Ultrafiltration of Micellar Solutions in the Presence of Electrolytes

Ultrafiltration of Micellar Solutions in the Presence of Electrolytes

JOURNAL OF COLLOID AND INTERFACE SCIENCE ARTICLE NO. 183, 484–490 (1996) 0571 Ultrafiltration of Micellar Solutions in the Presence of Electrolytes...

135KB Sizes 2 Downloads 48 Views

JOURNAL OF COLLOID AND INTERFACE SCIENCE ARTICLE NO.

183, 484–490 (1996)

0571

Ultrafiltration of Micellar Solutions in the Presence of Electrolytes PHILIPPE TOUNISSOU, MARC HEBRANT, LUDWIG RODEHUSER,

AND

CHRISTIAN TONDRE 1

Laboratorie d’Etude des Syste`mes Organiques et Colloı¨daux (LESOC), Unite´ associe´e au CNRS–No. 406, 2 Universite´ Henri Poincare´, Nancy I B.P. No. 239, 54506 Vandoeuvre-le`s-Nancy Ce´dex, France Received November 29, 1995; accepted July 2, 1996

Micellar extraction coupled with ultrafiltration techniques represents a potentially attractive tool for the removal of different kinds of contaminants from waste waters. Even though most industrial streams to be treated contain large amounts of electrolytes, very little is known about the behavior of micellar solutions in ultrafiltration when large amounts of salts are present. This paper is concerned with an investigation of two cationic surfactants (cetyltrimethylammonium bromide (CTAB) and cetylpyridinium chloride (CPC)) and one anionic surfactant (sodium dodecylsulfate (SDS)) in the presence of several salts occurring in specific industrial processes (carbonates and hydrogenocarbonates, nitrites, nitrates). The ultrafiltration behavior of these systems, with individual salt concentrations up to 0.9 M, is studied from the viewpoint of the fluxes obtained, the amount of surfactant leakage and of the membrane ‘‘gel point.’’ Two types of polymeric membranes with molecular weight cutoff of 10,000 Da are considered (cellulose and polysulfone), which show significantly different behaviors. q 1996 Academic Press, Inc. Key Words: micelles; ultrafiltration; surfactant; polymeric membrane.

INTRODUCTION

During the past few years there has been a constantly increasing number of publications dealing with micellebased separation processes (1–3). One of the promising developments of these new processes concerns the removal of metal ions by ultrafiltration techniques using semi-permeable membranes with molecular weight cutoff in the range 2,000 to 30,000 Da, so as to retain the micelles in which the metal ion species are themselves trapped (4–16). When ionic micelles are used the method is analogous in principle to the method using polyelectrolytes (17–19), which can also preferentially bind multivalent metal ions. In both cases a colloidal particle is used to enhance the removal of the metal ions normally too small to be retained by classical ultrafiltration membranes. The abbreviation MEUF has been 1 2

To whom correspondence should be addressed. Institut Nance´ien de Chimie Mole´culaire (INCM).

484

0021-9797/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

AID

JCIS 4452

/

6g17$$$$61

used by several authors to denote ‘‘micellar enhanced ultrafiltration’’ processes (20). The advantage of using micelles instead of polyelectrolytes may be found in the large variety of surfactants commercially available, with all kinds of different properties, and also in their ability to solubilize hydrophobic substances such as extractant molecules (21, 22) or synergistic agents (23). On the other hand one cannot ignore the drawbacks of micellar systems for practical applications. The major drawback is due to the existence of monomeric surfactant (at a concentration close to the cmc), which is not retained by the membranes. Biodegradable surfactants or low cmc surfactants can be considered in order to solve this problem. In addition, most studies reported so far, with a few exceptions, were concerned with surfactant solutions with low metal ion content, whereas the industrial streams to be treated often contain also large amounts of electrolytes. This situation can induce two kinds of problems: (i) the stability of the micellar solution may be affected by the presence of salts, and it is important to know what salt concentrations are tolerable; (ii) the presence of salts is known to alter the properties of the micellar solutions, with possible transitions from spherical micelles to rodlike or even wormlike micelles and correlative changes in viscosity. These changes may affect the filterability of the micellar solution in reducing the flux of the permeate. The work presented in this paper was aimed at investigating the behavior of different micellar solutions in ultrafiltration when large amounts of salts are present. These salts have been chosen for their role in specific industrial processes: carbonates and hydrogenocarbonates are present in some basic industrial streams, whereas nitrates and nitrites are often found in metal extraction. We will consider two cationic surfactants (cetyltrimethyl ammonium bromide (CTAB) and cetylpyridinium chloride (CPC), the latter being very convenient for UV-visible detection) and one anionic surfactant (sodium dodecyl sulfate (SDS)), which have been extensively studied in the literature. The flux of permeate and the leakage of monomeric surfactant will be studied with two types of membranes mounted on a classical

09-30-96 14:00:48

coidas

AP: Colloid

ULTRAFILTRATION OF MICELLAR SOLUTIONS

485

stirred cell: a hydrophilic membrane (cellulose) and a hydrophobic one (polysulfone). EXPERIMENTAL PART

Chemicals The surfactants used had the following origins: hexadecyltrimethyl ammonium bromide (CTAB) and hexadecylpyridinium chloride (CPC) were obtained from Fluka (purum); sodium dodecylsulfate (SDS) was obtained from Roth (Germany). They were used as received. The salts were bought from Fluka (NaHCO3 ) or from Prolabo, France (Na2CO3 , NaNO2 , NaNO3 ). They were of analytical grade. Deionized doubly distilled water was used throughout. Techniques The ultrafiltration experiments were carried out at room temperature with a stirred cell Amicon 8010 having a volume of 10 ml. The stirring was regulated at 420 r.p.m. and a nitrogen pressure of 3.5 bar was applied. The Millipore membranes with a molecular weight cutoff of 10,000 Da were made either of cellulose (YM10) or polysulfone (PM10). Their surface of filtration was 4.1 cm2 . The amount of filtrate collected versus time was measured with a graduated flask. For the comparison of the fluxes measured in different conditions to be rigorously valid, the flux with pure water was systematically checked between two experiments to ensure that there was no flux decline due to partial plugging. In case of flux decline the cleaning procedure was pursued until the reference flux was obtained, or a new membrane was used. For some experiments the exit of the Ami-

FIG. 1. Filtered volume as a function of time for 2 1 10 02 M CPC solutions in the absence of salts ( h ) and with different concentrations of hydrogenocarbonates (solution A in the text): A/4 ( l ); A/2 ( L ); 3A/ 4 ( m ); A ( n ). Pure water standard ( j ). Ultrafiltration membrane: cellulose (YM 10).

AID

JCIS 4452

/

6g17$$$$62

09-30-96 14:00:48

FIG. 2. Filtered volume as a function of time for 2 1 10 02 M CPC solutions in the absence of salts ( h ) and with different concentrations of salts (solution C in the text): C/4 ( l ); C/2 ( L ); 3C/4 ( m ); C ( n ). Pure water standard ( j ). Ultrafiltration membrane: cellulose (YM 10).

con cell was directly connected to a circulating UV-visible cell (optical path 0.1 cm) in order to monitor in a continuous way the concentration of surfactant in the filtrate (this was done at 260 nm). A Varian DMS 100 UV-visible spectrophotometer was used for the absorbance measurements. Electrolyte Solutions The compositions of the solutions in which the surfactants CTAB and CPC were added are indicated below. Each solution is referred by a letter (A, B, C, and D) which will be used throughout the paper, including figures and tables. In order to determine the influence of the salt concentrations, we have also used dilutions of the total compositions indi-

FIG. 3. Flux of permeate as a function of percent volume filtered (total volume 10 mls). Same conditions and symbols as in Fig. 1.

coidas

AP: Colloid

486

TOUNISSOU ET AL.

FIG. 4. Flux of permeate as a function of percent volume filtered (total volume 10 mls). Same conditions and symbols as in Fig. 2.

cated below: one quarter, one half, and three quarter (this will be referred for instance as A/4, A/2, 3A/4). Solutions (A): NaHCO3 1 M, pH 8.5–8.6; (B): NaHCO3 / Na2CO3 1.1 M, pH 9.5–9.6; (C): NaHCO3 /Na2CO3 1.1 M, NaNO2 0.9 M, pH 9.5–9.6; (D): NaHCO3 /Na2CO3 1.1 M, NaNO3 0.9 M, pH 9.5–9.6; (E) NaHCO3 /Na2CO3 1.1 M, NaNO2 0.9 M, NaNO3 0.9 M, pH 9.5–9.6. RESULTS AND DISCUSSION

The solubility of CPC solutions (2 1 10 02 M) containing different kinds of salts was tested first. It was demonstrated that there were no real solubility problems in the presence

of solutions A, B, and C whatever the dilution factor considered (in case of solutions C the more concentrated solutions had to be heated just above room temperature). Solutions D and E were only soluble at 407C at any dilution, with the exception of the more concentrated E solution, which remained turbid under these conditions. For all these solutions except the last one we have carried out ultrafiltration experiments in order to measure the volumes of solution filtered as a function of time. The time was limited to 30 min, because we considered that longer filtration times have no practical interest. Figures 1 and 2 report, as examples, the results obtained with solutions A and C, respectively. With solutions D and E the filtration was extremely slow (even when membranes with molecular weight cutoff of 30,000 Da were used). The presence of nitrates in these solutions makes the filtration very difficult even though the solutions look totally homogeneous and transparent. The results concerning solutions D and E will not be reported for this reason. Figures 3 and 4 represent the variation of the fluxes (in liters h 01 m02 ) as a function of the filtered volume (in percent of the initial volume), in connection with the data reported in Figs. 1 and 2, respectively. The latter representation clearly shows the position of the ‘‘gel point,’’ i.e. the point where the flux becomes zero. Note that concentration polarization is known to be responsible for the flux decline (24–26). This phenomenon added to the adsorption of the surfactant (27–29) on the polymeric membrane finally leads to a complete plugging, especially in case of stirred cell ultrafiltration. When only carbonates or hydrogenocarbonates are present in addition to the cationic surfactant, ‘‘the gel point’’ occurs for a filtered volume larger

TABLE 1 Fluxes (liter h01 m02) Measured with Cellulose Membranes for CPC 0.02 M in the Presence of Different Concentrations of Electrolytes (See Text for Definitions of Solutions A, B, and C and Their Dilutions) at Different Stages of Ultrafiltration and Filtered Volume (%) at ‘‘Gel Point’’

Pure H2O (reference) CPC 0.02 M alone A 3A/4 A/2 A/4 B 3B/4 B/2 B/4 C 3C/4 C/2 C/4

AID

JCIS 4452

/

Initial stage

At 20% volume filtered

At 50% volume filtered

At 75% volume filtered

Filtered volume% at ‘‘gel point’’

98 91 80 82 88 88 59 61 74 85 10 16 43 89

100 84 65 67 77 80 53 60 67 73 10 17 39 87

98 85 65 65 73 77 51 56 63 71 — — 37 90

95 69 46 48 52 54 40 44 50 55 — — 22 73

— 98 96 97 97 98 96 96 96 97 — — — 96

6g17$$$$62

09-30-96 14:00:48

coidas

AP: Colloid

487

ULTRAFILTRATION OF MICELLAR SOLUTIONS

TABLE 2 Fluxes (liter h01 m02) Measured with Cellulose Membranes for CTAB 0.02 M in the Presence of Different Concentrations of Electrolytes at Different Stages of Ultrafiltration (See Text for Definitions of Solutions A and B and Their Dilutions) and Filtered Volume (%) at ‘‘Gel Point’’

Pure H2O (reference) CTAB 0.02 M alone A 3A/4 A/2 A/4 B 3B/4 B/2 B/4

Initial stage

At 20% volume filtered

At 50% volume filtered

At 75% volume filtered

Filtered volume % at ‘‘gel point’’

98 94 74 82 91 90 66 64 71 76

100 82 66 72 76 84 55 59 61 67

98 88 62 65 67 81 52 55 58 64

95 73 47 48 43 63 42 44 46 51

— 98 96 96 97 98 96 96 96 97

than 90% of the initial one, allowing concentration factors larger than 10. We have collected in Tables 1–3 the data obtained in a similar way for a large number of surfactant/electrolytes systems. These tables give the fluxes measured at different stages of ultrafiltration and the filtered volume at the ‘‘gel point.’’ They show that (i) for a certain type of electrolyte composition the fluxes measured increase as expected when the dilution factor is increased; (ii) for a fixed dilution, the flux value is little affected by a change of pH when only carbonates and hydrogenocarbonates are present (solutions A and B and their dilutions), but the addition of NaNO2 has a drastic effect in reducing the flux (solution C); (iii) the results obtained with CTAB and CPC (Tables 1 and 2) are very similar, indicating that the nature of the cationic head does not play an important part in controlling the flux. We can conclude that, as far as cationic surfactants are concerned, the presence of large amounts of carbonates or hydrogenocarbonates is not a problem for what regards the fluxes obtained in ultrafiltration. When nitrites or nitrates are also present in the preceding solutions the flux decline is significant, which may induce problems in practical applications.

Table 3 is concerned with an anionic surfactant, SDS. The fluxes were determined at two surfactant concentrations (5 1 10 02 and 0.1 M) and in the presence of 0.3 M NaNO3 . Under these conditions the addition of nitrate has quite a small influence on the flux since the reducing effect of adding 0.3 M nitrate to a 5 1 10 02 M SDS solution, is less important than that obtained by simply doubling the SDS concentration in the absence of nitrate. All the preceding data were obtained with a hydrophilic membrane made of cellulose. Some additional experiments were performed with a hydrophobic membrane made of polysulfone in order to compare the fluxes obtained in both cases. The results are reported in Fig. 5 for CPC 2 1 10 02 M alone and in the presence of electrolyte solutions A and B, respectively (the full symbols refer to the polysulfone membrane and the empty ones to cellulose). The flux appears to be significantly larger with the polysulfone membrane in the cases of pure water (not reported in the figure) and surfactant alone. When the electrolytes are present either the fluxes become comparable (solution A) or the polysulfone membrane shows a reduced flux (solution B). So even though pure water is crossing polysulfone membranes more

TABLE 3 Fluxes (liters h01 m02) Measured with Cellulose Membranes for SDS at Different Concentrations and in the Presence of 0, 3 M Na NO3 at Different Stages of Ultrafiltration and Filtered Volume (%) at ‘‘Gel Point’’

Pure SDS SDS SDS

Initial stage

At 20% volume filtered

At 50% volume filtered

At 75% volume filtered

Filtered volume % at ‘‘gel point’’

98 90 73 81

100 79 65 71

98 70 50 63

95 39 19 42

— 96 92 94

H2O (reference) 0, 05 M 0, 1 M 0, 05 M / 0, 3 M Na NO3

AID

JCIS 4452

/

6g17$$$$62

09-30-96 14:00:48

coidas

AP: Colloid

488

TOUNISSOU ET AL.

FIG. 5. Comparison of the fluxes of permeate versus percent filtered volume obtained with polysulfone (PS) and cellulose (CL) membranes respectively. Full symbols Å PS; empty symbols Å CL. CPC 2 1 10 02 M in the absence of salts ( l, L ) and with 1 M NaHCO3 (solution A) ( m, n ) or 1.1 M NaHCO3 /Na2CO3 (solution B) ( l, s ).

easily than cellulose membranes, the flux reduction upon salt addition is even larger. This indicates that, in addition to the viscosity effect, there must be a modification of the interactions between the membrane and the surfactant. After considering the flux behavior, we turn now to the problem of leakage of monomeric surfactant in the filtrate with special attention on the effect of electrolytes. It has been demonstrated in many instances that when micellar solutions are treated by ultrafiltration, the surfactant concentration found in the filtrate is equal or close to the cmc (this assumes of course that the molecular weight cutoff of the membrane has been adequately chosen so as to retain the micellar particles) (11, 30). This may seem surprising when put in relation with the known dynamic nature of micelles (31). In fact, the dynamic processes being very fast compared to the ultrafiltration time, the number of micelles is not changed. On the other hand the presence of salts is known to decrease the cmc of ionic surfactants, due to the electrostatic shielding effect (32): the repulsive forces between the head groups are normally fighting against the aggregation, which becomes easier in the presence of electrolyte. An empirical law of the form (33)

FIG. 6. Recording of absorbance change versus time during ultrafiltration of 2 1 10 02 M micellar solutions of CPC in the absence of salt and with different concentrations of hydrogenocarbonates ( l Å 260 nm, optical path 0.1 cm). (1) without salt; (2) A/4; (3) A/2; (4) 3A/4; (5) A. Ultrafiltration membrane: cellulose (YM 10).

measured as a function of time (see Experimental Part). This was done for CPC 10 02 M alone and in the presence of salts (solutions A, B, and C at different dilutions). An example of the curves showing the absorbance of the filtrate versus time is given in Fig. 6 for the surfactant alone and with solutions A/4, A/2, 3A/4, and A. The curve for CPC alone shows an almost instantaneous increase of absorbance until a plateau value is reached at 0.45 Absorbance unit. The extinction coefficient of CPC at 260 nm was determined to be 4364 M 01 cm01 and was found to be independent of the presence of the electrolytes used. The plateau thus corresponds to a CPC concentration of 1.03 1 10 03 M, whereas

log cmc Å 0 a log Cc / b has been proposed to take into account the salt effect observed with different kinds of surfactants (Cc is the total counterion concentration and a, b are constants for a particular ionic head group). The expected decrease of the cmc upon salt addition is confirmed by the ultrafiltration experiments with CPC, whose concentration in the filtrate could be continuously

AID

JCIS 4452

/

6g17$$$$62

09-30-96 14:00:48

FIG. 7. CPC concentration found in the permeate as a function of the concentration of different salts: solution A ( h ); solution B ( n ); solution C ( j ), with dilutions 1/4, 1/2, and 3/4. Factor 1 corresponds to no dilution of the salt solution and Factor 0 to the situation in the absence of added salt. Ultrafiltration membrane: cellulose (YM 10).

coidas

AP: Colloid

489

ULTRAFILTRATION OF MICELLAR SOLUTIONS

FIG. 8. CPC concentration as a function of percent filtered volume. Comparison of polysulfone (PS) and cellulose (CL) membranes. Full symbols Å PS; empty symbols Å CL. CPC 2 1 10 02 M in the absence of salts ( l, L ) and with 1 M NaHCO3 (solution A) ( m, n ) or 1.1 M NaHCO3 / Na2CO3 (solution B) ( l, s ). 04

the cmc is equal to 9.0 1 10 M. The concentration in the filtrate during the first 15–20 min can thus be considered to be close to the cmc. A significant increase of the surfactant concentration in the filtrate is occuring when the ‘‘gel point’’ is approached, suggesting that this gel is partly pushed through the membrane pores. This last effect no longer occurs in the presence of salts, but in that case the main effect is a drastic decrease of the monomer concentration which is divided by a factor of almost 10 with solution A/4 and by a factor even larger for solutions A/2, 3A/4, and A. The same kind of experiments was also performed in the presence of solutions B and C, with results very similar to those reported in Fig. 6. All the data have been collected in Fig. 7, in which we have plotted the changes of the surfactant concentration in the filtrate versus the concentration of the electrolyte solutions. The drop of the surfactant concentration with salt appears to increase in the order: solution A õ solution B õ solution C. The level of CPC concentration for the higher concentrations of electrolytes is so low that its precise measurement is not easy. For this reason we cannot certify the accuracy of the corresponding determinations. The important point is that a high salinity can be considered favorable to reduce the contamination of the filtrate by surfactants. We have finally tested the leakage of surfactant in the case of polysulfone membranes. Figure 8 shows the variation of the CPC concentration versus the percent of filtered solution. The passage of the surfactant in the absence of salt can be seen to be significantly reduced comparatively to the cellulose membrane (remember that the cmc in that case corresponds to 0.9 mM). On the other hand the effect of the added salt appears to be very weak, contrary to what

AID

JCIS 4452

/

6g17$$$$62

09-30-96 14:00:48

was observed with cellulose membranes. The best choice of the membrane material will thus depend on the nature of the solution to be treated and it cannot be made in absolute terms. By choosing some systems of practical interest we have tried in this work to determine how some of the parameters governing the ultrafiltration of micellar solutions may be influenced by the presence of salts. The fluxes at different stages of the ultrafiltration, the ‘‘gel point,’’ the amount of surfactant in the permeate have been studied for this purpose as a function of different salt concentrations and by varying the nature of the surfactants and of the material constituting the membranes. Some limits have appeared in the use of stirred cell ultrafiltration. We have been investigating at the same time the behavior of similar systems in tangential ultrafiltration, using mineral membranes. The results will be reported in our next paper. REFERENCES 1. Scamehorn, J. F., and Harwell, J. H. (Eds.), ‘‘Surfactant-Based Separation Processes: Surfactant Science Series,’’ Vol. 33. Dekker, New York, 1989. 2. Pramauro, E., and Bianco Prevot, A., Pure Appl. Chem. 67, 551 (1995). 3. Tondre, C., Hebrant, M., Ismael, M., and Son, S.-G., ‘‘Proceedings 1st World Congress on Emulsions, Vol. 3, Comm. 4-30-047, Paris, 1993.’’ 4. Dunn, R. O., Scamehorn, J. F., and Christian, S. D., Colloid Surf. 35, 49 (1989). 5. Scamehorn, J. F., Christian, S. D., and Ellington, R. T., in ‘‘SurfactantBased Separation Processes: Surfactant Science Series’’ (J. F. Scamehorn and J. H. Harwell, Eds.), Vol. 33, p. 29. Dekker, New York, 1989. 6. Dharmawardana, U. R., Christian, S. D., Taylor, R. W., and Scamehorn, J. F., Langmuir 8, 414 (1992). 7. Scamehorn, J. F., Christian, S. D., El-Sayed, D. A., Uchiyama, H., and Younis, S. S., Sep. Sci. Technol. 29, 809 (1994). 8. Pramauro, E., Bianco Prevot, A., Pelizzeti, E., Marchelli, R., Dossena, A., and Biancardi, A., Anal. Chim. Acta 264, 303 (1992). 9. Pramauro, E., Bianco, A., Barni, E., Viscardi, G., and Hinze, W. L., Colloid Surf. 63, 291 (1992). 10. Reiller, P., Lemordant, D., Moulin, C., and Beaucaire, C., J. Colloid Interface Sci. 163, 81 (1994). 11. Hafiane, A., Issid, I., and Lemordant, D., J. Colloid Interface Sci. 142, 167 (1991). 12. Ismael, M., and Tondre, C., Langmuir 8, 1039 (1992). 13. Tondre, C., Son, S.-G., Hebrant, M., Scrimin, P., and Tecilla, P., Langmuir 9, 950 (1993). 14. Ismael, M., and Tondre, C., J. Colloid Interface Sci. 160, 252 (1993). 15. Hebrant, M., Bouraine, A., Tondre, C., Brembilla, A., and Lochon, P., Langmuir 10, 3994 (1994). 16. Richmond, W., Tondre, C., Krzyzanowska, E., and Szymanowski, J., J. Chem. Soc. Faraday Trans. 91, 657 (1995). 17. Scamehorn, J. F., Christian, S. D., Tucker, E. E., and Tan, B. I., Colloid Surf. 49, 259 (1990). 18. Mundkur, S. D., and Watters, J. C., Sep. Sci. Technol. 28, 1157 (1993). 19. Volchek, K., Krentsel, E., Zhilin, Y., Shtereva, G., and Dytnersky, Y., J. Membr. Sci. 79, 253 (1993). 20. Dunn, R. O., Scamehorn, J. F., and Christian, S. D., Sep. Sci. Technol. 20, 257 (1985). 21. Son, S.-G., Hebrant, M., Tecilla, P., Scrimin, P., and Tondre, C., J. Phys. Chem. 96, 11072 (1992).

coidas

AP: Colloid

490

TOUNISSOU ET AL.

22. Tondre, C., and Hebrant, M., J. Phys. Chem. 96, 11079 (1992). 23. Cierpiszewski, R., Hebrant, M., Szymanowski, J., and Tondre, C., J . Chem. Soc., Faraday Trans. 92, 249 (1996). 24. Grieves, R. B., Bhattacharyya, D., Schomp, W. G., and Bewley, J. L., AIChE J. 19, 766 (1973). 25. Koltuniewicz, A., and Noworyta, A., Ind. Eng. Chem. Res. 33, 1771 (1994). 26. Van Den Berg, G. B., and Smolders, C. A., Filtration & Separation 115 (1988). 27. Field, R., Hang, S., and Arnot, T., J. Membr. Sci. 86, 291 (1994). 28. Jonsson, A.-S., and Jonsson, B., J. Membr. Sci. 56, 49 (1991).

AID

JCIS 4452

/

6g17$$$$63

09-30-96 14:00:48

29. Markels, J. H., Lynn, S., and Radke, C. J., J. Membr. Sci. 86, 241 (1994). 30. Scamehorn, J. F., and Harwell, J. H., in ‘‘Surfactants in Chemical/ Process Engineering: Surfactant Science Series’’ (D. Wasan, M. E. Ginn, and D. O. Shah, Eds.), Vol. 28, p. 77. Dekker, New York, 1988. 31. Aniansson, E. A. G., Wall, S. N., Almgren, M., Hoffmann, H., Kielmann, I., Ulbricht, W., Zana, R., Lang, J., and Tondre, C., J. Phys. Chem. 80, 905 (1976). 32. Atwood, D., and Florence, A. T. (Eds.), ‘‘Surfactant Systems,’’ p. 92. Chapman and Hall, New York, 1983. 33. Corrin, M. L., and Harkins, W. D., J. Am. Chem. Soc. 69, 683 (1947).

coidas

AP: Colloid