Why Are Saccharides Dehydrated in the Presence of Electrolytes? Insights from Molecular Modeling and Thermodynamic Measurements.

The mechanisms governing the interactions of neutral polar solutes with ions in aqueous solutions are still poorly understood, despite the importance of this phenomenon in many fields (chemistry, physicochemistry, biology, food industries). In order to go further through the understanding of the molecular mechanisms governing the ions' specific effects, this paper presents a generic method dealing with the characterization and understanding of interactions between saccharides and ions in aqueous systems. For that, an original approach combining a computational technique and experimental measurements (thermodynamic properties) is proposed to explain and rationalize the relationship between the solute hydration and the physical chemistry of the ions in solution (cation/anion, charge, size, and hydration). These relationships make it possible to evaluate the hydration state of a saccharide, a polar neutral molecule, according to the ionic composition, from the knowledge of the ions' hydration properties. This work proposes new insight into molecular mechanisms governing the polar neutral solute/ion interactions and a new understanding of the hydration phenomenon in electrolytic solutions.

More than a century ago, Hofmeister found a particular ordering of ions in the ability of salts to precipitate egg white.[1] His main conclusion was that the salts’ effectiveness in precipitating proteins is related to the ions’ capacity to order water. Today this phenomenon refers to hydration. An important point that he highlights is that ions cannot be considered as isolated species in water. The salting-out effect is a ubiquitous phenomenon in chemistry, biochemistry, biophysics, and food.[2] Despite the understanding of a crucial specific effect of ions in the development of such areas, the influence of cations and anions on polar neutral solutes is still poorly understood.

In order to contribute to the understanding of the molecular level mechanisms governing the ions’ specific effects on biocompounds, this paper deals with the characterization and the understanding of interactions between saccharides and electrolytes. Indeed, studies of the interactions between saccharides and electrolytes have been attracting renewed interest over the last 20 years.[3−9]

Up to now, the influence of the electrolyte on the saccharide hydration state has been interpreted using the structural hydration model,[10] based on the cosphere concept developed by Gurney.[11] According to this model, the influence of the solute on the structure of water is the solute is represented by a shell of water surrounding the solute. The overlap of the hydration cosphere of the saccharide and of the ions (saccharide/ion interactions) then leads to the saccharide dehydration. More recently, the influence of the electrolyte on the saccharide hydration has been characterized by the molar volume. The study has highlighted that the dehydration mechanism results from multiple interactions taking place at the molecular scale (saccharide/ions; ions/water).[12] Therefore, the link between the saccharide hydration state and the ions’ properties (cation/anion, charge, size, and hydration) needs to be deepened.

Effects observed at the macroscopic scale, such as changes in apparent molar volume, result from nanoscale phenomena. Thus, it is necessary to develop new complementary approaches to characterize the systems at the microscopic scale. Molecular modeling, and more precisely quantum mechanics, are particularly adapted for describing systems at this scale.

In this context, the objective of this work is to go further through the understanding of the mechanisms that govern polar neutral solutes hydration as a function of the ionic composition. The original methodology envisaged is based on experimental and theoretical approaches at different scales. The fundamental approach proposed consists, on the one hand, of characterizing the hydration properties of the ions by using quantum mechanics. On the other hand, it relies on the study of the relationship between the hydration properties of saccharides, based on experimental measurements, and the ion properties obtained.

In our previous work, a systematic volumetric study of saccharides (xylose, glucose, and sucrose) in the presence of various electrolytes (LiCl, NaCl, KCl, Na2SO4, K2SO4, CaCl2, MgCl2, MgSO4) has been carried out from density measurements at 298.15 K.[12] From this data, the saccharide hydration number at infinite dilution nH (expressed in molecules of water per molecule of solute), which explicitly reveals the solute hydration degree, has been calculated for different ionic compositions. The results are given in Supporting Information, in Table S1. First of all, for an infinite dilution in water, saccharide hydration numbers are 3.9, 4.7, and 7.9 for the xylose, glucose, and sucrose respectively (Table S1). These values match the number of hydroxyl groups present in each saccharide (4, 5, and 8 for the xylose, glucose, and sucrose respectively). On the other hand, for an infinite dilution in the various electrolytes, the saccharide hydration number decreases with the increase of the electrolyte molality (for instance, see sucrose in Figures S1 and S2 in Supporting Information). This phenomenon depends on the electrolyte nature. With regard to the impact of the cation, the saccharides dehydration increases according to the following sequence (Figure S1):

Concerning the impact of the anion, the saccharides dehydration increases according to the following sequence (Figure S2):

Whatever the ion (cation or anion), saccharides are further dehydrated in the presence of divalent ions than in the presence of monovalent ones. That can be attributed to stronger attractive electrostatic interactions between the hydrophilic groups of the saccharides and the divalent ions. These interactions govern the dehydration phenomenon. Moreover, for a given ion valence, previous work has highlighted that saccharides dehydration increases with the decrease of the ion free energy of hydration, i.e., lower ion–water interactions. Thus, ion–water interactions have to be considered for a better understanding of the saccharide hydration properties in the presence of various ions.[12]

Ions hydration properties were carried out at the density functional theory (DFT) level of theory, the B3PW91 functional[13] and 6-31++G** basis sets with the Gaussian 09 suite of programs (see Supporting Information). The water-shell around the solutes is obtained using the methodology proposed by Castro et al.[14] for mercury compounds and also by Zhao et al.[15] for samarium complexes. In summary, the solvation shell is constructed in a stepwise manner by adding one water molecule at a time. At each step, a full geometry optimization of the system is performed without any symmetry constraints. The coordinates of the optimized hydrated structures and figures showing the spatial water molecules arrangement are provided in Supporting Information.

The solvation enthalpy of the ion ΔHw is calculated as the difference between the enthalpy of the complex (Hcomplex) and those of the two separated fragments, namely, the ion (Hion) and the water cluster (Hhyd.shell).

The ion first-hydration shell is considered to be complete when the interaction energy between the ion and the water reaches a plateau, i.e., is stable with respect to water addition, as defined by eq . On these structures, the ion coordination number is thus defined as the number of water molecules directly bonded to the central ion in its first layer of solvation sphere. In the following, a water molecule is considered in direct interaction with the ion, if the bond length is shorter than the sum of the van der Waals radii of the atom in interaction.[16] In addition, in the case of anions, hydrogen bonds have to be established between the hydrogen of the water molecule and the anion.

The structural parameters of the ions hydration shell, i.e., the ion coordination number, its optimized coordinated water structure, and the average distance between the ion and the water molecules, are listed in Table . For comparison, other results from molecular modeling are also reported.[17−26] One can state that our values are consistent with the literature, indicating the validity of our approach and assumptions.

Table 1

Structural Parameters of the Ion Hydration Shella

Values obtained in the present work are reported in the first row of the table. CN: coordination number, optimized structure: water cluster coordinated to the ion, d: average distance between ion and coordinated water molecules.

Concerning the cation, the number of water molecules belonging to the first solvation shell depends on the nature of the cation. The optimized building blocks, corresponding to the first observed plateau of the ΔHw = f (water number) curve for the different cations, reveal that 4, 5, 6, 6, and 8 water molecules are respectively involved in the first hydration sphere of Li+, Na+, K+, Ca2+, and Mg2+ (Table ). Interestingly, the coordination number of the monovalent cations can be ranked according to the sequence: Li+ < Na+ < K+ and for the divalent cations: Mg2+ < Ca2+. For a monovalent cation, the K+ coordination number is larger than those of Li+ and Na+, as expected from the larger size of K+ compared to Li+ and Na+. Indeed, because of steric hindrance, smaller ions, such as Li+ for monovalent cations (Mg2+ for divalent cations), have fewer coordinated water molecules than other ions of identical valence. Moreover, the shorter M+–water distance is found for Li+. This shorter distance is associated with the radial expansion of the valence orbital of the cation and indicates that the cation–water interaction involves atomic orbital overlap contributions. Therefore, the shortest is the distance, the largest is the overlap, and as a result, the strongest is the water–cation interaction. Likewise, both the coordination number and the cation–water distance of Mg2+ are lower than those of Ca2+. Comparing monovalent and divalent cations having the same coordination number, for instance, K+ and Mg2+, the Mg2+–water distance is shorter than the K+–water one. This is again associated with the lowest radial expansion of the orbital of divalent metals with respect to monovalent ones. Using the same argument as before, this means that the cation–water interactions are stronger with Mg2+ than with K+.

Concerning the anion hydration properties, the coordination numbers of Cl– and SO42– are respectively 5 and 9. The sulfate anion, which is a polyatomic molecule, is a highly hydrated anion, with a very compact water shell, compared to Cl–. Unlike the water–cation interactions where bonding is essentially due to orbital overlap, the water–anion interactions are mainly hydrogen bonds. For monatomic anions such as Cl–, that leads to a competition between the water network formation and the solvation of the anion. The consequence of this competition is that the structure of the water network around Cl– is relaxed.

To evaluate the role of the electrolyte on the saccharide hydration state, the influence of the cations and anions will be first quantified separately.

Considering first the effect of the cation, the possible existence of correlations between the cation coordination number and the saccharide hydration number in electrolyte solutions was investigated. The experimental saccharide hydration numbers, calculated for different electrolytes containing either Cl– or SO42– and for a given cation molality (1 mol·kg–1), are represented as a function of the computed coordination number of the cations (Figure ). The saccharides are all the more dehydrated as the cations coordination number (i.e., the number of water molecules directly interacting with the cation) increases (Figure ). The large saccharide dehydration observed in the presence of divalent ions can be attributed to the lowest radial expansion of the orbital of divalent metals with respect to monovalent ones, as previously mentioned. This means that the interactions between cations and the oxygen of the saccharide hydroxyl groups are stronger with divalent cations than with monovalent ones. As a consequence, saccharides are more dehydrated. Moreover, focusing on the electrolyte bearing a Cl– counteranion, the saccharide hydration number decreases linearly with the increase of the cation coordination number (linear regression coefficient >0.95 for the three saccharides). The existence of this linear relation indicates that the presence of Cl– has a weak impact on the saccharide hydration. This is further highlighted by the extrapolation of the saccharide coordination number from this linear relationship. Indeed, it is noteworthy that the saccharide hydration numbers, obtained by extrapolation of the straight line to CN = 0, which represents the hydration state of each saccharide in pure water, are very close to the values determined from the saccharide apparent molar volumes. For instance, the glucose hydration number, obtained by extrapolation is 4.75 (electrolyte at 0.15 mol·kg–1), compared to the value obtained from the saccharide apparent molar volumes which is 4.72. Furthermore, the saccharide hydration numbers, for cations having the same coordination number (K+ and Mg2+), are close. For example, the hydration number of xylose with KCl is 3.56, which is very close to the value of 3.51 determined in the presence of MgCl2. A stronger difference between the sucrose hydration number in the presence of KCl (nH = 7.27) and of MgCl2 (nH = 7.19) is, however, observable. Indeed, sucrose is more sensitive to the dehydration caused by the presence of the divalent cations because of its large number of hydroxyl groups. Besides, the saccharide dehydration increases with the number of hydrophilic groups contained in each saccharide. Indeed, the straight slopes for xylose, glucose, and sucrose are −0.08, −0.11, and −0.13, respectively (Figure ).

Figure 1

Saccharide hydration numbers (nH saccharide) versus the coordination numbers (CN cation) of different electrolytes containing Cl– (filled symbols) and SO42– (empty symbols) for a given cation molality (1 mol·kg–1).

Hence, in the presence of Cl–, the fact that the saccharides are all the more dehydrated than the cation coordination number is important (i.e., increase of the number of water molecules contained in the first coordination shell of the cation). That means that the greater the cation coordination number is, the more it can create interactions with the saccharide. The saccharide–water interactions that hydrated the hydrophilic groups of the saccharides are then replaced by saccharide–cation interactions. Equally, the cation–water interactions are replaced by cation-hydrophilic group interactions. Both saccharide and cation are dehydrated. It appears that the anion Cl– has a weak impact on the saccharide hydration. In the presence of the chloride anion, interactions in solution are then governed by the cation–saccharide interactions. To our best knowledge, such a conclusion has never been reported to date.

In a second step, the anion influence was studied. Figure shows that the saccharide dehydration effect is much larger with electrolyte containing the sulfate ion than the chloride anion. The sulfate anion promotes saccharide dehydration. On the one hand, the larger influence of the sulfate ion is linked to its higher charge (divalent anion) that creates more hydrogen bonds with the saccharides. On the other hand, it can be correlated with its greater coordination number (9) compared to the chloride ion (5) and the shorter anion–water distance (Table ). For sulfate electrolytes with monovalent cations (Na+ and K+), the saccharide hydration number also decreases linearly with the increase of the coordination numbers of the cations. Furthermore, the divalent cations remain the most dehydrating component compared to monovalent cation. However, for cations having the same coordination number (K+ and Mg2+), unlike the results obtained in the presence of Cl–, the saccharide hydration numbers are different. Indeed, the sucrose hydration number is 7.00 in the presence of K2SO4 and 6.65 in the presence of MgSO4. That shows the strong impact of the sulfate anion on the saccharide hydration. Hence, in order to evaluate possible synergistic effects of anion and cation on the saccharide hydration, the possible existence of a correlation between the saccharide hydration numbers (nH) and the total coordination numbers of the ions of the electrolyte (CNtotal = CNcation + CNanion) is studied (Figure ). The CNtotal takes into account the stoichiometric coefficients of the electrolyte ions. For example, the CNtot of Na2SO4 is twice the CN of Na+ to which the CN of SO42– is added (see Table ), meaning a total coordination number (CNtotal) of 19.

Figure 2

Saccharide hydration numbers (nH saccharide) versus the total coordination numbers of electrolytes (CNtotal) containing SO42– for different electrolyte molalities (0.25 and 0.5 mol·kg–1).

In the presence of SO42–, the saccharides hydration numbers decrease linearly with the increase of the total coordination number of the ions (linear regression coefficients > 0.94). This is indicative of the effect of SO42– on the saccharide dehydration. As in the presence of Cl–, the saccharide dehydration increases with the saccharides’ hydrophilic groups number, which is consistent. Indeed, the straight slopes for xylose, glucose, and sucrose are −0.02, −0.03, and −0.05, respectively.

Finally, in order to take into account the influence of the electrolyte concentration and since it was demonstrated that Cl– exhibits a weak influence on the dehydration, a generalized coordination number is defined by including the molality of ions in electrolyte as follows:with CNcation or CNanion: cation or anion coordination numbers, respectively, and mcation or manion: cation or anion molality (mol·kg–1).

The evolution of the saccharide hydration number in sulfate electrolytes, for various ionic composition, is represented in Figure as a function of . The one concerning Cl– is presented in Figure S3.

Figure 3

Saccharide hydration numbers (nH saccharide) versus the total coordination numbers of electrolytes containing SO42– (CNtotal) for different electrolyte molalities (0.25, 0.5, 1 mol·kg–1).

For each saccharide and regardless of the electrolyte concentration, saccharide hydration numbers decrease linearly with , which is remarkable. The total cation coordination number reflects then the ions’ ability to form direct interactions with the saccharide hydroxyl groups. It takes into account the proportions of the cations and anions of the solution and the number of opportunities for ions to interact with saccharides.

Moreover, the relationships established enable the determination of the saccharide hydration numbers from the knowledge of the ions properties.

This work constitutes a new approach, which allows the understanding of mechanisms governing the ions’ specific effects on polar neutral molecules. More precisely, it deals with phenomena governing saccharide hydration, as a function of the physical chemistry properties of ions in solution (cation or anion charge, size, and hydration). To this end, an original methodology based on the combination of experimental measurements (molar volume) and computational technique has been used.

From experimental data, the saccharide hydration number has been calculated in various electrolytes, for different ionic compositions. Ions hydration properties were computed using theoretical methods (DFT). Saccharide and ions properties obtained from these two complementary methods were then put in parallel, in order to rationalize the relation between the saccharide hydration number in electrolyte solution and the hydration properties of ions.

The influence of the cations and anions on the saccharide hydration has been quantified separately.

The study of the influence of cations highlights that saccharide dehydration increases with the increase of the cation coordination number (i.e., the number of water molecules directly interacting with the cation), whatever the cation valence. Divalent cations have a larger impact on the saccharide dehydration than monovalent ones. That can be attributed to the lowest radial expansion of the orbital of divalent metal with respect to monovalent ones. Interactions between cations and the oxygen of the saccharide hydroxyl groups (the cation–oxygen interaction involving in atomic orbital overlap) are therefore stronger with divalent cations than with monovalent ones. This causes higher saccharide dehydration. For a given cation valence, the cation coordination number increases with the size of the cation. That allows the cation to create more saccharide–cation interactions and then to further dehydrate the saccharide. The cation coordination number reflects the ions’ ability to form direct interactions with the saccharide hydroxyl groups.

Concerning the influence of the anion on the saccharide hydration, this work has highlighted that Cl– has weak interactions with saccharides. Then, the saccharide hydration number only depends on the cation coordination number. In sharp contrast, SO42– has strong interactions with saccharides. In this case, the saccharide hydration number depends on the coordination number of both cations and of anions of the electrolyte.

Relationships established in this work make possible to evaluate the hydration state of a polar neutral solute, a saccharide, as a function of the ionic composition, from the knowledge of the ions’ hydration properties.

The results of the scaling up may be transferable to other polar and neutral biomolecules. It would be interesting to study what would become of the relationships obtained with molecules having a charged or hydrophobic part.

Methods

Apparent molar volumes of saccharides, VΦ,S, which characterize the hydration state of the solute, have been determined in ref (12). At infinite dilution, the apparent molar volumes, Vϕ,S0, are equal in value to the standard partial molar volumes. The partial molar volume of a nonelectrolyte can be divided in two terms: the intrinsic molar volume of the nonhydrated solute, Vint, and the electrostriction partial molar volume, Velect, due to the interaction of the nonelectrolyte with water (shrinkage in volume):[27]

The positive intrinsic volume is given by the size of the solute molecule, and for a given temperature, it is considered that it remains at the same magnitude in water and aqueous salt solutions. The interaction term Velect is variable and depends on the interactions between the solute molecule and surrounding the water molecules (hydration shell).

The hydration number nH (or solvation number, expressed in molecules of water per molecule of solute) explicitly reveals the hydration degree of a solute in water. The hydration number can be calculated from the volumetric properties using the following equation:[28,29]where Ve0 is the molar volume of electrostricted water, and Vb0 is the molar volume of bulk water. The value of (Ve0 – Vb0) is about −3.3 cm3·mol–1 at 298.15 K.[28,29] The Velec values can be obtained from rearrangement of eq as follows:

The intrinsic volume Vint for saccharides is calculated from the crystal molar volume (eq ).where M is the molar mass of saccharide, 0.7 is the packing density for molecules in organic crystals, and 0.634 is the packing density for random packed spheres. The crystal density (dcryst) determined by single-crystal X-ray diffraction is 1.52 g·cm–3 for the xylose,[30] 1.56 g·cm–3 for the glucose,[31] and 1.59 g·cm–3 for the sucrose.[32] The obtained nH values are reported in Table S1 (Supporting Information).

Safety

No unexpected or unusually high safety hazards were encountered.