pubs.acs.org/Langmuir © 2009 American Chemical Society
Inhibition of Bubble Coalescence by Osmolytes: Sucrose, Other Sugars, and Urea Christine L. Henry and Vincent S. J. Craig* Department of Applied Maths, Research School of Physics, The Australian National University, Canberra ACT 0200, Australia Received April 30, 2009. Revised Manuscript Received June 8, 2009 We report on bubble coalescence inhibition by non-surface-active, nonelectrolytes urea and sucrose, and other small sugars, in aqueous solution. Urea has no effect on bubble stability up to high concentrations>1 M, while sucrose inhibits coalescence in the range 0.01-0.3 M, similar to inhibiting electrolytes. Urea and sucrose both increase bubble coalescence inhibition in inhibiting and noninhibiting electrolytes in a cooperative manner, but urea decreases the efficacy of sucrose in mixed solutions. Several mono- and disaccharides also inhibit bubble coalescence at ∼ 0.1 M, and the sugars vary in effectiveness. Disaccharides are more effective than the sum of their individual monosaccharide constituents, and sugars with very similar structures (for instance, diastereomers galactose and mannose) can show large differences in coalescence inhibition and hence thin film stability. We conclude that solute charge is not required for bubble coalescence inhibition, which indicates that the mechanism is not one of electrostatic surface repulsion and instead an effect on dynamic film thinning other than Gibbs-Marangoni elasticity is implicated. Solute structure is important in determining coalescence.
Introduction Electrolytes can inhibit bubble coalescence in water. An easily observed example is the persistent froth seen in seawater but not in fresh water.1 The effect requires electrolyte on the order of 0.1 M, and bubble lifetime is on the order of 10 s. Stabilization of bubbles by electrolytes is therefore far less efficient than surfactant coalescence inhibition, which can occur at concentrations as low as 10-7 M and can create longer-lasting foams. Furthermore, electrolyte coalescence inhibition is ion specific. While some salts inhibit coalescence at 0.1 M, others show no effect at 0.5 M or higher. The ion specificity depends on the combination of cation and anion present. Craig et al. showed previously that cations and anions may each be grouped into two categories R and β, and while RR and ββ electrolytes inhibit coalescence, the “crossproducts” of Rβ and βR salts have no effect.2,3 A subset of the ions studied is shown in Table 1 to illustrate the assignments and the combining rules. In a separate publication currently in preparation, we will propose an explanation for the physical meaning of R and β assignments; however, the exact mechanism by which inhibiting electrolytes stabilize bubbles against coalescence remains unclear. One of the key questions arising in connection with this phenomenon is how the ion specificity observed in bubble coalescence (a gas-solution system) is connected to the wider arena of specific ion effects. Such effects are ubiquitous in high salt environments, including biological systems.4,5 This general ion specificity was first rigorously described by Hofmeister in his *To whom correspondence should be addressed. E-mail: vince.craig@anu. edu.au. (1) Winkel, E. S.; Ceccio, S. L.; Dowling, D. R.; Perlin, M. Exp. Fluids 2004, 37, 802–810. (2) Craig, V. S. J.; Ninham, B. W.; Pashley, R. M. J. Phys. Chem. 1993, 97, 10192–10197. (3) Craig, V. S. J.; Ninham, B. W.; Pashley, R. M. Nature 1993, 364, 317–319. (4) Kunz, W.; Lo Nostro, P.; Ninham, B. W. Curr. Opin. Colloid Interface Sci. 2004, 9, 1–18. (5) Lo Nostro, P.; Lo Nostro, A.; Ninham, B. W.; Pesavento, G.; Fratoni, L.; Baglioni, P. Curr. Opin. Colloid Interface Sci. 2004, 9, 97–101.
11406 DOI: 10.1021/la9015355
studies on proteins in solution.6 “Hofmeister series” of cations and anions can each be ordered from salting-in to salting-out for any protein, depending on whether the ion increases or decreases the protein’s solubility. The ion-combining rules observed in bubble coalescence inhibition do not correlate with the Hofmeister series; that is, the R and β categorizations of ions are found not to correspond to series of cations and anions associated with Hofmeister effects, and protein salting-in or salting-out by ions does not depend on the combination of cation and anion, with the effect of each ion being independent of its counterion. However, protein dissolution and bubble coalescence inhibition are not entirely dissimilar processes, because both cases are concerned with the stability of a hydrophobic interface in solution. In fact, protein solvation and dissolution has recently been linked to ion affinity for the airwater interface, with Pegram and Record showing that ions that are more withdrawn from the hydrophobic air-water surface are “salters out” of hydrophobic proteins (and vice versa).7 Protein solubility can be controlled not only by Hofmeister ions but also by organic solutes sometimes known as “osmolytes” because of their biological role in controlling osmotic pressure.8 To further probe the link between Hofmeister ion specificity and ion specific bubble coalescence, we have measured bubble coalescence inhibition by osmolytes with differing effects on protein solubility. Urea is a denaturant and enhances protein solvation and dissolution, while sucrose (and other sugars) constricts solvent and excludes hydrophobic proteins from solution.9,10 Inhibition of bubble coalescence in urea and sucrose solutions, in the presence and absence of electrolytes, is here reported and compared with salt effects. Coalescence inhibition by small sugars other than sucrose is also studied, to determine the importance of (6) Kunz, W.; Henle, J.; Ninham, B. W. Curr. Opin. Colloid Interface Sci. 2004, 9 (1-2), 19–37. (7) Pegram, L. M.; Record, M. T.Jr. J. Phys. Chem. B 2008, 112, 9428–9436. (8) Street, T. O. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 13997–14002. (9) Bolen, D. W. Methods 2004, 34, 312–322. (10) Pegram, L. M.; Record, M. T.Jr. J. Phys. Chem. C 2009, 113, 2171–2174.
Published on Web 07/02/2009
Langmuir 2009, 25(19), 11406–11412
Henry and Craig
Article
Table 1. r and β ion Assignments of Single Electrolytes in Watera ions Naþ Kþ Hþ (CH3)4Nþ assignment r r β β √ √ r Cl√ √ r Br√ β ClO4√ √ β CH3COOa Based on Craig et al.3 Y = inhibit coalescence. RR, ββ = Y. N = no inhibition. Rβ, βR = N. ions
solute structure. We note that this extends an earlier publication that reported bubble coalescence inhibition in sucrose, fructose, and glucose solutions.2 In the past, the action on protein solubility of urea, sucrose, and Hofmeister ions has been described in terms of their supposed effects on solvent structure. Sucrose was known as a “structuremaker” that imposed solvent structure (increased hydrogenbonding) and excluded proteins, while urea and other “structure-breakers” disrupt existing structures and allow proteins to enter. It is now generally accepted that this bulk water structure mechanism is not how solutes affect protein solubility.11 Instead of an ion inserting into and disrupting a large-scale water network, its effects are local and confined to the surrounding solvent molecules.12,13 Under this model, changes to protein solubility will be driven by how the ion (or other osmolyte) interacts directly with the local protein surface.7,11,14-17 Analogously, it was hypothesized that osmolytes might affect bubble stability via their local interactions with water and with the gas-water interface. Determining how these solutes affect bubble coalescence may give further information about the mechanism of thin film stabilization in salt solutions, which is not yet known.
Materials and Methods The change in bubble coalescence due to addition of solute was measured using the change in turbidity in a custom-built bubble column.2 The experimental arrangement is depicted in Figure 1. In each experiment, a quantity of pure water (41.0 mL) was placed into a cylindrical glass tube with a sintered glass frit at the bottom and a constriction some 30 mm above the frit. A stable flow of N2 gas (12 mL/s) into the tube through the frit results in a bubble stream rising in the column. A concentrated stock solution is then titrated into the water. A beam of light from a diode laser is expanded and passed through the bubbling solvent before being condensed onto a photodiode detector. When coalescence is readily taking place, the amount of light reaching the detector is high, as is the measured voltage. For pure water, this is defined as the 100% coalescence level. As bubble coalescence inhibition occurs, the small bubbles created at the frit remain small and more light is scattered, lowering the voltage at the detector. At a sufficiently high concentration of a coalescence-inhibiting solute, the voltage stabilizes at a minimum level. This is defined as 0% coalescence. Intermediate voltages are converted to bubble coalescence using a linear conversion. The transition concentration, which can be used to compare different inhibiting solutes, is defined as 50% coalescence. The baseline level in inhibiting electrolytes is used to determine percentage coalescence for (11) Ball, P. Chem. Rev. 2008, 108(1), 74–108. (12) Zhang, Y.; Cremer, P. S. Curr. Opin. Chem. Biol. 2006, 10, 658–663. (13) Collins, K. D.; Neilson, G. W.; Enderby, J. E. Biophys. Chem. 2007, 129, 95–104. (14) Lo Nostro, P.; Ninham, B. W.; Milani, S.; Fratoni, L.; Baglioni, P. Biopolymers 2006, 81, 136–148. (15) Batchelor, J. D.; Olteanu, A.; Tripathy, A.; Pielak, G. J. J. Am. Chem. Soc. 2004, 126, 1958–1961. (16) Moreira, L. A.; Bostr€om, M.; Ninham, B. W.; Biscaia, E. C.; Tavares, F. W. Colloids Surf., A 2005, 282-283, 457–463. (17) Pegram, L. M.; Record, M. T.Jr. Chem. Phys. Lett. 2008, 467, 1–8. (18) Henry, C. L.; Dalton, C. N.; Scruton, L.; Craig, V. S. J. J. Phys. Chem. C 2007, 111, 1015–1023.
Langmuir 2009, 25(19), 11406–11412
Figure 1. Schematic of bubble coalescence apparatus. Small bubbles are produced at the frit and rise in the column, passing through the constriction that promotes bubble collisions. In pure water they rapidly coalesce such that the bubbles breaking the laser beam are large and few in number. In this case, most of the light is transmitted and strikes the photodiode detector. When coalescence is inhibited, the bubbles remain small and numerous, resulting in a large amount of scattering and low light intensity at the detector. noninhibiting or partially inhibiting solutes. The apparatus has been previously described.2,18 All experiments were done at room temperature (∼23 °C). All water used in the bubble coalescence studies was purified using a Milli-Q gradient system. Electrolyte NaClO4 was used as received; NMe4Br was dried at 250 °C for several hours to drive off moisture, and NaCl and KCl were roasted at 500 °C for several hours to remove organic contaminants (all electrolytes from SigmaAldrich). Urea (BDH, AnalaR grade) was freeze-dried before use. Sugars were used as received: monosaccharides arabinose (Kerfoot’s), fructose, galactose, mannose (all SigmaAldrich), and glucose (BDH) and disaccharides lactose (SigmaAldrich), maltose (M&B), and sucrose (Merck). Sugar structures are depicted with their coalescence inhibition results in Table 2. The lactose was obtained as the monohydrate; all other sugars were anhydrous. Stock solutions were added to 41.0 mL water. In the experiments where the ratio of cosolutes was varied, one component was added first before introducing a stock solution of the second solute; otherwise, equimolar mixtures were used.
Results Coalescence Inhibition in Sucrose and Electolytes. The effect of sucrose on bubble coalescence in water is shown in Figure 2 (with typical 1:1 inhibiting electrolyte NaCl shown for comparison). Sucrose acts to inhibit bubble coalescence, over the concentration range 0.01-0.3 M, very similar to the electrolyte effect. This finding is consistent with the earlier results of Craig et al.2 The transition concentration is in the region of 0.075-0.09 M. Mixtures of sucrose and electrolytes were also studied, to determine how this solute might change bubble coalescence in the presence of ions. The results are presented in Figure 3 for inhibiting electrolyte NaCl and for two noninhibiting electrolytes NaClO4 and (CH3)4NBr. A cooperative effect is seen between sucrose and electrolytes. Equimolar mixtures of sucrose and both inhibiting and noninhibiting salts show lower coalescence (greater inhibition) than the sum of their individual contributions at a given concentration. This is clearly observed in the case of DOI: 10.1021/la9015355
11407
Article
Henry and Craig
Table 2. Bubble Coalescence Inhibiting Power of Sugars, Ranked from Most Inhibiting to Least Inhibiting at a Concentration of 0.15 M
Figure 3. Sucrose and electrolyte coalescence inhibition separately and in mixed solutions. Sucrose ([), inhibiting RR salt NaCl (O), noninhibiting Rβ salt NaClO4 (0), and noninhibiting βR salt (CH3)4NBr (2); and equimolar mixtures of NaCl þ sucrose (b), NaClO4 þ sucrose (9), and (CH3)4NBr þ sucrose (4). 100% coalescence is defined in pure water; 0% is a stable low value in inhibiting electrolytes. Sucrose and electrolytes are cooperative in effect, with the mixture inhibiting at lower concentrations than the sum of individual solute contributions would produce.
Figure 4. Bubble coalescence inhibition in urea solution ([). Structure of urea is shown in inset. 100% coalescence is defined in pure water, and the baseline (0% coalescence) is the stable low voltage value in inhibiting electrolytes. Urea shows no inhibition up to ∼1 M, which makes it similar to “noninhibiting” or lessinhibiting electrolytes in effect. a Arabinose continues trending downward rather than flattening, and thus has a coalescence percentage lower than that of galactose at higher concentrations.
Figure 2. Coalescence inhibition by sucrose (2) as a function of
concentration, shown on a log scale. Typical inhibiting univalent RR electrolyte NaCl (O) is shown for comparison. The inhibition effect shows similar concentration-dependence. 100% coalescence is defined in pure water; 0% is a stable low value in inhibiting solutions.
noninhibiting salts, with mixtures of NaClO4 or (CH3)4NBr and sucrose showing bubble coalescence inhibition greater than that in sucrose alone. The inhibition in the NaCl þ sucrose mixture is also greater than the sum of the individual contributions. An equimolar mix of sucrose and KCl (not shown) also demonstrates 11408 DOI: 10.1021/la9015355
this cooperative effect. Varying the ratio of the solute concentrations reveals no anomalous effect. Coalescence Inhibition in Urea and Electrolytes. Urea is a nonsurfactant with very little effect on surface tension.19 Urea increases protein solubility at concentrations above 3 M.20 Coalescence inhibition in urea was measured to a concentration of 3.10 M. The concentration was not increased beyond this because of the stock solution that had to be added to pure water in the bubble column, so that a baseline signal could be obtained. As shown in Figure 4, urea has no effect on bubble coalescence much below 1 M, after which bubble coalescence decreases slowly with increasing concentration. In equimolar mixtures of urea and electrolyte, coalescence inhibition is indistinguishable from the electrolyte solution in the absence of urea up to 0.3 M of each species (not shown). However, when urea is introduced at high concentrations (2-3 M) before addition of electrolytes, a cooperative effect is observable in both noninhibiting electrolyte NaClO4 (not shown) and inhibiting electrolyte NaCl (Figure 5). Coalescence Inhibition in Sucrose and Urea. We also tested the effect of mixing sucrose and urea on bubble coalescence in water. It was speculated that the “salting-in” (19) Washburn, E. D. International Critical Tables of numerical data, physics, chemistry and technology; McGraw-Hill: New York, 1930. (20) Bian, L.-J.; Yang, X.-Y. Chin. J. Chem. 2006, 24, 653–659.
Langmuir 2009, 25(19), 11406–11412
Henry and Craig
Figure 5. Urea ([) enhances the effect on bubble coalescence of RR electrolyte NaCl. NaCl alone (O) requires a higher concentration for full inhibition (0% coalescence) than does NaCl þ 3.10 M urea (b). 100% coalescence is defined in pure water; 0% is a stable low value in inhibiting electrolytes.
Figure 6. The presence of urea reduces the effectiveness of sucrose. After urea ([) is added to 2 M concentration, sucrose þ 2 M urea (4) requires higher concentration to reach the same level of coalescence inhibition as observed in sucrose in the absence of urea (2).
Figure 7. Coalescence inhibition by sugars. 0% coalescence is set by the baseline in sucrose (2). Also shown are disaccharides lactose (b) and maltose ([) and monosaccharide hexoses glucose (]), galactose (O), fructose (4), and mannose (0), and pentose arabinose (9). The sugars show a wide variation in their power to inhibit bubble coalescence relative to pure water.
(urea) and “salting-out” (sucrose) effects might cancel out or that the opposing interfacial solvation effects of the two solutes might interact in a nonlinear fashion. As shown in Figure 6, this hypothesis is satisfied. After urea reached 2 M concentration (so that a low level of inhibition was observed), sucrose was added to the solution in the column. A higher concentration of sucrose was required, to elicit the same coalescence inhibition as observed in the absence of urea. Urea thus appears to reduce the effectiveness of sucrose as an inhibitor of bubble coalescence. This result is in contrast to the cooperative effects of urea and sucrose with electrolytes. Langmuir 2009, 25(19), 11406–11412
Article
Bubble Coalescence Inhibition in Several Sugars. Bubble coalescence inhibition is here reported for a wide range of monoand disaccharides, as shown in Figure 7. All of them display some coalescence inhibition with increasing concentration, and the degree of inhibition is different with different sugars. Inhibition is observed at around 0.1 M in all sugars (comparable to the effective concentration in inhibiting electrolytes). In comparing the efficacy of the sugars as inhibitors of bubble coalescence, we find that the transition concentration (50% coalescence) is not a meaningful measure, as it is in electrolytes. All inhibiting salts reduce coalescence by (roughly) the same degree, so that the final column turbidity is roughly equivalent and hence the detector photodiode voltages are comparable across all salts. In contrast, with increasing concentration, some sugars show a sigmoidal curve of decreasing coalescence to a stable coalescence value, but the turbidity at stable coalescence is not consistent across all sugars. For this reason, the results in Figure 7 are plotted using the most inhibiting sugar, sucrose, to set the “0% coalescence” column turbidity baseline. We note that the turbidity at 0% coalescence obtained with sucrose is very similar to that obtained with salts. To make a quantitative comparison of coalescence inhibition, in Table 2, we have used the change in photodiode voltage for each sugar at a sugar concentration of 0.15 M (chosen because all sugars have been measured at this concentration). A larger value of voltage change indicates a larger change in column turbidity and is hence associated with more powerful inhibition of bubble coalescence. The sugar structures are also given. Some of the sugars appear to contain some surface-active contaminants. In particular, coalescence inhibition in lactose showed changes over time after each addition of stock solution. This is a behavior that commonly indicates the presence of a surface-active contaminant that is being carried to the top of the column and deposited onto the glass as the system self-cleans. In such cases, it may not be practical to allow equilibration at each concentration change, and so the intermediate inhibition values in lactose must be treated with caution. At the final concentration, the solution is left for a long time to ensure that a stable coalescence regime is reached. Adhikari et al. also noted the presence of surface-active contaminant in commercially available lactose.21 Inhibition in the disaccharides has been compared with inhibition in their monosaccharide components. Sucrose, for instance, comprises one glucose and one fructose monomer, and its bubble coalescence inhibition is compared with glucose and fructose, and with a 1:1 (glucose þ fructose) mixture, as shown in Figure 8. Results are also shown for lactose (glucose þ galactose) in Figure 9. Both sucrose and lactose are significantly better inhibitors than their component monosaccharides, and are also better than an unreacted equimolar mixture of their two separate components. Maltose (a diglucose) is also a somewhat better inhibitor than glucose (see Table 2 and Figure 7).
Discussion The findings from the reported experiments can be grouped under two major themes. The first theme is the relationship between protein solubility (Hofmeister effects) and bubble coalescence inhibition. The link has not been clear in the case of ion specificity in electrolyte solutions, and so has been here studied through the prism of other osmolytes, sucrose and urea, which have opposing effects on protein dissolution. (21) Adhikari, B.; Howes, T.; Shrestha, A.; Bhandari, B. R. J. Food Eng. 2007, 79(4), 1136–1143.
DOI: 10.1021/la9015355
11409
Article
Figure 8. Coalescence inhibition by sucrose (2); its component monosaccharides glucose (]) and fructose (4); and an equimolar mixture of glucose þ fructose ([). The mixture is graphed as the concentration of each monosaccharide so it can be compared directly to sucrose. 0% coalescence is set at a stable low coalescence value in sucrose. Sucrose is stronger than each of its monomers and the equimolar mixture of them.
Figure 9. Coalescence inhibition by lactose (b); its component monosaccharides glucose (]) and galactose (O); and an equimolar mixture of glucose þ galactose ([). The mixture is graphed as the concentration of each monosaccharide so it can be compared directly to lactose. 0% coalescence is set at a stable low coalescence value in sucrose. Lactose is a better inhibitor than each of its monomers and an equimolar mixture of them. Lactose is noticeably impure, and the coalescence changes with time as the solution self-cleans.
The second theme is the inhibition of bubble coalescence, or thin film stabilization, by small organic molecules which are neither surfactants nor electrolytes;the sugars. While this was first reported some time ago by one of us,2 the details and implications had not been followed up. In particular, the solute specificity, shown in the variation of inhibition among sugar molecules, is interesting and unexpected. Osmolytes and Electrolytes. We observe a difference between the effects of sucrose and of urea on bubble coalescence inhibition. Sucrose inhibits coalescence over a concentration range 0.01-0.3 M, while urea shows no effect on coalescence inhibition below 1 M concentration, and inhibition does not match that in electrolytes even at 3 M. These solutes also have opposite effects on hydrophobic protein solubility. Sucrose decreases solubility and stabilizes protein structure, while urea increases solubility and denatures proteins. This difference in bubble coalescence inhibition is consistent with the existence of some link between Hofmeister effects on protein solubility and ion specific bubble coalescence inhibition in electrolyte solutions. Up until now, the lack of correlation between Hofmeister ions and the R and β assignments of bubble coalescence inhibition has prevented the connection between protein solubility and bubble coalescence from being revealed. However, in a separate forthcoming publication, we show that the behavior of ions in bubble 11410 DOI: 10.1021/la9015355
Henry and Craig
coalescence can indeed be related to Hofmeister effects. Each is related to ion partitioning (as determined by Pegram and Record using the solute partitioning model.7,22) at either the air-water or protein-water interface.23 A recent publication by Pegram and Record suggests that osmolytes partition at the air-water interface such that the depletion is sucrose>glucose>urea.10 We note that this ordering correlates with the effectiveness of these osmolytes in bubble coalescence inhibition. Unfortunately, the other sugars studied here were not investigated. It is suggested that coalescence inhibition may be related to the partitioning of the osmolytes at the air-water interface, with those that are most depleted at the interface being most effective at inhibiting bubble coalescence. This suggests that the connection between effects at biological interfaces and effects on the air-water interface is the interfacial partitioning; however, more work needs to be done to confirm the observed correlation. Both urea and sucrose, when at high enough concentrations to partially inhibit bubble coalescence, appear to act to increase the effect of added electrolyte on coalescence inhibition. Notably, the addition of noninhibiting electrolyte to a sucrose or urea solution in which coalescence is partially inhibited causes a further decrease in coalescence, although adding the electrolyte alone leads to no change. Inhibiting salt also requires a lower concentration to elicit a given amount of coalescence inhibition. This finding suggests that there is some cooperative process between electrolyte and nonelectrolyte that can stabilize the thin liquid film. A previous study with mixed electrolytes showed no evidence of cooperativity between multiple ions; that is, the coalescence inhibition of mixtures was nowhere greater than the individual salt contributions, and could be less.18,24 There is an opposing effect with the mixture of sucrose and urea, as urea acts to make sucrose less effective as a coalescence inhibitor. Both sucrose and urea readily form hydrogen bonds with water, and it is possible that the “turning off” of the sucrose inhibitory effect in the presence of urea is due to association of the two species or to changes in the sucrose hydration shell; however, no information was found in the literature on such behavior of sucrose-urea mixtures. Small Molecule Specificity in Bubble Coalescence Inhibition. The inhibitory effect of sucrose on bubble coalescence, discussed above, is notable as a possible link between coalescence inhibition and stability of biological interfaces. Bubble persistence in sucrose solutions is also of interest in its own right. This could provide a means to stabilize gas-liquid interfaces without introducing surfactant or changing ionic strength in solution; for example, where bubbles and solid particles such as silica are present in a system, the surface forces could be controlled independently. In order to investigate the phenomenon of nonelectrolyte bubble coalescence inhibition, a range of other sugars were also investigated. Our results show that sugars can act to inhibit bubble coalescence in water, and the inhibition is “sugar specific”. A range of behaviors is observed between sucrose, which inhibits bubble coalescence to a degree comparable with 1:1 inhibiting electrolytes and a transition concentration
