K. Curr. Opin. Struct. Biol. 1991, 1, 732–740.
(7) Gabius, H.-J.; Siebert, H.-C.; Andre´, S.; Jime´nez-Barbero, J.; Ru¨diger,
H. ChemBioChem 2004, 5, 740–764.
9
10.1021/ja903950t  2009 American Chemical Society
J. AM. CHEM. SOC. 2009, 131, 18129–18138 18129

A R T I C L E S
Ram´ırez-Gualito et al.
as the molecular region where it is produced.17 Only with this
knowledge it is possible to avoid speculation. Thermodynamic
data may allow researchers to establish the magnitude of
interactions but do not provide direct evidence of their specific-
ity. The nature and structure of the solvent molecules, behaving
as individuals or as a bulk, are also of paramount relevance to
determine the specificity or lack thereof of the interaction
between two different entities, or between any entity and the
solvent.
Experimental determination of thermodynamic binding data
is essential for understanding the process itself, but also for
evaluating the adequacy of the theoretical methods employed
to model and quantify molecular interactions. Theoretical data
strongly depend on the level of theory as well as on the
correction of intrinsic failures in methodology such as the basis
set superposition or the scaling of the zero-point energy. Some
of the corrections required have their origin in the fact that the
energy is usually calculated considering complex molecules as
ideal gases.
Simple models are very useful to understand molecular
recognition processes. On this basis, we have decided to focus
on the investigation of isolated CH/π interactions between sugars
and aromatic moieties by eliminating the possibility of hydrogen
bonding by using O-substituted sugars as well as the proper
solvents. In the study presented herein, the sugar hydroxyls have
been substituted by two common protecting groups: O-methyl
and O-acetyl moieties. As solvent, we have chosen benzene to
mimic the nonpolar, hydrophobic, and aromatic character of the
lectin binding sites. Even if these modifications do not reproduce
the conditions of the biological systems, it is useful to isolate
the interactions to determine if the potential interaction between
carbohydrates and aromatic substrates has an enthalpic origin
or if it is purely entropic.
A key question arises: Can an aromatic molecule (benzene)
differentiate between two isomeric carbohydrates? It is generally
accepted that, besides solvation, two factors govern molecular
recognition: complementarity and preorganization.24,25 Our study
shows that, in fact, the interacting groups should complement
each other but preorganization is not strictly essential for the
CH/π interaction to take place. Since weak interactions have
their origin in dispersion forces, they also have an impact on
the enthalpic term of the free energy; the existence of the CH/π
interaction can be demonstrated by measuring the energy
involved, which should be stabilizing and dependent on the
chemical nature of the interacting molecules, to be specific.16,26,27
To the best of our knowledge, this is the first time the interplay
of calorimetric solvation data, NMR shielding effects, intermo-
lecular NOEs, and theoretical calculations is used to draw
conclusions about specific interactions between carbohydrates
and proteins, and this could be established as an adequate
methodology for this type of study.
Figure 1. Orientation of the three hydrogen atoms required to interact with
the aromatic molecule through the corresponding CH/π interactions. The
hydrogen atoms marked with an asterisk establish the molecular region
capable of stacking. The disposition indicated in a, R-galactose (positions
3, 4, 5) corresponds to a consecutive arrangement; that in b, ꢀ-glucose,
(positions 1, 3, 5) belongs to an alternated arrangement.
is necessary for interaction. This is very relevant when
carbohydrates interact with aromatic compounds in the recogni-
tion site of a protein. Due to the lack of acidity of the hydrogen
atoms of the carbohydrate’s pyranose ring, at least three
hydrogen atoms must be located in the same molecular region,
and there must be enough space to allow their proximity to the
aromatic compound (Figure 1). This orientation can be of two
types: consecutive, as in R-galactose (a, Figure 1), or alternated,
as in ꢀ-glucose (b, Figure 1). Previous studies of the CH/π
interaction have been based on the individual use of diverse
analytical, biochemical, and biophysical techniques, such as
calorimetry,11 X-ray analysis,12-14 NMR,15 IR,16 and molecular
modeling.17,18 These studies have been published as different
attempts to properly describe the interactions.19-22 However,
the method presented herein, with the combination of NMR,
calorimety, and theoretical calculations to adequately describe
the nature of the interaction, has not been reported before.
Knowledge of all the factors that control this interaction
would be of fundamental importance for modifing and modulat-
ing the affinity of natural receptors, depending on the process
under study, and should be instrumental in improving the design
of artificial receptors, as well as for the rational design of
carbohydrate- or glycomimetic-based pharmaceuticals.23
Undoubtedly, a rigorous way to establish the nature of a weak
interaction would be to determine its associated energy, as well
(11) Bautista-Iban˜ez, L.; Ram´ırez-Gualito, K.; Quiro´z-Garc´ıa, B.; Rojas-
Aguilar, A.; Cuevas, G. J. Org. Chem. 2008, 73, 849–857.
(12) Iitaka, Y.; Kodama, Y.; Nishihata, K.; Nishio, M. J. Chem. Soc., Chem.
Commun. 1974, 389–390.
(13) Kodama, K.; Nishihata, K.; Nishio, M.; Iitaka, Y. J. Chem. Soc., Perkin
Trans. 2 1976, 1490–1495.
(14) Nishihata, K.; Nishio, M. Tetrahedron Lett. 1977, 18, 1041–1044.
(15) Jime´nez-Barbero, J.; Asensio, J. L.; Can˜ada, F. J.; Poveda, A. Curr.
Opin. Struct. Biol. 1999, 9, 549–555.
(16) Stanca-Kaposta, E. C.; Gamblin, D. P.; Screen, J.; Liu, B.; Snoek,
L. C.; Davis, B. G.; Simons, J. P. Phys. Chem. Chem. Phys. 2007, 8,
4444–4451.
(17) Ferna´ndez-Alonso, M. C.; Can˜ada, F. J.; Jime´nez-Barbero, J.; Cuevas,
G. J. Am. Chem. Soc. 2005, 127, 7379–7386.
(18) Raju, R. K.; Ramraj, A.; Vincent, N. A.; Hillier, I. H.; Burton, N. A.
Phys. Chem. Chem. Phys. 2008, 10, 6500–6508.
Results and Discussion
(19) Sauter, N. K.; Hanson, J. E.; Glick, G. D.; Brown, J. H.; Crowther,
R. L.; Park, S. J.; Skehel, J. J.; Wiley, D. C. Biochemistry 1992, 31,
9609–9621.
a. Theoretical Calculations. Benzene was chosen as a good
model to perform computational studies, since this molecule
maintains the essence of the interaction between an aromatic
moiety and a carbohydrate molecule. It is the simplest possible
model and could help answer a key question: What is the
(20) Emsley, P.; Fotinov, C.; Black, I.; Fairweather, N. F.; Charles, I. G.;
Watts, C.; Hewitt, E.; Isaacs, N. W. J. Biol. Chem. 2000, 275, 8889–
8894.
(21) Pascal, J. M.; Day, P. J.; Monzingo, A. F.; Ernst, S. R.; Robertus,
J. D.; Iglesias, R.; Pe´rez, Y.; Fe´rreras, J. M.; Citores, L.; Girbe´s, T.
Proteins: Struct., Funct. Genet. 2001, 43, 319–326.
(22) Vandenbussche, S.; Dolores D´ıaz, D.; Ferna´ndez-Alonso, M. C.; Pan,
W.; Vincent, S. P.; Cuevas, G.; Ca˜nada, F. J; Jime´nez-Barbero, J.;
K., K. B. Chem.sEur. J. 2008, 14, 7570–7578.
(24) Koshland, D. E. J. Proc. Natl. Acad. Sci. U.S.A. 1958, 44, 98–123.
(25) Fischer, E. Ber. Deutsch. Chem. Ges. 1894, 27, 2985–2993.
(26) Eblinger, F.; Schneider, H.-J. Angew. Chem., Int. Ed. 1998, 37, 826–
829.
(23) Terraneo, G.; Potenza, D.; Canales, A.; Jime´nez-Barbero, J.; Baldridge,
K. K.; Bernardi, A. J. Am. Chem. Soc. 2007, 129, 2890–2900.
(27) Hossain, M. A.; Schneider, H.-J. Chem.sEur. J. 1999, 5, 1284–1290.
9
18130 J. AM. CHEM. SOC. VOL. 131, NO. 50, 2009

Enthalpic Nature of the CH/π Interaction
A R T I C L E S
formed. Therefore, substitutions at the hydroxyl groups were
introduced. In principle, the chosen substituent should not
interact with benzene to a significant extent; thus, methyl groups
were chosen as a first approximation. Moreover, this structural
modification has the advantage of removing the possibility of
hydrogen bonding, which could compete with the CH/π
interaction.
To make the chemical modifications to the compounds of
interest, permethylation was done according to the conditions
described by Wang et al.28 using iodomethane as the alkylating
agent. Compounds were purified and characterized by different
techniques (see Supporting Information). Compound ꢀ-Me5Gal
(4) was crystallized, and its X-ray diffraction-based structure
is shown in Figure 3.
One more important issue that could arise is the fact that the
methyl groups could generate a congested environment that
would interfere with the interaction. Figure 3 shows the two
conformers present in the crystalline network of ꢀ-Me5Gal. In
the solid state, ꢀ-Me5Gal has an extended conformation where
there is enough space for the approach of a benzene molecule.
In fact, there are four C-H bonds on the R face of the
carbohydrate available for providing CH/π interactions. Mobility
of the O-methyl groups could challenge the approach of the
aromatic ring, and establishment of the interaction could, in
addition, constrain the expected motion of these pendant groups
in solution with the concomitant conformational restrictions and
the corresponding entropic penalty.
Figure 2. Structures of the benzene/ꢀ-fucose (1) complex17 and benzene/
R-fucose (2) complex at the MP2/6-31G(d,p) level of theory with BSSE
correction incorporated during the full geometry optimization procedure.
relationship between molecular complementarities and the
specificity of the interaction?
At the beginning of this study, a nonsubstituted regular sugar
was employed. Using computer simulations, it is possible to
orient the aromatic group (benzene) toward the nonpolar region
of the carbohydrate, thus avoiding the participation of hydrogen
bonds, which prevails over CH/π interactions in Vacuo. To
simplify the conformational problems posed by the presence
of the sugar hydroxymethyl group, it was changed to a simple
methyl group, as in 6-deoxy sugars. Therefore, we used fucose
as the model for galactose. Moreover, to restrict the degrees of
freedom of the hydroxyl groups, intramolecular OH · · ·OH
hydrogen bonds were forced by orienting the hydroxyls in a
cooperative counterclockwise (from the perspective of the
carbohydrate’s ꢀ face) manner.
The supramolecular structures of the ꢀ-(1) and R-fucose (2)
associated with benzene obtained at the MP2/6-31G(d,p) level
with the basis set superposition error (BSSE) correction
incorporated during the full optimization process of the geometry
are presented in Figure 2. The interaction energy reported for
the benzene/1 complex was previously described17 and is -12.5
kJ mol-1, while we calculated that for epimer 2 in this study at
the same level of theory to be -11.3 kJ mol-1. Even when in
both supramolecules the benzene is oriented laterally, the
presence of a hydroxyl group in the anomeric position (C1) of
2 generates a repulsive environment that, on average, places
the benzene moiety more separated from 2 than from 1,
especially at this region. Following are the distances for
complexes 1 and 2, respectively: C1-C′1, 4.52 and 4.94 Å;
C2-C′2, 4.82 and 5.04 Å; C3-C′3, both 3.96 Å; C4-C′4, 4.05
and 3.97 Å; and C5-C′5, 3.89 and 4.02 Å.
b. Thermochemistry. Although factors such as the disruption
of the packing forces in the presence of a solvent (i.e., benzene),
the differences in the solvation of the solutes, and the degree
of solvation, among others, may play a role in the enthalpic
changes measured in an experiment, it is possible to estimate
the energy associated with the interaction generated by van der
Waals complexes because the system used as the reference (6)
lacks the interaction of interest but it can be assumed, through
the design of the experiment, that all the other conditions are
similar. Thus, by measuring the amount of heat exchanged when
benzene binds to the carbohydrate and comparing the systems,
it is possible to assume that the difference in energy is related
to the CH/π interaction.
Nevertheless, according to the X-ray structure for the putative
complex with benzene, in the key sugar regions where three
C-H vectors exist close in space to generate the patch for
benzene interaction, the existing oxygen atoms are located in
such a way that they are not able to generate repulsive
environment to form the van der Waals complex.
Rigorous analysis of the interaction energy should also
consider the solvation energy. In the present context, the term
“solvation process” refers to the energetic and structural changes
occurring in a system upon transferring solute molecules into
the solvent. These changes could result in the disruption of
intramolecular interactions within the solvent and solute, as well
as the formation of new intermolecular interactions between the
molecules of solute and solvent.29 In our case, when comparing
the different carbohydrates, the existing differences in the
solvation enthalpies should reflect the magnitude of the interac-
tion of each sugar isomer with the solvent molecules. Scheme
1 shows the terms required to determine this thermodynamic
property. The enthalpy of phase change as well as the dissolution
enthalpy can be determined by direct measurement in a
calorimeter.11 The dissolution enthalpy values of R-Me5Gal (3),
ꢀ-Me5Gal (4), ꢀ-Ac5Gal (5), and R-Me5Man (6) solutes in
benzene (∆disH°) were then determined.
m
When considering the obtained ∆disH° values, it is necessary
m
to take into account the heat of the broken intramolecular
interactions and the heat released when each molecule of solute
gets in the bulk of the solution and is surrounded by a shell of
solvent molecules.30 It could be deduced that, for the solid
compounds 4 and 5, a positive value of this quantity was
observed, indicating that these compounds are more stable in
(28) Wang, H.; Sun, L.; Glazebnik, S.; Zhao, K. Tetrahedron Lett. 1995,
36, 2953–2956.
(29) Krestov, G. A. Thermodynamics of solVation; Ellis Horwood: New
York, 1991.
Nevertheless, since it is essential that the carbohydrate be
soluble in benzene, the direct experiment could not be per-
(30) Berry, R. S.; Rice, S. A.; Ross, J. Physical Chemistry; John Wiley
and Sons: New York, 1980.
9
J. AM. CHEM. SOC. VOL. 131, NO. 50, 2009 18131

A R T I C L E S
Ram´ırez-Gualito et al.
Figure 3. Conformation of ꢀ-Me5Gal (4) in the solid state.
Scheme 1. Thermodynamic Determination of the Heat of Solvation
chemical nature in comparison to those of the O-methyl group
(as in 3, 4, and 6).
The phase change enthalpies reflect the cohesive energy of
the crystalline lattice in the case of the solid compounds. Strong
dipolar intermolecular interactions must be predominant in
compound 5 due to the O-acetyl groups. In fact, the heat of
sublimation of this compound is the highest of the four
substances studied. In pure compounds 3, 4, and 6, intermo-
lecular interactions between O-methyl groups must be generated
by dispersion forces, thus significantly reducing the heat of
sublimation of ꢀ-Me5Gal (4) with respect to that of ꢀ-Ac5Gal
(5), and even making compounds 3 and 6 liquids.
Table 1. Enthalpies of Solvation (in kJ mol-1) of Methyl
2,3,4,6-Tetra-O-methyl-R-D-galactopyranoside (R-Me5Gal, 3),
Methyl 2,3,4,6-Tetra-O-methyl-ꢀ-D-galactopyranoside (ꢀ-Me5Gal,
4), 1,2,3,4,6-Penta-O-acetyl-ꢀ-D-galactopyranose (ꢀ-Ac5Gal, 5),
and Methyl 2,3,4,6-Tetra-O-methyl-R-D-mannopyranoside
(R-Me5Man, 6) in benzene (1:10, mol:mol)a
It should be noted that, for solid compounds 4 and 5, the
heat of sublimation ∆gcrH° is reported. On the other hand, for
m
liquid compounds 3 and 6, the heat of vaporization ∆gl H° is
m
b
c
∆
disHm°
∆gcrHm° or ∆gl Hm°
∆
solvHm°
σtot, (
reported. The solvation energy ∆solvH° of the mannose deriva-
m
tive, 6 (-78.8 ( 3.9 kJ mol-1), indicates that this is the least
3
4
5
6
-0.67
16.54
12.10
-3.65
88.37
105.25
144.64
75.18
-89.04
-88.71
-132.54
-78.83
2.00
5.54
6.19
3.91
stable system in benzene solution. As expected from the
stereochemistry of the pyranose ring, the ∆solvH° values are
m
higher for compounds 3 (-89.0 ( 2.0 kJ mol-1) and 4 (-88.7
( 5.5 kJ mol-1). In principle, we can infer that the energy
difference between galactose derivatives 3 and 4, and the
mannose analogue, 6, is due mainly to the CH/π energy. This
can be estimated through eq 1.11
a Calculated from the respective enthalpies of dissolution, determined
by Calvet microcalorimetry at 303.15 K, and the enthalpies of phase
change, determined by scanning calorimetry. b For solid compounds 4
and 5 it is reported the heat of sublimation ∆gcrH°, and for liquid
m
c
compounds 3 and 6 it is reported the heat of vaporization ∆gl H°. σtot is
m
the overall uncertainty on the enthalpy of solvation, calculated from the
uncertainties on the enthalpies of dissolution and phase change (see
Supporting Information).
CH/π energy ) ∆solvHom(3 or 4) - ∆solvH°m(6)
(1)
At this point, it is possible to assume that the ∆solvH° term
m
of 6 only includes the interactions between the O-methyl groups
and the solvent, while the same term ∆solvH° for 3 and 4 also
m
pure form than in solution. It can be anticipated that breaking
the crystalline form of these molecules is energy demanding.
In contrast, a negative, exothermic dissolution enthalpy can be
observed for the two liquid compounds, 3 and 6. The results
are gathered in Table 1.
includes the CH/π interaction. According to this approximation,
the CH/π interaction between the methyl galactopyranoside
epimers and benzene amounts to 10.2 and 9.9 kJ mol-1 for 3
and 4, respectively. Although only an approximation, these
values are strikingly in agreement with the values calculated at
the MP2/6-31G(d,p) level when the BSSE correction is included
during the optimization process, 12.5 kJ mol-1 for model
complex 117 and 11.3 kJ mol-1 for 2.
The mannose derivative 6 can be used as a reference since,
according to our previous studies, the pyranose ring does not
have the substitution pattern required for establishing a CH/π
interaction11 (at least three C-H bonds exposed in the same
molecular region).17 Thus, compound 6 cannot interact with
benzene as a van der Waals complex. On the other hand, the
data for 5 (ꢀ-Ac5Gal) were employed to estimate the effect of
the methyl group on the interaction. In this compound, the more
acidic hydrogen atoms of the O-acetyl group show a distinct
Carbohydrate-aromatic interactions have been widely studied
using computational methods. The obtained energy values
oscillate between 10.5 and 20.9 kJ mol-1 when the MP2 method
is used. Since the different theoretical methods include different
levels of approximation,31-33 with diverse considerations of the
repulsive and attractive, it is of paramount importance to access
9
18132 J. AM. CHEM. SOC. VOL. 131, NO. 50, 2009

Enthalpic Nature of the CH/π Interaction
A R T I C L E S
the experimental reference values around 10.0 kJ mol-1 in order
to estimate the quality of the different theoretical treatments.
For the penta-O-acetyl compound 5, the heat of sublimation
is 144.6 kJ mol-1, in concordance with its vapor pressure.
Interestingly, the interaction of this molecule with benzene
amounts to -132.5 kJ mol-1, the most stabilizing one observed
in the present study. If there were no stabilization mediated by
the CH/π interaction with the hydrogen atoms at the ring, the
observed stabilization energy would have its origin in the
independent interaction of the acetyl groups with the benzene
molecules. If the solvation value of 5 is subtracted from the
solvation of compound 6, there is a difference of -53.7 kJ
mol-1. It is possible to contemplate that, on average, each of
the O-acetyl group interactions contributes with -10.7 kJ mol-1,
a value expected for a CH/π interaction.
Table 2. Chemical Shift Difference, ∆δ ) δCDCl3 - δC6D6 (in
ppm), for Compounds 3-6
R-Me5Gal, 3
ꢀ-Me5Gal, 4
ꢀ-Ac5Gal, 5
R-Me5Man, 6
H1
H2
H3
H4
0.01
-0.18
-0.08
0.17
-0.05
-0.01
-0.04
0.22
0.25
0.25
0.10
0.27
0.01
-0.37
0.18
0.20
0.24
0.04
-0.02
0.13
0.02
0.25
0.10
0.30
-0.13
0.15
0.13
-0.10
-0.20
-0.15
0.01
0.06
0.24
0.24
0.26
-0.38
-0.09
0.01
H5
0.60
0.06
0.06
0.56
H6proS
H6proR
Me(1)
Me(2)
Me(3)
Me(4)
Me(5)
0.42
0.41 or 0.29a
0.61
0.10
0.19
0.46 or 0.34a
a Chemical shifts of methyl(3) and methyl(5) are interchangeable.
c. NMR Determinations. As mentioned above, thermody-
namic data can be used to establish the existence of the
interaction but not the region of the molecule where it happens.
NMR can be safely employed for this purpose. This technique
has been nicely used for the detection of intermolecular NOEs
to study interactions of specific complexes such as cyclodextrins
and trifluoroethanol34 or cyclodextrins and aromatic com-
pounds.35 In this context, the use of an anisotropic titration
procedure has established that ꢀ-galactose and benzene11 or
phenol17 interact in the region of the C3-C6 segment of the
carbohydrate, basically as described by the theoretical calculations.
Saccharides are amphipathic substances.36 Substitution of the
carbohydrate hydroxyl with methyl groups increases its hydro-
phobic properties and, in principle, should facilitate the access
of aromatic molecules to the proximity of the pyranose ring.
However, a pattern originated by specific and stabilizing CH/π
interactions would decrease the translational and rotational
degrees of freedom of the molecule. On the other hand, when
considering solvation, in the absence of specific interactions,
the approach of benzene molecules to the sugar would be
random.
If the ring protons are now considered for ꢀ-galactoside 4,
the hydrogen atoms at positions 3, 4, and 5 are strongly shielded,
with values that range between 0.18 and 0.24 ppm. This
experimental fact contrasts with the observations for the
R-mannoside 6, where there is a lack of protection for the same
hydrogen atoms. Indeed, the observed differences are the
opposite, with values ranging between -0.10 and -0.20 ppm.
This is expected since the presence of an equatorial methoxyl
group at position 4 of mannose, just in the region of approach
of benzene, generates a repulsive environment, and thus, the
CH/π interaction does not take place.
Between these two cases, for 3 and 5 there are intermediate
observations. For 3, the protection of H-4 is high (∆δ ) 0.17
ppm), while there is lack of protection of H-3 and H-5 (-0.08
and -0.05 ppm, respectively). The presence of one axial
methoxy group at the anomeric position should account for these
values.
1
The H NMR data of compound 5 are particularly relevant.
The annular protons appear downfield from the methylated
compounds, although this is a regular behavior for O-acetylated
sugars. Besides the electron-withdrawing effect of the acetyl
groups, there are some structural features that deserve attention.
In the solid-state X-ray structures of both ꢀ-Ac5Gal (5)37,38 and
R- Ac5Gal39 (Figure 3), the OCOC segments (where the initial
O is that of the carbonyl group and the final C is part of the
carbohydrate ring of the ꢀ-Ac5Gal) have values of 1.61° for
C1, 7.19° for C2, 4.99° for C3, 3.13° for C4, and 1.47° for C5.
This shows that the different carbonyl groups eclipse the protons
that are part of the galactose ring; these protons are paramag-
netically shifted because they are located in the deshielding
region of the carbonyl group. These conformational preferences
have been studied before,40,41 and it has been reported that the
preferred conformation is that where the C-O bond of the
carbonyl is parallel to the C-H bond of the pyranose ring.
Indeed, our observations, including the chemical shifts, are in
accordance with those studies, suggesting that the conformations
in solution and in the solid state are similar. At the same time,
1
Table 2 presents the differences in the H NMR chemical
shifts of the compounds under study in chloroform-d and
benzene-d6. A positive value indicates shielding, thus protection,
and the possibility that the corresponding proton interacts with
the benzene face. In other words, it is within the anisotropic
cone of the aromatic entity and allows the location of one
benzene molecule close to a specific region of the carbohydrate
ring.
A factor common to all the carbohydrates studied herein is
that all the methyl groups are shielded in the presence of
benzene. For the O-methyl sugars, the observed values oscillate
between 0.10 and 0.30 ppm, while for the acetylated systems,
the values are much larger, ranging from 0.34 to 0.61 ppm.
These observations can be correlated with the different chemical
nature of the O-acetyl with respect to the O-methyl group and
its tendency to interact in a much stronger manner with benzene
molecules.
(31) Spiwok, V.; Lipovova´, P.; Ska´lova´, T.; Vondra´cˇkova´, E.; Dohna´lek,
J.; Hasˇek, J.; Kra´lova´, B. J. Comput.-Aided Mol. Des. 2005, 19, 887–
901.
(37) Thibodeaux, D. P.; Johnson, G. P.; Stevens, E. D.; French, A. D.
Carbohydr. Res. 2002, 337, 2301–2310.
(32) Sujatha, M. S.; Sasidhar, Y. U.; Balaji, P. V. Biochemistry 2005, 44,
8554–8562.
(38) Kumar, R.; Tiwari, P.; Maulik, P. R.; Misra, A. K. Carbohydr. Res.
2005, 340, 2335–2339.
(33) Sujatha, M. S.; Sasidhar, Y. U.; Balaji, P. V. Protein Sci. 2004, 13,
2502–2514.
(39) Roslund, M. U.; Klika, K. D.; Lehtila¨, R. L.; Ta¨htinen, P.; Sillanpa¨a¨,
R.; Leino, R. J. Org. Chem. 2004, 69, 18–25.
(34) Guerrero-Mart´ınez, A.; Berger, S.; Tardajos, G. ChemPhysChem 2006,
7, 2074–2076.
(40) Lopez-Calahorra, F.; Velasco, D.; Castells, J.; Jaime, C. J. Org. Chem.
1990, 55, 3530–3536.
(35) Ribeiro, J. P.; Bacchi, S.; Dell’Anna, G.; Morando, M.; Canada, F. J.;
Cozzi, F.; Jimenez-Barbero, J. Eur. J. Org. Chem. 2008, 5891–5898.
(36) Quiocho, F. A.; Vyas, N. K. Nature 1984, 310, 381–386.
(41) Osawa, E.; Imai, K.; Fujiyoshi-Yoneda, T.; Jaime, C.; Ma, P.;
Masamune, S. Tetrahedron 1991, 47, 4579–4590.
9
J. AM. CHEM. SOC. VOL. 131, NO. 50, 2009 18133

A R T I C L E S
Ram´ırez-Gualito et al.
hydrates 3-5 at the same concentration (ca. 0.24 M), similar
experimental parameters were employed (Figure 5). The dif-
ferent behavior observed for those molecules is notorious. As
expected, for all the molecules, the methyl group showed
significant NOEs, indicating its direct interactions with benzene.
In contrast, the NOEs observed for the ring protons of the
different molecules are rather different. For ꢀ-Me5Gal (4) there
are clear intermolecular NOEs on H-3, H-4, and H-5 based upon
the difference of the steady NOE spectra, supporting the
existence of interactions between the carbohydrate and the
aromatic moiety. The observed NOEs for R-Me5Gal 3 are
smaller than those for 4, and those observed for ꢀ-Ac5Gal (5)
are just above the noise signal.
Therefore, the experimental NOE data (supported by the
chemical shift perturbation analysis) seem to indicate that
benzene interacts with the different sugars, but much more
strongly with the bottom ring face of ꢀ-Me5Gal (4), as also
indicated by the theoretical calculations. It is noticeable that
the ring protons of mannopyranoside 6 have very little interac-
tion with benzene, although it has been observed that this
interaction depends on the sample concentration (the study of
this effect is in progress, see Supporting Information). Finally,
for the peracetylated compound 5, very weak NOEs are
observed. This observation may be safely explained by the lack
of direct interaction between benzene and the sugar ring protons.
The tiny effects observed could be due to indirect reorientation
effects provoked by the interaction of benzene molecules with
the methyl groups of the O-acetyl substituents. The different
intensities of the NOEs observed in Figure 5 for each carbo-
hydrate can be associated with its proximity to the benzene ring.
In all cases, the methyl groups that are necessarily solvated and
close to the solvent show intense signals. The intensity of the
signals follows the order of approximation established by the
theoretical calculations: higher intensities and thus closer
proximity of the ꢀ-galactose derivative, less for the R isomer,
and very small for the acetylated compound (Figure 6). For the
peracetylated compound ꢀ-Ac5Gal (5), the preferred interaction
of the benzene molecules with the acetyl groups could be
observed through NOESY cross peaks (see Supporting Informa-
tion).
Figure 4. Superposition of ꢀ-Me5Gal (4, ball and sticks) with the R-Ac5Gal
(top) and ꢀ-Ac5Gal 5 analogues (bottom) (wire frame).
this conformational arrangement puts the oxygen atom in the
region needed for the interaction with benzene. Similar results
are observed in the R-Ac5Gal anomer. Therefore, the interaction
of the ring protons with the benzene moieties is geometrically
disfavored by the repulsive environment generated by the
carbonyl orientation. Figure 4 shows the superposition of one
of the ꢀ-Me5Gal conformers, 4 (balls and sticks), with the
anomers ꢀ-Ac5Gal, 5 (Figure 3a), and R-Ac5Gal (Figure 3b)
(wire frame), suggesting the difficulty of establishing proper
interactions between the ring protons of the acetylated sugars
and an approaching aromatic moiety.
A scheme representing the observed interactions is shown in
Figure 6. Very few experimental observations on the existence
of aromatic-sugar interactions when exploring acetylated or
methylated carbohydrate molecules have been reported and
systematically compared. Recently, Waters et al.46-48 studied
the attractive interaction between one tryptophan residue and a
diagonally cross-stand tetra-O-acetyl-Glc-Ser, fundamental to
stabilize the folding of a ꢀ-hairpin.48 NMR data were used to
demonstrate the existence of the interaction between the R face
of the carbohydrate and the tryptophan residue. The authors
stated that, when the acetyl groups are eliminated, the interaction
with tryptophan is lost. This result is not expected since CH/π
interactions do, indeed, take place for natural carbohydrates
(unsubstituted) in aqueous solutions.17 Thus, the absence of the
acetyl groups should maintain the CH/π interaction if it exists.
Given the particular geometrical arrangement of the studied
glycopeptides, it could be possible that instead of (or besides)
Indeed, for 5, only the region around H-5 on the side of the
hydroxymethyl group is available for the interaction with the
aromatic moiety. This explains the observed protection and thus
shielding of H-5.
The NMR results may also be satisfactorily integrated with
the calorimetry data. In fact, although in compound 5 the
hydrogen atoms of the ring are not capable of participating in
CH/π interactions, the molecule can interact with the aromatic
compound through the O-acetyl groups. Again, on this basis, it
can be assessed that the nature of the O-methyl and O-acetyl
groups is rather different.
NOE experiments were also performed to try to detect the
possibility of close proximity between the carbohydrate and
benzene. The detection of intermolecular NOEs between solvent
and small organic molecules has been used to establish the way
in which solutes are solvated.34,42-45 For the modified carbo-
(42) Berger, S.; Diaz, M. D.; Hawat, C. Pol. J. Chem. 1999, 73, 193–197.
(43) Bagno, A.; Rastrelli, F.; Scorrano, G. J. Magn. Reson. 2004, 167, 31–
35.
(46) Kiehna, S. E.; Laughrey, Z. R.; Waters, M. L. Chem. Commun. 2007,
4026–4028.
(44) Angulo, M.; Hawat, C.; Hofmann, H.-J.; Berger, S. Org. Biomol. Chem.
2003, 1, 1049–1052.
(47) Tatko, C. D.; Waters, M. L. J. Am. Chem. Soc. 2004, 126, 2028–
2034.
(45) Brand, T.; Cabrita, E. J.; Berger, S. Prog. Nucl. Magn. Reson.
Spectrosc. 2005, 46, 159–196.
(48) Laughrey, Z. R.; Kiehna, S. E.; Riemen, A. J.; Waters, M. L. J. Am.
Chem. Soc. 2008, 130, 14625–14633.
9
18134 J. AM. CHEM. SOC. VOL. 131, NO. 50, 2009

Enthalpic Nature of the CH/π Interaction
A R T I C L E S
Figure 5. Comparative NOEs on compounds 3, 4, and 5 at 298 K upon inversion of the solvent signal (benzene). The intensity scale is the same for all
cases.
Figure 6. Origin of the observed NOE for compounds 3 and 4 (top), 5 and 6 (bottom).
9
J. AM. CHEM. SOC. VOL. 131, NO. 50, 2009 18135

A R T I C L E S
Ram´ırez-Gualito et al.
an increase in the cost of solvation of the unsubstituted glucose
suggested by the authors, additional dominant interactions
between the methyl group of one of the acetyl groups and
tryptophan take place. The energy associated with the phenom-
enon would certainly point toward this interaction and would
be in line with our results. A study of the interaction between
the carbohydrate and the aromatic in systems free of other
interactions would be needed to better identify its energy and
dynamics.
Obviously, in complex molecules, such as those glycopeptides
mentioned above, the interaction phenomenon would be a
consequence of different processes, including the solvation/
desolvation process. However, it has also been shown that
nonsubstituted sugars may also effectively interact with aromatic
moieties, in a nonambiguous manner.17 Thus, for each particular
case, many factors should be considered, including sugar ring-
to-aromatic interactions, methyl-to-aromatic interactions, and
the role of solvation, especially for complex biomolecules. In
any case, from our perspective, it is sound to systematically
study the interaction between carbohydrates and aromatic
moieties in systems free of additional interactions, which may
substantially modify and modulate the structural observations
and energy measurements.
Today, structures of various complexes formed by carbohy-
drates bound to the recognition sites of proteins determined by
X-ray are available in different databases. Many lectin/sugar
complexes relevant to this study have been described.49 For
example, the complex formed by evulin I (a nontoxic protein
that inactivates ribosomes) and ꢀ-galactose, where the relevant
interaction occurs with a tryptophan residue of the protein, is
in perfect agreement with the results described here.21 The same
is true for the interaction between the nonreducing residue of
the disaccharide RGal1,3ꢀGalOMe and a tryptophan residue of
isolectin 1-B4 from Griffonia simplicifolia.50 A good example
where the aromatic residue is tyrosine is fucosyllactose
(RFuc1,2ꢀGal1,4Glc), when this amino acid interacts with lectin
II of Ulex europaeus.51 In light of the study presented, it is
possible to establish that the interactions between the carbohy-
drate and the aromatic residues of the proteins mentioned above
have a stabilizing nature and an effect on the enthalpic term of
the Gibbs free energy. It is noticeable that the CH/π interaction
is a key contribution, in addition to rest of the typical interactions
of the biological media that include hydrogen bonds, among
others.
detected. For acetylated sugars, chemical shifts and NOE data
demonstrate that the methyl groups of the O-acetyl substituents
effectively interact with benzene. These methyl groups generate
a repulsive environment around the sugar ring that precludes
efficient CH/π interaction with the R-face of the sugar. Calo-
rimetry measurements indicate that every methyl group con-
tributes to the stability of the system with an energy of ca. -10.7
kJ mol-1. Moreover, methyl groups in O-acetyl substituents
behave differently from those in O-Me moieties when interacting
with benzene. Although not completely substantiated, this fact
may have its origin in the polarization of the methyl hydrogens
in the O-acetyl group, which contributes to its well-known
acidity. Benzene can recognize and differentiate the R- and
ꢀ-anomers of permethylated galactose from the mannose
analogue and from the corresponding per-O-acetylated galactose
derivatives. For the simple per-O-acetylated monosaccharide
models, such as those studied herein, the acetyl groups do
interact intensely with benzene and, in addition, block its access
to the (R-) R-face of the carbohydrate. Previous reports
suggesting that a methyl from an acetyl group is similar to a
methyl group of a methoxy group were reinterpreted in terms
of our findings. The observed effect could be a consequence of
the methyl of the acetyl group-aromatic interaction instead of
a carbohydrate-aromatic one.
Our novel use of a combination of NMR measurements with
theoretical calculations and calorimetric data has satisfactorily
supportedaninteractionmodelforsubstitutedcarbohydrate-aromatic
interactions in simple models and could be adopted for other
similar studies. The role of solvent and osmolites for modulating
the observed interaction is now under study in our laboratories.
Experimental Section
General Procedure To Carry Out the Permethylation of
28
Carbohydrates.
Methyl-D-pyranoside (1 g, 5.15 mmol, 1.0
equiv) was dissolved in dimethyl sulfoxide (7 mL). NaOH aqueous
solution (50%, w/w) (10 mL, 50 mmol, 9.70 equiv) was added
slowly. The mixture was stirred to form a gel suspension, and CH3I
(3 mL, 48.18 mmol, 9.35 equiv) was added dropwise. The reaction
mixture was stirred for 24 h. Water (100 mL) was added, and the
aqueous phase was extracted with CH2Cl2 (3 × 100 mL). The
combined organic phases were dried (Na2SO4), filtered, and
concentrated in Vacuo. All compounds were purified by flash
column chromatography (silica gel, hexane/ethyl acetate 9:1). The
product, methyl 2,3,4,6-tetra-O-methyl-ꢀ-D-galactopyranoside (3),
was dissolved in EtOAc (30 mL), activated charcoal was added
(20%, w/w), and the black suspension was stirred for 20 min. The
final product was recrystallized from hexane/CH2Cl2 (g99.9%).
Methyl 2,3,4,6-tetra-O-methyl-R-D-galactopyranoside (4) and
methyl 2,3,4,6-tetra-O-methyl-R-D-mannopyranoside (5) were puri-
fied by flash column chromatography (silica gel, hexane/ethyl
acetate 9:1) and by activated charcoal column (CH2Cl2) to get
g99.9% purity for each compound. NMR signals assignments for
compound 4 were done using HSQC, HMBC, NOESY, and COSY.
2D experiments are included in the Supporting Information.
Methyl 2,3,4,6-Tetra-O-methyl-ꢀ-D-galactopyranoside (3).52-54
Yield: 875 mg (68%). White solid, mp 45-47 °C. IR (KBr): 2979,
2921, 2836, 1448, 1373, 1334, 1305, 1183, 1073, 1106, 957, 997,
957, 908, 876, 740, 660. 1H NMR (500 MHz, C6D6): 2.97 (dd, J )
9.5, 3.0, 1H3); 3.11 (s, 3H, Me5); 3.27 (s, 3H, Me3); 3.33 (ddd, J
) 1.5, 5.5, 7.5, 1H5); 3.37 (s, H, Me1); 3.46 (dd, J ) 1.0, 3.0,
Conclusions
The interaction energies of R-Me5Gal (3), ꢀ-Me5Gal (4),
ꢀ-Ac5Gal (5), and R-Me5Man (6) with benzene, computed as
the enthalpy of solvation determined using Calvet microcalo-
rimetry and differential scanning calorimetry, are -89.0 ( 2.0,
-88.7 ( 5.5, -132.5 ( 6.2, and -78.8 ( 3.9 kJ mol-1
,
respectively. This confirms that the nature of these interactions
is enthalpic.
From the structural viewpoint, the interacting regions have
been determined by using NMR experiments. While for
ꢀ-Me5Gal, the interaction takes primarily place using the (R-)
R-face of the carbohydrate for establishing CH/π interactions
with benzene, for R-Me5Man no interaction of this type is
(49) http://webenligne.cermav.cnrs.fr/lectines.
(50) Templel, W.; Tschampel, S.; Woods, R. J. J. Biol. Chem. 2002, 277,
6615–6621.
(52) Asres, D. D.; Perreault, H. Can. J. Chem. 1999, 77, 319–325.
(53) Rathbone, E. B.; Stephen, A. M.; Pachler, K. G. R. Carbohydr. Res.
1971, 20, 141–150.
(54) Rathbone, E. B.; Stephen, A. M.; Pachler, K. G. R. Carbohydr. Res.
1972, 21, 73–81.
(51) Loris, R. D. G. H.; Dao-Thi, M.-H.; Messens, J.; Imberty, A.; Wyns,
L. J. Mol. Biol. 2000, 301, 987–1002.
9
18136 J. AM. CHEM. SOC. VOL. 131, NO. 50, 2009

Enthalpic Nature of the CH/π Interaction
A R T I C L E S
1H4); 3.50 (dd, J ) 5.5, 9.0, 1H6a); 3.57 (s, 3H, Me2); 3.64 (dd, J
) 8.0, 9.5, 1H6b); 3.68 (dd, J ) 7.5, 10.0, 1H2); 4.15 (d, J ) 7.5,
1H1). 13C NMR (125 MHz, C6D6): 56.3 (Me1); 58.3 (Me3); 58.8
(Me5); 60.7 (Me2); 60.9 (Me4); 71.4 (C6); 73.4 (C5); 75.6 (C4); 81.3
(C2); 84.9 (C3); 105.3 (C1). CI-MS: 251 ([M+H]+; [C11H22O6]H+).
Methyl 2,3,4,6-Tetra-O-methyl-r-D-galactopyranoside (4).52-54
Yield: 759 mg (59%). Yellow liquid. IR (film): 2979, 2910, 2829,
1450, 1359, 1252, 1199, 1098, 1053, 986, 954, 883, 764, 665.1H
NMR (500 MHz, C6D6): 3.14 (s, 3H, Me5); 3.20 (s, 3H, Me1); 3.26
(s, 3H, Me2); 3.27 (s, 3H, Me3); 3.52 (m, 1H6a); 3.53 (m, 1H4);
3.60 (dd, J ) 2.5, 9.0, 1H6b); 3.62 (m, 1H3); 3.82 (dd, J ) 4.0,
10.5, 1H6b); 3.90 (m, 1H5); 4.77 (d, J ) 3.0, 1H1). 13C NMR (125
MHz, C6D6): 55.0 (Me1); 58.2 (Me3); 58.4 (Me2); 58.8 (Me5); 61.0
(Me4); 70.0 (C5); 71.8 (C6); 77.0 (C4); 78.8 (C2); 81.0 (C3); 98.8
(C1). CI-MS: 251 ([M+H]+; [C11H22O6]H+).
involved in each dissolution experiment are provided in the
Supporting Information.
General Procedure To Measure the Enthalpies of Sublimation
and Vaporization by Differential Scanning Calorimetry. The
calorimetric measurements of the enthalpy associated with the change
from condensed to gas phase were performed using the isothermal
or scanning operation of a modified Perkin -Elmer DSC7 differential
scanning calorimeter.56,57 The sensitive element of this device is a
DSC7 calorimetric holder assembly, located inside of a vacuum
chamber and connected to the DSC7 analyzer by an electrical feed.
The vacuum chamber is evacuated with a rotary vacuum pump,
and residual pressure is monitored by a pressure gauge relayed to
a Pirani gauge control. Perkin-Elmer 0219-0041 open standard
aluminum pans were utilized as vaporization or sublimation cells.
For isothermal vaporization experiments on the methyl 2,3,4,6-
tetra-O-methyl-ꢀ-D-galactopyranoside, a temperature of 323.15 K
was established as the most suitable from preliminary tests. For
the measurement experiments, samples of around 10 mg of the
liquid substance were placed inside the vaporization pan and
weighed on a Sartorius 4503 microbalance sensitive to 1.0 µg. Once
the prepared sample pans were loaded in the calorimetric sensor,
temperature and heat flux were stabilized and data acquisition began.
Three minutes was enough to get a good initial baseline, and then
a valve relaying the vacuum chamber to the vacuum pump was
opened, the pressure inside the chamber was downloaded quickly
to promote the vaporization process, and the complete calorimetric
curve was registered in a lapse of 30 min.
For the phase change experiments with the solid 1,2,3,4,6-penta-
O-acetyl-ꢀ-D-galactopyranose, from preliminary tests, scanning opera-
tion and a range of temperature from 403.15 to 473.15 were established
as the most appropriate. For the measurement experiments, samples
of around 11 mg of the solid carbohydrates were placed inside the
sublimation pan, and the set was weighed to 1.0 µg of sensitivity. The
prepared sample pans were loaded in the calorimetric sensor, and then
the pressure inside the vacuum chamber was fixed in the range of
100-150 Pa, while the temperature of the sample was held at 403.15
K. After 3 min for stabilizing the calorimeter’s heat flux signal,
scanning the temperature at a rate of 10 K/min started. In the range of
428-443 K, each calorimetric curve showed a sharp peak due to the
melting of the sample, immediately followed by a wide rounded signal
in the interval of 443-455 K, due to the vaporization of the melted
substance.
In isothermal as well as in scanning methodology, throughout
loading and thermal stabilization of the calorimetric system, a small
fraction of the substance vaporizes or sublimes; therefore, an
accurate quantification of the mass lost in this part of the
experimental procedure is necessary and was performed by
independent experiments as previously described.46,47
Methyl 2,3,4,6-Tetra-O-methyl-r-D-mannopyranoside (6).55
Yield: 797 mg (62%). Yellow liquid. IR (film): 2980, 2909, 2829,
1
1451, 1377, 1291, 1191, 1114, 1065, 997, 971, 871, 795, 662. H
NMR (500 MHz, C6D6): 3.13 (s, 3H, Me1); 3.12 (s, 3H, Me5); 3.22
(s, 3H, Me3); 3.23 (s, 3H, Me4); 3.42 (s, H4); 3.43 (dd, J ) 2.0,
3.0, 1H2); 3.53 (dd, J ) 2.0, 11.0, 1H6a); 3.58 (m, 1H6b); 3.60 (m,
1H3); 3.62 (m, 1H4); 3.70 (m, 1H5); 3.68 (dd, J ) 7.5, 10.0, 1H2);
4.64 (d, J ) 2.0, 1H1). 13C NMR (125 MHz, C6D6): 54.4 (Me1);
57.2 (Me3); 58.9 (Me2); 59.0 (Me5); 60.4 (Me4); 72.4 (C6); 72.5
(C5); 77.0 (C4); 77.4 (C2); 82.6 (C3); 99.0 (C1). CI-MS: 251
([M+H]+; [C11H22O6]H+).
Compound 1,2,3,4,6-penta-O-acetyl-ꢀ-D-galactopyranose (5,
g99.0%) was purchased from Aldrich and used without further
purification.
NMR Experiments. Modified methylpyranoside (15 mg, 0.06
mmol) was dissolved in a mixture of C6D6-C6H6 (0.25 mL-0.25
mL) to measure the NOE effect using a Bruker 500 MHz
spectrometer. 1D NOE difference spectra were obtained upon
irradiation of the benzene signal and internal subtraction of data
acquired by on-resonance and off-resonance selective excitation on
alternate scans.
X-ray Determination. Solid-state structure was resolved in a
Bruker Smart Apex CCD X-ray diffractometer. See Supporting
Information for details.
General Procedure To Measure the Enthalpies of Dissolution
by Heat Flux Calorimetry. Dissolution experiments were performed
by heat flux calorimetry, using a differential Setaram C80 Calvet
calorimeter working in isothermal mode at 303.15 K. Sensitivity
and temperature control of the calorimetric device has been
described elsewhere.10 For dissolution experiments, stainless steel
mixing with membrane vessels were employed, and the mass of
each pyranoside and aromatic solvent was calculated in order to
generate the maximal possible thermal signal but with a resulting
molar relation carbohydrate-solvent, after the dissolution process,
as near as possible to 1:10. The masses of the substances involved
in each dissolution experiment were measured in an MC210 P
Sartorius balance sensitive to 10 µg. After loading of the mixing
vessels into the fluxmeters of the Calvet calorimeter, temperature
and heat flux were stabilized by 60 min, and then data acquisition
was started. Five minutes was enough to get a good initial baseline,
and then the aluminum membrane that separates carbohydrate from
the solvent inside of the mixing vessel was broken, and reversing
of the C80 calorimeter was performed during 15 min to promote a
total dissolution process. Analysis of the amplified dissolution
curves, generated by the data treatment software of the C80
calorimeter, showed that a lapse of 115 min was enough for a total
heat transfer from the dissolution cell to the fluxmeters. The C80
calorimeter works at constant pressure; consequently, integration
of the curve of heat flux as a function of time releases directly the
enthalpy of dissolution of each carbohydrate in the respective
aromatic solvent. Tables including all data of mass and heat
Data acquisition and integration of the area under each calori-
metric curve were performed using the Pyris software of the DSC-7
calorimeter. Prior to the measurements, the calorimetric system was
calibrated for energy and temperature using high-purity samples
of indium and zinc. Tables providing detailed experimental data
and the procedure to calculate the enthalpy of the phase change
are supplied in the Supporting Information.
Acknowledgment. This paper is dedicated to the memory of
Prof. Irma González-Bravo. K.R.-G. thanks the Consejo Nacional
de Ciencia y Tecnolog´ıa (CONACYT) for scholarships. This work
was supported by CONACYT through grants 49921-Q and 47679-
Q, and by DGAPA-UNAM grant IN-209606. J.J.-B. Thanks
“Ramón y Cajal” programme and MICINN (Spain), grant CTQ2006-
10374-C02-01. We are much indebted to Dr. R. Alfredo Toscano
for providing single-crystal X-ray diffraction data. We are grateful
(56) Rojas, A.; Orozco, E. Thermochim. Acta 2003, 405, 93–107.
(57) Rojas-Aguilar, A.; Orozco-Guaren˜o, E.; Mart´ınez-Herrera, M. J. Chem.
Thermodyn. 2001, 33, 1405–1418.
(55) Handa, N.; Montgomery, R. Carbohydr. Res. 1969, 11, 467–484.
9
J. AM. CHEM. SOC. VOL. 131, NO. 50, 2009 18137

A R T I C L E S
Ram´ırez-Gualito et al.
to Ing. Luis Velasco and Dr. Javier Perez Flores for technical
support and to Rebeca Lo´pez-Garc´ıa for the revision of the English
version of this manuscript. We also thank the Coordinación de
Comunicación y Sistemas de la Universidad La Salle and DGSCA-
UNAM for computer support.
and sublimation of compound 5 (Tables S5 and S7, pp S8, S10);
NOESY 500 MHz NMR spectrum of compound 5 (Figure S24,
p S34); and CIF file for 4. This material is available free of
charge via the Internet at http://pubs.acs.org. CCDC-746171
ꢀ-Me5Gal (4) contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from
the Cambridge Crystallographic Data Center via www.ccdc.
cam.ac-uk/data_request/cif.
Supporting Information Available: Full optimized geometries
of the supramolecular complexes; complete ref 26; 1H 500 MHz
NMR spectra of compounds 3-6 (Figures S13-S16, pp
S12-S15); enthalpies of dissolution and vaporization of com-
pound 3 (Tables S4 and S6, pp S8, S9); enthalpies of dissolution
JA903950T
9
18138 J. AM. CHEM. SOC. VOL. 131, NO. 50, 2009