Ice/Water Interface: Zeta Potential, Point of Zero ...

31 downloads 0 Views 67KB Size Report
the zeta potential vs pH data points were significantly scattered, it was determined that the isoelectric point (iep) of D2O ice particles in water at 3.5°C containing ...
Journal of Colloid and Interface Science 220, 229 –234 (1999) Article ID jcis.1999.6528, available online at http://www.idealibrary.com on

Ice/Water Interface: Zeta Potential, Point of Zero Charge, and Hydrophobicity Jan Drzymala,* ,1 Zygmunt Sadowski,* Lucyna Holysz,† and Emil Chibowski† *Institute of Mining Engineering, Technical University of Wroclaw, Wybrzeze Wyspianskiego 27, 50-370 Wroclaw, Poland; and †Department of Physical Chemistry, Faculty of Chemistry, Maria Curie-Sklodowska University, 20-031 Lublin, Poland Received December 24, 1998; accepted September 10, 1999

INTRODUCTION The ice/water interface is a common and important part of many biological, environmental, and technological systems. In contrast to its importance, the system has not been extensively studied and is not well understood. Therefore, in this paper the properties of the H 2O ice/water and D 2O ice/water interfaces were investigated. Although the zeta potential vs pH data points were significantly scattered, it was determined that the isoelectric point (iep) of D 2O ice particles in water at 3.5°C containing 10 23 M NaCl occurs at about pH 3.0. The negative values of the zeta potential, calculated from the electrophoretic mobility, seem to decrease with decreasing content of NaCl, while the iep shifts to a higher pH. The point of zero charge (pzc) of D 2O ice and H 2O ice, determined by changes in pH of 10 24 M NaCl aqueous solution at 0.5°C after the ice particle addition, was found to be very different from the iep and equal to pH 7.0 6 0.5. The shift of the iep with NaCl concentration and the difference in the positions of the iep and pzc on the pH scale point to complex specific adsorption of ions at the interface. Interestingly, similar values of iep and pzc were found for very different systems, such as hydrophilic ice and highly hydrophobic hexadecane droplets in water. A comparison of the zeta potential vs pH curves for hydrophilic ice and hydrophobic materials that do not possess dissociative functional groups at the interface (diamond, air bubbles, bacteria, and hexadecane) indicated that all of them have an iep near pH 3.5. These results indicate that the zeta potential and surface charge data alone cannot be used to delineate the electrochemical properties of a given water/moiety interface because similar electrical properties do not necessary mean a similar structure of the interfacial region. A good example is the aliphatic hydrocarbon/water interface in comparison to the ice/water interface. Although the experiments were carried out with care, both the zeta potential, measured with a precise ZetaPlus meter, and DpH values (a measure of surface charge) vs pH were significantly scattered, and the origin of dissemination of the data points was not established. Differently charged ice particles and not fully equilibrium conditions at the ice/water interface may have been responsible for the dissemination of the data. © 1999 Academic Press Key Words: zeta potential; surface charge; water; ice; heavy ice; heavy water; deuterium oxide; Nocardia sp., potentiometric titration; electrophoretic mobility; hydrophobicity; hydrophilicity.

The ice/water interface is a common and important part of many biological, environmental, and technological systems. In contrast to its importance, the system has not been extensively studied and therefore is not well understood. For instance, there is no agreement regarding the natural hydrophobicity of ice understood as an ability to form stable three-phase contacts of water/air/ice. Some researchers believe that ice is hydrophobic (1) while others consider it a hydrophilic material (2) because a drop of water placed on an ice surface spreads spontaneously. Other properties of the ice/water interface, such as the surface charge and zeta potential, which are routinely measured for solid/water systems, have not been measured for ice because of experimental difficulties. Schulman and Perreira (3) were perhaps the first to recognize the importance of the zeta potential of the ice/water interface not only for this particular system but also as a model of the interface for other interfacial systems. However, their effort to measure the zeta potential of ice by the streaming potential failed due to experimental difficulties. Recent progress in the design of instruments for electrophoretic mobility determination has removed most of the obstacles to performing reliable and accurate measurements of the zeta potential. The remaining problem is the difficulty of keeping ice/water systems at a suitably low temperature. In our work, this problem has been overcome, at least partially, with the idea that the properties of the ice/water interface can be investigated by using D 2O ice instead of H 2O ice. Since the melting point of D 2O is 3.8°C (4), there is enough room between 0 and 3.8°C to keep the D 2O ice/H 2O water system at equilibrium and stable. We believe that our research on the properties of the ice/ water interface will create an interest in studying this system to gain a more complete understanding of this interface, which is important for the many reasons mentioned above. EXPERIMENTAL SECTION

Zeta Potential Measurements The zeta potential measurements were carried out in water in the absence and the presence of either 10 24 or 10 23 M NaCl. In

1

To whom correspondence should be addressed. E-mail: jd@ig. pwr.wroc.pl. 229

0021-9797/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

230

DRZYMALA ET AL.

the series of measurements in 10 24 M NaCl using deionized distilled water, portions of about 10 cm 3 each were used to prepare solutions of different pH. The pH was regulated with either 0.1 M HCl or NaOH aqueous solutions and measured with a Precision Digital OP-208/1 pH meter. The glass electrode of the pH meter was calibrated against pH standards at 3.5°C. The prepared solutions were stored in closed plastic tubes until use after about 24 h. Shortly before use, the solutions were cooled to between 0.5 and 0.8°C in an icy water bath. A sample of heavy water containing 1.5 cm 3 of D 2O was placed in a plastic test tube and dipped into liquid nitrogen for freezing. The cooling time was about 30 s. The chunk of heavy ice obtained was removed from the test tube, wrapped in a polyethylene foil, and crushed with a hammer. The resulting small particles of heavy ice were introduced into a cooled NaCl aqueous solution and shaken vigorously manually. The final suspension was transferred to the measuring cell of the zetameter. The same procedure was applied for the series of experiments with D 2O ice in water containing 10 23 M NaCl, as well for the measurements in pure water. The zeta potential was measured with a ZetaPlus meter, manufactured by Brookhaven Instruments Corp. (U.S.A.). The ZetaPlus meter utilizes the shift in frequency of scattered light when, due to an applied electric field, a particle moves perpendicularly to the laser beam (Doppler effect). The frequency shift is proportional to the electrophoretic mobility. The zeta potential, in turn, is usually proportional to the electrophoretic mobility of particles, and in our measurements the proportionality resulting from the Smoluchowski equation was utilized by the software of the ZetaPlus apparatus to calculate the zeta potential. According to the manual, the ZetaPlus meter is very accurate because the electroosmotic effects are completely eliminated while broadening of the measured mobility distribution caused by the Brownian motion of particles is small due to a low scattering angle equal to 15°. During measurements, the ZetaPlus meter was set up to keep a constant and lowest possible temperature in the measuring cell. At this setting, the temperature in the cell, measured with an external thermometer, was 3.5°C. Thus, the measurements of the zeta potential of heavy ice were carried out at temperatures slightly below its melting point, which is equal to 3.8°C (4). From 5 to 10 readings of the zeta potential were carried out for each sample containing heavy ice in water in the absence or presence of NaCl at a desired pH and at 3.5°C. It took about 3– 6 min to accomplish one series of measurements. After that, the tests were terminated even though the zetameter was ready to continue the determinations. The described procedure was repeated from three to five times, each time using a new heavy ice suspension at a given pH. Thus, the final experimental value of the zeta potential was calculated as an average of three to five measurements with 5 to 10 readings for each sample. The zeta potential of Nocardia sp. was also measured with a

ZetaPlus zetameter, and details of the measurements are given elsewhere (5). Point of Zero Charge The pH-shift method was used to determine the sign and magnitude of the surface charge at the ice/water and heavy ice/water interface in terms of DpH. This method relies on monitoring the pH change caused by the introduction of ice into water. To measure the pH change, 40 cm 3 of the solution containing 10 24 M NaCl was cooled to 0°C. The solution kept in a 50-cm 3 beaker was placed in an icy water bath. The glass electrode and thermometer were inserted into the measuring cell and kept for about 30 min for equilibration until the measurements began. The double-distilled water was frozen in a polyethylene foil bag in a refrigerator at 27°C. Next, ice was crushed in a mortar to produce pieces smaller then 0.5 cm, and the sample was kept at room temperature for 30 min to warm the ice to about 0°C. Before ice was introduced into the water solution for pH measurements, it was crushed again to get particles smaller than 0.5 mm in diameter while the excess water was removed with a filter paper. A 5-cm 3 sample of ice particles was introduced into the aqueous solution of NaCl at 0.5°C, and the pH was measured and read after 30 s of equilibration. The same procedure was used for both ice and heavy ice. A similar procedure was used to determine the sign and magnitude of the surface charge of hexadecane droplets in 10 23 M NaCl at 20°C. Hexadecane was dispersed mechanically in 750 cm 3 of aqueous NaCl with a high-speed impeller. The total concentration of C 16H 34 in the aqueous solution was 10 23 M. The pH was measured before and after 5 min of dispersion of the oil phase in water. Materials Heavy water (D 2O) was purchased from the Institute for Nuclear Research, Radioisotope Production and Distribution Center, S´wierk near Warsaw. The D 2O content in the heavy water was 98.84%. Heavy ice was produced by freezing heavy water in a refrigerator at about 27°C or otherwise stated. Hexadecane was purchased from Reachim Co. (Russia). It was labeled as pure but the impurities were not specified. The gas chromatography test showed no polar impurities. RESULTS AND DISCUSSION

Heavy ice, similarly to other materials, forms an electrical double layer at the solid/water interface. The electrical double layer (edl) consists of molecules and ions coming from the ice, the water, the electrolytes used for regulation of the ionic strength, and the chemicals use for pH adjustment. Therefore, in the ice/water system, the electrical double layer contained OH 2, H 1, Cl 2, and Na 1 ions as well as water dipoles. The structure of the edl depends on many factors, including adsorption of potential determining ions (OH 2 and H 1, i.e., pH of the

ICE/WATER INTERFACE

solution), specific adsorption of other ions (Cl 2 and Na 1), and the hydrophobicity of the surface responsible for water dipole arrangements. Two characteristic features of the edl are (6): (a) the isoelectric point (iep), indicating that the potential at the slipping plane (that is, the zeta potential) is zero; and (b) the point of zero charge (pzc) when the surface charge caused by the potential-determining ions is zero. When the iep coincides with the pzc there is either a lack of adsorption of the supporting electrolyte ions or the adsorption of cations and anions of the supporting electrolyte is the same. According to Fig. 1a, the iep of D 2O ice in the absence of any supporting electrolyte is somewhere between 3 and 4.6, that is, pH 3.8 6 0.8. The iep near pH 4.0 indicates that down to about this pH a negative net charge exists at the slipping plane of the ice/water interface, which may result from preferential adsorption of negatively charged species on the ice surface. However, when the aqueous solutions contains 10 24 M NaCl (Fig. 1b), the iep seems to shift to a pH below about 3.6, pointing to some specific adsorption of Cl 2 ions. Although the pzc in Fig. 1b is 3.3 6 0.2 for 10 24 NaCl, this value is uncertain because below about pH 4 the number of Cl 2 ions introduced into the solution with HCl is greater than that when NaCl is used as the electrolyte. Even a greater adsorption of Cl 2 ions seems to occur when the NaCl concentration is increased to 10 23 M, because the zeta potential curve becomes more negative and the iep decreases to a pH near 3.0 (Fig. 1c). All of the graphs in Figs. 1a–1c show that the zeta potential as a function of pH for the D 2O ice/water interface is a very complex relationship. Not only is the zeta potential in 10 23 M NaCl, in contrast to most systems, more negative than that in 10 24 M NaCl, but also between pH 6.5 and 7.7 the zeta potential values in 10 24 M NaCl are either around 227 6 2 mV or close to 0 mV, and the reason for this phenomenon is unknown. Some of the ice particles may be negatively charged (Cl 2 adsorption) and others uncharged. Incomplete equilibrium at the ice/water interface may also contribute to the observed scatter of the zeta potential data points. It is interesting that the electrokinetic pH characteristic of D 2O ice is similar to that of other highly hydrated surfaces such as K 2SO 4 and Na 2SO 4 z 10 H 2O (7). Although the electrophoretic mobilities of these salts were measured at nonequilibrium conditions, the structures of D 2O ice/water and highly hydrated solids/water systems may be similar and hence the D 2O ice/water interface may serve as a model system. To learn more about the electrical double layer at the ice/ water interface, other experiments were carried out in which the sign and magnitude of the surface charge caused by H 1 and OH 2 ions adsorption were measured by the pH-shift method. The results are presented in Fig. 2. It appears from this figure that the point of zero charge, that is, the pH at which the adsorption of potential determining H 1 and OH 2 ions is identical, lies in the vicinity of pH 7 for both ice and heavy ice. Even though the experimental points shown in Fig. 2 are scattered, certainly the pzc is in the range of pH 7.0 6 0.5. This value indicates that there is not much preferential adsorption of

231

FIG. 1. The zeta potential of heavy ice (D 2O ice) in water at 3.5°C. pH was regulated with 0.1 M NaOH and HCl. (a) In the absence of NaCl, (b) in 10 24 M NaCl, and (c) in 10 23 M NaCl.

OH 2 over H 1 ions at the ice/water interface because the ion product for water at 0°C is 1.2 3 10 215 (8), meaning that equal concentrations of H 1 and OH 2 ions in the solution occur at pH

232

DRZYMALA ET AL.

FIG. 2. The shift of pH (DpH 5 pH final 2 pH initial) in 10 24 M NaCl aqueous solution caused by introduction of D 2O ice or H 2O ice vs final pH of the solution. Equilibration time was 30 s; temperature 1.5 6 0.5°C. The pzc of both solids is equal to 7.0 6 0.5.

7.47. The facts that for ice the iep depends on ionic strength (the pzc and iep do not coincide) and the zeta potential increases with the ionic strength point to specific adsorption of ions of the supporting electrolyte (6) as well to complexity of the structure of the electrical double layer at the ice/water interface. It is also possible that the values of the zeta potential and the surface charge of ice determined are in error due to lack of true equilibrium between ice and the solutions during the measurements. Other reasons may be involved as well. It has been known since the work of Ney (9) and James (10) that certain groups of solids tend to have similar iep values and similar zeta potentials vs pH. Such similarity occurs, for instance, with materials with carboxylic groups at the interface. Another example is a family of nonionogenic surfaces for which the zeta potential is negative for a very wide pH range. A negative zeta potential of the ice surface at all pH values above pH 3– 4 and the very weak acid/base character of the water molecules justify a comparison of the zeta potential of ice with that of other nonionogenic solids. We will start the comparison with the aliphatic hydrocarbons for which the iep lies, similarly to the iep of ice, at a pH of about 3.3 (11–13), while the pzc occurs at much higher pH. Our results for the pzc of hexadecane obtained by the “pH-shift method” (Fig. 3) indicated that the pzc for hexadecane is near pH 6.3. Moreover, as can be seen in Fig. 4, the zeta potential curves for both ice and hexadecane in 10 23 M NaCl are similar. Unfortunately, no zeta potentials for hexadecane or other aliphatic hydrocarbons are available for ionic strength less than 10 23 M. Therefore, further exploration of the similarities of both systems is not possible. However, as for the surface hydrophobicity, there is a great difference between the systems discussed. The hydrocarbon/water system is very hydrophobic, with a contact angle approaching 110°, one of the highest contact angles in natural systems (14). One the other hand, based on flotation tests (15) and contact angle measurements (2), ice is hydrophilic, or as it

FIG. 3. Changes of pH (DpH) caused by dispersing hexadecane in water containing 10 22 M NaCl vs final pH of the solution. The total concentration of hexadecane was 10 23 M. The pzc of hexadecane is at pH 6.25.

was claimed by van Oss et al. (1), it is only slightly hydrophobic. It is widely accepted (16, 17) that highly hydrophobic surfaces, in contrast to hydrophilic surfaces, do not contain a layer of oriented water dipoles on the surface. Therefore, the structure of the interfacial region of the ice/water should be different from that of hexadecane/water, and their iep and zeta potential curves should also be different, but Fig. 4 suggests that the difference is not significant. Perhaps the similarity of iep and pzc values of hydrocarbons with ice is incidental because ice acquires its high negative zeta potential from significant Cl 2 ion adsorption and because the zeta potential without chloride ion adsorption is less negative (Fig. 1a) than that observed for 10 24 M NaCl. Therefore, a comparison of the zeta potential curves for ice with other nonionogenic materials of lower hydrophobicity, such as diamond, air bubbles, and bacteria, would be helpful. Their zeta potentials in aqueous solutions containing 10 23 M NaCl, along with zeta potential of hexadecane and D 2O ice, are also shown in Fig. 4.

FIG. 4. A comparison of zeta potentials of D 2O ice with those of diamond (20), air bubbles (27) (interpolated data), hexadecane (11, 12), and Nocardia sp. (5). All data were determined in the presence of 10 23 M NaCl. The zeta potential of D 2O ice in the presence of 10 24 M NaCl is also given for comparison.

233

ICE/WATER INTERFACE

The materials shown in Fig. 4 are of different hydrophobicities because ice is hydrophilic or only slightly hydrophobic, hexadecane is highly hydrophobic with a contact angle equal to 110°, and the air bubble, due to the similarity to the oil droplet (flotation, the presence of structural hydrophobic forces) (18), is also considered a highly hydrophobic medium. Diamonds and bacteria are also hydrophobic but their hydrophobicity is much lower than that of hydrocarbons because the contact angle of bacteria is usually only between 10° and 20° (19) and for the diamond sample shown in Fig. 4, it is between 2° and 12° (20). It can be seen from Fig. 4 that the slope of the zeta potential curves as well as their plateau levels increases with the hydrophobicity of the system. For highly hydrophobic hexadecane and air, between the iep and pH 6.5, the slope is about 20 mV/pH, while for the Nocardia sp. and diamond the slope is smaller and amounts to about 10 mV/pH. The hydrophilic ice/water interface would follow this pattern if it did not have an essential specific adsorption of chloride ions. In the absence of the specific adsorption, the zeta potential in 10 23 M NaCl should be lower than that in 10 24 M NaCl, and it should be lower than the zeta potentials of slightly hydrophobic diamond, which in fact is the case (see Fig. 1a). The fact that increasing hydrophobicity makes the zeta potential more negative is known in related literature. For instance, it was found that after methylation diamond exhibits more negative zeta potential values (20). Chibowski and Waksmundzki (21, 22) also found that the negative zeta potential of sulfur and Teflon in water increased with a progressively increasing precoverage of the surface with n-heptane or nhexane, respectively. For bare sulfur, the zeta potential was 225 mV, while for the surface precovered with three statistical monolayers of heptane, it increased to 280 mV; and for a layer as thick as 14 monolayers deposited on the surface, it reached 2110 mV. In the case of Teflon, the negative zeta potential increased from 246 V (bare surface) to 267 mV for a four statistical monolayer-precovered Teflon surface. Sometimes the methylation does not change the zeta potential of the modified material much as happens with silica (23). The reason for this might be that the methylated silica surface still was not uniformly hydrophobic, and the hydrophilic sites (siliceous acid groups, -SiO 2) determined its surface properties. Now the question arises why hydrophobicity makes the zeta potential more negative. A possible explanation can be found by taking into account the surface conductance effect. According to Wal et al. (24), zeta potentials calculated from electrophoretic mobility are underestimated due to surface conduction behind the shear plane. For instance, for Bacillus brevis, the slope of the zeta potential vs pH changes from about 15 mV/pH when calculated from the classic Smoluchowski equation to about 30 mV/pH from the equation which takes into account the surface conductance. If we assume that the surface conductance decreases with the hydrophobicity of the system because of the decreasing polarity of the surface, then very likely all of the materials shown in Fig. 4 would have similar

true zeta potential values. Thus, we can conclude that materials without strongly dissociative functional groups at the surface and without excessive specific adsorption of ions may have similar iep and comparable true zeta potential values. On the other hand, it is well known that hydrophobic hydration takes place at a hydrophobic surface (25, 26). Hence, the oriented water dipoles in such a specific system may play a role in the creation of the electrokinetic phenomena that appear in the measured zeta potential. CONCLUSIONS

The results presented in this work indicate that the pzc of ice is at pH 7.0 6 0.5 while its iep is between 3 and 4.6 and depends slightly on the ionic strength of NaCl. There is a significant specific adsorption of chloride ions at the ice/water interface because the zeta potential values in 10 23 M NaCl are more negative than the values in 10 24 M NaCl. Also scattering of zeta potential data points, especially in 10 23 NaCl solutions, was observed. In addition, the zeta potential values in 10 24 M NaCl at a pH between 6.5 and 7.7 were either around 227 6 2 mV or close to 0 mV. Unfortunately, the reasons for both effects were not established. There are similarities between the interfacial properties of the ice/water system and nonionogenic materials having different hydrophobicities because their iep and pzc values are comparable while their zeta potential curves become more negative with increasing hydrophobicity of the system, very likely caused by decreasing surface conductance and/or water dipole structuring. These facts indicate that either the structure of the ice/water interface is similar to other nonionogenic surfaces or the data presented for ice are in error due to, for example, incomplete equilibrium conditions reached during the measurements. The data presented here on the ice/water interface should be considered preliminary. More well-designed tests are needed to obtain a better picture of the ice/water interface. ACKNOWLEDGMENTS This work was financed by the Technical University of Wroclaw and Maria Curie-Sklodowska University, Poland. Determination of the pzc of ice by Eng. Alicja Szulmanowicz and hexadecane by Jadwiga Brzezniakiewicz, M.Sc., is greatly acknowledged. The assistance of Bozena Dzikowska, M.Sc., in the zeta potential determinations is also very appreciated.

REFERENCES 1. van Oss, C. J., Giese, R. F., Wentzek, R., Norris, J., and Chuvilin, J., Adhesion Sci. Technol. 6(4), 503 (1992). 2. Drzymala, J., Adv. Colloid Interface Sci. 50, 143 (1994). 3. Schulman, J. H., and Perreira, H. C., Technical Report No.1, NONR Report Contract 266(64), June 1964. 4. “CRC Handbook of Chemistry and Physics,” 67th ed. (R. C. Weast, Ed.). CRC Press, Boca Raton, FL, 1986/87. 5. Sadowski, Z., submitted for publication. 6. Lyklema, J., Croat. Chim. Acta 60, 371 (1987).

234

DRZYMALA ET AL.

7. Hancer, M., Hu, Y., Fuerstenau, M. C., and Miller, J. D., “Proc. XX Int. Mineral Processing Congress,” Aachen, GDMB Clausthal-Zellerfeld, Vol. 3, p. 715, 1977. 8. Stumm, W., and Morgan, J. J., “Aquatic Chemistry.” Wiley, New York, 1970. 9. Ney, P., “Zeta-Potentiale und Flotierbarkeit von Mineralen.” SpringerVerlag, Wien/New York, 1973. 10. James, A. J., in “Surface and Colloid Science” (R. J. Good, and R. R. Stromberg, Eds.), Vol. 11, p. 121. Plenum New York/London, 1979. 11. Stachurski, J., and Michałek, M., J. Colloid Interface Sci. 184, 433 (1996). 12. Stachurski, J., and Michałek, M., Colloids Surf. 15, 255 (1985). 13. Fuerstenau, D. W., and Pradip, Physicochem. Prob. Miner. Process. 12, 3 (1980) [in Polish]. 14. Adamson, A. W., “Physical Chemistry of Surfaces,” 5th ed. Wiley, New York, 1990. 15. Drzymala, J., Miner. Eng. 12, 329 (1999). 16. Drost-Hansen, W., Ind. Eng. Chem. 61(11), 10 (1969).

17. Liu, S-T., Colloid J. USSR 52, 743 (1990). 18. Yoon, R-H., and Aksoy, B. S., “Proc. XX Int. Mineral Processing Congress,” Vol. 3, p. 107, GDMB Clausthal-Zellerfeld, Aachen 1977. 19. Smith, R. W., Miner. Proc. Extractive Metal. Rev. 4, 277 (1989). 20. Shergold, H. L., and Hartley, C. J., Int. J. Miner. Process 9, 219 (1982). 21. Chibowski, E., and Waksmundzki, A., J. Colloid Interface Sci. 64, 380 (1978). 22. Chibowski, E., and Waksmundzki, A., J. Colloid Interface Sci. 66, 213 (1978). 23. Laskowski, J., and Kitchener, J. A., J. Colloid Interface Sci. 29, 671 (1969). 24. Wal, A., Minor, M., Norde, W., Zehnder, A. J. B., and Lyklema, J., Langmuir 13, 165 (1997). 25. Israelachvili, J. N., “Intermolecular and Surface Forces.” Academic Press, San Diego, 1985. 26. Israelachvili, J. N., and Wennerstrom, H., Nature 379, 219 (1996). 27. Li, C., and Somasundaran, P., J. Colloid Interface Sci. 148, 587 (1992).