Citation
Tailoring micellar stability to control interfacial properties and behavior of dispersed systems

Material Information

Title:
Tailoring micellar stability to control interfacial properties and behavior of dispersed systems
Creator:
Patist, Alexander, 1971-
Publication Date:
Language:
English
Physical Description:
xv, 143 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Alcohols ( jstor )
Colloids ( jstor )
Dyes ( jstor )
Foams ( jstor )
Interfacial tension ( jstor )
Micelles ( jstor )
Molecules ( jstor )
Monomers ( jstor )
Relaxation time ( jstor )
Surfactants ( jstor )
Chemical Engineering thesis, Ph. D ( lcsh )
Dissertations, Academic -- Chemical Engineering -- UF ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1999.
Bibliography:
Includes bibliographical references (leaves 134-142).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Alexander Patist.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
030479719 ( ALEPH )
43334795 ( OCLC )

Downloads

This item has the following downloads:


Full Text










TAILORING MICELLAR STABILITY TO CONTROL INTERFACIAL PROPERTIES
AND BEHAVIOR OF DISPERSED SYSTEMS








By

ALEXANDER PATIST


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1999


















I dedicate this dissertation to my parents, to my sister, Carla, and to my late grandmother, Oma Patist, who would have been so proud...














ACKNOWLEDGMENTS


I would like to express my sincere thanks and appreciation to my advisor, Professor Dinesh Shah, chairman of my supervisory committee, for his guidance, motivation, inspiration, and many non-scientific discussions relevant to happiness and fulfillment in life. Thanks also to the other supervisory committee members, Professors Brij Moudgil, Ben Koopman, Raj Rajagopalan, and Spyros Svoronos for their valuable time and suggestions.

Many thanks go to Dr. Patricia Aikens and Dr. Kevin Penfield from ICI Surfactants for their financial support of the research as well as several national and international surfactant symposia, which I was able to attend through their support.

Dr. Sunil Bhagwat is gratefully acknowledged for his input in the development of a new method for the determination of the slow micellar relaxation time for nonionic surfactants and Professor Wilhelm Knoche from the University of Bielefeld, Germany, for use of his pressure-jump equipment with optical detection.

I have greatly appreciated the opportunity to collaborate with the Engineering Research Center (ERC) for Particle Science and Technology at the University of Florida. Especially, Dr. Yakov Rabonovich, Joshua Adler, and Pankaj Singh for their successful help and implementation of the Atomic Force Microscopy and dispersion stability measurements. I would also like to acknowledge all the undergraduate students sponsored by the ERC for their experimental contributions to this thesis: Teri Axelberd, Brenda









Daneka, Nithya Desikan, Andrew Howes, Jonathan Johnson, Jr., Janardhan Lavakumar, Patrick Nguyen, Rahul Pagidipati, Stephen Patrick, Neemesh Shah, Rishita Shah, Franz Wakefield, and Seamus Wedge. It was great to work with you all.

I wish to thank my colleagues from the Department of Chemical Engineering and the Center for Surface Science and Engineering for their help and cooperation. Dr. Paul Huibers, Dr. Vishal Chhabra, Dr. Brajesh Jha, Dr. Patanjali, Pavan Shukla (Buddy) and last but not least, Steve and Vikki Truesdail for all their help, support, and the good times we shared together.

Finally, I owe a deep sense of gratitude to Dr. Sami Karabomi and the late Dr. Nico van Os at the Shell Research and Technology Centre in Amsterdam, for bringing me in touch with Professor Shah, which eventually resulted in this dissertation.















TABLE OF CONTENTS



ACKNOW LEDGM ENTS .................................................................................................. iii

LIST OF TABLES .............................................................................................................. ix

LIST OF FIGURES ........................................................................................................ x

ABSTRACT ..................................................................................................................... xiv

CHAPTERS

1 INTRODUCTION ................................................................................................... 1

1.1 Surfactants ............................................................................................... 1
1.2 M icellization ........................................................................................... 3
1.3 Structure of M icelles ............................................................................... 6
1.4 Dynamic Properties of Surfactant Solutions .......................................... 8
1.5 Importance of Micellar Relaxation Time on
Technological Processes ............................................................. 10
1.6 Rationale of the Proposed Research ...................................................... 15

2 DETERMINATION OF CRITICAL MICELLE CONCENTRATION
OF NONIONIC SURFACTANTS ........................................................ 19

2.1 Introduction .......................................................................................... 19
2.2 CMC Determination by the Dye Micellization Method ......................... 20
2.3 CMC Determination by the Surface Tension Method ........................... 23
2.4 Experimental Procedure ........................................................................ 24
2.4.1 M aterials .................................................................................... 24
2.4.2 Surface Tension ........................................................................ 24
2.4.3 Spectrophotometry ................................................................... 24
2.4.4 Foam Fractionation .................................................................... 25
2.5 Results and Discussion ........................................................................... 25
2.6 Conclusions .......................................................................................... 31









3 MICELLAR KINETICS OF NONIONIC SURFACTANTS ............................ 33

3.1 Introduction .......................................................................................... 33
3.2 Measurement of Nonionic Micellar Relaxation Time by
Stopped-Flow ............................................................................ 34
3.3 Experimental Procedure ........................................................................ 37
3.3.1 M aterials .................................................................................... 37
3.3.2 Relaxation Time Measurement by the Stopped-Flow Method ...... 37
3.3.3 Relaxation Time Measurement by Pressure-Jump with
Optical Detection ........................................................... 38
3.4 Results and Discussion .......................................................................... 38
3.5 Validation of the Slow Relaxation Time by Pressure-Jump with
Optical Detection ...................................................................... 40
3.6 C onclusions .......................................................................................... 44

4 MICELLAR KINETICS AND DYNAMIC SURFACE TENSION ................ 45

4.1 Introduction .......................................................................................... 45
4.2 Dynamic Surface Tension by the Maximum Bubble Pressure Method .... 46 4.3 Dynamic Surface Tension and Micellar Stability .................................. 48
4.4 Experimental Procedure ........................................................................ 50
4.4.1 M aterials ................................................................................... 50
4.4.2 Dynamic Surface Tension ........................................................ 51
4.5 Results and Discussion .......................................................................... 52
4.6 Importance of Diffusion Time of Nonionic Surfactants in DST
Measurements ........................................................................... 56
4.7 C onclusions .......................................................................................... 61

5 EFFECT OF TAILORING MICELLAR STABILITY ON FOAMING
PROPERTIES AND FOAMING METHODOLOGY .......................... 62

5.1 Introduction .......................................................................................... 62
5.2 Experimental Procedure ........................................................................ 66
5.2.1 M aterials ................................................................................... 66
5.2.2 Relaxation Time Measurement by Pressure-Jump with
Electrical Conductivity Detection ................................. 66
5.2.3 Surface Tension ........................................................................ 67
5.2.4 Surface Viscosity ...................................................................... 67
5.2.5 Foamability by Shaking ............................................................. 68
5.2.6 Foamability by Air Bubbling .................................................... 68
5.2.7 Foam Stability .......................................................................... 68
5.2.8 Dynamic Surface Tension ....................................................... 69
5.3 Foaming Properties of SDS/CnTAB Mixtures ..................................... 69
5.4 Foaming Properties of SDS/COH Mixtures ....................................... 74
5.5 C onclusions .......................................................................................... 82









6 IMPORTANCE OF MICELLAR STABILITY ON ANTIFOAMING
A C T IO N ............................................................................................... 84

6.1 Introduction .......................................................................................... 84
6.2 Experimental Procedure ........................................................................ 87
6.2.1 M aterials ................................................................................... 87
6.2.2 M ethods ...................................................................................... 87
6.3 Results and Discussion .......................................................................... 88
6.4 Comparison between Electrolyte and Antifoaming Agent on Micellar
Stability ..................................................................................... 95
6.5 C onclusions .......................................................................................... 97

7 TAILORING INTRAMICELLAR FORCES TO CONTROL
DISPERSION STABILITY ................................................................. 99

7.1 Introduction ......................................................................................... 99
7.2 Experimental Procedure .......................................................................... 102
7.2.1 M aterials ...................................................................................... 102
7.2.2 M icroscopy .................................................................................. 103
7.2.3 Pressure-Conductivity Measurements ......................................... 103
7.2.4 Turbidity Measurements .............................................................. 104
7.3 Results and Discussion ............................................................................ 104
7.3.1 A FM Im ages ................................................................................ 104
7.3.2 Force-Distance Curves ................................................................ 106
7.3.3 Relation between Micellar Stability and Dispersion Stability ..... 112 7.3.4 Mechanical Properties of the Adsorbed Micellar Layer .............. 116
7.4 C onclusions ............................................................................................. 119

8 SUMMARY AND RECOMMENDATIONS FOR FUTURE WORK ............... 120

8.1 Measurement of CMC of Nonionic Surfactants ...................................... 120
8.2 Micellar Relaxation Time of Nonionic Surfactants ................................ 121
8.3 Tailoring Micellar Stability to Control Foaming and Antifoaming
A ction .......................................................................................... 123
8.4 Tailoring Intramicellar Forces to Control Dispersion Stability ............... 125
8.5 Recommendations for Future Work ........................................................ 125


APPENDIX A HISTORICAL PERSPECTIVE ON MICELLAR K IN E T IC S ....................................................................... 127

APPENDIX B MATHEMATICAL PROOF OF CONSTANT CHARACTERISTIC DIFFUSION TIME FOR LARGE BUBBLE RADII ............................................. 132









REFEREN CES ................................................................................................................ 134

BIO G RA PH ICA L SKETCH ........................................................................................... 143














LIST OF TABLES


Table pge

2-1 Comparison of critical micelle concentrations as determined by surface
tension and dye micellization methods ................................................. 26

3-1 Micellar relaxation constants, t2, measured by the stopped-flow dilution
technique .............................................................................................. 39

4-1 Equilibrium surface tension, CMC and the slow micellar relaxation
time "2 as measured by the stopped-flow dilution technique ................ 52

4-2 Constants used in the calculation of the characteristic diffusion time, td .............. 57

4-3 Comparison of the characteristic diffusion time, td and the slow micellar
relaxation time, T2 in the dynamic surface tension measurement
for R l> 1 m m ........................................................................................ 60

7-1 CMC values of alkyltrimethylammonium bromides ........................................... 104














LIST OF FIGURES


Figure pg 1-1 Schematic representation of the three states in which surfactant molecules
reside in w ater ........................................................................................ 4

1-2 Mechanisms for two relaxation times, ri and T2, involved in a surfactant
solution above CM C ............................................................................... 9

1-3 At equilibrium, the rate of micelle formation equals the rate of micelle
disintegration ........................................................................................ 10

1-4 Various liquid/gas phenomena exhibiting minima and maxima at 200 mM
SD S concentration ................................................................................. 12

1-5 Various liquid/liquid and solid/liquid phenomena exhibiting minima
and maxima at 200 mM SDS concentration ........................................... 13

1-6 Correlation between molecular properties and macroscopic phenomena ...... 15 2-1 Scan of a UV-VIS absorbance spectrum .......................................................... 20

2-2 CMC determination of Tween 20 (CMC = 0.042 mM) using the dye
micellization method (absorbance at 542 nm) ..................................... 22

2-3 Critical micelle concentration of Tween 20 determined by surface
tension and dye micellization methods ................................................. 27

2-4 Critical micelle concentration of Tween 20 determined after foam
fractionation by surface tension and dye micellization methods ........... 28

2-5 Critical micelle concentration of pure C12(EO)5 determined by surface
tension and dye micellization methods ................................................ 29

2-6 Schematic diagram showing how the surface tension method would
suggest a lower CMC than the dye micellization method because
of the saturated air/liquid interface ........................................................ 30









3-1 Absorbance spectra of Eosin Y in water and 2 mM Triton X-100
solution (Eosin Y concentration: 0.019 mM) ........................................ 34

3-2 Stopped-flow apparatus used for determination of the slow micellar
relaxation constant, t2, for nonionic surfactants .................................... 35

3-3 Schematic diagram showing the micellar relaxation process involved
in a stopped-flow dilution experiment ................................................. 36

3-4 Typical relaxation curve obtained in a stopped-flow dilution experiment ........ 36 3-5 Schematic diagram showing the micellar relaxation process involved
in a pressure-jump experiment ............................................................ 40

3-6 Validation of relaxation constants, T2, by pressure-jump and
stopped-flow dilution techniques, both with optical detection .............. 41

4-1 Characteristic bubble pressure vs. time curve in the maximum bubble
pressure method to determine dynamic surface tension ....................... 48

4-2 Effect of micellar stability on dynamic surface tension ................................... 49

4-3 Setup for the measurement of dynamic surface tension by means of the
maximum bubble pressure method ..................................................... 51

4-4 Dimensionless dynamic surface tension vs. bubble lifetime for 2 mM
solutions of Synperonic A7, Brij 35 and Synperonic AS0 ................... 53

4-5 Dimensionless dynamic surface tension vs. bubble lifetime for 2 mM
solutions of C12(EO)5 and C12(EO)8 ..................................................... 53

4-6 Dynamic surface tension vs. bubble lifetime for 2 mM solutions of
C12(EO )5 and C 12(EO )8 .......................................................................... 55

4-7 Characteristic diffusion time td as function of bubble size in stationary
solutions of Synperonic A7, Brij 35 and Synperonic A50 ................... 58

4-8 Characteristic diffusion time td as a function of bubble size in a stationary
solution of CI2(EO)5 and CI2(EO)8 ........................................................ 59

5-1 Schematic diagram for the adsorption of surfactant molecules onto
new surface area due to the disintegration of micelles ......................... 64

5-2 Tailoring SDS micellar stability by the addition of 1-dodecanol (C120H)
or dodecyltrimethylammonium bromide (C12TAB) .............................. 65









5-3 Effect of 5 mM CTAB on foaming properties of 100 mM SDS solutions ..... 71 5-4 Schematic diagram representing the effect of chain length compatibility
on molecular packing at the air/water interface .................................... 72

5-5 The effect of 5 mol% COH on the slow micellar relaxation time, t2, in
SD S solutions ........................................................................................ 74

5-6 The effect of 5 mol% COH on the slow micellar relaxation time, "2, in
25 and 200 mM SDS solutions ............................................................. 76

5-7 The effect of 5 mol% C120H on foaming properties of SDS solutions ........... 77

5-8 Dynamic surface tension (yD) of SDS and SDS/C120H solutions (15 mM
SD S, 5m ol% C120H ) ............................................................................. 80

5-9 Dimensionless dynamic surface tension (0) of SDS and SDS/C120H
mixtures (15 mM SDS, 5 mol% C120H) .............................................. 82

6-1 Structures of the antifoaming agents used in the present study ........................ 88

6-2 Effect of antifoaming agents on the slow micellar relaxation time, r2 and
foamability of 150 mM SDS solutions ................................................. 89

6-3 Effect of tetraalkylammonium chloride (TCnAC, for n = 1, 2, 3 and 4)
on the slow micellar relaxation time, T2 in 150 mM SDS solutions .................. 91

6-4 Effect of tetraethylammonium chloride on foaming properties of
150 m M SDS solutions ....................................................................... 93

6-5 Schematic representation of the micro-structural changes in the SDS
micellar packing upon addition of antifoaming agents ......................... 95

6-6 Effect of Na counterions on the micellar stability of 150 mM SDS
solutions ................................................................................................ 96

7-1 Schematic diagram showing the possible structures of surfactants
adsorbed at the solid/liquid interface ....................................................... 102

7-2 Top view AFM image of mica immersed in water (pH 6) .................................. 105

7-3 Top view AFM image of mica immersed in a 2 CMC CI4TAB
solution (pH 6) ........................................................................................ 106









7-4 Force-distance curve for a 2 CMC C12TAB solution showing the
different stages as the tip approaches the mica surface ........................... 107

7-5 Force-distance curves for 2 CMC solutions of CnTAB
(for n = 10, 12, 14 and 16) on mica (pH 6) ............................................. 108

7-6 Maximum compressive force as function of alkyl chain length for
CMC solutions of CnTAB (for n = 10, 12, 14 and 16) on mica .............. 108

7-7 Force-distance curves for a 2 CMC CI2TAB solution on mica
containing increasing amounts of SDS (pH 6) ........................................ 109

7-8 Maximum compressive force as function of SDS concentration for a
2 CM C C12TAB solution on mica ........................................................... 110

7-9 Change in electrical conductivity due to micelle break-up in the
pressure-jum p experim ent ....................................................................... 111

7-10 Turbidity of silica dispersions and the maximum compressive force vs.
concentration of C12TAB at pH 4 after 60 min .................. 113

7-11 Turbidity of silica dispersions and the maximum compressive force vs.
concentration of C12TAB in the presence of 100 mM NaCl at pH 4 ...... 114 7-12 Turbidity and the maximum compressive force vs. concentration of
SDS in the presence of 5 mM CI2TAB, 100 mM NaCI at pH 4 ............. 115

7-13 Calculation of the actual tip radius from the micellar profile obtained
by the A F M .............................................................................................. 117

7-14 Interaction parameters between the AFM tip and the adsorbed layer
of micelles for the calculation of the elasticity ........................................ 117

A-1 Typical size distribution curve of aggregates in a micellar solution ................... 128














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the degree of Doctor of Philosophy


TAILORING MICELLAR STABILITY TO CONTROL INTERFACIAL PROPERTIES AND BEHAVIOR OF DISPERSED SYSTEMS


By

Alexander Patist

August 1999


Chairman: Dinesh 0. Shah
Major Department: Chemical Engineering


The association of many classes of surface-active molecules into micellar aggregates is a well-known phenomenon. Micelles are in dynamic equilibrium, constantly disintegrating and reforming. This relaxation process is characterized by the slow micellar relaxation time constant, t2, which is directly related to the micellar stability. The micellar stability of sodium dodecyl sulfate (SDS) micelles has been shown to significantly influence technological processes involving a rapid increase in interfacial area, such as foaming, antifoaming, wetting, emulsification, solubilization, and detergency. First, the available monomers adsorb onto the freshly created interface. Then, additional monomers must be provided by the breakup of micelles. Especially when the free monomer concentration is low, the micellar breakup time is a rate-limiting step in the









supply of monomers, which is the case for many nonionic surfactant solutions. However, at present, no paper on the importance of nonionic micellar lifetime on dynamic interfacial processes is available in the literature. The aim of this study is to develop a method to measure the slow micellar relaxation time of nonionic surfactants and to relate the relaxation data to dynamic interfacial processes. In the present study, stopped-flow and pressure-jump techniques were used to determine the slow micellar relaxation time

(t2) of nonionic surfactants. Both techniques gave the same time constants, within the experimental error. The slow relaxation times are much longer for nonionic surfactants (from seconds to minutes) than for ionic surfactants (usually milliseconds) because of the absence of ionic repulsion between the head groups. The observed relaxation time was related to dynamic surface tension and foaming experiments. A slow breakup of micelles (i.e., a long relaxation time '2) corresponds to a high dynamic surface tension and low foamability, whereas a fast breakup of micelles leads to a lower dynamic surface tension and higher foamability.

Relaxation time data of surfactant solutions correlate with the (dynamic) properties of a given surfactant solution. Moreover, the results suggest that appropriate micelles with specific stability or T2 can be designed by controlling the surfactant structure, concentration, and physicochemical conditions, as well as by mixing anionic/cationic or ionic/nonionic surfactants for a desired technological application.














CHAPTER 1
INTRODUCTION


1.1 Surfactants

Surfactants are more commonly known as soaps, which have traditionally consisted of sodium salts of naturally occurring fatty acids. The advent of petrochemicals, however, has brought numerous synthetic detergents, which are typically sulfates, sulfonates, or trimethylammonium salts of long chain hydrocarbons [Miller and Neogi, 1985]. Today, surfactants are among the most versatile products of the chemical industry, appearing in motor oils, detergents to clean laundry and homes, cosmetics, foods, oil recovery, pharmaceuticals, mineral flotation, and contact lenses [Rosen, 1989]. The last decade has seen the extension of surfactant applications to such high-technology areas as electronic printing, magnetic recording, enhanced filtration systems, biotechnology, and microelectronics [Holmberg, 1998]. Recently, a whole new area of research was opened by the introduction of biodegradable, sugar-based surfactants [Holmberg 1998; von Rybinski and Stoll, 1997].

A surfactant is a surface-active substance that has the property of adsorbing onto surfaces or interfaces (gas/liquid, liquid/liquid or solid/liquid) to lower the surface or interfacial free energy. Surfactants have a characteristic molecular structure consisting of a nonpolar group that has very little attraction for the water, known as the hydrophobic group, and a polar group that has strong attraction for the water, called the hydrophilic









group. The adsorption of surfactant molecules at the air/water interface can be described by the Gibbs equation [Hiemenz and Rajagopalan, 1997]. As early as 1878, Gibbs [1948] derived a differential equation relating the surface tension, the number of moles and the chemical potentials of the components at the interface, dy=-X du, (1.1)



where dy is the change in interfacial tension of the solvent, F1 is the surface excess concentration, which can be approximated by the number of moles per unit area and dyi is the change in chemical potential of the components in the system. The Gibbs equation can be used to calculate the surfactant concentration at the interface and hence the area per molecule from the simple measurement of surface tension. For dilute solutions of a nonionic surfactant or a 1:1 ionic surfactant in the presence of electrolyte, equation (1.1) can be written as [Hiemenz and Rajagopalan, 1997],


F= r) (1.2) RT ydlnC


where R is the gas constant, T the absolute temperature and C the concentration of surfactant. The surface excess concentration, F, can be obtained from the slope of a plot of the surface tension y versus lnC at constant temperature, which then can be used for the calculation of the area per molecule a (in squared Angstrom),



1020 (1.3)
N avog17









where Navog is the Avogadro number and F is the surface excess concentration. An extensive list of areas per molecule for a variety of surfactants and counterions is given by Rosen [ 1989] and Oh and Shah [ 1993b].



1.2 Micellization

Since the beginning of the study of surfactant solutions, it was recognized that the physical properties of surfactant solutions, such as surface tension, osmotic pressure, electrical conductivity and solubility (as function of temperature) show an abrupt change in the neighborhood of a critical concentration. The unusual properties of fatty acid salts in dilute aqueous solution were first investigated by McBain [1913, 1920] and later by Hartley [1936]. Other evidence for molecular aggregation was obtained from vapor pressure measurements and the solubility of organic material. The formation of colloidalsized clusters of individual surfactant molecules in solution is now better known as micellization. Although first suggested by McBain [1913], the earliest concrete model for spherical micelles is attributed to Hartley et al. [1936]. Figure 1-1 schematically shows the three environments in which surfactant molecules reside in a typical surfactant solution. Surfactant molecules disperse as monomers in the aqueous phase, form aggregates (micelles), or adsorb as a film at the air/water interface. The surfactant is in dynamic equilibrium between these states. Thus, at given temperature, pressure and concentration, the number of monomers, micelles and monomers adsorbed at the air/water interface is fixed under equilibrium conditions.


















ose





Figure 1-1. Schematic representation of the three states in which surfactant molecules
reside in water (monomers adsorbed at the air/water interface, monomers
in the bulk solution and micelles).



The process of surfactant clustering or micellization is primarily an entropy driven process. When surfactants are dissolved in water, the hydrophobic group disrupts the structure of water and therefore increases the free energy of the system. Surfactant molecules therefore concentrate at interfaces, so that their hydrophobic groups are directed away from the water and the free energy of the solution is minimized. The distortion of the water structure can also be decreased (and the free energy of the solution reduced) by the aggregation of surface-active molecules into clusters (micelles) with their hydrophobic groups directed towards the interior of the cluster and their hydrophilic groups directed toward the water. However, the surfactant molecules transferred from the solution to the micelle may experience some loss of freedom from being confined to the micelle. In addition, they may experience an electrostatic repulsion from other similarly charged surfactant molecules in the case of surfactants with ionic head groups. These forces increase the free energy of the system and oppose micellization. Hence, micelle









formation depends on the force balance between the factors favoring micellization (van der Waals and hydrophobic forces) and those opposing it (kinetic energy of the molecules and electrostatic repulsion). The explanation for the entropy-dominated association of surfactant molecules is called the "hydrophobic effect" or "hydrophobic bonding" [Tanford, 1980].

The concentration at which micelles first appear in solution is called the critical micelle concentration (CMC). Representing the surfactant by S, the micellization process can be described by the reaction,


nS �* S,, (1.4)


in which S, is the micelle with a degree of aggregation n. The aggregation number n increases with increasing length of the hydrophobic group and decreases with increasing size of the hydrophilic group [Rosen, 1989]. In general, the greater the dissimilarity between the surfactant molecules and the solvent, the greater the aggregation number.

Experimentally, the CMC is determined from the discontinuity or inflection point in the plot of a physical property of the solution as a function of surfactant concentration. A wide variety of techniques involving the measurement of such physical properties as the surface tension, conductivity, light scattering intensity and osmotic pressure have been used to determine CMC values [Preston, 1948; Shinoda and Nakagawa, 1963; Mukerjee and Mysels, 1971]. In general, those factors that increase the aggregation number tend to decrease the CMC. For example, increasing the alkyl chain length of the surfactant, decreases the CMC. The presence of electrolyte also decreases the CMC, due to the so-called salting out effect. The work needed to create the volume in water required









to accommodate a nonpolar solute is increased in electrolyte solution because of the strong water-ion interactions. When the surfactant monomers are salted out by the presence of electrolyte, micellization is favored and the CMC is decreased. Another factor favoring micellization in electrolyte solutions is the shielding of charges between the ionic head groups (in case of ionic surfactants) [Rosen, 1989]



1.3 Structure of Micelles

In the last couple of decades, the recognition that surfactant structures can mimic biological structures has gained substantial interest [Bergethon and Simons, 1990]. Enzymes, for example, are protein molecules into which a reactant molecule fits to form a reactive intermediate. The highly efficient and specific catalytic effect of enzymes makes their investigation an interesting area of biomedical and detergent research [Gloxhuber and Kunstler, 1992; van Ee et al., 1997]. Likewise, cell membranes not only compartmentalize biological systems but also play a variety of functions in the life of the cell. Surfactant structures can be used as model systems to mimic both enzymes and membranes. Lipid aggregates known as liposomes are common in physiological systems, for example the use of specially designed liposomes as drug-delivery vehicles [Lasic, 1993]. Self-assembled structures such as micelles or reversed micelles also play an increasingly important role in catalysis and separation processes in engineering and environmental science and technology [Fendler, 1975; Myers, 1991; Gratzel and Kalyanasundaram, 1991 ].

A theory of micellar structure, based upon the geometry of various micellar shapes and the space occupied by the hydrophilic and hydrophobic groups of the surfactant








molecules, has been developed by Israelachvili et al. [1976] and Mitchell and Ninham [1981]. In aqueous media, for example, surfactants with bulky or loosely packed hydrophilic groups and long, thin hydrophobic groups tend to form spherical micelles, while those with short, bulky hydrophobic groups and small, close packed hydrophilic groups tend to form lamellar or cylindrical micelles. At concentrations slightly above the CMC, micelles are considered of spherical shape. Changes in temperature, surfactant concentration or additives in the solution may change the size, shape, aggregation number and stability of the micelles. The structure of a micelle could vary from spherical to rod or disc-like to lamellar in shape. In concentrated solutions (much higher than the CMC), lamellar micelles form, such that the water molecules occupy the region between parallel sheets of surfactants. Micelles may also form long cylinders packed together (known as lyotropic mesomorphs or liquid crystalline phases) at high surfactant concentrations [Miller and Neogi, 1985; Ekwall, 1967].

The adsorption of surfactants at solid-liquid interfaces is very important in controlling large scale industrial processes, such as liquid/liquid dispersion stability in paints, detergency, water purification, oil recovery and ore flotation [Rosen, 1989; Adamson and Gast, 1997]. Many studies have been undertaken over the last three decades to investigate the characteristics of aggregates formed by a surfactant on a solid surface, using calorimetric studies, neutron-scattering, ellipsometry and the surface force apparatus developed by Israelachvili [1991]. All those techniques have provided quantitative measures on adsorption but little information on aggregate structure. Much is known about structures of aggregates in bulk solution, leading to spherical or cylindrical micelles, bilayers and bicontinous phases in bulk solution. At interfaces however, the









self-assembly process is influenced by additional surfactant-surface and solvent-surface interactions, including surface roughness, heterogeneity and charge behavior [Manne et al., 1994]. Recently, the structure and shape of those adsorbed aggregates have been revealed by atomic force microscopy (AFM) [Manne and Gaub, 1995]. The results suggest that the solid surface can alter the micellar structure of adsorbed aggregates. The studies presented in this dissertation (Chapter 7), indicate that dispersion stability can be modified by the strength of the adsorbed surfactant layer on the solid surface.



1.4 Dynamic Properties of Surfactant Solutions

The association of many classes of surface-active molecules into micellar aggregates is a well-known phenomenon. Micelles are often drawn as static structures of spherical aggregates of oriented surfactant molecules. However, micelles are in dynamic equilibrium with individual surfactant molecules that are constantly being exchanged between the bulk and the micelles. Additionally, the micelles themselves are continuously disintegrating and reassembling. There are two relaxation processes involved in micellar solutions. The first one is the fast relaxation process referred to as 't, (generally on the order of microseconds), which is associated with the fast exchange of monomers between micelles and the surrounding bulk phase. This process is considered as the collision between surfactant monomers and micelles. The second relaxation time 't2 (usually of the order of milliseconds to minutes) is attributed to the micelle formation and dissolution process. Figure 1-2 shows the two characteristic relaxation times, "r1 and t2, associated with micellar solutions.














T+I 00






"12


Figure 1-2. Mechanisms for the two relaxation times, "t and t2, involved in a
surfactant solution above CMC.



Micellar relaxation kinetics show dependence on temperature, pressure and concentration and have been studied by various techniques such as stopped-flow, temperature-jump, pressure-jump and ultrasonic absorption [James et al., 1977; Tondre et al., 1975; Hoffmann et al., 1976; Frindi et al., 1994; Kato et al., 1995; Lang, 1987; WynJones, 1975]. Micelle formation and disintegration can be compared to the equilibrium between water and water vapor at a given temperature and pressure. For a closed system containing water and water vapor in equilibrium, one can assume that the number of water molecules per unit area per second evaporating from the surface is the same as the number of water molecules condensing at the surface. Thus, the total number of molecules in the vapor phase or in the liquid does not change with time. So, the rate of condensation is equal to the rate of evaporation. The same principle holds for a micellar solution. Figure 1-3 shows schematically the formation and disintegration of micelles. It is evident that at equilibrium the number of micelles formed in a given time, is equal to






















Figure 1-3.


t=O t>O

At equilibrium, the rate of micelle formation equals the rate of micelle disintegration.


the number of micelles disintegrated in the same time period as shown by the broken circles in Figure 1-3 for t > 0. In this study more than one experimental technique was used to confirm that indeed both processes (i.e. micelle formation and disintegration) occur at the same rate (Chapter 3).

The two relaxation times can be used to calculate two important parameters of a micellar solution: (1) the residence time of a surfactant molecule in a micelle and (2) the average lifetime or stability of a micelle. The kinetics of this process have been evaluated by Aniansson et al. [1976] and Kahlweit [1981, 1982]. A brief historical perspective on micellar kinetics is given in Appendix A.



1.5 Importance of Micellar Relaxation Time on Technological Processes

The importance of micelle breakup on processes involving an increase in interfacial area was first reported by Mijnlieff et al. [1965]. For several years, researchers have tried to correlate the slow micellar relaxation time, t2, with equilibrium properties,









such as surface tension and surface viscosity, with no success. However, a strong correlation was found between the T2 of sodium dodecyl sulfate (SDS) micelles and various dynamic interfacial processes such as foamability, wetting time of textiles, bubble volume, emulsion droplet size and solubilization rate of benzene [Shah, 1998]. The micellar stability of SDS solutions was determined earlier by Lessner et al. [1981a,b] and later by Oh and Shah [1993a] using pressure-jump with electrical conductivity detection. This technique is described in detail by Huibers et al. [1996]. Maximum micellar stability was found at 200 mM (5 seconds). Figure 1-4 presents the various phenomena exhibiting minima and maxima at the liquid/gas interface. At 200 mM SDS, minimum foamability, maximum single film stability, maximum single bubble volume and a minimum frequency of bubble generation were found. These phenomena were explained based upon the monomer flux to newly created interfaces. If the micelles in solution are very stable, they cannot provide monomers fast enough to the interface and thus the interfacial tension remains higher. Therefore, lower foamability, larger single bubble foam volumes and a minimum frequency of bubble generation were found [Oh and Shah, 1991; Oh et al. 1992]. Very unstable micelles, however, provide monomers fast enough to the surface resulting in lower interfacial tensions. Maximum single film stability was found at 200 mM, i.e., when the micelles are most stable [Patel et al., 1996].




























2 200 mM


Frequency of Bubble Generation Volume of Single Single Film Stability Foamability




Slow Micellar Relaxation Time, T2


SDS CONCENTRATION


Various liquid/gas phenomena exhibiting minima and maxima at 200 mM SDS concentration [Shah, 1998].


Various interfacial phenomena occurring at the liquid/liquid and solid/liquid interface in SDS solutions are shown in Figure 1-5. The wetting time and droplet size in emulsions exhibit maxima at 200 mM. When micelles are very stable, the flux of monomers decreases and hence the wetting process slows down.


Figure 1-4.










Time to Reach Saturation of SDS Solution by Benzene Detergency, Removal of Orange OT Solubilization Rate of Benzene



Droplet Size in Emulsions Wetting Time


200 mM

SDS CONCENTRATION


Various liquid/liquid and solid/liquid phenomena exhibiting minima and maxima at 200 mM SDS concentration [Shah, 1998].


Different types of fabrics, such as polyesters, Dacron, Nylon, cotton, and silk were investigated. The maximum wetting time of the investigated fabrics occurs at 200 mM SDS concentration. Although the absolute magnitude of the wetting time depends on the fabric, the maximum occurring at 200 mM is a property of the SDS solution and not of


Figure 1-5.









the fabric. The liquid/liquid and solid/liquid phenomena can also be explained based upon the monomer flux necessary to stabilize newly created interfaces. Very stable micelles result in high dynamic surface tensions and hence larger droplet sizes and longer wetting times are obtained [Oh et al., 1993; Oh and Shah, 1992]. The solubilization rate of benzene in SDS solutions, as well as, the detergency or removal of orange OT dye from fabric surface, show maxima at 200 mM concentration. The time required to reach saturation of the SDS solution upon the addition of benzene is minimum at 200 mM SDS concentration. This suggests that very stable micelles (i.e., tightly packed micelles) are more effective in the solubilization of oil [Oh and Shah, 1993a]. This can be explained based upon the interior of the micelles. The interior of rigid (i.e., tightly packed) micelles is more hydrophobic as compared to that of loosely packed micelles and hence the stronger hydrophobic core causes more rapid partitioning or solubilization of benzene and Orange OT into the micelles at 200 mM SDS concentration. The maximum micellar stability occuring at 200 mM of SDS can be explained by the Intermicellar Coulombic Repuslion Model (ICRM). This model is based upon electrostatic repulsion, reduction in intermicellar distance and, in case of SDS micelles, a structural change from spherical to cylindrical micelles [Oh et al., 1993a; Reiss-Hudson and Luzzati, 1964; Ekwall, 1967].

In summary, SDS solutions exhibit maxima and minima for various properties at 200 mM concentration due to maximum stability of SDS micelles at this concentration. The more stable micelles lead to less monomer flux and hence to a higher dynamic surface tension. First, the available monomers adsorb onto the freshly created interface. Then, additional monomers must be provided by the breakup of micelles. Especially when the free monomer concentration is low, as indicated by a low CMC, the micellar









breakup time is the rate-limiting step in the supply of monomers, which is the case for many nonionic surfactant solutions. However, at present, no paper on the importance of nonionic micellar relaxation time on dynamic interfacial processes is available in the literature.



1.5 Rationale of the Proposed Research

From the previous sections, it has become clear that micelles are important in a variety of technological processes, such as foaming, wetting, solubilization, emulsification, wetting and detergency [Rosen, 1989; Miller and Neogi, 1985; Davies and Rideal, 1963]. Figure 1-6 illustrates how molecular properties of surfactants are related to the performance of technological processes.


I Surfactant Molecules


Correlation between molecular properties and macroscopic phenomena.


Adsorbed Films
(Surface Tension, Surface Viscosity)


Micelles
(Micellar Relaxation Time, Dynamic Surface Tension)


Applications
(Foams, Emulsions, Solubilization,
Lubrication, Wetting, etc.)


Figure 1-6.








The structure of surfactant molecules influences the equilibrium and dynamic properties of adsorbed films as well as those of micelles. Both of these, in turn, influence the performance of the technological processes mentioned above. In the present dissertation, the relation between SDS micellar stability and technological processes is extended to nonionic surfactants. Does the stability of nonionic surfactant micelles also influence dynamic interfacial properties, such as dynamic surface tension and foaming? Since the monomer concentration (i.e., the CMC) of nonionic surfactants is much lower than for ionic surfactants, it is expected that micelle breakup plays an even more important role in nonionic surfactant solutions. Relative little has been published on the kinetics of nonionic surfactants [Lang, 1975; Strey and Pakusch, 1986; Michels et al., 1997] and the results are often contradictory. Relaxation times in the order of microseconds have been reported using stopped-flow and temperature-jump techniques. In that case the mixing time of the stopped-flow apparatus (usually on the order of milliseconds) and the presence of salt in the temperature jump sample (to have a reasonable electrical conductivity) can easily lead to erroneous results. Moreover, no direct correlation between T2 and technological processes has been reported.

In Chapter 2, the importance of the method to be employed to determine the CMC of nonionic surfactants is discussed. The surface tension method, which is most commonly used for the determination of the CMC is compared to the dye micellization method for a wide range of technical grade as well as pure monodisperse nonionic surfactants. Commercially available, technical grade nonionic surfactants have a wide distribution in the degree of ethoxylation, which can lead to an erroneous interpretation of the surface tension versus surfactant concentration plot.









Chapter 3 presents a new method, which was developed to measure the slow micellar relaxation time of nonionic surfactants. The relaxation data for a variety of technical grade as well as pure monodisperse surfactants was determined.

In Chapter 4, the relaxation data as obtained in Chapter 3 was related to interfacial processes, such as dynamic surface tension and foaming. The ability to determine relaxation data of ionic as well as nonionic surfactant solutions allows for the prediction of the performance of a given surfactant solution. Moreover, the results suggest that one can design or tailor micelles with a specific stability by controlling the surfactant structure, concentration and physicochemical conditions, as well as by mixing anionic/cationic or ionic/nonionic surfactants for a desired technological application.

Chapter 5 demonstrates how the stability of SDS micelles can be tailored by the addition of oppositely charged surfactant, long chain alcohols or electrolyte, to control interfacial properties, such as surface tension, surface viscosity, foamability, foam stability and dynamic surface tension. Tailoring the SDS micellar stability by the addition of long chain alcohols gives rise to the fact that different methods of producing foam can result in opposite foaming behavior. The amount of foam produced depends on the time scale at which new interfacial area is produced. This is explained based upon dynamic surface tension and micellar stability.

If foaming behavior can be predicted by micellar stability, it is expected that antifoaming action also relates to micellar stability. Chapter 6 discusses the effect of micellar stability on antifoaming behavior for SDS solutions in the presence of a variety of commonly used antifoaming agents. A new way of controlling antifoaming action by tailoring micellar stability is presented.








In Chapter 7 an attempt has been made to correlate the effect of molecular interactions between surfactant molecules in bulk with the stability of micellar films adsorbed onto mica and silica. Atomic Force Microscopy (AFM) was used to reveal the structures of alkyltrimethylammonium bromides on mica and to measure the forcedistance curves between the AFM tip and the surface. The results are used to delineate the molecular mechanism by which surfactants influence dispersion stability. Since at high electrolyte concentrations the electrical double layer repulsion is negligible, the repulsive force must come from the steric stabilization by either surfactants, micelles, surfaceactive polymers or a mixture of both. Chapter 8 then summarizes the conclusions of this dissertation and the scope of the future extension of these studies.














CHAPTER 2
DETERMINATION OF CRITICAL MICELLE CONCENTRATION OF NONIONIC SURFACTANTS


2.1 Introduction

Ionic surfactants such as sodium dodecyl sulfate (SDS) or alkyl trimethylammonium bromides (CTAB) are usually easier to obtain in pure form than the ethoxylated nonionic surfactants. The commercially available (technical grade) ethoxylated surfactants not only have a size distribution in the hydrophobic part, but also in the degree of ethoxylation. The distribution of ethoxylation has been a subject of study for long time [Wanka et al., 1990]. However, the importance of this aspect in the method to be employed for the measurement of critical micelle concentration (CMC) has not been brought out in the literature satisfactorily. Large differences are often observed between CMC values of nonionic surfactants as determined by different methods. This is attributed to a broad molecular weight distribution and the presence of impurities [Alexandris et al., 1994; Mysels and Stafford, 1990].

There are a number of methods which have been employed for the determination of CMC of surface-active agents [Shinoda and Nakagawa, 1963; Hunter, 1987]. An excellent evaluation of the methods for determining the CMC is included in the comprehensive compilation of CMC's in aqueous solution by Mukerjee and Mysels [1971]. In this study the effect of size distribution in the ethoxylation of nonionic









surfactants will be discussed for two commonly used methods for the determination of CMC, namely dye micellization and the surface tension methods.



2.2 CMC Determination by the Dye Micellization Method

Dyes can be used in many ways to measure CMC. Water-soluble dyes, such as Merocyanine, Eosin, Rhodamine and Sudan are known to show a shift in the maximum wavelength (Xn) due to the presence of micelles [Shinoda and Nakagawa, 1963; Hunter, 1987]. This shift for Eosin Y is shown in Figure 2-1.



� ....... �Increasing surfactant '": y-concentration




Absorbance


Ct4
4.Increasing surfactant
concentration
SI I

518 542 Wavelength (nm)


Figure 2-1. Scan of a UV-VIS absorbance spectrum. The shift in the maximum
absorbance from 518 to 538 nm is due to the presence of micelles. The maximum difference is observed at 542 nm. Surfactant: Triton X-100;
dye: Eosin Y, 0.019 mM.



Eosin Y in water shows a maximum absorbance at 518 nm. Increasing the surfactant concentration, however, results in an increase of the absorbance at 538 nm. The









maximum difference is observed at 542 nm. This shift is either followed as a change in X ,mx or a change in absorbance of the micellized dye at a fixed wavelength (542 nm in Figure 2-1) as a function of surfactant concentration. It has been suggested that the inflection point in X.. should be treated as the CMC [Shinoda and Nakagawa, 1963]. However, not all dyes show a distinct shift in Xma (e.g. Merocyanine 540). Therefore, if Xn. of the micellized dye is sufficiently different from the aqueous dye, the absorbance at this wavelength (for Eosin Y at 542 nm) can be followed as a function of surfactant concentration to measure the extent of dye uptake. Below the CMC the rise in absorbance is very small, whereas above the CMC the rise is sharp. At the point where roughly half the dye is in the continuous phase, the maximum absorbance shows the sharp shift (in Figure 2-1 from 518 to 538 nm). Since the micellization process is known to be much less sharp for nonionic surfactants than for ionic surfactants, the rise in absorbance varies strongly over a range of surfactant concentration. At high enough surfactant concentrations, the absorbance vs. concentration curve will flatten again as most of the dye shifts to the micelles depleting the dye in the continuous phase. The linear portion near the inflection point is extrapolated to the point where the absorbance matches that of the dye in the absence of any surfactant and this concentration is defined as the CMC of the surfactant, see Figure 2-2.

A second method using dye involves the solubilization of a water-insoluble (hydrophobic) dye in micellar solutions. This solubilized dye can then be measured by its UV-visible absorbance. This method yields satisfactory results for some surfactants.



























Figure 2-2.


0.00 -..... .
0.001 0.01 0.1 1

Concentration of Tween 20 (mM)

CMC determination of Tween 20 (CMC = 0.042 mM) using the dye micellization method (absorbance at 542 nm). Eosin Y concentration: 0.019 mM. The horizontal dashed line represents the dye absorbance in water in the absence of surfactant.


However, sometimes the solubilization of such a dye is so high that the micellar structure is affected by the dye. If the solubilization is weaker, the dye taken up by the micelles is often insufficient to produce a reasonable absorbance signal. Especially, when the CMC is very low, the micellar phase at concentrations just above the CMC is far too small to solubilize a significant amount of dye. Therefore, it is difficult to define a clear CMC for nonionic surfactants using this method and hence this approach was not pursued in the present study.








2.3 CMC Determination by the Surface Tension Method

The surface tension of aqueous solutions of surface-active agents decreases rapidly until the CMC is reached and then stays constant above the CMC. An important point has to be made while using the surface tension method, irrespective of the instrument employed for measuring it. The surface tension method is based upon the fact that the free surfactant concentration remains almost unchanged after the onset of micellization even if the total surfactant concentration is increased. However, due to surface-active impurities, the surface may get saturated with highly surface-active molecules, although the actual onset of micellization may take place at a higher surfactant concentration. The surface tension may thus become invariant with surfactant concentration even below the CMC. Therefore, the saturation of the surface, identified as the CMC in the surface tension vs. concentration of surfactant plot, does not always reflect the presence of micelles in the bulk solution. In the present study, a comparison is made of the CMC of a few commercial surfactants (Tween 20, 22, 40, 60 and 80, Triton X-100, Brij 35, 58 and 78) as measured by the surface tension method (Wilhelmy plate) and the dye micellization method. When the impurities are selectively removed from a commercial surfactant sample, the value of the CMC as measured by the surface tension method is expected to be closer to the value measured by the dye micellization method. The present study compares the surface tension and dye micellization behavior of pure (monodisperse) and commercial (technical grade) nonionic surfactants.









2.4 Experimental Procedure

2.4.1 Materials

Tween 20, 22, 40, 60, and 80, and Brij 35, 58, and 78 were supplied by ICI Aricas, Inc. (Wilmington, DE). Triton X-100 was obtained from Aldrich Chemical Company (Milwaukee, WI). The pure nonionics penta-ethyleneglycol mono n-dodecyl ether (C12(EO)5) and octa-ethyleneglycol mono n-dodecyl ether (C12(EO)8) were purchased from Nikko Chemicals Co. (Tokyo, Japan). The high purity grade dyes Eosin Y (C2oH6Br4Na2O5, anionic) and Merocyanine 540 (C26H32N3NaO6S2, anionic) were supplied by Acros Organics (Fair Lawn, NJ). Deionized, distilled water was used in all experiments. Surfactant concentrations were calculated using their molecular weights disregarding the presence of any probable impurities.



2.4.2 Surface Tension

Equilibrium surface tensions were measured for freshly prepared solutions by the Wilhelmy plate method at 22�C. The platinum plate was always cleaned and heated to a red/orange color with a Bunsen burner before use.



2.4.3 Spectrophotometry

Eosin Y was used for the CMC determination of the nine commercial surfactants using the dye micellization method. Merocyanine 540 was used for the two ultra pure nonionic surfactants. Absorbance spectra were taken using a Hewlett Packard UV-VIS spectrophotometer (model 8453) with temperature control. All spectra were taken at 220C.









2.4.4 Foam Fractionation

Foam fractionation was carried out by shaking 25 mL of the surfactant solution vigorously such that the volume of the foam and the liquid together was four times the volume (100 mL) of the initial liquid. One half of the initial liquid was then separated from the foam as foam-fractionated sample and used for the surface tension and dyemicellization studies. The sample was expected to contain a lower amount of the highly surface-active impurities after foam fractionation. The surfactant loss due to foam fractionation was determined by a Tekmar-Dohrmann Phoenix 8000 Total Organic Carbon (TOC) Analyzer.



2.5 Results and Discussion

Table 2-1 shows the CMC values obtained by surface tension and dye micellization methods as well as the ratio of the two values. The CMC values obtained from the dye micellization method are approximately 1.6 to 6.5 times higher than the values obtained by the surface tension method. The smallest difference between CMCs as measured for Tween 80 (Table 2-1), indicates that this sample does not contain a large amount of surface-active impurities or a wide distribution of ethoxylation. The largest difference was observed for Tween 22. This is probably due to the large ethoxylation number (80) which makes a wide distribution of ethoxylation more likely. Figure 2-3 shows the different CMCs obtained for Tween 20. The break in the y vs. logC curve observed for these surfactants is thus not an indication of the CMC.









Table 2-1. Comparison of critical micelle concentations as determined by surface
tension (S.T.) and dye micellization (D.M.) methods. Dye concentration:
Eosin Y, 0.019 mM.

Surfactant Structure CMC by CMC by (CMCr"
S.T. D.M. CMCS.)

(mM) (mM)

Tween 20 Sorbitan Laurate Ester (E020) 0.011 0.042 3.8 Tween 22 Sorbitan Laurate Ester (E080) 0.013 0.084 6.5 Tween 40 Sorbitan Palmitate Ester (E020) 0.0067 0.024 3.6 Tween 60 Sorbitan Stearate Ester (E020) 0.0055 0.022 4.0 Tween 80 Sorbitan Oleate Ester (E020) 0.018 0.028 1.6 Triton X-100 Octyl Phenol Ether (E01o) 0.080 0.20 2.5 Brij 35 Lauryl Alcohol Ether (E023) 0.030 0.068 2.3 Brij 58 Cetyl Alcohol Ether (E020) 0.0028 0.01 3.6 Brij 78 Stearyl Alcohol Ether (E020) 0.0018 0.0071 3.9
C12(EO)5 ' Lauryl Alcohol Ether (EOs) 0.058 0.060 0.98 C12(EO)8 a Lauryl Alcohol Ether (E08) 0.070 0.072 0.97 a Pure nonionic surfactant. Merocyanine 540 dye was used for the CMC determination.




It may be expected that if the impurities are removed, the two methods yield results which are closer to each other. It has been suggested in literature [Tharapiwattananon et al., 1996; Chang et al., 1992; Chiu and Huang, 1991], that foaming and consequently skimming the foam away from the solution purifies the surfactant solution. This process, called foam fractionation, is a process in which solute species are adsorbed at a gas-liquid interface between a dispersed phase (gas bubble) and a continuous phase (bulk liquid). Foam fractionation processes have been used to remove surface-active agents from aqueous solutions [Elving, 1982].










0.30

0.25 0.20 0.15 0.10 005


z




F -3 Figure 2-3.


In order to confirm this, a solution of Tween 20 was subjected to foam fractionation and the remaining solution was used for the CMC determination. Total Organic Carbon analysis showed that less than 5% of the total surfactant concentration was removed and therefore the surfactant concentration was taken to be the same as before foam fractionation. Figure 2-4 shows a graph of the CMC determined by the dye micellization and surface tension methods for the foam fractionated Tween 20 sample. It is clear that the two methods show values much closer to each other (0.051 vs. 0.057 mM), indicating that the surface-active impurities as well as the lower ethoxylated molecules were removed from the original solution.


30 . . .. . ... J 0.00
0.001 0.01 0.1 1

Concentration of Tween 20 (mM)

Critical micelle concentration of Tween 20 determined by surface tension
(A, CMC = 0.011 mM) and dye micellization methods (+, CMC = 0.042 mM). Eosin Y concentration: 0.019 mM. The horizontal dashed line represents the dye absorbance in water in the absence of surfactant as well
as the equilibrium surface tension.









60 0.30 55 0.25 50 _ 0.20
0
45 '0.15


030
S40 0.1 0




30 0.00
0.001 0.01 0.1 1

Concentration of Tween 20 (mM)


Figure 2-4. Critical micelle concentration of Tween 20 determined after foam
fractionation by surface tension (A, CMC = 0.051 mM) and dye micellization methods (+, CMC = 0.057 mM). The horizontal dashed line represents the dye absorbance in water in the absence of surfactant as well
as the equilibrium surface tension.



Figure 2-5 shows the CMC curves for the pure (monodisperse) nonionic surfactant C12(EO)5. In this case, a different dye was used (Merocyanine 540), which shows a micellized dye peak at 575 nm. It is clear that in the absence of impurities, both surface tension and dye micellization methods yield the same result (see also Table 2-1).

This study shows that the surface tension method can be erroneous for commercial (technical grade) surfactants under common conditions. The onset of micellization in the case of a pure surfactant causes the free (non-micellized) surfactant concentration to remain constant when the total surfactant concentration is increased. This is correctly interpreted as the CMC of the surfactant. In case of impure surfactants or
























27 _


Figure 2-5.


0.50 0.40 0.30


0.20 0.10 0.00


0.01 0.1 1

Concentration of C12(EO)5 (mM)

Critical micelle concentration of pure C12(EO)5 determined by surface tension (A, CMC = 0.058 mM) and dye micellization methods (+, CMC = 0.060 mM). Merocyanine 540 concentration: 0.019 mM. The horizontal dashed line represents the dye absorbance in water in the absence of surfactant as well as the equilibrium surface tension.


surfactant mixtures, however, the situation is different. An impure sample of sodium dodecyl sulfate (SDS) for example, invariably contains lauryl alcohol, which is much more hydrophobic than SDS and thus has a much higher affinity for adsorption [Fruhner and Czichocki, 1996; Chiu and Wang, 1990]. Thus, the lauryl alcohol can saturate the surface and exhibit constant surface tension without any micelle formation in the bulk solution. Recently, Goebel and Lunkenheimer [1997] reported the importance of purity in the measurement of interfacial tension. As was shown in their study, using water/n-alkane interfaces, trace impurities significantly influence the interfacial tension. For nonionic surfactants, the lower ethoxylated species, which are always present in a commercial (technical grade) ethoxylated nonionic surfactant, have a higher affinity for adsorption









than the higher ethoxylated species. At concentrations much below the true CMC value, the air/liquid interface can be already saturated with the more surface-active species before any micelle formation in the bulk solution. This is visualized in Figure 2-6A.


A: below CMC


B: at CMC


t = Nonionic Surfactant


Surface Tension


CMCS.T.


CMCDye


Surfactant Concentration


Schematic diagram showing how the surface tension method would suggest a lower CMC than the dye micellization method because of the saturated air/liquid interface (solution A). Micelles start to form at a higher surfactant concentration as determined by the dye micellization method (solution B). Every open circle represents, for example, 4 (EO) groups.


Figure 2-6.









An increase in the bulk surfactant concentration from this point on will ultimately result in the formation of micelles at a specific surfactant concentration, which is the true CMC of the solution (Figure 2-6B). Although the CMC is indeed lowered by the presence of more surface-active species, the lowering is not as significant as suggested by the results from the surface tension method. The dye micellization method in such a situation would certainly yield a higher CMC than the surface tension method, as demonstrated in this study.



2.6 Conclusions

1. The surface tension method (Wilhelmy plate) for the determination of the CMC of

commercial (technical grade) nonionic surfactants is very sensitive to the presence of

molecular species with higher surface-activity.

2. The shortcoming of the surface tension method for determining the CMC of

technical grade nonionic surfactants was demonstrated. In the presence of highly surface-active impurities, the air/liquid interface gets saturated at concentrations much below the true CMC leading to a wrong interpretation of the break in the ', vs. logC curve. The micellized dye method gives reliable results for micelle formation in bulk

solution, even in the presence of such impurities.

3. Foam fractionation of a solution of technical grade nonionic surfactant can selectively

remove the species with higher surface-activity. The CMC values, as measured by the surface tension and dye micellization method, are in close agreement with each other

after foam fractionation.






32


4. The CMC values obtained for the pure monodisperse nonionic surfactants C12(EO)5

and C12(EO)8 are the same using surface tension and dye micellization methods,

indicating the absence of highly surface-active impurities.














CHAPTER 3
MICELLAR KINETICS OF NONIONIC SURFACTANTS


3.1 Introduction

As mentioned in Chapter 1, the micellar relaxation time measured for ionic surfactants (e.g., SDS) can be determined by the pressure-jump technique with electrical conductivity detection. This technique takes advantage of the fact that the CMC shifts to higher concentration when a surfactant solution is pressurized [Attwood and Florence, 1983; Kaneshina et al., 1983]. In case of ionic surfactants, the electrical conductivity increases with pressure. When the pressure is instantaneously released to atmospheric, monomers will reassociate to form new micelles, which can be followed as an exponential decay in the electrical conductivity with time [Huibers et al., 1996]. The slow micellar relaxation constant, t2, can be calculated from the first order reaction constant, k (t2 = 1/k). The pressure-jump technique with electrical conductivity detection is a very powerful tool that allows for the measurement of t2 for mixed micelles or micellar solutions in presence of additives. Chapters 5 and 6 demonstrate how this technique can be used to measure T2 for tailored micelles with a specific stability to control dynamic surface tension and hence technological processes, such as foaming and antifoaming. This chapter discusses a method developed to determine the slow micellar relaxation constant of a variety of nonionic surfactants. In the next chapter, the obtained relaxation data are compared to dynamic surface tension and foaming.








3.2 Measurement of Nonionic Micellar Relaxation Time by Stopped-Flow

The measurement of nonionic micellar relaxation time is more complicated than for ionic surfactants, since the electrical conductivity is not a sensitive parameter. Therefore, the use of a dye is necessary to obtain information about the micellar kinetics of nonionic surfactants. A number of dyes or fluorescent compounds, such as Merocyanine, Eosin, Rhodamine and Sudan show an appreciable change of extinction coefficient depending on whether the dye resides in, or outside the micelle in aqueous phase. This effect is often used to determine the CMC, as discussed in Chapter 2 [Shinoda and Nakagawa, 1963; Hunter, 1987], but it also provides a way of following the relaxation kinetics upon a fast temperature, pressure or concentration jump by employing spectrophotometric detection methods. As explained earlier in Chapter 2, Eosin Y in water shows a maximum absorbance (Xmax) at 518 nm.





Dye/surfactant
Dye/water solution

Absorbance s i Due to ,4 micelle / \ formation






518 542
Wavelength (nm)


Figure 3-1. Absorbance spectra of Eosin Y in water and 2 mM Triton X- 100 solution
(Eosin Y concentration: 0.019 mM).








Increasing the surfactant concentration, however, causes the dye to partition between the water and the micelles, causing the maximum absorbance to shift to approximately 538 nm. The maximum shift in absorbance occurs at 542 nm. (Figure 3-1). The concept of change in absorbance due to the presence of micelles can be used in the determination of the slow relaxation constant, T2, for nonionic surfactants using the stopped-flow dilution technique. Stopped-flow (Figure 3-2) is a method designed to measure the kinetics of fast reactions [James, 1977].





Chamber 1: U Light Source (UV-VIS)
Surfactant/Dye

~~~~Mixing Chabr b

SPMT Detector 0

Chamber 2: - Tm
Water/Dye l J "





Figure 3-2. Stopped-flow apparatus used for determination of the slow micellar
relaxation constant, Tr2, for nonionic surfactants.



The apparatus employs two separate syringes, which can be filled with reactants, which are pushed instantaneously into a transparent cell. The change in absorbance can be detected with a very sensitive photomultiplier detector as the reaction progresses. When one solution containing micelles and dye is instantaneously diluted with another solution








containing water and dye of the same dye concentration, the absorbance of dye in micelles will decrease as micelles breakup, indicating the relaxation time of micelles. This is schematically shown in Figure 3-3. Figure 3-4 shows the decrease in intensity (absorbance at 542 nm) as a function of time in a typical stopped-flow dilution experiment. The exponential decay shown in Figure 3-4 can be fit to first order reaction kinetics, resulting in the associated time constant 'r2.


S -


Figure 3-3. Schematic diagram showing the micellar relaxation process involved in a
stopped-flow dilution experiment. Micelles disintegrate after mixing,
thereby releasing dye into the water.


Signal
(V)


10


Time (s)


Figure 3-4.


Typical relaxation curve obtained in a stopped-flow dilution experiment. Mixing is induced at t = 0.


--4
lee.









3.3 Experimental Procedure

3.3.1 Materials

Tween 20, 22 and 80, Brij 35 and Synperonic A7 and A50 were supplied by ICI Americas, Inc. (Wilmington, DE). Triton X-100 was supplied by Aldrich Chemical Company (Milwaukee, WI). The pure nonionics penta-ethyleneglycol mono n-dodecyl ether (C12(EO)5) and octa-ethyleneglycol mono n-dodecyl ether (C12(EO)8) were purchased from Nikko Chemicals Co. (Tokyo, Japan). The high purity grade dyes Eosin Y (C2oH6Br4Na2Os, anionic) and Merocyanine 540 (C26H32N3NaO6S2, anionic) were supplied by Acros Organics (Fair Lawn, NJ). Deionized, distilled water was used in all experiments. Surfactant concentrations were calculated using their molecular weights disregarding the presence of any probable impurities.



3.3.2 Relaxation Time Measurement by the Stopped-Flow Method

The slow relaxation time T2 of the nonionic surfactants was measured by the stopped-flow dilution method. The stopped flow apparatus (Dionex Corporation, Houston, TX) mixes two solutions in approximately two milliseconds in a 20 mm light path cuvette. Two detectors allow for the simultaneous measurement of transmitted, fluorescent or scattered light, or the same light at different wavelengths (200-1000 nm). The monochromator includes a deuterium and tungsten light source. In this study the absorbance at 542 nm was measured as function of time. The system was connected to a Durrum (Palo Alto, CA) temperature jump system used here only to amplify the signal coming from the photomultiplier. The signal was stored using a Tektronix 340A oscilloscope.








3.3.3 Relaxation Time Measurement by Pressure-Jump with Optical Detection

The pressure-jump technique with optical detection was used to confirm the

relaxation data obtained by the stopped-flow apparatus. Experiments were performed at the University of Bielefeld, Germany, using the setup described in detail by Knoche and Wiese [1976]. All relaxation data were obtained at 22�C. The pressure-jump with optical detection is technically the same as the pressure-jump apparatus with electrical conductivity detection, except the detection method.



3.4 Results and Discussion

Table 3-1 shows the slow relaxation times T2 measured for a variety of nonionic surfactants using the stopped-flow dilution technique. A dye concentration of 0.019 mM was found to be the minimum concentration to still observe a significant absorbance signal. Tondre et al. [1975] found that a surfactant/dye molar ratio as low as 20 does not influence the relaxation time. With the exception of Tween 22, all other investigated surfactant samples have surfactant/dye ratios larger than 20.

It is clear that the relaxation time can vary from 2 to 150 seconds depending upon the molecular structure of the nonionic surfactant. The long relaxation time of 150 seconds for Synperonic A7, can be described as a frozen micelle as compared to those exhibiting relaxation times on the order of milliseconds (usually ionic surfactants). Nonionic surfactants show a much longer relaxation time (T2) than ionic surfactants, because of the absence of ionic repulsion between the head groups. The surfactants Synperonic A7, Brij 35 and Synperonic A50 have comparable alkyl chain lengths (C12C15) but increasing degree of ethoxylation. It is clear that increasing the number of









Table 3-1. Micellar relaxation constants, T2, measured by the stopped-flow dilution
technique.

Surfactant Structure Conc. CMCDye T2
(mM) (mM) (s)

Tween 20 Sorbitan Laurate Ester (E020) 0.47 0.042 6 Tween 22 Sorbitan Laurate Ester (E080) 0.37 0.084 2
Tween 80 Sorbitan Oleate Ester (E020) 0.49 0.028 8-10
Triton X-100 Octyl Phenol Ether (E010) 0.40 0.20 3.5 Synperonic A7 C12-C15 Alkanol Ether (E07) 0.80 0.050 150 Brij 35 Lauryl Alcohol Ether (EO23) 0.50 0.068 80 Synperonic A50 C12-C15 Alkanol Ether (E050) 0.40 0.084 40 CIAEO) Lauryl Alcohol Ether (EOs) 0.80 0.060 10 C12(EO)8" Lauryl Alcohol Ether (EO8) 0.40 0.072 4

a Pure (monodisperse) nonionic surfactant. Merocyanine 540 dye was used for the CMC and T2 determination. Both dyes resulted in the same CMC and T2 data.



ethylene oxide units decreases the relaxation time, which was also observed for octylphenyl polyoxyethylenes by Lang and Eyring [1972]. The relaxation times obtained for the ultra pure nonionic surfactants C]2(EO)5 and C12(EO)8 are relatively small as compared to the Synperonics (respectively 10 and 4 seconds as compared to 150 seconds). The difference might be attributed to the broad molecular weight distribution and the presence of impurities (Chapter 2). It is known [Greenshields, 1998], that Synperonic A7 contains a significant amount of long chain alcohols that apparently contributes to the stability of the micelles.









3.5 Validation of the Slow Relaxation Time by Pressure-Jump with Optical Detection

In the stopped-flow dilution technique the number of micelles decreases and thus the kinetics of micelle breakup is measured (Figure 3-3). However, in the pressure-jump technique (with either electrical conductivity or optical detection), the kinetics of micelle formation is measured after the pressure is released to atmospheric. This process is schematically shown in Figure 3-5. For a surfactant solution in equilibrium, the rate of micellar dissociation equals the rate of association. Therefore, both the stopped-flow and pressure-jump techniques should yield the same relaxation constant T2, if the perturbation is small enough.


High Pressure, High CMC


Low Pressure, Low CMC


Figure 3-5.


Schematic diagram showing the micellar relaxation process involved in a pressure-jump experiment. After the pressure is released to atmospheric, monomers reassociate to micelles.


In order to show that both the micelle breakup and micelle formation rates exhibit the same time constants, pressure-jump studies with optical detection were performed at the University of Bielefeld, Germany. Figure 3-6 shows the relaxation time t2 of Triton X100 and Brij 35 measured by stopped-flow and pressure-jump with optical detection


k+









techniques (absorbance at 542 nm). It is evident that the relaxation time measured for both surfactants is the same for both techniques within the experimental error. This suggests that the relaxation time, "E2, for micellization and demicellization are indeed the same.



100
90 0 Stopped-Flow
80 Ml Pressure-Jump 9
E
70

C
.o _ _. 6 0
'"50

-z 40
30
0
20
10 3.5 4.3
0
Triton X-1 00 Brij 35


Figure 3-6. Validation of relaxation constants, T2, by pressure-jump and stopped-flow
dilution techniques, both with optical detection (absorbance at 542 nm,
Eosin Y concentration: 0.019 mM).


As mentioned earlier, the relaxation constants as measured by pressure-jump and stopped-flow techniques will only be the same if the perturbation of the system is small enough. The following derivation proofs that for small perturbations in the concentration of micelles or monomers, the obtained relaxation constant is indeed the reaction constant k at equilibrium.

Consider micellization as the following simplified reaction, where A represents the monomers and B the micelles,









k2
A-- (3.1) ki

So, at equilibrium,

kICAeq =k2CB'eq (3.2) The net change in concentration of A, dCA /dt, can be given as,



dC = -klCA + k2CB (3.3) dt


where the first term represents the decrease in concentration of A due to the forward reaction (k,) and the second term represents the increase in concentration of A due to the backward reaction (k2). The pressure-jump and stopped-flow dilution methods impose a small perturbation in CA, ACA, so that, CA = CA'eq + ACA (3.4)


From experimental observation and as derived by Aniansson and Wall [1974], ACA decays according to first order reaction kinetics back to CAeq, ACA = ACaoe-aAt

(3.5)

where ACA,O is the perturbation at t = 0. Hence, the expression for CA becomes, CA = CAeq + ACA,Oe-CAI (3.6) Consequently,

dCA - AC oaae-OrAl (3.7) dt








The rate equation for reaction (3.1) can be equated by substituting (3.4) and (3.7) in (3.3),


dCA = _k1 (CAeq + ACA)+ k2CB = -ACAOUAe-, A


(3.8)


rewriting (3.8) yields,


- k [(cA, + ACA)--2 CB AC,Oa A


(3.9)


and substituting for k2/kl-CA,eq/CB,eq,


k l[(cA'a +AcCA')- ,eq C81= ACAO� Ae 'A
kI[(C~qC8 B,eq J






k l (C Aeq C Beq )+ (A C A CB, eq)- (C Aeq CB )] = A C AOA e -Ar
C B ,eq


Hence,


(3.10)






(3.11)


If the perturbation is small enough we can assume that CB;; CBeq and thus CA,eqCB,eq " CA,eqCB, resulting in,


(3.12)


kACA = AC A,OcAe-Al Equation (3.5) cancels out from (3.12) resulting in,


k- -OA


(3.13)









This proof shows that the relaxation constant GA, as determined by the pressure-jump and stopped-flow methods, is the reaction constant of micelle formation at equilibrium, ki, provided the change in monomer or micelle concentration is small enough.



3.6 Conclusions

1. Two spectroscopic techniques (stopped-flow and pressure-jump, both with optical

detection) were used to measure the slow micellar relaxation time of nonionic surfactants. Nonionic surfactants show a much longer relaxation process than ionic

surfactants because of the absence of ionic repulsion between the head groups.

2. Increasing the degree of ethoxylation leads to shorter relaxation times in case of

nonionic surfactants.

3. The slow micellar relaxation time, T2, for the technical grade nonionic surfactants is

significantly higher than the relaxation times for the pure, monodisperse surfactants.

Apparently, the distribution in degree of ethoxylation and the presence of impurities (such as long chain alcohols) contribute to the stability of nonionic surfactant

micelles.

4. Eosin Y and Merocyanine 540 dye resulted in the same CMC and slow micellar

relaxation constants, indicating that at surfactant/dye ratios larger than 20, no negative

influence of the dye on the micellar stability exists.

5. The measurement of T2 by stopped-flow or pressure-jump with optical detection

resulted in the same time constants (within the experimental error), indicating that the

processes of micellization and demicellization occur at the same rate.














CHAPTER 4
MICELLAR KINETICS AND DYNAMIC SURFACE TENSION


4.1 Introduction

Surface tension is the most direct measure of the surface-activity of a surfactant in solution. Equilibrium measurements of surface tension have been performed for many surfactant systems under various solution conditions. By comparison, relatively little data exists for the dynamic surface tension (DST) of these solutions. The understanding of dynamic surface tension is important to any technological application where a new interface is rapidly created in the presence of a surfactant solution [Miller et al., 1995]. Such new interfaces may be of various types, for example new air/water interfaces in foaming or film formation processes, new liquid/liquid interfaces in emulsification or new solid/liquid interfaces in detergency or fabric wetting applications. In most cases, the equilibrium surface tension is never reached, and the actual surface tension experienced by the interface is much higher. In these cases, the dynamic surface tension plays a more important role than the equilibrium surface tension [Borchardt and Yates, 1993; Rosen, 1989].

The effect of dynamic surface tension is particularly important in solutions containing large surfactants, which can be expected to have a slower rate of diffusion to the interface. Another important parameter in the measurement of dynamic surface tension is micellar stability. As discussed in chapter 1, very stable micelles cannot









breakup fast enough to provide additional monomers resulting in higher interfacial tensions. Hence, the minima and maxima in solid/liquid, liquid/liquid and gas/liquid phenomena were obtained. This chapter discusses the importance of micellar stability and diffusivity on dynamic surface tension for a variety of surfactants studied in the previous Chapter.



4.2 Dynamic Surface Tension by the Maximum Bubble Pressure Method

Since the equilibrium surface tension is not reached in many dynamic interfacial processes, it is the dynamic surface tension that must be studied and correlated with processes of interest. The dynamic surface tension can be measured by the drop weight [Jho and Burke 1983; Miller et al. 1993], oscillating jet [Thomas and Hall, 1975], capillary wave [Kelvin, 1871], growing drop technique [MacLeod and Radke, 1993], and the maximum bubble pressure method [Mysels, 1986; Ross et al., 1992]. The maximum bubble pressure (MBP) technique is the most commonly used technique for the measurement of dynamic surface tension, and has been applied to a variety of anionic and nonionic surfactant solutions [Garrett and Ward, 1989; Fainerman et al., 1993, 1994; Tamura et al. 1995; Hua and Rosen, 1991]. The technique was first developed over a century ago as reviewed by Mysels [1990], but only became practical in recent decades with the availability of fast pressure transducers and the electronics necessary to accurately monitor the rapidly changing pressure signal.

The dynamic surface tension of values for short interface lifetimes can vary greatly from the equilibrium values. For pure water, a newly formed interface should have a surface tension approaching 72 mN/m, the equilibrium value as measured by, for









example, the Wilhelmy plate static surface tension method. The physical principle behind the MBP measurement is the Laplace pressure; the pressure inside a curved liquid interface is higher than the ambient. This excess pressure (P) can be calculated from the Laplace equation [Adamson and Gast, 1997], P= - + pgh (4.1)
r

where y is the surface tension, r is the radius of curvature of the bubble, p is the liquid density, g is the gravitational constant and h is the depth of the bubble in the liquid. The first term expresses the Laplace pressure due to the curved gas/liquid interface, and the second term is the hydrostatic pressure due to the liquid height above the forming bubble. The first term will vary during the life cycle of the bubble, while the second term will remain constant. Figure 4-1 shows a typical life cycle of a bubble and the resulting pressure within the capillary. After a bubble breaks off from the capillary tip, the pressure is the lowest. As more gas flows into the capillary, the pressure builds up as the gas is pushed out of the capillary and the radius of curvature at the tip decreases. During this expansion process, surfactant is populating the new interface and acting to lower the surface tension. At the point of minimum interface radius of curvature, where a hemisphere of gas is formed at the capillary tip, the pressure is maximum. Both the minimum radius of curvature and the surface tension, as described by the Laplace equation, govern the maximum pressure experienced. On the incremental addition of gas, the bubble expands out of the capillary tip and the radius of curvature gradually increases. This results in a drop in the Laplace pressure, and a resulting rapid expansion of the









bubble. At some point, the bubble breaks off from the capillary tip and the whole process starts again.


A B
A' Pressure


L- Time

Capillary
immersed in
& liquid


uu0
Point of Pm


Figure 4-1. Characteristic bubble pressure vs. time curve in the maximum bubble
pressure method to determine dynamic surface tension. The maximum pressure is reached when the bubble is a perfect hemisphere at the tip of the capillary. Interface lifetime, A, is controlled by adjusting the bubble
rate; B represents the dead time.



4.3 Dynamic Surface Tension and Micellar Stability

Dynamic surface tension depends on several factors: monomer concentration (CMC), micellar stability, diffusion rate of the surfactant molecule to the interface, and the total surfactant concentration. During the formation of bubbles, surfactant monomers adsorb onto the freshly created interface from the bulk solution. The bubble dynamics are controlled mainly by diffusion below the CMC [Rillaerts and Joos, 1982]. Above CMC, however, the diffusion of monomers is augmented by the spontaneous breakdown of micelles. If the monomer is depleted by the adsorption process, micelles must breakup to provide additional monomers. If the micelles in solution are very stable, they cannot









provide monomer fast enough and the dynamic surface tension remains higher. However, if the micelles are relatively unstable, their disintegration resupplies the depleted monomer and lower dynamic surface tensions are obtained. This is illustrated in Figure 42.


Very unstable micelles


Low D.S.T. High D.S.T.


c39


- 0"

Very stable micelles


Figure 4-2.


Effect of micellar stability on dynamic surface tension.


In summary, for long bubble lifetimes, the equilibrium surface tension determines the interfacial tension at the air/water interface. However, when the bubble lifetime decreases, more and more monomer is depleted from the bulk solution and thus micelles must breakup in order to provide additional monomers. In that case, the breakup of micelles and thus the micellar stability determines the surface tension lowering. In order to show the importance of micellar breakup in the dynamic surface tension measurement, a dimensionless parameter 0 was introduced [Engels et al., 1998],


' )D - 'eq

Yw - eq


(4.2)


Q









where yD is the dynamic surface tension, yeq the equilibrium surface tension as measured by the Wilhelmy plate method, and yw the surface tension of pure water. This equation normalizes the surface tension with respect to the surface-activity of the solution. The denominator (yw - yeq) can be considered as the effectiveness of the surfactant [Rosen, 1989]. When yD = 7yeq, 0 = 0, which indicates that the surfactant concentration at the surface of the bubble is the same as that under equilibrium conditions. However, when yD = yw, 0 = 1, indicating that no surfactant is present at the interface of the bubble. Values between 0 and 1 are a measure for the surfactant concentration at the surface and hence, the stability of micelles, assuming the diffusion time of monomers to be negligible. This is known to hold for ionic surfactants [Oh et al., 1992], however, the validity of this assumption for nonionic surfactants will be discussed in section 4.5. In this study, the dynamic surface tension behavior of three technical grade nonionic (Brij 35, Synperonic A7 and A50) and two pure nonionic surfactants was studied (C12(EO)5 and CI2(EO)8).



4.4 Experimental Procedure

4.4.1 Materials

Brij 35, Synperonic A7 and A50 were supplied by ICI Americas, Inc. (Wilmington, DE). Monodisperse penta-ethyleneglycol mono n-dodecyl ether (C12(EO)5) and octa-ethyleneglycol mono n-dodecyl ether (CI2(EO)8) were purchased from Nikko Chemicals Co. (Tokyo, Japan). Deionized, distilled water was used in all experiments. Surfactant concentrations were calculated using their molecular weights disregarding the presence of any probable impurities. All experiments were performed at 22�C.









4.4.2 Dynamic Surface Tension

The maximum bubble pressure apparatus was constructed using a differential pressure transducer purchased from Omega Engineering, Inc. (Stanford, CT), with a sensitivity of 0 to 10 in. (25 cm) H20 (0 to 2500 Pa). A #23 steel needle was used as a capillary, with a nominal 0.025 in. (0.64 mm) external diameter, 0.013 in. (0.33 mm) internal diameter, and a flush cut tip. The capillary diameter was chosen so that the viscous resistance of water to bubble growth could be ignored. Such internal and viscous effects are a potential source of error in these measurements that need to be taken into consideration [Garrett and Ward, 1989]. All measurements were conducted with the capillary tip 1 cm beneath the liquid surface. Compressed air was used as the bubbling gas and an oscilloscope connected to the pressure transducer was used to determine the bubble frequency and the dynamic surface tension. A schematic diagram of the setup is shown in Figure 4-3.




Junction tube (enlarged scale)
Pressure transducer
Rotameter


(pressure) (voltage)
AOscilloscope ir flow Capillary tube (ID 0.013") . .1.0 cm depth

Surfactant solution Bubble formation



Figure 4-3. Setup for the measurement of dynamic surface tension by means of the
maximum bubble pressure method.








4.5 Results and Discussion

Some properties used in the calculation of 0 are shown in Table 4-1. The three commercial (technical grade) surfactants have similar structures and CMCs (Table 4-1).



Table 4-1. Equilibrium surface tension, CMC, and the slow micellar relaxation
time, "t2, as measured by the stopped-flow dilution technique.

Surfactant Structure Yeq CMCDye T2 (mN/m) (mM) (s)

Synperonic A7 C12-C15 Alkanol Ether (E07) 29.0 0.050 150 Brij 35 Lauryl Alcohol Ether (E023) 38.7 0.068 80 Synperonic A50 C12-C15 Alkanol Ether (E050) 49.5 0.084 40 C12(EO)5" a Lauryl Alcohol Ether (EO5) 30.0 0.060 10 C12(EO)8" Lauryl Alcohol Ether (E08) 34.7 0.072 4 a Pure (monodisperse) nonionic surfactant.




The only difference between the molecules is an increase in degree of ethoxylation. Figure 4-4 shows the dimensionless parameter 0 versus the bubble lifetime for 2 mM solutions of Synperonic A7, Brij 35 and Synperonic A50. Figure 4-5 shows similar curves for the pure nonionics C12(EO)5 and C12(EO)8. The slow micellar relaxation constant, t2, did not show an appreciable change as function of surfactant concentration and therefore the relaxation kinetics (determined at the concentrations given in Table 3-1) are similar at 2 mM concentration [Eastoe, 1997].









1
0.9 0.8 0.7 0.6 0 0.5
0.4 0.3 0.2 0.1
0


0 0.5 1 1.5
Bubble Lifetime (s)


Dimensionless dynamic surface tension vs. bubble lifetime for 2 mM solutions of Synperonic A7, Brij 35 and Synperonic A50.


1
0.9 0.8 0.7 0.6 00.5
0.4 0.3 0.2 0.1
0


0.5 1 1.5
Bubble Lifetime (s)


Dimensionless dynamic surface tension vs. bubble lifetime for 2 mM solutions of C12(EO)5 and C12(EO)8.


Figure 4-4.


Figure 4-5.








It is clear that Synperonic A7 shows the slowest rate of adsorption of surfactant molecules due to the stability of micelles, resulting in 0 values close to 1. On the other hand, Synperonic A50 shows a faster adsorption of surfactant molecules, indicated by the lower 0 values. The same trend is observed for the pure nonionic surfactants. More stable micelles result in lower 0 values. Thus, increasing the degree of ethoxylation destabilizes micelles, resulting in lower dynamic surface tensions.

A mathematical model which relates the slow micellar relaxation time (T2) to dynamic surface tension has been proposed by Fainerman and Makievski [1993], =rq+RTF2 -2
2(t),_ = 7", + Rct D (4.3) where yeq is the equilibrium surface tension, R is the gas constant, F is the surface excess concentration, c the total surfactant concentration, t is the bubble lifetime and D the diffusion coefficient. It is clear that for long bubble lifetimes (t--oo), the dynamic surface tension approaches the equilibrium surface tension (yeq). Furthermore, Fainerman [19921 found that the dynamic surface tension of micellar solutions at high surfactant concentration does not depend on the total surfactant concentration. Rather, it depends on the slow dissociation process of micelles. Equation 4.3 can be used to determine the slow micellar relaxation constant "2 from the dynamic surface tension measurements in the present study. Plotting the dynamic surface tension yuD vs. tt results in curve with slope (dytDdt) proportional to (t2) 1/2 for very large t. The resulting values for "r2, however, showed very poor agreement with the values obtained in this study. This can be attributed to the slope, which is almost zero at very long bubble lifetimes. Therefore, the error in the








calculation becomes very large. For example, Figure 4-6 shows the dynamic surface tension data for the two, monodisperse nonionic surfactants. The absolute value of U2 is hard to obtain from the curve at long bubble lifetimes, however, it is clear that (d'D/dt) for C12(EO)5 is larger than for C!2(EO)8, and hence T2 of C12(EO)5 > C12(EO)8.


65.0 60.0 1-.55.0
E
z 50.0
E
45.0
40.0

35.0 30.0


0.5 1 1.5 2
Bubble Lifetime (s)


Figure 4-6. Dynamic surface tension vs. bubble lifetime for 2 mM solutions of
C12(EO)5 and C]2(EO)8.



A number of alternative theoretical approaches have also been proposed to account for the effects of adsorption barriers and micellar breakdown kinetics on dynamic surface tension [Filippov, 1994a,b; Rillaerts and Joos, 1982]. Although these treatments have met with some success for certain surfactants, they do not appear to be generally applicable. In general, dynamic surface tension is a useful tool to qualitatively confirm the









slow micellar relaxation times, T2, of nonionic surfactants, obtained by the stopped-flow dilution or pressure-jump techniques with optical detection.



4.6 Importance of Diffusion Time of Nonionic Surfactants in DST Measurements

It is desirable to compare the diffusion time of surfactant monomers from the bulk to the air/water interface with the relaxation time of micelles during the bubble process. In this section, a simple model will be presented and applied to the nonionic surfactants studied in the dynamic surface tension measurements (Synperonic A7, Brij 35 and Synperonic A50 and C12(EO)5 and C12(EO)8). The model assumes that micelles do not adsorb at the air/water interface [Horozov et al., 1997]. Consequently, the number of monomers (Ni) needed to saturate the bubble surface of radius R, can be calculated according to,


NI - 47rR2 (4.4)
A


where A is the area per molecule of the surfactant at the air/water interface. The number of molecules per mL solution (N2) at the CMC (in mM) is given by, N 2 = CMC * N Avg (4.5)


where Nvog is the Avogadro number. The volume around the bubble, which contains N monomers, is given by NI/N2 mL. Therefore, a shell of radius R2 (containing NJ monomers) can be calculated by,

NI = 4 ).(R2 - R3) (4.6)








The unknown quantity R2 can be determined by using the values of NI, N2 and R, calculated previously. The average characteristic diffusion length Ld is (R2-RI)/2, which can be used to calculate the characteristic diffusion time td, according to Overbeek [1977],


td d (4.7) 2D


This is the diffusion time when the solution is assumed to be stationary. In fact, the diffusion time in real situations is much shorter than this value due to the convective motion of the solution induced by bubble motion. In order to calculate td for the three commercial and the two pure nonionic surfactants (Synperonic A7, A50, Brij 35 and C12(EO)5 and C12(EO)8 respectively), the diffusion coefficient D and the molecular area need to be known. Consequently, the characteristic diffusion time can be calculated as function of bubble radius R1. Table 4-2 shows the constants used in the calculation of td of the nonionic surfactants investigated in this study.



Table 4-2. Constants used in the calculation of the characteristic diffusion time td.

Surfactant Mw CMC Dx 106 Area/molecule T2 (g/mol) (mM) (cmE/s) (A) (s) Synperonic A7 522 0.050 3.67 50.3 150 Brij 35 1198 0.068 2.43 69.2 80 Synperonic A50 2414 0.084 1.71 210 40 C2(EO) 406 0.058 4.17 51' 10 CI2(EO)8a 538 0.070 3.62 66' 4 a Pure (monodisperse) nonionic surfactant. b In good agreement with Rosen [1982].








For example, for Synperonic A7 the following values are found assuming a bubble diameter of R, = 1 mm: N, = 2.5x103, N2 = 3.0x1022, R2 = 1.06x10-3 m, Ld = 3.11x10-5 m, and td = 1.31 s. Monomer diffusion coefficients, obtained for C8(EO)4 by Faucompre and Lindmann [1987], were used to approximate the other D values using the molecular weight (Mw), according to [Eastoe et al., 1996],


D , ( Mw 2 D2 MW I


(4.8)


The areas per molecule were calculated from the surface excess concentration F using the Gibbs adsorption isotherm, equations (1.2) and (1.3). The results of the diffusion model are shown in Figures 4-7 and 4-8.



1.6
1.4
1.2-- Syeonic A7
1.0
-0.8 Bj 35
0.6 0.4
0.2 Syrperonic A50
0.0 i ::I.t


Figure 4-7.


0 0.5 1 1.5 2 Bubble Radius (mrm)

Characteristic diffusion time td as function of bubble size in stationary solutions of Synperonic A7, Brij 35 and Synperonic A50.


The characteristic diffusion time increases as the bubble radius increases and then flattens off. A similar trend was observed for the pure nonionic surfactants. Beyond a









1.6 1.4 1.2 1.0
0.8
0.8 C12(E0)5

0.6
0.4-C
0.4 C12(E0)8
0.2
0.0 I
0 0.5 1 1.5 2 Bubble Radius (mim)

Figure 4-8. Characteristic diffusion time td as a function of bubble size in a stationary
solution of CI2(EO)5 and C12(EO)8.



critical bubble radius RI, td does not further increase anymore. This can be attributed to the fact that AR (R2-RI) remains constant at larger bubble size. This constancy is determined by the CMC of the surfactant, and hence, higher CMC's result in constant td values at smaller bubble radii. The mathematical proof is given in Appendix B. Figures 47 and 4-8 assume a stationary solution, however, due to the convective motion induced by the bubbling of the gas, the real diffusion time is much shorter.

It is clear that, for any bubble size, the characteristic diffusion times of the technical grade surfactants are much smaller than the slow micellar relaxation time t2. This indicates that the micellar breakup time is a rate-limiting step in the supply of monomers to newly created interface. For the pure nonionic surfactants, however, the micellar breakup time is much smaller and hence, diffusion becomes a more competitive process. Table 4-3 lists the average characteristic diffusion times for bubble radii R > 1









mm. In case of the pure nonionic surfactants, diffusion accounts for approximately 10% of the adsorption process.



Table 4-3. Comparison of the characteristic diffusion time, td, and slow micellar
relaxation time, T2, in the dynamic surface tension measurement for
R1> 1 mm.

Surfactant Dx 106 td T2 (cm 2/s) (s) (s)

Synperonic A7 3.67 1.3 150 Brij 35 2.43 0.55 80 Synperonic A50 1.71 0.05 40 C12(EO)5 4.17 0.8 10 C12EO) 3.62 0.4 4 a Pure (monodisperse) nonionic surfactant.




Another observation is that the size of the molecule does not seem to play an important role in the diffusion process. Earlier, it was expected that Synperonic A50, the largest molecule investigated in this study, would adsorb much slower to the interface due to its size. Still, a very small characteristic diffusion time was obtained. This can be attributed to; 1) higher CMC and 2) larger the area per molecule as compared to other nonionic surfactant molecules. Even though the diffusion coefficient of Synperonic A50 is small (Table 4-2), the CMC as well as the area per molecule are significantly larger than for the other surfactants. Thus, for Synperonic A50 relatively more monomers are available in the bulk solution (higher CMC) and, moreover, less molecules are required to cover the bubble surface (larger area per molecule).








4.7 Conclusions

1. The dynamic surface tension behavior of three technical grade nonionic (Brij 35,

Synperonic A7 and A50) and two pure nonionic surfactants (C12(EO)5 and C12(EO)8) was studied by means of the maximum bubble pressure method and related to the slow micellar relaxation time 'r2. A slow breakup of micelles (i.e. a long relaxation time t2), corresponds to high dynamic surface tensions, whereas

very labile micelles result in lower dynamic surface tensions.

2. A dimensionless dynamic surface tension parameter, 0, was introduced indicating

the importance of micellar stability in processes involving an increase in interfacial area. 0 values close to 0 indicate a very fast breakup of micelles, resulting in low surface tensions. 0 values close to 1 indicate a very slow breakup of micelles,

resulting in relatively high surface tensions.

3. Increasing degree of ethoxylation leads to shorter relaxation times for nonionic

surfactants and hence lower dynamic surface tensions.

4. The characteristic diffusion time, td, of surfactant monomers is small as compared to

the micellar relaxation time t2. However, for the pure nonionic surfactants, diffusion accounts for approximately 10% of the adsorption process, ignoring

convective motion due to bubble formation.

5. The size of the surfactant molecule does not seem to influence the characteristic

diffusion time of monomers. Synperonic A50, the largest molecule investigated in this study, showed much faster adsorption than the other nonionic surfactants. This

can be attributed to its high CMC and large head group area.














CHAPTER 5
EFFECT OF TAILORING MICELLAR STABILITY ON FOAMING PROPERTIES AND FOAMING METHODOLOGY


5.1 Introduction

Foams are dispersions of gas in a liquid in which the volume of the dispersed phase is so high that the system can be regarded as a network of interconnected films [Vold, 1983]. Foam is produced when air or some other gas is introduced beneath the surface of a liquid that expands to enclose gas with a film of liquid. Foam has a stable honeycomb structure of gas cells whose walls consist of thin liquid films with approximately plane parallel sides. Foams cannot be formed from pure liquids: a surfaceactive specie needs to be present in the system. This can be confirmed by a thermodynamic consideration. The Helmholtz free energy of a foam contains the surface area as an extensive variable, according to [Everett, 1988], dF = -pdV- SdT +)dA + pdni (5.1)



where y is the surface tension of the liquid, A is the total surface area, and p and n refer to the chemical potential and concentration of the components in the system, respectively. At constant temperature T and concentration n, equation 5.1 yields after integration,



AF =yAA-fpdV (5.2)









From equation (5.2) it is clear that a decrease in Helmholtz free energy results both from a loss of area and from the expansion of a gas. This can only occur in a coalescence process and hence, foam composed of gas in a pure liquid is thermodynamically unstable. Therefore, a third component is necessary to produce a stable foam.

Foamability and foam stability primarily depend on chemical composition and properties of the adsorbed surfactant molecules. These, in turn, influence numerous factors, such as diffusion rate, micelle breakup time, rheology of the adsorbed layer, gaseous diffusion out of and into bubbles, size distribution of the bubbles, surface tension, bulk and surface viscosity and microstructure of the foam. Also, the presence of electrolyte, temperature and pressure influences foam behavior [Bikerman, 1973; Rosen, 1989; Ross, 1958, 1980; Pugh, 1996].

As shown earlier by Oh and Shah [1991], the rate of adsorption of surfactant monomers onto the newly created surface during the foam generation process is primarily dependent upon the disintegration of micelles. Very stable micelles (for SDS at 200 mM concentration) cannot breakup fast enough to supply monomers necessary to stabilize the newly created area. Hence, higher dynamic surface tensions and thus less foam is generated. On the other hand, very labile micelles breakup quickly enough to provide additional monomers and thus more foam is generated. Therefore, by tailoring the micellar stability, one can control dynamic interfacial processes, such as foaming, wetting emulsification and solubilization. A schematic representation of the adsorption of surfactant monomers to the expanding interface due to the disintegration of micelles during foam generation is shown in Figure 5-1.











Air




Surfactant 7777777777777777777777
Solution Air

Thin Liquid Film


Air

Figure 5-1. Schematic diagram showing the adsorption of surfactant molecules onto
new surface area due to the disintegration of micelles. Very stable micelles
result in a lower foamability.



Another way of altering the micellar stability (other than changing the surfactant concentration) is the addition of long chain alcohols or oppositely charged surfactants. An example of tailoring SDS micellar stability by the addition of 1-dodecanol (C120H) or dodecyltrimethylammonium bromide (C12TAB), a cationic surfactant, is given in Figure 5-2. It is clear that the micellar stability (T2) can be increased significantly by the addition of 1-dodecanol (for a 25 mM SDS solution from 1 ms to 230 ms). The introduction of ion-dipole interactions causes the micelle to pack closer and hence, longer relaxation times are observed. Even more significant is the addition of CI2TAB. In that case, ion-ion electrostatic interactions are responsible for the three orders of magnitude increase in t2 (from 1 ms to 5000 ms). Lessner et al. [ 1981 a,b] studied the influence of salt (NaCIO4) on the micellar stability of SDS. An increase in micellar stability was observed and a shift of the maximum relaxation time to lower surfactant concentrations.










rem 25 mM SDS, T2 1 IMs.




25 mM SDS + 1.25 mM C120H, 2 = 230 ms.



S100 mM SDS+ 10 mM C12TAB,2 = 5000 ms(!)


Figure 5-2. Tailoring SDS micellar stability by the addition of 1-dodecanol
(C120H) or dodecyltrimethylammonium bromide (C 2TAB).



The study of the effect of alcohols on SDS micellar stability is not new. Earlier investigations by Leung and Shah [1986] and Yiv et al. [1981] showed that short chain alcohols (C1 to C5) labilize SDS micelles, which was explained on the basis of the Aniansson and Wall theory [ 1976]. A similar trend was found for the addition of glycerol [Huibers, 1996]. Glycerol increases the bulk viscosity, but decreases the slow micellar relaxation time T2.

In summary, the addition of alcohols or alkyltrimethylammonium bromides allows for the tailoring of micelles with specific stability, which in turn determines the dynamic surface tension and hence dynamic interfacial processes, for example foaming. In this study the effect of micellar kinetics on foaming properties of mixtures of sodium dodecyl sulfate (SDS), alkyltrimethylammonium bromides (CnTAB for n = 8, 10, 12, 14 and 16) and long chain alcohols (CnOH, for n = 8, 10, 12, 14 and 16) was investigated. Varying the chain length of the cosurfactant gives rise to the concept of chain length compatibility. In addition, the effect of the foaming methodology on foamability was investigated. The









rate at which foam is produced can result in opposite foaming behavior, which will be explained based upon micellar relaxation time and dynamic surface tension.



5.2 Experimental Procedure

5.2.1 Materials

Sodium dodecyl sulfate (99% purity) was supplied by Sigma Chemical Co. (St. Louis, MO). The following chemicals were also used without further purification: alkyltrimethylammonium bromides, C8TAB (Lancaster Inc., Windham, NH), CIOTAB (Acros Organics, Pittsburgh, PA), CI2TAB, C14TAB (Sigma Chemical Co.) and C16TAB (Aldrich Chemical Company, Milwaukee, WI). C8OH was supplied by Fisher Scientific (Fair Lawn, NJ). CIO0H, and C120H were supplied by Aldrich Chemical Company, Inc. (Milwaukee, WI). C140H was supplied by Eastman Kodak Company (Rochester, NY) and C160H was supplied by Sigma Chemical Co. All solutions were prepared using water that was both deionized and distilled. All experiments were carried out at 22�C.



5.2.2 Relaxation Time Measurement by Pressure-Jump with Electrical Conductivity
Detection

The slow micellar relaxation time, 12, was measured using a pressure-jump apparatus from Dia-Log Corporation (Dilsseldorf, Germany) by means of change in conductivity that results from micelle formation or disintegration. The surfactant solution was pressurized to 100-120 atmospheres and the solution was allowed to reach its new equilibrium state (at high CMC). Subsequently, the pressure was suddenly released to atmospheric (initial CMC) by rupture of a thin metal diaphragm. To reduce the monomer









concentration after the pressure drop, some monomers will enter micelles already present, which can be seen as the fast relaxation process, commonly referred to as rI. A much slower relaxation process is the formation of new micelles (referred to as r2) and can be observed as an exponential drop in electrical conductivity. The relaxation time T2 can then be calculated from the exponential decay in electrical conductivity [Huibers et al., 1996].



5.2.3 Surface Tension

Equilibrium surface tensions were measured from freshly prepared solutions by the Wilhelmy plate method. The platinum plate was always cleaned and heated to a red/orange color with a Bunsen burner before use.



5.2.4 Surface Viscosity

A deep-channel surface viscometer [Wasan et al., 1971; Chattopadhyay et al., 1992] was used to measure the surface viscosity of each solution. In the deep-channel surface viscometer the channel walls are stationary concentric cylinders, while the floor moves with a constant angular velocity. To measure the centerline velocity of the air/water interface, a small Teflon particle was placed at the interface, and the time for that particle to make one complete revolution was recorded from visual observation. After measuring the centerline velocity of the air/water interface, the surface viscosity can be calculated using,


[ Vb 1 (5.3) )r L;rve










where E is the surface viscosity, il the bulk viscosity of the solution, yo the channel width, Vb the plate rotational speed, V the centerline velocity of the air/water interface and D the ratio of depth to width of the liquid channel.



5.2.5 Foamability by Shaking

A 100 mL graduated cylinder was used for the foamability measurements by vigorously shaking. The measurements were carried out by shaking 15 mL of the sample solution for 10 s. The foamability was recorded as the volume of foam produced immediately after shaking. Each solution was tested at least seven times.



5.2.6 Foamability by Air Bubbling

A quartz column, 3.5 cm in diameter, was used to acquire the foamability by air bubbling measurements. At the base of the cylinder a single capillary, 2.5 mm in diameter, was used to generate the bubbles. Fifty mL of the sample solution was poured into the column using a long funnel that reached to the bottom to avoid any initial foam formation. The air was then turned on at a constant flow rate of 158.2 cc/min. The foam volume produced after 2 minutes was recorded. The measurement was repeated at least five times for each sample.



5.2.7 Foam Stability

A quartz cylinder (3.5 cm diameter) was used for the foam stability measurements. The cylinder contained a single capillary (2.5 mm diameter) to generate








the bubbles. 50 mL solution was poured into the cylinder using a long tube reaching the bottom, thereby avoiding contact with the walls, since any solution on the walls could act as an additional supply of surfactant molecules increasing the foam stability. After the foam height reached 50 cm, the airflow was stopped and the time recorded to collapse to half of its initial height.



5.2.8 Dynamic Surface Tension

Dynamic surface tensions were measured for the SDS and SDS/C120H mixtures using the setup described earlier in Chapter 4. All measurements were taken with the capillary tip 1 cm beneath the liquid surface.



5.3 Foaming Properties of SDS/C_.TAB Mixtures

Foaming properties, such as surface tension, surface viscosity, micellar relaxation time (t2), foam stability and foamability (by shaking) were measured for mixtures of SDS and CnTAB (for n = 8, 10, 12, 14 and 16). A low molecular ratio SDS/CnTAB (2 1/1) was used to avoid precipitation. The SDS concentration was chosen to be 100 mM, which is considerably higher than the CMC (CMC = 8.3 mM), because at SDS concentrations close to CMC the relaxation time is too small (on the order of milliseconds) to observe a significant change after the addition of CnTAB. Moreover, a high SDS concentration minimizes the effect of possible trace impurities. Figure 5-3 shows a master diagram of all the foaming properties of the SDS/CTAB solutions. The 'C2 of pure SDS at 100 mM was determined to be 0.15 s, in perfect agreement with the value measured by Lessner et al. [1981b] and Oh and Shah [1991]. The t2 increases to 1.35 seconds when 5 mM









C12TAB is added to 100 mM SDS solutions, which indicates the formation of more stable micelles as compared to pure SDS micelles. The maximum relaxation time was observed when the chain length of both the surfactant and cosurfactant were the same. The equilibrium surface tension (y) was lowest when the chain length of both surfactants was the same (SDS/C12TAB). Mixtures of anionic and cationic surfactants show a lower surface tension than pure anionic or cationic surfactant solutions alone, because of the electrostatic interaction between the ionic head groups. The lowest surface tension as observed for SDS/C12TAB, however, can be explained on the basis of chain length compatibility of both the surfactant molecules. Another parameter indicating the close packing of molecules at the air/water interface is surface viscosity. Measurements have shown that the maximum surface viscosity is obtained when the foaming agents possessed similar chain lengths. So, apparently the optimum packing of surfactant molecules does not only appear at the air/water interface but in micelles as well. A schematic diagram showing the proposed explanation for the molecular packing at the air/water interface is presented in Figure 5-4. In mixtures of anionic and cationic surfactants there are two factors contributing to the micellar stability. First is the electrostatic interaction between the ionic head groups. This will results in a closer packing as compared to pure SDS micelles. Second is the chain length of the surfactants. Since the electrostatic interaction between the head groups plays a role in all the solutions, only the difference in chain length can be observed.
















00.8

-


z32
E
30


CL

0.05


0 16 15 E 14

13


12 1

160

. 120

(/).,80

0 40

0

SDS SDS
+ +
C8TAB C1oTAB


SDS
+
C12TAB


SDS SDS + +
C14TAB C16TAB


Figure 5-3. Effect of 5 mM CnTAB on foaming properties of 100 mM SDS solutions.
The dashed lines represent 100 mM pure SDS solution.












&4 &4

Large Intermolecular Smallest Intermolecular Large Intermolecular
Distance Distance Distance
C12S04- + C8TAB C12804- + CI2TAB+ C12S04 + CI6TAB+


Figure 5-4. Schematic diagram representing the effect of chain length compatibility on
molecular packing at the air/water interface. When the chain length of adjacent hydrocarbon chains is matched, the hydrocarbon layer has more
order, tighter packing and greater stability [Shiao, 1976].



The effect of chain length compatibility is an important factor in systems involving interfacial films. As surface-active molecules as well as other hydrocarbon molecules are aligned at interfaces, the properties of the interface are impacted to a large extent upon the matching or mismatching of the alkyl chain lengths. In general, the chain length of surfactants used in a given system must be the same to maximize lateral molecular interactions that stabilize the surfactant film at the interface. If chain length mismatching is present in a surfactant film, the excess segment of the hydrocarbon tail has more freedom to disrupt the molecular packing through increased tail motion. This thermal disturbance presumably propagates along the chain at a considerable length towards the polar group of the molecule causing an increase in the area/molecule [Shah and Schulman, 1967]. The effect of chain length compatibility is also important to other interfacial properties and technologies, such as lubrication, contact angle, bubble size, environmental remediation, enhanced oil recovery, water solubilization in microemulsions, and microemulsion stability [Shiao et al., 1976; 1998].









In addition to the surface measurements described above, foam stability as well as foamability measurements were performed. Foamability was lowest for solutions containing SDS and CI2TAB. This can be explained based upon rate of micellar breakup. Micelles must be broken up into monomers for adsorption onto newly created surface of bubbles. Without this process, foam can not be generated. If the micelles in solution are very stable, they can not rapidly provide surfactant monomers to the newly created surface. Hence, foamability is poor. However, if the micelles are relatively unstable, their disintegration provides the surfactant monomers, which can rapidly adsorb to the newly created surface. This enhances foamability of the micellar solutions.

A relation between micellar stability and foam stability is less pronounced. Patel et al. [1996] found that stable SDS micelles form very stable foam films. Also, it is known that foam stability increases with bulk viscosity or surface viscosity [Davies and Rideal, 1963; Brown et al., 1953; Shah et al., 1978]. Normally, counterions decrease the repulsion between adjacent surfactant head groups, causing a more condensed film of higher surface viscosity [Chattopadhyay, 1992], thereby increasing foam stability. Another important factor influencing foam stability is the micellar structure inside the thin liquid film, which has been investigated by Nikolov and Wasan [1989a,b]. The stratification of thin liquid films can be explained as a layer by layer thinning of ordered structures of micelles inside the film. This structured phenomenon is affected by micellar effective volume fraction, their stability, interaction and polydispersity.








5.4 Foaming Properties of SDS/C OH Mixtures

In this section, the influence of long chain alcohols (CnOH for n = 8, 10, 12, 14 and 16) on the SDS micellar stability will be discussed. Furthermore, the effect of ndodecanol (C120H) on the micellar stability was related to foaming properties, such as foamability (by two different methods), dynamic and equilibrium surface tension and surface viscosity.

The slow micellar relaxation time t2 of the SDS/long chain alcohol mixtures, which is directly related to the micellar stability, was determined using the pressure-jump technique with electrical conductivity detection. Figure 5-5 shows the slow relaxation time T2 as function of SDS concentration in the presence of 5 mol% COH for n = 8, 10, 12, 14 and 16 (higher concentrations of COH lead to solubility problems at 22�C).


1





C,)
%0.


0.0


0.00




Figure 5-5.


0 .1


)1 )1


0 25 50 75 100 125 150 175 200

SDS Cornentration (mM) The effect of 5 mol% CnOH on the slow micellar relaxation time, T2, in SDS solutions.


I









It is clear that at lower SDS concentrations (< -150 mM) the micellar stability can be maximized by the addition of C120H, where the alcohol chain length is equal to the surfactant chain length. The presence of shorter or longer chain alcohols also results in an increase in micellar stability. Apparently, the introduction of ion-dipole interactions between the hydroxyl group of the alcohol and the sulfate group of the SDS cause a tighter packing of the micelle, resulting in greater micellar stability. Beyond approximately 150 mM, depending on the alcohol chain length, long chain alcohols other than C120H start destabilizing micelles, due to mismatching of the chains resulting in a disruption of the molecular packing causing the micelles to destabilize and hence, lower micellar relaxation times are obtained. The effect of 5 mol% COH on the micellar stability of 25 and 200 mM SDS solutions is presented in Figure 5-6. At a concentration of 25 mM of SDS or at approximately three times the CMC (8.2 mM), the stabilizing effect of C120H is much more significant than it is at 200 mM as shown in Figure 5-5 and 5-6. This indicates that the micellar stability of relatively low concentration SDS solutions can be greatly enhanced by the addition of C120H (at 25 mM almost 230 times). The stabilizing effect of alcohol may be due to the shielding of negative charges of SDS with hydroxyl groups of the alcohol molecules and a stabilizing effect of the hydrocarbon tails, resulting in tightly packed micelles. However, at 200 mM the enhanced micellar stability resulting from the CI20H addition is very small, showing that the contribution of C120H to already stable micelles is not significant. The dependence of T2 on the concentration of pure SDS solutions was discussed earlier in Chapter 1 [also Oh and Shah, 1993a].













0.1


T2 (s)


0.01



0.001

10 T2 (s)


Figure 5-6.


I 25 mM SDS + 5 M01% CnOH








25 mM SDS




200 mM SDS + 200 m SDS5 M01% CnOH
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -


C8OH Co0H C120H C140H 0160H Long Chain Alcohol

The effect of 5 mol% COH on the slow micellar relaxation time, T2, in 25 and 200 mM SDS solutions.


The effect of C120H on the SDS micellar stability was related to the following interfacial properties: surface viscosity, equilibrium surface tension and foamability (by shaking and air bubbling through a single capillary). Figure 5-7B shows the results of the

























































1I I I


Single Capillary Method




--SDS E
"--SDS + 5 mol % C12OH

0 50 100 150 2(
SDS Concentration (rM)


The effect of 5 mol% C120H on foaming properties of SDS solutions. Different methods of foaming can result in opposite foaming behavior.


10

1


0.1 0.01 0.001 0.25

0.2 0.15

0.1 0.05

0 35


B


E E;:: CO
0 z.



U.


I I I




C




Shaking Method




D


Figure 5-7.









surface viscosity measurements in the presence and absence of C 20H. From this graph it becomes clear that molecular packing is an important phenomenon in micelles as well as at the air/water interface (Figure 5-7A). At low SDS concentration (25 mM) the effect of C120H is much more pronounced than it is at 200 mM. The alcohol causes the molecules to pack tighter resulting in a high surface viscosity. However, at 200 mM the SDS molecules are already tightly packed, which does not allow the alcohol to increase the surface viscosity significantly. The equilibrium surface tension of 25-200 mM SDS solutions in the presence and absence of 5 mol% C120H is shown in Figure 5-7C. It is clear that in the presence of C120H the surface tension is approximately lowered by 7 mN/m due to closer packing of molecules at the air/water interface. The alcohol decreases the molecular area due to the ion-dipole interactions between the SDS head group and the hydroxyl group of the long chain alcohol and a maximum hydrophobic interaction between the carbon chains.

Two different methods were applied for the foamability experiments: vigorous hand shaking and air bubbling through a single capillary. Both methods show different results. In the first case, larger foam volumes were obtained for pure SDS solutions than SDS/C]20H mixtures as shown in Figure 5-7D. Especially at low SDS concentrations (25mM) the SDS/C120H mixtures produced significantly less foam than the pure SDS samples. This can be explained based on the ability of micelles to breakup in order to provide monomers to stabilize newly created interface. Very stable micelles cannot breakup fast enough to augment the flux of monomers necessary to stabilize the new air/water interface, resulting in higher interfacial tensions and hence, foaming ability is low (Figure 5-7D). So, apparently the breakup of micelles is a rate-limiting step in the









supply of monomers to rapidly created air/water interface. At 200 mM the SDS and SDS/C12OH micelles are equally stable (see Figures 5-5 and 5-7A) and hence, equal foam volumes are produced (Figure 5-7D).

Foamability measurements performed by blowing air through a single capillary yielded different results. Figure 5-7E shows that the foam volume of the SDS/C12OH mixtures is consistently higher than pure SDS solutions, irrespective of the SDS concentration. Apparently, by using a single capillary enough time is given to the surfactant molecules to diffuse and stabilize the newly created air/water interface. Therefore a lower dynamic surface tension is obtained. The relation between the dynamic surface tension and the amount of foam created can be given by [Adamson and Gast, 1997],

W = 2 AA (5.4) where W is the work done, y the interfacial tension at the air/water interface and AA the change in interfacial area. The same relation holds for emulsification processes [Walstra, 1983]. Obviously, when the same amount of work is applied, lower surface tension results in more interfacial area (either by decreasing the bubble size or by increasing foam volume), provided the presence of a surface-active specie. Since the surface tension of the SDS/C120H mixtures is significantly lower than for pure SDS (Figure 5-7C), the former one will produce more foam using the single capillary foam column.

The results of the foamability measurements were evaluated by a more quantitative experiment, namely dynamic surface tension by the maximum bubble pressure method (Chapter 4). The understanding of dynamic surface tension is important in any technological application where a new gas/liquid interface is rapidly being created









in a surfactant solution. In most cases the equilibrium surface tension is never reached and the actual surface tension experienced at the air/water interface is much higher. In this study two solutions of 15 mM SDS and one containing SDS + 5 mol% C120H were investigated. A concentration of 15 mM was chosen, since at too high surfactant concentrations, deviations from equilibrium surface tension are negligible at the bubble frequencies accessible with the current setup. The dynamic surface tension (DST) as a function of bubble lifetime (reciprocal value of bubble frequency) is given in Figure 5-8. The effect of 5 mol% C120H is clearly visible. At higher bubble lifetimes, the SDS/C120H mixture shows a significant lower surface tension than pure SDS, indicating that the equilibrium surface tension of the SDS/C120H solution is much lower, just as observed earlier in Figure 5-7C.



45
-0-SDS

42 ---SDS + C120H


, 39
z

u 36


33


30
0 0.5 1 1.5 2 Bubble Lifetime (s)


Figure 5-8. Dynamic surface tension (yD) of SDS and SDS/C120H solutions (15 mM
SDS, 5 mol% C120H).








However, when the bubble lifetime decreases and thus the bubble frequency increases, the curves are getting closer to each other up to a bubble lifetime of approximately 0.15 s where they actually cross. This means that at bubble lifetimes smaller than 0.15 s (frequencies higher than -7 s), the SDS/C120H mixed micelles are not able to breakup quickly enough to augment the flux of monomers necessary to stabilize the newly formed bubbles as compared to pure SDS. In this region the micellar stability and thus the ability of micelles to breakup fast enough, determines surface tension reduction.

In order to show the importance of micellar breakup in the dynamic surface tension measurement, a dimensionless parameter 0 was introduced, as discussed in Chapter 4, equation (4.2). Values between 0 and 1 are a measure for the surfactant concentration at the surface and hence, the stability of micelles, assuming the diffusion time of monomers to be negligible [Oh et al., 1992]. The more stable the micelles, the less monomer flux and hence 0 values closer to 1 will be obtained. Figure 5-9 shows the dimensionless parameter 0 versus the bubble lifetime for SDS and SDS/C120H solutions of 15 mM SDS and 5 mol% C120H. In this graph the 0 values are consistently higher for SDS/C120H than for pure SDS over all bubble lifetimes. Apparently, when accounted for the surface-activity of SDS and SDS/C120H, the breakup of SDS/C120H mixed micelles is a rate-limiting step in high- speed dynamic processes. Therefore, it is very important to consider the time scale of generating newly created interfaces in industrial processes, since that determines whether the breakup of micelles and the dynamic surface tension are the dominant factors in foaming, emulsification, wetting and solubilization processes.










1.0
0.9 --w-SDS + C12OH

0.8 --SDS
0.7
0.6
0 0.5
0.4
0.3
0.2
0.1 * S Sep
0.0
0 0.5 1 1.5 2 Bubble Lifetime (s)


Figure 5-9. Dimensionless dynamic surface tension (0) of SDS and SDS/C120H
mixtures (15 mM SDS, 5 mol% C120H).



5.5 Conclusions

1. The SDS micellar stability can be greatly enhanced by the introduction of ion-ion

(SDS/CnTAB) or ion-dipole interactions (SDS/CnOH).

2. For mixed solutions of anionic and cationic surfactants or anionic surfactants and long

chain alcohols, the foaming properties depend on the chain length of the individual molecules. In general, the chain length of the surfactant and cosurfactant must be the same to maximize lateral molecular interactions, resulting in minimum surface tension, maximum surface viscosity, maximum micellar stability, minimum

foamability (by the shaking method) and maximum foam stability.

3. Long chain alcohols (CnOH for n = 8, 10, 12, 14 and 16) stabilize SDS micelles, up to

approximately 150 mM SDS (depending on the carbon chain length of the alcohol)









due to the strong ion-dipole interaction between the negatively charged SDS head group and the hydroxyl group of the alcohol. Beyond this critical concentration the chain length compatibility starts playing a role. Therefore, only C120H will cause a further increase in micellar stability, whereas the mismatch in chain length between the other alcohols and the SDS results in a disruption of the molecular packing in the

micelle, thereby decreasing the stability.

4. The effect of adding C120H is most pronounced when the stability of pure SDS

micelles is very low, i.e., at low SDS concentrations (25 mM). At higher SDS concentrations, the micellar stability of SDS alone increases, which makes the effect

of C120H less pronounced.

5. The effect of micellar stability plays an important role in processes involving a rapid

increase in surface area. If enough time is allowed for the interface to form, the dynamic surface tension approaches the equilibrium surface tension and thus more foam is generated (more in case of SDS/C120H mixtures). However, in very highspeed processes, the micellar stability and thus the time it takes for micelles to breakup determines the rate of adsorption of surfactant molecules and therefore higher surface tensions will be attained for SDS/C120H solutions. In that case less foam is generated, even though the equilibrium surface tension of the SDS/C120H system is lower. In conclusion, different methods of foaming can produce opposite results as

illustrated by the foamability measurements in this study.














CHAPTER 6
IMPORTANCE OF MICELLAR STABILITY ON ANTIFOAMING ACTION


6.1 Introduction

Foams are used for many different purposes, such as mineral flotation, food processing, purification (foam fractionation), processing of textiles, personal care, enhanced oil recovery and fire fighting [Bikerman, 1973; Prud'homme and Khan, 1996]. In some processes, however, the formation of foam is not desired and its presence can cause serious problems [Kroshwitz, 1993; Garrett, 1993]. Foaming properties, such as surface tension, surface viscosity, micellar stability and film elasticity can be greatly modified by the addition of organic materials. The antifoaming mechanisms by foam breakers can be summarized as follows [Rosen, 1989; Ross, 1958]:



1. Removing surface-active materials from the air/water interface.

Surfactant molecules are removed from the surface by adsorption onto or dissolution in the soil. Finely divided hydrophobic silica particles break the foam by adsorbing surfactant molecules from the bubble surface and carrying them into solution. The presence of certain types of soil in a surfactant solution shows decreased foaming due

to this mechanism.









2. Replacing surfactant molecules with other molecules at the surface.

The surfactant molecules at the bubble surface are replaced by adding non-cohesive molecules of limited solubility in the solution. The tertiary acetylenic glycols, ethyl

ethers and isoamyl alcohols break foam in this manner.

3. Converting the surfactant film into a solid brittle film with no elasticity.

Calcium salts of long chain fatty acids break foams of SDS or sodium

dodecylbenzene sulfate by this mechanism.

4. Reducing the surface viscosity of the film.

Tributyl phosphate has a large cross sectional area at the air/water interface. This reduces the cohesive forces between the surfactant molecules and consequently

reduces the surface viscosity, which leads to increased drainage of the liquid film.



The nature of the electrolyte is of course very important. Electrolytes containing multivalent ions can be anticipated to cause larger effects as compared to those with monovalent ions, especially for ionic surfactants [Ross and Bramfitt, 1957]. The stabilization of foam films containing high surfactant concentrations caused by stratification of long range ordered microstructures in thin films, was shown by Ivan and Dimitrov [1988]. Nikolov and Wasan [1992] have theoretically and experimentally shown that the stepwise thinning of a foam film formed from a SDS micellar solution is governed by a long-range electrostatic repulsion by ionic micelles and a restricted volume effect in the film. Bergeron and Radke [1992] determined the disjoining pressure isotherms for single isolated foam films stabilized by SDS above the CMC. Studies of the oscillatory form of the disjoining pressure permitted quantitative interpretation of this




Full Text

PAGE 1

TAILORING MICELLAR STABILITY TO CONTROL INTERFACIAL PROPERTIES AND BEHAVIOR OF DISPERSED SYSTEMS By ALEXANDER PATIST A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1999

PAGE 2

I dedicate this dissertation to my parents, to my sister, Carla, and to my late grandmother, Oma Patist, who would have been so proud. . .

PAGE 3

ACKNOWLEDGMENTS I would like to express my sincere thanks and appreciation to my advisor, Professor Dinesh Shah, chairman of my supervisory committee, for his guidance, motivation, inspiration, and many non-scientific discussions relevant to happiness and fulfillment in life. Thanks also to the other supervisory committee members. Professors Brij Moudgil, Ben Koopman, Raj Rajagopalan, and Spyros Svoronos for their valuable time and suggestions. Many thanks go to Dr. Patricia Aikens and Dr. Kevin Penfield from ICI Surfactants for their financial support of the research as well as several national and international surfactant symposia, which I was able to attend through their support. Dr. Sunil Bhagwat is gratefully acknowledged for his input in the development of a new method for the determination of the slow micellar relaxation time for nonionic surfactants and Professor Wilhelm Knoche from the University of Bielefeld, Germany, for use of his pressure-jump equipment with optical detection. I have greatly appreciated the opportunity to collaborate with the Engineering Research Center (ERC) for Particle Science and Technology at the University of Florida. Especially, Dr. Yakov Rabonovich, Joshua Adler, and Pankaj Singh for their successful help and implementation of the Atomic Force Microscopy and dispersion stability measurements. I would also like to acknowledge all the undergraduate students sponsored by the ERC for their experimental contributions to this thesis: Teri Axelberd, Brenda iii

PAGE 4

Daneka, Nithya Desikan, Andrew Howes, Jonathan Johnson, Jr., Janardhan Lavakumar, Patrick Nguyen, Rahul Pagidipati, Stephen Patrick, Neemesh Shah, Rishita Shah, Franz Wakefield, and Seamus Wedge. It was great to work with you all. I wish to thank my colleagues from the Department of Chemical Engineering and the Center for Surface Science and Engineering for their help and cooperation. Dr. Paul Huibers, Dr. Vishal Chhabra, Dr. Brajesh Jha, Dr. Patanjali, Pavan Shukla (Buddy) and last but not least, Steve and Vikki Truesdail for all their help, support, and the good times we shared together. Finally, I owe a deep sense of gratitude to Dr. Sami Karabomi and the late Dr. Nico van Os at the Shell Research and Technology Centre in Amsterdam, for bringing me in touch with Professor Shah, which eventually resulted in this dissertation. iv

PAGE 5

TABLE OF CONTENTS Eage ACKNOWLEDGMENTS iii UST OF TABLES ix UST OF FIGURES x ABSTRACT xiv CHAPTERS 1 INTRODUCTION 1 1.1 Surfactants 1 1.2 Micellization 3 1 .3 Structure of Micelles 6 1.4 Dynamic Properties of Surfactant Solutions 8 1 .5 Importance of Micellar Relaxation Time on Technological Processes 10 1 .6 Rationale of the Proposed Research 15 2 DETERMINATION OF CRITICAL MICELLE CONCENTRATION OF NONIONIC SURFACTANTS 19 2.1 Introduction 19 2.2 CMC Determination by the Dye Micellization Method 20 2.3 CMC Determination by the Surface Tension Method 23 2.4 Experimental Procedure 24 2.4.1 Materials 24 2.4.2 Surface Tension 24 2.4.3 Spectrophotometry 24 2.4.4 Foam Fractionation 25 2.5 Results and Discussion 25 2.6 Conclusions 31 V

PAGE 6

3 MICELLAR KINETICS OF NONIONIC SURFACTANTS 33 3.1 Introduction 33 3.2 Measurement of Nonionic Micellar Relaxation Time by Stopped-Flow 34 3.3 Experimental Procedure 37 3.3.1 Materials 37 3.3.2 Relaxation Time Measurement by the Stopped-Flow Method 37 3.3.3 Relaxation Time Measurement by Pressure-Jump with Optical Detection 38 3.4 Results and Discussion 38 3.5 Validation of the Slow Relaxation Time by Pressure-Jump with Optical Detection 40 3.6 Conclusions 44 4 MICELLAR KINETICS AND DYNAMIC SURFACE TENSION 45 4. 1 Introduction 45 4.2 Dynamic Surface Tension by the Maximum Bubble Pressure Method .... 46 4.3 Dynamic Surface Tension and Micellar Stability 48 4.4 Experimental Procedure 50 4.4.1 Materials 50 4.4.2 Dynamic Surface Tension 51 4.5 Results and Discussion 52 4.6 Importance of Diffusion Time of Nonionic Surfactants in DST Measurements 56 4.7 Conclusions 61 5 EFFECT OF TAILORING MICELLAR STABILITY ON FOAMING PROPERTIES AND FOAMING METHODOLOGY 62 5.1 Introduction 62 5.2 Experimental Procedure 66 5.2.1 Materials 66 5.2.2 Relaxation Time Measurement by Pressure-Jump with Electrical Conductivity Detection 66 5.2.3 Surface Tension 67 5.2.4 Surface Viscosity 67 5.2.5 Foamabihty by Shaking 68 5.2.6 Foamability by Air Bubbhng 68 5.2.7 Foam Stability 68 5.2.8 Dynamic Surface Tension 69 5.3 Foaming Properties of SDS/CnTAB Mixtures 69 5.4 Foaming Properties of SDS/CnOH Mixtures 74 5.5 Conclusions 82 vi

PAGE 7

6 IMPORTANCE OF MICELLAR STABILITY ON ANTIFOAMING ACTION 84 6. 1 Introduction 84 6.2 Experimental Procedure 87 6.2.1 Materials 87 6.2.2 Methods 87 6.3 Results and Discussion 88 6.4 Comparison between Electrolyte and Antifoaming Agent on Micellar Stability 95 6.5 Conclusions 97 7 TAILORING INTRAMICELLAR FORCES TO CONTROL DISPERSION STABILITY 99 7.1 Introduction 99 7.2 Experimental Procedure 102 7.2.1 Materials 102 7.2.2 Microscopy 103 7.2.3 Pressure-Conductivity Measurements 103 7.2.4 Turbidity Measurements 104 7.3 Results and Discussion 104 7.3.1 AFM Images 104 7.3.2 Force-Distance Curves 106 7.3.3 Relation between Micellar Stability and Dispersion Stability 1 12 7.3.4 Mechanical Properties of the Adsorbed Micellar Layer 1 16 7.4 Conclusions 119 8 SUMMARY AND RECOMMENDATIONS FOR FUTURE WORK 1 20 8. 1 Measurement of CMC of Nonionic Surfactants 120 8.2 Micellar Relaxation Time of Nonionic Surfactants 121 8.3 Tailoring Micellar Stability to Control Foaming and Antifoaming Action 123 8.4 Tailoring Intramicellar Forces to Control Dispersion Stability 125 8.5 Reconmiendations for Future Work 125 APPENDIX A HISTORICAL PERSPECTIVE ON MICELLAR KINETICS 127 APPENDIX B MATHEMATICAL PROOF OF CONSTANT CHARACTERISTIC DIFFUSION TIME FOR LARGE BUBBLE RADH 132 vii

PAGE 8

REFERENCES BIOGRAPHICAL SKETCH

PAGE 9

] 1 ] LIST OF TABLES Table Eage 21 Comparison of critical micelle concentrations as determined by surface tension and dye micellization methods 26 31 Micellar relaxation constants, X2, measured by the stopped-flow dilution technique 39 41 Equilibrium surface tension, CMC and the slow micellar relaxation time T2 as measured by the stopped-flow dilution technique 52 4-2 Constants used in the calculation of the characteristic diffusion time, td 57 4-3 Comparison of the characteristic diffusion time, td and the slow micellar relaxation time, T2 in the dynamic surface tension measurement for/?/> 1 mm 60 71 CMC values of alkyltrimethylammonium bromides 104 ix

PAGE 10

LIST OF HGURES Figure page 1-1 Schematic representation of the three states in which surfactant molecules reside in water 4 1-2 Mechanisms for two relaxation times, Xi and T2, involved in a surfactant solution above CMC 9 1-3 At equilibrium, the rate of micelle formation equals the rate of micelle disintegration 10 1-4 Various liquid/gas phenomena exhibiting minima and maxima at 200 mM SDS concentration 12 1-5 Various liquid/liquid and solid/liquid phenomena exhibiting minima and maxima at 200 mM SDS concentration 13 16 Correlation between molecular properties and macroscopic phenomena 15 21 Scan of a UV-VIS absorbance spectrum 20 2-2 CMC determination of Tween 20 (CMC = 0.042 mM) using the dye micellization method (absorbance at 542 nm) 22 2-3 Critical micelle concentration of Tween 20 determined by surface tension and dye micellization methods 27 2-4 Critical micelle concentration of Tween 20 determined after foam fractionation by surface tension and dye micellization methods 28 2-5 Critical micelle concentration of pure C 12(60)5 determined by surface tension and dye micellization methods 29 2-6 Schematic diagram showing how the surface tension method would suggest a lower CMC than the dye micellization method because of the saturated air/liquid interface 30 X

PAGE 11

1 j 31 Absorbance spectra of Eosin Y in water and 2 mM Triton X100 solution (Eosin Y concentration: 0.019 mM) 34 3-2 Stopped-flow apparatus used for determination of the slow micellar relaxation constant, T2, for nonionic surfactants 35 3-3 Schematic diagram showing the micellar relaxation process involved in a stopped-flow dilution experiment 36 3-4 Typical relaxation curve obtained in a stopped-flow dilution experiment 36 ^ 3-5 Schematic diagram showing the micellar relaxation process involved in a pressure-jump experiment 40 '-) 36 Validation of relaxation constants, X2, by pressure-jump and stopped-flow dilution techniques, both with optical detection 41 41 Characteristic bubble pressure vs. time curve in the maximum bubble pressure method to determine dynamic surface tension 48 4-2 Effect of micellar stability on dynamic surface tension 49 4-3 Setup for the measurement of dynamic surface tension by means of the maximum bubble pressure method 51 4-4 Dimensionless dynamic surface tension vs. bubble lifetime for 2 mM solutions of Synperonic A7, Brij 35 and Synperonic A50 53 4-5 Dimensionless dynamic surface tension vs. bubble lifetime for 2 mM solutions of Ci2(EO)5 and Ci2(EO)8 53 4-6 Dynamic surface tension vs. bubble lifetime for 2 mM solutions of C,2(EO)5andCi2(EO)8 55 ' 4-7 Characteristic diffusion time td as function of bubble size in stationary solutions of Synperonic A7, Brij 35 and Synperonic A50 58 48 Characteristic diffusion time td as a function of bubble size in a stationary '\ solution of Ci2(EO)5and Ci2(EO)8 59 51 Schematic diagram for the adsorption of surfactant molecules onto new surface area due to the disintegration of micelles 64 5-2 Tailoring SDS micellar stability by the addition of 1-dodecanol (C12OH) or dodecyltrimethylammonium bromide (C12TAB) 65 xi t

PAGE 12

5-3 Effect of 5 mM CnTAB on foaming properties of 100 mM SDS solutions 7 1 5-4 Schematic diagram representing the effect of chain length compatibility on molecular packing at the air/water interface 72 5-5 The effect of 5 mol% CnOH on the slow micellar relaxation time, Xj, in SDS solutions 74 5-6 The effect of 5 mol% CnOH on the slow micellar relaxation time, Xj, in 25 and 200 mM SDS solutions 76 5-7 The effect of 5 mol% C12OH on foaming properties of SDS solutions 77 5-8 Dynamic surface tension (Yd) of SDS and SDS/C12OH solutions (15 mM SDS, 5mol% C12OH) 80 59 Dimensionless dynamic surface tension (0) of SDS and SDS/C12OH mixtures (15 mM SDS, 5 mol% C,20H) 82 61 Structures of the antifoaming agents used in the present study 88 6-2 Effect of antifoaming agents on the slow micellar relaxation time, T2 and foamability of 150 mM SDS solutions 89 6-3 Effect of tetraalkylammonium chloride (TCnAC, for n = 1, 2, 3 and 4) on the slow micellar relaxation time, T2 in 150 mM SDS solutions 91 6-4 Effect of tetraethylammonium chloride on foaming properties of 150 mM SDS solutions 93 6-5 Schematic representation of the micro-structural changes in the SDS micellar packing upon addition of antifoaming agents 95 66 Effect of Na* counterions on the micellar stability of 150 mM SDS solutions 96 71 Schematic diagram showing the possible structures of surfactants adsorbed at the solid/liquid interface 102 7-2 Top view AFM image of mica immersed in water (pH 6) 105 7-3 Top view AFM image of mica immersed in a 2 CMC C14TAB solution (pH 6) 106 xii

PAGE 13

7-4 Force-distance curve for a 2 CMC CuTAB solution showing the different stages as the tip approaches the mica surface 107 7-5 Force-distance curves for 2 CMC solutions of CpTAB (for n = 10, 12, 14 and 16) on mica (pH 6) 108 7-6 Maximum compressive force as function of alkyl chain length for CMC solutions of CnTAB (for n = 10, 12, 14 and 16) on mica 108 7-7 Force-distance curves for a 2 CMC C12TAB solution on mica containing increasing amounts of SDS (pH 6) 109 7-8 Maximum compressive force as function of SDS concentration for a 2 CMC C12TAB solution on mica 1 10 7-9 Change in electrical conductivity due to micelle break-up in the pressure-jump experiment Ill 7-10 Turbidity of silica dispersions and the maximum compressive force vs. concentration of C12TAB at pH 4 after 60 min 1 13 7-11 Turbidity of silica dispersions and the maximum compressive force vs. concentration of C12TAB in the presence of 100 mM NaCl at pH 4 1 14 7-12 Turbidity and the maximum compressive force vs. concentration of SDS in the presence of 5 mM C12TAB, 100 mM NaCl at pH 4 1 15 7-13 Calculation of the actual tip radius from the micellar profile obtained by the AFM 117 7-14 Interaction parameters between the AFM tip and the adsorbed layer of micelles for the calculation of the elasticity 1 17 A-1 Typical size distribution curve of aggregates in a micellar solution 128 xiii

PAGE 14

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the degree of Doctor of Philosophy TAILORING MICELLAR STABILITY TO CONTROL INTERFACIAL PROPERTIES AND BEHAVIOR OF DISPERSED SYSTEMS By Alexander Patist August 1999 Chairman: Dinesh O. Shah Major Department: Chemical Engineering The association of many classes of surface-active molecules into micellar aggregates is a well-known phenomenon. Micelles are in dynamic equilibrium, constantly disintegrating and reforming. This relaxation process is characterized by the slow micellar relaxation time constant, X2, which is directly related to the micellar stability. The micellar stability of sodium dodecyl sulfate (SDS) micelles has been shown to significantly influence technological processes involving a rapid increase in interfacial area, such as foaming, antifoaming, wetting, emulsification, solubilization, and detergency. First, the available monomers adsorb onto the freshly created interface. Then, additional monomers must be provided by the breakup of micelles. Especially when the free monomer concentration is low, the micellar breakup time is a rate-limiting step in the xiv

PAGE 15

supply of monomers, which is the case for many nonionic surfactant solutions. However, at present, no paper on the importance of nonionic micellar lifetime on dynamic interfacial processes is available in the literature. The aim of this study is to develop a method to measure the slow micellar relaxation time of nonionic surfactants and to relate the relaxation data to dynamic interfacial processes. In the present study, stopped-flow and pressure-jump techniques were used to determine the slow micellar relaxation time (T2) of nonionic surfactants. Both techniques gave the same time constants, within the experimental error. The slow relaxation times are much longer for nonionic surfactants (from seconds to minutes) than for ionic surfactants (usually milliseconds) because of the absence of ionic repulsion between the head groups. The observed relaxation time was related to dynamic surface tension and foaming experiments. A slow breakup of micelles (i.e., a long relaxation time x^) corresponds to a high dynamic surface tension and low foamability, whereas a fast breakup of micelles leads to a lower dynamic surface tension and higher foamability. Relaxation time data of surfactant solutions correlate with the (dynamic) properties of a given surfactant solution. Moreover, the results suggest that appropriate micelles with specific stability or Xj can be designed by controlling the surfactant structure, concentration, and physicochemical conditions, as well as by mixing anionic/cationic or ionic/nonionic surfactants for a desired technological application. XV

PAGE 16

CHAPTER 1 INTRODUCTION 1.1 Surfactants Surfactants are more commonly known as soaps, which have traditionally consisted of sodium salts of naturally occurring fatty acids. The advent of petrochemicals, however, has brought numerous synthetic detergents, which are typically sulfates, sulfonates, or trimethylammonium salts of long chain hydrocarbons [Miller and Neogi, 1985]. Today, surfactants are among the most versatile products of the chemical industry, appearing in motor oils, detergents to clean laundry and homes, cosmetics, foods, oil recovery, pharmaceuticals, mineral flotation, and contact lenses [Rosen, 1989]. The last decade has seen the extension of surfactant applications to such high-technology areas as electronic printing, magnetic recording, enhanced filtration systems, biotechnology, and microelectronics [Holmberg, 1998]. Recently, a whole new area of research was opened by the introduction of biodegradable, sugar-based surfactants [Holmberg 1998; von Rybinski and Stoll, 1997]. A surfactant is a surface-active substance that has the property of adsorbing onto surfaces or interfaces (gas/liquid, liquid/liquid or solid/liquid) to lower the surface or interfacial free energy. Surfactants have a characteristic molecular structure consisting of a nonpolar group that has very little attraction for the water, known as the hydrophobic group, and a polar group that has strong attraction for the water, called the hydrophilic 1

PAGE 17

group. The adsorption of surfactant molecules at the air/water interface can be described by the Gibbs equation [Hiemenz and Rajagopalan, 1997]. As early as 1878, Gibbs [1948] derived a differential equation relating the surface tension, the number of moles and the chemical potentials of the components at the interface, dy = -J^r,dM, (1.1) where dy is the change in interfacial tension of the solvent, F, is the surface excess concentration, which can be approximated by the number of moles per unit area and dfii is the change in chemical potential of the components in the system. The Gibbs equation can be used to calculate the surfactant concentration at the interface and hence the area per molecule from the simple measurement of surface tension. For dilute solutions of a nonionic surfactant or a 1:1 ionic surfactant in the presence of electrolyte, equation (1.1) can be written as [Hiemenz and Rajagopalan, 1997], Y = ^ ( ] RT ydlnC jj (1.2) where R is the gas constant, T the absolute temperature and C the concentration of surfactant. The surface excess concentration, F, can be obtained from the slope of a plot of the surface tension y versus InC at constant temperature, which then can be used for the calculation of the area per molecule a (in squared Angstrom), 10'° «= (13) N F

PAGE 18

3 where Navog is the Avogadro number and T is the surface excess concentration. An extensive list of areas per molecule for a variety of surfactants and counterions is given by Rosen [1989] and Oh and Shah [1993b]. 1.2 Micellization Since the beginning of the study of surfactant solutions, it was recognized that the physical properties of surfactant solutions, such as surface tension, osmotic pressure, electrical conductivity and solubility (as function of temperature) show an abrupt change in the neighborhood of a critical concentration. The unusual properties of fatty acid salts in dilute aqueous solution were first investigated by McBain [1913, 1920] and later by Hartley [1936]. Other evidence for molecular aggregation was obtained from vapor pressure measurements and the solubility of organic material. The formation of colloidalsized clusters of individual surfactant molecules in solution is now better known as micellization. Although first suggested by McBain [1913], the earliest concrete model for spherical micelles is attributed to Hartley et al. [1936]. Figure 1-1 schematically shows the three environments in which surfactant molecules reside in a typical surfactant solution. Surfactant molecules disperse as monomers in the aqueous phase, form aggregates (micelles), or adsorb as a film at the air/water interface. The surfactant is in dynamic equilibrium between these states. Thus, at given temperature, pressure and concentration, the number of monomers, micelles and monomers adsorbed at the air/water interface is fixed under equilibrium conditions.

PAGE 19

4 Figure 1-1. Schematic representation of the three states in which surfactant molecules reside in water (monomers adsorbed at the air/water interface, monomers in the bulk solution and micelles). The process of surfactant clustering or micellization is primarily an entropy driven process. When surfactants are dissolved in water, the hydrophobic group disrupts the structure of water and therefore increases the free energy of the system. Surfactant molecules therefore concentrate at interfaces, so that their hydrophobic groups are directed away from the water and the free energy of the solution is minimized. The distortion of the water structure can also be decreased (and the free energy of the solution reduced) by the aggregation of surface-active molecules into clusters (micelles) with their hydrophobic groups directed towards the interior of the cluster and their hydrophilic groups directed toward the water. However, the surfactant molecules transferred from the solution to the micelle may experience some loss of freedom from being confined to the micelle. In addition, they may experience an electrostatic repulsion from other similarly charged surfactant molecules in the case of surfactants with ionic head groups. These forces increase the free energy of the system and oppose micellization. Hence, micelle

PAGE 20

5 formation depends on the force balance between the factors favoring micelhzation (van der Waals and hydrophobic forces) and those opposing it (kinetic energy of the molecules and electrostatic repulsion). The explanation for the entropy-dominated association of surfactant molecules is called the "hydrophobic effect" or "hydrophobic bonding" [Tanford, 1980]. The concentration at which micelles first appear in solution is called the critical micelle concentration (CMC). Representing the surfactant by S, the micelhzation process can be described by the reaction, in which 5„ is the micelle with a degree of aggregation n. The aggregation number n increases with increasing length of the hydrophobic group and decreases with increasing size of the hydrophilic group [Rosen, 1989]. hi general, the greater the dissimilarity between the surfactant molecules and the solvent, the greater the aggregation number. Experimentally, the CMC is determined from the discontinuity or inflection point in the plot of a physical property of the solution as a function of surfactant concentration. A wide variety of techniques involving the measurement of such physical properties as the surface tension, conductivity, light scattering intensity and osmotic pressure have been used to determine CMC values [Preston, 1948; Shinoda and Nakagawa, 1963; Mukerjee and Mysels, 1971]. In general, those factors that increase the aggregation number tend to decrease the CMC. For example, increasing the alkyl chain length of the surfactant, decreases the CMC. The presence of electrolyte also decreases the CMC, due to the so-called salting out effect. The work needed to create the volume in water required

PAGE 21

6 to accommodate a nonpolar solute is increased in electrolyte solution because of the strong water-ion interactions. When the surfactant monomers are salted out by the presence of electrolyte, micellization is favored and the CMC is decreased. Another factor favoring micellization in electrolyte solutions is the shielding of charges between the ionic head groups (in case of ionic surfactants) [Rosen, 1989] 1.3 Structure of Micelles In the last couple of decades, the recognition that surfactant structures can mimic biological structures has gained substantial interest [Bergethon and Simons, 1990]. Enzymes, for example, are protein molecules into which a reactant molecule fits to form a reactive intermediate. The highly efficient and specific catalytic effect of enzymes makes their investigation an interesting area of biomedical and detergent research [Gloxhuber and Kunstler, 1992; van Ee et al., 1997]. Likewise, cell membranes not only compartmentalize biological systems but also play a variety of functions in the life of the cell. Surfactant structures can be used as model systems to mimic both enzymes and membranes. Lipid aggregates known as liposomes are coimnon in physiological systems, for example the use of specially designed liposomes as drug-delivery vehicles [Lasic, 1993]. Self-assembled structures such as micelles or reversed micelles also play an increasingly important role in catalysis and separation processes in engineering and environmental science and technology [Fendler, 1975; Myers, 1991; Gratzel and Kalyanasundaram, 1991]. A theory of micellar structure, based upon the geometry of various micellar shapes and the space occupied by the hydrophilic and hydrophobic groups of the surfactant

PAGE 22

7 molecules, has been developed by Israelachvili et al. [1976] and Mitchell and Ninham [1981]. In aqueous media, for example, surfactants with bulky or loosely packed hydrophilic groups and long, thin hydrophobic groups tend to form spherical micelles, while those with short, bulky hydrophobic groups and small, close packed hydrophilic groups tend to form lamellar or cylindrical micelles. At concentrations slightly above the CMC, micelles are considered of spherical shape. Changes in temperature, surfactant concentration or additives in the solution may change the size, shape, aggregation number and stability of the micelles. The structure of a micelle could vary from spherical to rod or disc-like to lamellar in shape. In concentrated solutions (much higher than the CMC), lamellar micelles form, such that the water molecules occupy the region between parallel sheets of surfactants. Micelles may also form long cylinders packed together (known as lyotropic mesomorphs or liquid crystalline phases) at high surfactant concentrations [Miller and Neogi, 1985; Ekwall, 1967]. The adsorption of surfactants at solid-liquid interfaces is very important in controlling large scale industrial processes, such as liquid/liquid dispersion stability in paints, detergency, water purification, oil recovery and ore flotation [Rosen, 1989; Adamson and Cast, 1997]. Many studies have been undertaken over the last three decades to investigate the characteristics of aggregates formed by a surfactant on a solid surface, using calorimetric studies, neutron-scattering, ellipsometry and the surface force apparatus developed by Israelachvili [1991]. All those techniques have provided quantitative measures on adsorption but little information on aggregate structure. Much is known about structures of aggregates in bulk solution, leading to spherical or cylindrical micelles, bilayers and bicontinous phases in bulk solution. At interfaces however, the

PAGE 23

8 self-assembly process is influenced by additional surfactant-surface and solvent-surface interactions, including surface roughness, heterogeneity and charge behavior [Manne et al., 1994]. Recently, the structure and shape of those adsorbed aggregates have been revealed by atomic force microscopy (AFM) [Manne and Gaub, 1995]. The results suggest that the solid surface can alter the micellar structure of adsorbed aggregates. The studies presented in this dissertation (Chapter 7), indicate that dispersion stability can be modified by the strength of the adsorbed surfactant layer on the solid surface. 1 .4 Dynamic Properties of Surfactant Solutions The association of many classes of surface-active molecules into micellar aggregates is a well-known phenomenon. Micelles are often drawn as static structures of spherical aggregates of oriented surfactant molecules. However, micelles are in dynamic equilibrium with individual surfactant molecules that are constantly being exchanged between the bulk and the micelles. Additionally, the micelles themselves are continuously disintegrating and reassembling. There are two relaxation processes involved in micellar solutions. The first one is the fast relaxation process referred to as Ti (generally on the order of microseconds), which is associated with the fast exchange of monomers between micelles and the surrounding bulk phase. This process is considered as the collision between surfactant monomers and micelles. The second relaxation time X2 (usually of the order of milliseconds to minutes) is attributed to the micelle formation and dissolution process. Figure 1-2 shows the two characteristic relaxation times, Xi and X2, associated with micellar solutions.

PAGE 24

+ ^ + 1:2 Figure 1-2. Mechanisms for the two relaxation times, Xi and T2, involved in a surfactant solution above CMC. Micellar relaxation kinetics show dependence on temperature, pressure and concentration and have been studied by various techniques such as stopped-flow, temperature-jump, pressure-jump and ultrasonic absorption [James et al., 1977; Tondre et al., 1975; Hoffmann et al., 1976; Frindi et al., 1994; Kato et al., 1995; Lang, 1987; WynJones, 1975]. Micelle formation and disintegration can be compared to the equilibrium between water and water vapor at a given temperature and pressure. For a closed system containing water and water vapor in equilibrium, one can assume that the number of water molecules per unit area per second evaporating from the surface is the same as the number of water molecules condensing at the surface. Thus, the total number of molecules in the vapor phase or in the liquid does not change with time. So, the rate of condensation is equal to the rate of evaporation. The same principle holds for a micellar solution. Figure 1-3 shows schematically the formation and disintegration of micelles. It is evident that at equilibrium the number of micelles formed in a given time, is equal to

PAGE 25

10 t \ a. 1 ^ s _/ ^ o V J t = 0 t>0 Figure 1-3. At equilibrium, the rate of micelle formation equals the rate of micelle disintegration. the number of micelles disintegrated in the same time period as shown by the broken circles in Figure 1-3 for t > 0. In this study more than one experimental technique was used to confirm that indeed both processes (i.e. micelle formation and disintegration) occur at the same rate (Chapter 3). The two relaxation times can be used to calculate two important parameters of a micellar solution: (1) the residence time of a surfactant molecule in a micelle and (2) the average lifetime or stability of a micelle. The kinetics of this process have been evaluated by Aniansson et al. [1976] and Kahlweit [1981, 1982]. A brief historical perspective on micellar kinetics is given in Appendix A. 1.5 Importance of Micellar Relaxation Time on Technological Processes The importance of micelle breakup on processes involving an increase in interfacial area was first reported by Mijnlieff et al. [1965]. For several years, researchers have tried to correlate the slow micellar relaxation time, X2, with equilibrium properties,

PAGE 26

1 11 such as surface tension and surface viscosity, with no success. However, a strong correlation was found between the T2 of sodium dodecyl sulfate (SDS) micelles and various dynamic interfacial processes such as foamability, wetting time of textiles, bubble volume, emulsion droplet size and solubilization rate of benzene [Shah, 1998]. The micellar stability of SDS solutions was determined earlier by Lessner et al. [1981a,b] and later by Oh and Shah [1993a] using pressure-jump with electrical conductivity detection. This technique is described in detail by Huibers et al. [1996]. Maximum micellar stability was found at 200 mM (5 seconds). Figure 1-4 presents the various phenomena exhibiting minima and maxima at the liquid/gas interface. At 200 mM SDS, minimum foamability, maximum single film stability, maximum single bubble volume and a minimum frequency of bubble generation were found. These phenomena were explained based upon the monomer flux to newly created interfaces. If the micelles in solution are very stable, they cannot provide monomers fast enough to the interface and thus the interfacial tension remains higher. Therefore, lower foamability, larger single bubble foam volumes and a minimum frequency of bubble generation were found [Oh and Shah, 1991; Oh et al. 1992]. Very unstable micelles, however, provide monomers fast enough to the surface resulting in lower interfacial tensions. Maximum single film stability was found at 200 mM, i.e., when the micelles are most stable [Patel et al., 1996].

PAGE 27

12 Frequency of Bubble Generation Volume of Single Single Film Stability Foamability Slow Micellar Relaxation Time, T2 I 200 mM SDS CONCENTRATION Figure 1-4. Various liquid/gas phenomena exhibiting minima and maxima at 200 mM SDS concentration [Shah, 1998]. Various interfacial phenomena occurring at the liquid/liquid and solid/liquid interface in SDS solutions are shown in Figure 1-5. The wetting time and droplet size in emulsions exhibit maxima at 200 mM. When micelles are very stable, the flux of monomers decreases and hence the wetting process slows down.

PAGE 28

13 Time to Reach Saturation of SDS Solution by Benzene Detergency, Removal of Orange OT Solubilization Rate of Benzene Droplet Size in Emulsions Wetting Time I 200 mM SDS CONCENTRATION Figure 1-5. Various liquid/liquid and solid/liquid phenomena exhibiting minima and maxima at 200 mM SDS concentration [Shah, 1998]. Different types of fabrics, such as polyesters, Dacron, Nylon, cotton, and silk were investigated. The maximum wetting time of the investigated fabrics occurs at 200 mM SDS concentration. Although the absolute magnitude of the wetting time depends on the fabric, the maximum occurring at 200 mM is a property of the SDS solution and not of

PAGE 29

14 the fabric. The Hquid/liquid and solid/Hquid phenomena can also be explained based upon the monomer flux necessary to stabilize newly created interfaces. Very stable micelles result in high dynamic surface tensions and hence larger droplet sizes and longer wetting times are obtained [Oh et al., 1993; Oh and Shah, 1992]. The solubilization rate of benzene in SDS solutions, as well as, the detergency or removal of orange OT dye from fabric surface, show maxima at 200 mM concentration. The time required to reach saturation of the SDS solution upon the addition of benzene is minimum at 200 mM SDS concentration. This suggests that very stable micelles (i.e., tightly packed micelles) are more effective in the solubilization of oil [Oh and Shah, 1993a]. This can be explained based upon the interior of the micelles. The interior of rigid (i.e., tightly packed) micelles is more hydrophobic as compared to that of loosely packed micelles and hence the stronger hydrophobic core causes more rapid partitioning or solubilization of benzene and Orange OT into the micelles at 200 mM SDS concentration. The maximum micellar stability occuring at 200 mM of SDS can be explained by the Intermicellar Coulombic Repuslion Model (ICRM). This model is based upon electrostatic repulsion, reduction in intermicellar distance and, in case of SDS micelles, a structural change from spherical to cylindrical micelles [Oh et al., 1993a; Reiss-Hudson and Luzzati, 1964; Ekwall, 1967]. In summary, SDS solutions exhibit maxima and minima for various properties at 200 mM concentration due to maximum stability of SDS micelles at this concentration. The more stable micelles lead to less monomer flux and hence to a higher dynamic surface tension. First, the available monomers adsorb onto the freshly created interface. Then, additional monomers must be provided by the breakup of micelles. Especially when the free monomer concentration is low, as indicated by a low CMC, the micellar

PAGE 30

15 breakup time is the rate-limiting step in the supply of monomers, which is the case for many nonionic surfactant solutions. However, at present, no paper on the importance of nonionic micellar relaxation time on dynamic interfacial processes is available in the literature. 1.5 Rationale of the Proposed Research From the previous sections, it has become clear that micelles are important in a variety of technological processes, such as foaming, wetting, solubilization, emulsification, wetting and detergency [Rosen, 1989; Miller and Neogi, 1985; Davies and Rideal, 1963]. Figure 1-6 illustrates how molecular properties of surfactants are related to the performance of technological processes. Surfactant Molecules Adsorbed Films (Surface Tension, Surface Viscosity) Micelles (Micellar Relaxation Tinne, Dynamic Surface Tension) Applications (Foams, Emulsions, Solubilization, Lubrication, Wetting, etc.) Figure 1-6. Correlation between molecular properties and macroscopic phenomena.

PAGE 31

16 The structure of surfactant molecules influences the equilibrium and dynamic properties of adsorbed films as well as those of micelles. Both of these, in turn, influence the performance of the technological processes mentioned above. In the present dissertation, the relation between SDS micellar stability and technological processes is extended to nonionic surfactants. Does the stability of nonionic surfactant micelles also influence dynamic interfacial properties, such as dynamic surface tension and foaming? Since the monomer concentration (i.e., the CMC) of nonionic surfactants is much lower than for ionic surfactants, it is expected that micelle breakup plays an even more important role in nonionic surfactant solutions. Relative little has been published on the kinetics of nonionic surfactants [Lang, 1975; Strey and Pakusch, 1986; Michels et al., 1997] and the results are often contradictory. Relaxation times in the order of microseconds have been reported using stopped-flow and temperature -jump techniques. In that case the mixing time of the stopped-flow apparatus (usually on the order of milliseconds) and the presence of salt in the temperature jump sample (to have a reasonable electrical conductivity) can easily lead to erroneous results. Moreover, no direct correlation between Xj and technological processes has been reported. In Chapter 2, the importance of the method to be employed to determine the CMC of nonionic surfactants is discussed. The surface tension method, which is most conmionly used for the determination of the CMC is compared to the dye micellization method for a wide range of technical grade as well as pure monodisperse nonionic surfactants. Commercially available, technical grade nonionic surfactants have a wide distribution in the degree of ethoxylation, which can lead to an erroneous interpretation of the surface tension versus surfactant concentration plot.

PAGE 32

17 Chapter 3 presents a new method, which was developed to measure the slow micellar relaxation time of nonionic surfactants. The relaxation data for a variety of technical grade as well as pure monodisperse surfactants was determined. In Chapter 4, the relaxation data as obtained in Chapter 3 was related to interfacial processes, such as dynamic surface tension and foaming. The ability to determine relaxation data of ionic as well as nonionic surfactant solutions allows for the prediction of the performance of a given surfactant solution. Moreover, the results suggest that one can design or tailor micelles with a specific stability by controlling the surfactant structure, concentration and physicochemical conditions, as well as by mixing anionic/cationic or ionic/nonionic surfactants for a desired technological application. Chapter 5 demonstrates how the stability of SDS micelles can be tailored by the addition of oppositely charged surfactant, long chain alcohols or electrolyte, to control interfacial properties, such as surface tension, surface viscosity, foamability, foam stability and dynamic surface tension. Tailoring the SDS micellar stability by the addition of long chain alcohols gives rise to the fact that different methods of producing foam can result in opposite foaming behavior. The amount of foam produced depends on the time scale at which new interfacial area is produced. This is explained based upon dynamic surface tension and micellar stability. If foaming behavior can be predicted by micellar stability, it is expected that antifoaming action also relates to micellar stability. Chapter 6 discusses the effect of micellar stability on antifoaming behavior for SDS solutions in the presence of a variety of commonly used antifoaming agents. A new way of controlling antifoaming action by tailoring micellar stability is presented.

PAGE 33

In Chapter 7 an attempt has been made to correlate the effect of molecular interactions between surfactant molecules in bulk with the stability of micellar films adsorbed onto mica and silica. Atomic Force Microscopy (AFM) was used to reveal the structures of alkyltrimethylanmionium bromides on mica and to measure the forcedistance curves between the AFM tip and the surface. The results are used to delineate the molecular mechanism by which surfactants influence dispersion stability. Since at high electrolyte concentrations the electrical double layer repulsion is negligible, the repulsive force must come from the steric stabilization by either surfactants, micelles, surfaceactive polymers or a mixture of both. Chapter 8 then summarizes the conclusions of this dissertation and the scope of the future extension of these studies.

PAGE 34

CHAPTER 2 DETERMINATION OF CRITICAL MICELLE CONCENTRATION OF NONIONIC SURFACTANTS 2.1 Introduction Ionic surfactants such as sodium dodecyl sulfate (SDS) or alkyl trimethylammonium bromides (CnTAB) are usually easier to obtain in pure form than the ethoxylated nonionic surfactants. The commercially available (technical grade) ethoxylated surfactants not only have a size distribution in the hydrophobic part, but also in the degree of ethoxylation. The distribution of ethoxylation has been a subject of study for long time [Wanka et al., 1990]. However, the importance of this aspect in the method to be employed for the measurement of critical micelle concentration (CMC) has not been brought out in the literature satisfactorily. Large differences are often observed between CMC values of nonionic surfactants as determined by different methods. This is attributed to a broad molecular weight distribution and the presence of impurities [Alexandris et al., 1994; Mysels and Stafford, 1990]. There are a number of methods which have been employed for the determination of CMC of surface-active agents [Shinoda and Nakagawa, 1963; Hunter, 1987]. An excellent evaluation of the methods for determining the CMC is included in the comprehensive compilation of CMC's in aqueous solution by Mukerjee and Mysels [1971]. In this study the effect of size distribution in the ethoxylation of nonionic 19

PAGE 35

20 surfactants will be discussed for two commonly used methods for the determination of CMC, namely dye micellization and the surface tension methods. 2.2 CMC Determination by the Dye Micellization Method Dyes can be used in many ways to measure CMC. Water-soluble dyes, such as Merocyanine, Eosin, Rhodamine and Sudan are known to show a shift in the maximum wayelength (A^ax) due to the presence of micelles [Shinoda and Nakagawa, 1963; Hunter, 1987]. This shift for Eosin Y is shown in Figure 2-1. Absorbance 518 542 Wayelength (nm) Figure 2-1. Scan of a UV-VIS absorbance spectrum. The shift in the maximum absorbance from 518 to 538 nm is due to the presence of micelles. The maximum difference is obseryed at 542 nm. Surfactant: Triton X-100; dye: Eosin Y, 0.019 mM. Eosin Y in water shows a maximum absorbance at 518 nm. Dicreasing the surfactant concentration, howeyer, results in an increase of the absorbance at 538 nm. The

PAGE 36

21 maximum difference is observed at 542 nm. This shift is either followed as a change in Xroax or a change in absorbance of the micellized dye at a fixed wavelength (542 nm in Figure 2-1) as a function of surfactant concentration. It has been suggested that the inflection point in should be treated as the CMC [Shinoda and Nakagawa, 1963]. However, not all dyes show a distinct shift in X^ax (e.g. Merocyanine 540). Therefore, if Xmax of the micellized dye is sufficiently different from the aqueous dye, the absorbance at this wavelength (for Eosin Y at 542 nm) can be followed as a function of surfactant concentration to measure the extent of dye uptake. Below the CMC the rise in absorbance is very small, whereas above the CMC the rise is sharp. At the point where roughly half the dye is in the continuous phase, the maximum absorbance shows the sharp shift (in Figure 2-1 from 518 to 538 nm). Since the micellization process is known to be much less sharp for nonionic surfactants than for ionic surfactants, the rise in absorbance varies strongly over a range of surfactant concentration. At high enough surfactant concentrations, the absorbance vs. concentration curve will flatten again as most of the dye shifts to the micelles depleting the dye in the continuous phase. The linear portion near the inflection point is extrapolated to the point where the absorbance matches that of the dye in the absence of any surfactant and this concentration is defined as the CMC of the surfactant, see Figure 2-2. A second method using dye involves the solubilization of a water-insoluble (hydrophobic) dye in micellar solutions. This solubilized dye can then be measured by its UV-visible absorbance. This method yields satisfactory results for some surfactants.

PAGE 37

22 Concentration of Tween 20 (mM) Figure 2-2. CMC determination of Tween 20 (CMC = 0.042 mM) using the dye micellization method (absorbance at 542 nm). Eosin Y concentration: 0.019 mM. The horizontal dashed line represents the dye absorbance in water in the absence of surfactant. However, sometimes the solubilization of such a dye is so high that the micellar structure is affected by the dye. If the solubilization is weaker, the dye taken up by the micelles is often insufficient to produce a reasonable absorbance signal. Especially, when the CMC is very low, the micellar phase at concentrations just above the CMC is far too small to solubilize a significant amount of dye. Therefore, it is difficult to define a clear CMC for nonionic surfactants using this method and hence this approach was not pursued in the present study.

PAGE 38

23 2.3 CMC Determination by the Surface Tension Method The surface tension of aqueous solutions of surface-active agents decreases rapidly until the CMC is reached and then stays constant above the CMC. An important point has to be made while using the surface tension method, irrespective of the instrument employed for measuring it. The surface tension method is based upon the fact that the free surfactant concentration remains almost unchanged after the onset of micellization even if the total surfactant concentration is increased. However, due to surface-active impurities, the surface may get saturated with highly surface-active molecules, although the actual onset of micellization may take place at a higher surfactant concentration. The surface tension may thus become invariant with surfactant concentration even below the CMC. Therefore, the saturation of the surface, identified as the CMC in the surface tension vs. concentration of surfactant plot, does not always reflect the presence of micelles in the bulk solution. In the present study, a comparison is made of the CMC of a few commercial surfactants (Tween 20, 22, 40, 60 and 80, Triton X-100, Brij 35, 58 and 78) as measured by the surface tension method (Wilhelmy plate) and the dye micellization method. When the impurities are selectively removed from a commercial surfactant sample, the value of the CMC as measured by the surface tension method is expected to be closer to the value measured by the dye micellization method. The present study compares the surface tension and dye micellization behavior of pure (monodisperse) and commercial (technical grade) nonionic surfactants.

PAGE 39

24 2.4 Experimental Procedure 2.4.1 Materials Tween 20, 22, 40, 60, and 80, and Brij 35, 58, and 78 were supplied by ICI Aricas, Inc. (Wilmington, DE). Triton X-100 was obtained from Aldrich Chemical Company (Milwaukee, WI). The pure nonionics penta-ethyleneglycol mono n-dodecyl ether (Ci2(EO)5) and octa-ethyleneglycol mono n-dodecyl ether (Ci2(EO)8) were purchased from Nikko Chemicals Co. (Tokyo, Japan). The high purity grade dyes Eosin Y (C2oH6Br4Na205, anionic) and Merocyanine 540 (C26H32N3Na06S2, anionic) were supplied by Acros Organics (Fair Lawn, NJ). Deionized, distilled water was used in all experiments. Surfactant concentrations were calculated using their molecular weights disregarding the presence of any probable impurities. 2.4.2 Surface Tension Equilibrium surface tensions were measured for freshly prepared solutions by the Wilhelmy plate method at 22°C. The platinum plate was always cleaned and heated to a red/orange color with a Bunsen burner before use. 2.4.3 Spectrophotometry Eosin Y was used for the CMC determination of the nine commercial surfactants using the dye micellization method. Merocyanine 540 was used for the two ultra pure nonionic surfactants. Absorbance spectra were taken using a Hewlett Packard UV-VIS spectrophotometer (model 8453) with temperature control. All spectra were taken at 22°C.

PAGE 40

25 •| 2.4.4 Foam Fractionation Foam fractionation was carried out by shaking 25 mL of the surfactant solution vigorously such that the volume of the foam and the liquid together was four times the volume (100 mL) of the initial liquid. One half of the initial hquid was then separated from the foam as foam-fractionated sample and used for the surface tension and dyemicellization studies. The sample was expected to contain a lower amount of the highly surface-active impurities after foam fractionation. The surfactant loss due to foam 1 fractionation was determined by a Tekmar-Dohrmann Phoenix 8000 Total Organic Carbon (TOC) Analyzer. 2.5 Results and Discussion Table 2-1 shows the CMC values obtained by surface tension and dye micellization methods as well as the ratio of the two values. The CMC values obtained A from the dye micellization method are approximately 1.6 to 6.5 times higher than the values obtained by the surface tension method. The smallest difference between CMCs as measured for Tween 80 (Table 2-1), indicates that this sample does not contain a large amount of surface-active impurities or a wide distribution of ethoxylation. The largest difference was observed for Tween 22. This is probably due to the large ethoxylation number (80) which makes a wide distribution of ethoxylation more likely. Figure 2-3 ^ shows the different CMCs obtained for Tween 20. The break in the y vs. logC curve observed for these surfactants is thus not an indication of the CMC. i

PAGE 41

26 Table 2-1. Comparison of critical micelle concentations as determined by surface tension (S.T.) and dye micellization (D.M.) methods. Dye concentration: Eosin Y, 0.019 mM. Surfactant Structure CMC by CMC by ( CMC^ ^ Dye S.T. D.M. (mM) (mM) 1 ween zu Qi^frviton T qiii"q1"^* ThcI"/^!* i oOiDllan I^aUraie iJdiei \C4\J20) KJ.yJ I 1 0 049 J.O I ween Qr^rViitcin T ciiirQtf* Pctf^r /'Plinri i OUiUllaii l^Ul alC LZtalCl v,«^^80/ 0 on 0 084 \J.\JO'~r 6 5 1 ween *+u oUlUilail I ailllilalC JJfdLCl \L^\J20j \J.\J\J\J 1 0 094 1 ween uw 0 0055 0 099 1 ween ou oUlUllall \JlCalC JZalCl ylZVJjQ) 0 01 R 0 COR 1 fi Triton X-100 Octyl Phenol Ether (EOio) 0.080 0.20 2.5 Brij 35 Lauryl Alcohol Ether (EO23) 0.030 0.068 2.3 Brij 58 Cetyl Alcohol Ether (EO20) 0.0028 0.01 3.6 Brij 78 Stearyl Alcohol Ether (EO20) 0.0018 0.0071 3.9 C,2(EO)5" Lauryl Alcohol Ether (EO5) 0.058 0.060 0.98 C,2(EO)8" Lauryl Alcohol Ether (EOg) 0.070 0.072 0.97 " Pure nonionic surfactant. Merocyanine 540 dye was used for the CMC determination. It may be expected that if the impurities are removed, the two methods yield results which are closer to each other. It has been suggested in literature [Tharapiwattananon et al., 1996; Chang et al., 1992; Chiu and Huang, 1991], that foaming and consequently skimming the foam away from the solution purifies the surfactant solution. This process, called foam fractionation, is a process in which solute species are adsorbed at a gas-liquid interface between a dispersed phase (gas bubble) and a continuous phase (bulk liquid). Foam fractionation processes have been used to remove surface-active agents from aqueous solutions [Elving, 1982].

PAGE 42

27 0.001 0.01 0.1 1 Concentration of Tween 20 (mM) Figure 2-3. Critical micelle concentration of Tween 20 determined by surface tension (A, CMC = 0.011 mM) and dye micellization methods (+, CMC = 0.042 mM). Eosin Y concentration: 0.019 mM. The horizontal dashed line represents the dye absorbance in water in the absence of surfactant as well as the equilibrium surface tension. In order to confirm this, a solution of Tween 20 was subjected to foam fractionation and the remaining solution was used for the CMC determination. Total Organic Carbon analysis showed that less than 5% of the total surfactant concentration was removed and therefore the surfactant concentration was taken to be the same as before foam fractionation. Figure 2-4 shows a graph of the CMC determined by the dye micellization and surface tension methods for the foam fractionated Tween 20 sample. It is clear that the two methods show values much closer to each other (0.051 vs. 0.057 mM), indicating that the surface-active impurities as well as the lower ethoxylated molecules were removed from the original solution.

PAGE 43

28 0.001 0.01 0.1 1 Concentration of Tween 20 (mM) Figure 2-4. Critical micelle concentration of Tween 20 determined after foam fractionation by surface tension (A, CMC = 0.051 mM) and dye micellization methods (+, CMC = 0.057 mM). The horizontal dashed line represents the dye absorbance in water in the absence of surfactant as well as the equilibrium surface tension. Figure 2-5 shows the CMC curves for the pure (monodisperse) nonionic surfactant Ci2(EO)5. In this case, a different dye was used (Merocyanine 540), which shows a micellized dye peak at 575 nm. It is clear that in the absence of impurities, both surface tension and dye micellization methods yield the same result (see also Table 2-1). This study shows that the surface tension method can be erroneous for commercial (technical grade) surfactants under common conditions. The onset of micellization in the case of a pure surfactant causes the free (non-micellized) surfactant concentration to remain constant when the total surfactant concentration is increased. This is correctly interpreted as the CMC of the surfactant. In case of impure surfactants or

PAGE 44

29 Concentration of C12(EO)5 (mM) Figure 2-5. Critical micelle concentration of pure Ci2(EO)5 determined by surface tension (A, CMC = 0.058 mM) and dye micellization methods (+, CMC = 0.060 mM). Merocyanine 540 concentration: 0.019 mM. The horizontal dashed line represents the dye absorbance in water in the absence of surfactant as well as the equilibrium surface tension. surfactant mixtures, however, the situation is different. An impure sample of sodium dodecyl sulfate (SDS) for example, invariably contains lauryl alcohol, which is much more hydrophobic than SDS and thus has a much higher affinity for adsorption [Fruhner and Czichocki, 1996; Chiu and Wang, 1990]. Thus, the lauryl alcohol can saturate the surface and exhibit constant surface tension without any micelle formation in the bulk solution. Recently, Goebel and Lunkenheimer [1997] reported the importance of purity in the measurement of interfacial tension. As was shown in their study, using water/n-alkane interfaces, trace impurities significantly influence the interfacial tension. For nonionic surfactants, the lower ethoxylated species, which are always present in a commercial (technical grade) ethoxylated nonionic surfactant, have a higher affinity for adsorption

PAGE 45

30 than the higher ethoxylated species. At concentrations much below the true CMC value, the air/liquid interface can be already saturated with the more surface-active species before any micelle formation in the bulk solution. This is visualized in Figure 2-6A. A: below CMC B: at CMC = Nonionic Surfactant Surface Tension — r CMCsT CMCoye Surfactant Concentration Figure 2-6. Schematic diagram showing how the surface tension method would suggest a lower CMC than the dye micellization method because of the saturated air/liquid interface (solution A). Micelles start to form at a higher surfactant concentration as determined by the dye micellization method (solution B). Every open circle represents, for example, 4 (EO) groups.

PAGE 46

31 An increase in the bulk surfactant concentration from this point on will ultimately result in the formation of micelles at a specific surfactant concentration, which is the true CMC of the solution (Figure 2-6B). Although the CMC is indeed lowered by the presence of more surface-active species, the lowering is not as significant as suggested by the results from the surface tension method. The dye micellization method in such a situation would certainly yield a higher CMC than the surface tension method, as demonstrated in this study. 2.6 Conclusions 1. The surface tension method (Wilhelmy plate) for the determination of the CMC of commercial (technical grade) nonionic surfactants is very sensitive to the presence of molecular species with higher surface-activity. 2. The shortcoming of the surface tension method for determining the CMC of technical grade nonionic surfactants was demonstrated. In the presence of highly surface-active impurities, the air/liquid interface gets saturated at concentrations much below the true CMC leading to a wrong interpretation of the break in the y vs. logC curve. The micellized dye method gives reliable results for micelle formation in bulk solution, even in the presence of such impurities. 3. Foam fractionation of a solution of technical grade nonionic surfactant can selectively remove the species with higher surface-activity. The CMC values, as measured by the surface tension and dye micellization method, are in close agreement with each other after foam fractionation.

PAGE 47

32 The CMC values obtained for the pure monodisperse nonionic surfactants Ci2(EO)5 and Ci2(EO)8 are the same using surface tension and dye micellization methods, indicating the absence of highly surface-active impurities.

PAGE 48

CHAPTER 3 MICELLAR KINETICS OF NONIONIC SURFACTANTS 3.1 Introduction As mentioned in Chapter I, the micellar relaxation time measured for ionic surfactants (e.g., SDS) can be determined by the pressure-jump technique with electrical conductivity detection. This technique takes advantage of the fact that the CMC shifts to higher concentration when a surfactant solution is pressurized [Attwood and Florence, 1983; Kaneshina et al., 1983]. In case of ionic surfactants, the electrical conductivity increases with pressure. When the pressure is instantaneously released to atmospheric, monomers will reassociate to form new micelles, which can be followed as an exponential decay in the electrical conductivity with time [Huibers et al., 1996]. The slow micellar relaxation constant, T2, can be calculated from the first order reaction constant, k (X2 = 1/k). The pressure-jump technique with electrical conductivity detection is a very powerful tool that allows for the measurement of X2 for mixed micelles or micellar solutions in presence of additives. Chapters 5 and 6 demonstrate how this technique can be used to measure Xj for tailored micelles with a specific stability to control dynamic surface tension and hence technological processes, such as foaming and antifoaming. This chapter discusses a method developed to determine the slow micellar relaxation constant of a variety of nonionic surfactants. In the next chapter, the obtained relaxation data are compared to dynamic surface tension and foaming. 33

PAGE 49

34 3.2 Measurement of Nonionic Micellar Relaxation Time by Stopped-Flow The measurement of nonionic micellar relaxation time is more complicated than for ionic surfactants, since the electrical conductivity is not a sensitive parameter. Therefore, the use of a dye is necessary to obtain information about the micellar kinetics of nonionic surfactants. A number of dyes or fluorescent compounds, such as Merocyanine, Eosin, Rhodamine and Sudan show an appreciable change of extinction coefficient depending on whether the dye resides in, or outside the micelle in aqueous phase. This effect is often used to determine the CMC, as discussed in Chapter 2 [Shinoda and Nakagawa, 1963; Hunter, 1987], but it also provides a way of following the relaxation kinetics upon a fast temperature, pressure or concentration jump by employing spectrophotometric detection methods. As explained earlier in Chapter 2, Eosin Y in water shows a maximum absorbance (X,max) at 518 nm. Figure 31 . Absorbance spectra of Eosin Y in water and 2 mM Triton X100 solution (Eosin Y concentration: 0.019 mM).

PAGE 50

35 Increasing the surfactant concentration, however, causes the dye to partition between the water and the micelles, causing the maximum absorbance to shift to approximately 538 nm. The maximum shift in absorbance occurs at 542 nm. (Figure 3-1). The concept of change in absorbance due to the presence of micelles can be used in the determination of the slow relaxation constant, T2, for nonionic surfactants using the stopped-flow dilution technique. Stopped-flow (Figure 3-2) is a method designed to measure the kinetics of fast reactions [James, 1977]. Chamber 1: Surfactant/Dye U17 Light Source (UV-VIS) PMT Detector Chamber 2; Water/Dye Abs Time Figure 3-2. Stopped-flow apparatus used for determination of the slow micellar relaxation constant, T2, for nonionic surfactants. The apparatus employs two separate syringes, which can be filled with reactants, which are pushed instantaneously into a transparent cell. The change in absorbance can be detected with a very sensitive photomultiplier detector as the reaction progresses. When one solution containing micelles and dye is instantaneously diluted with another solution

PAGE 51

36 containing water and dye of the same dye concentration, the absorbance of dye in micelles will decrease as micelles breakup, indicating the relaxation time of micelles. This is schematically shown in Figure 3-3. Figure 3-4 shows the decrease in intensity (absorbance at 542 nm) as a function of time in a typical stopped-flow dilution experiment. The exponential decay shown in Figure 3-4 can be fit to first order reaction kinetics, resulting in the associated time constant X2Figure 3-3. Schematic diagram showing the micellar relaxation process involved in a stopped-flow dilution experiment. Micelles disintegrate after mixing, thereby releasing dye into the water. Figure 3-4. Typical relaxation curve obtained in a stopped-flow dilution experiment. Mixing is induced at t = 0.

PAGE 52

37 3.3 Experimental Procedure 3.3.1 Materials Tween 20, 22 and 80, Brij 35 and Synperonic A7 and A50 were supplied by ICI Americas, Inc. (Wilmington, DE). Triton X-100 was supplied by Aldrich Chemical Company (Milwaukee, WI). The pure nonionics penta-ethyleneglycol mono n-dodecyl ether (Ci2(EO)5) and octa-ethyleneglycol mono n-dodecyl ether (Ci2(EO)8) were purchased from Nikko Chemicals Co. (Tokyo, Japan). The high purity grade dyes Eosin Y (C2oH6Br4Na205, anionic) and Merocyanine 540 (C26H32N3Na06S2, anionic) were supplied by Acros Organics (Fair Lawn, NJ). Deionized, distilled water was used in all experiments. Surfactant concentrations were calculated using their molecular weights disregarding the presence of any probable impurities. 3.3.2 Relaxation Time Measurement by the Stopped-Flow Method The slow relaxation time X2 of the nonionic surfactants was measured by the stopped-flow dilution method. The stopped flow apparatus (Dionex Corporation, Houston, TX) mixes two solutions in approximately two milliseconds in a 20 mm light path cuvette. Two detectors allow for the simultaneous measurement of transmitted, fluorescent or scattered light, or the same light at different wavelengths (200-1000 nm). The monochromator includes a deuterium and tungsten light source. In this study the absorbance at 542 nm was measured as function of time. The system was connected to a Durrum (Palo Alto, CA) temperature jump system used here only to amplify the signal coming from the photomultiplier. The signal was stored using a Tektronix 340A oscilloscope.

PAGE 53

38 3.3.3 Relaxation Time Measurement by Pressure-Jump with Optical Detection The pressure-jump technique with optical detection was used to confirm the relaxation data obtained by the stopped-flow apparatus. Experiments were performed at the University of Bielefeld, Germany, using the setup described in detail by Knoche and Wiese [1976]. All relaxation data were obtained at 22°C. The pressure-jump with optical detection is technically the same as the pressure-jump apparatus with electrical conductivity detection, except the detection method. 3.4 Results and Discussion Table 3-1 shows the slow relaxation times measured for a variety of nonionic surfactants using the stopped-flow dilution technique. A dye concentration of 0.019 mM was found to be the minimum concentration to still observe a significant absorbance signal. Tondre et al. [1975] found that a surfactant/dye molar ratio as low as 20 does not influence the relaxation time. With the exception of Tween 22, all other investigated surfactant samples have surfactant/dye ratios larger than 20. It is clear that the relaxation time can vary from 2 to 150 seconds depending upon the molecular structure of the nonionic surfactant. The long relaxation time of 150 seconds for Synperonic A7, can be described as a frozen micelle as compared to those exhibiting relaxation times on the order of milliseconds (usually ionic surfactants). Nonionic surfactants show a much longer relaxation time (T2) than ionic surfactants, because of the absence of ionic repulsion between the head groups. The surfactants Synperonic A7, Brij 35 and Synperonic A50 have comparable alkyl chain lengths (C12C15) but increasing degree of ethoxylation. It is clear that increasing the number of

PAGE 54

39 Table 3-1. Micellar relaxation constants, T2, measured by the stopped-flow dilution technique. Surfactant Structure Cone. CMCoye X2 (mM) (mM) (s) Tween 20 Sorbitan Laurate Ester (EO20) 0.47 0.042 6 Tween 22 Sorbitan Laurate Ester (EOgo) 0.37 0.084 2 Tween 80 Sorbitan Oleate Ester (EO20) 0.49 0.028 8-10 Triton X-100 Octyl Phenol Ether (EOio) 0.40 0.20 3.5 Synperonic A7 C12-C15 Alkanol Ether (EO7) 0.80 0.050 150 Brij 35 Lauryl Alcohol Ether (EO23) 0.50 0.068 80 Synperonic A50 C12-C15 Alkanol Ether (EO50) 0.40 0.084 40 C,2(EO)5" Lauryl Alcohol Ether (EO5) 0.80 0.060 10 C,2(EO)8" Lauryl Alcohol Ether (EOs) 0.40 0.072 4 " Pure (monodisperse) nonionic surfactant. Merocyanine 540 dye was used for the CMC and T2 determination. Both dyes resulted in the same CMC and T2 data. ethylene oxide units decreases the relaxation time, which was also observed for octylphenyl polyoxyethylenes by Lang and Eyring [1972]. The relaxation times obtained for the ultra pure nonionic surfactants Ci2(EO)5 and Ci2(EO)8 are relatively small as compared to the Synperonics (respectively 10 and 4 seconds as compared to 150 seconds). The difference might be attributed to the broad molecular weight distribution and the presence of impurities (Chapter 2). It is known [Greenshields, 1998], that Synperonic A7 contains a significant amount of long chain alcohols that apparently contributes to the stability of the micelles.

PAGE 55

40 3.5 Validation of the Slow Relaxation Time by PressureJump with Optical Detection In the stopped-flow dilution technique the number of micelles decreases and thus the kinetics of micelle breakup is measured (Figure 3-3). However, in the pressure-jump technique (with either electrical conductivity or optical detection), the kinetics of micelle formation is measured after the pressure is released to atmospheric. This process is schematically shown in Figure 3-5. For a surfactant solution in equilibrium, the rate of micellar dissociation equals the rate of association. Therefore, both the stopped-flow and pressure-jump techniques should yield the same relaxation constant X2, if the perturbation is small enough. f_ High Pressure, High CMC Low Pressure, Low CMC Figure 3-5. Schematic diagram showing the micellar relaxation process involved in a pressure-jump experiment. After the pressure is released to atmospheric, monomers reassociate to micelles. In order to show that both the micelle breakup and micelle formation rates exhibit the same time constants, pressure-jump studies with optical detection were performed at the University of Bielefeld, Germany. Figure 3-6 shows the relaxation time \ of Triton X1(X) and Brij 35 measured by stopped-flow and pressure-jump with optical detection

PAGE 56

41 techniques (absorbance at 542 nm). It is evident that the relaxation time measured for both surfactants is the same for both techniques within the experimental error. This suggests that the relaxation time, Xj, for micellization and demicellization are indeed the same. (1) E ic o ^ 03 — X CM m H DC _o 100 90 80 70 60 50 40 30 20 10 Stopped-Flow B PressureJump 3.5 4.3 Triton X-100 sao 74.9 Brij 35 Figure 3-6. Validation of relaxation constants, X2, by pressure-jump and stopped-flow dilution techniques, both with optical detection (absorbance at 542 nm, Eosin Y concentration: 0.019 mM). As mentioned earlier, the relaxation constants as measured by pressure-jump and stopped-flow techniques will only be the same if the perturbation of the system is small enough. The following derivation proofs that for small perturbations in the concentration of micelles or monomers, the obtained relaxation constant is indeed the reaction constant k at equilibrium. Consider micellization as the following simplified reaction, where A represents the monomers and B the micelles.

PAGE 57

So, at equilibrium. k2 A 4 ^ B k, k C =k C 42 (3.1) (3.2) The net change in concentration of A, dCA/dt, can be given as, dC -^--kiC^+kjCg ^2.3) where the first term represents the decrease in concentration of A due to the forward reaction (ki) and the second term represents the increase in concentration of A due to the backward reaction (^2)The pressure-jump and stopped-flow dilution methods impose a small perturbation in Ca, AC a, so that, Q = ^A.eq + ^^A ^3_4-j From experimental observation and as derived by Aniansson and Wall [1974], ACa decays according to first order reaction kinetics back to CA.eq, AC, = AC,,o^-'^^' (3.5) where ACa.o is the perturbation at t = 0. Hence, the expression for Ca becomes. C, =C,,,+AC,,oe-'^^' (3.6) Consequently, d£ dt ' = -AC,.of7,e-'^^' (3.7)

PAGE 58

43 The rate equation for reaction (3.1) can be equated by substituting (3.4) and (3.7) in (3.3), dt (3.8) rewriting (3.8) yields, -k, -AC (J e'"*' (3.9) and substituting for k2Ai=CA.eq/CB.eq, B ,eq (3.10) Hence, . l{CA.e,Cs.e, )+ (AC,C,,^ )(C,,^C, )J _^ , ''^i ^ ~ A,a^ A^ B,eq (3.11) If the perturbation is small enough we can assume that Cb « Ca.eq and thus CA,eqCB,eq « CA.eqCa, TCSUlting in. (3.12) Equation (3.5) cancels out from (3.12) resulting in. ^1 (3.13)

PAGE 59

44 This proof shows that the relaxation constant Ga, as determined by the pressure-jump and stopped-flow methods, is the reaction constant of micelle formation at equilibrium, kj, provided the change in monomer or micelle concentration is small enough. 3.6 Conclusions 1. Two spectroscopic techniques (stopped-flow and pressure-jump, both with optical detection) were used to measure the slow micellar relaxation time of nonionic surfactants. Nonionic surfactants show a much longer relaxation process than ionic surfactants because of the absence of ionic repulsion between the head groups. 2. Increasing the degree of ethoxylation leads to shorter relaxation times in case of nonionic surfactants. 3. The slow micellar relaxation time, T2, for the technical grade nonionic surfactants is significantly higher than the relaxation times for the pure, monodisperse surfactants. Apparently, the distribution in degree of ethoxylation and the presence of impurities (such as long chain alcohols) contribute to the stability of nonionic surfactant micelles. 4. Eosin Y and Merocyanine 540 dye resulted in the same CMC and slow micellar relaxation constants, indicating that at surfactant/dye ratios larger than 20, no negative influence of the dye on the micellar stability exists. 5. The measurement of X2 by stopped-flow or pressure-jump with optical detection resulted in the same time constants (within the experimental error), indicating that the processes of micellization and demicellization occur at the same rate.

PAGE 60

CHAPTER 4 MICELLAR KINETICS AND DYNAMIC SURFACE TENSION 4.1 Introduction Surface tension is the most direct measure of the surface-activity of a surfactant in solution. Equilibrium measurements of surface tension have been performed for many surfactant systems under various solution conditions. By comparison, relatively little data exists for the dynamic surface tension (DST) of these solutions. The understanding of dynamic surface tension is important to any technological application where a new interface is rapidly created in the presence of a surfactant solution [Miller et al., 1995]. Such new interfaces may be of various types, for example new air/water interfaces in foaming or film formation processes, new liquid/liquid interfaces in emulsification or new solid/liquid interfaces in detergency or fabric wetting applications. In most cases, the equilibrium surface tension is never reached, and the actual surface tension experienced by the interface is much higher. In these cases, the dynamic surface tension plays a more important role than the equilibrium surface tension [Borchardt and Yates, 1993; Rosen, 1989]. The effect of dynamic surface tension is particularly important in solutions containing large surfactants, which can be expected to have a slower rate of diffusion to the interface. Another important parameter in the measurement of dynamic surface tension is micellar stability. As discussed in chapter 1, very stable micelles cannot 45

PAGE 61

46 breakup fast enough to provide additional monomers resulting in higher interfacial tensions. Hence, the minima and maxima in solid/liquid, liquid/liquid and gas/liquid phenomena were obtained. This chapter discusses the importance of micellar stability and diffusivity on dynamic surface tension for a variety of surfactants studied in the previous Chapter. 4.2 Dynamic Surface Tension bv the Maximum Bubble Pressure Method Since the equilibrium surface tension is not reached in many dynamic interfacial processes, it is the dynamic surface tension that must be studied and correlated with processes of interest. The dynamic surface tension can be measured by the drop weight [Jho and Burke 1983; Miller et al. 1993], oscillating jet [Thomas and Hall, 1975], capillary wave [Kelvin, 1871], growing drop technique [MacLeod and Radke, 1993], and the maximum bubble pressure method [Mysels, 1986; Ross et al., 1992]. The maximum bubble pressure (MBP) technique is the most commonly used technique for the measurement of dynamic surface tension, and has been applied to a variety of anionic and nonionic surfactant solutions [Garrett and Ward, 1989; Fainerman et al., 1993, 1994; Tamura et al. 1995; Hua and Rosen, 1991]. The technique was first developed over a century ago as reviewed by Mysels [1990], but only became practical in recent decades with the availabiUty of fast pressure transducers and the electronics necessary to accurately monitor the rapidly changing pressure signal. The dynamic surface tension of values for short interface lifetimes can vary greatly from the equilibrium values. For pure water, a newly formed interface should have a surface tension approaching 72 mN/m, the equilibrium value as measured by, for

PAGE 62

47 example, the Wilhelmy plate static surface tension method. The physical principle behind the MBP measurement is the Laplace pressure; the pressure inside a curved liquid interface is higher than the ambient. This excess pressure (P) can be calculated from the Laplace equation [Adamson and Gast, 1997], P = ^+pgh (4 1) where y is the surface tension, r is the radius of curvature of the bubble, p is the liquid density, g is the gravitational constant and h is the depth of the bubble in the liquid. The first term expresses the Laplace pressure due to the curved gas/liquid interface, and the second term is the hydrostatic pressure due to the liquid height above the forming bubble. The first term will vary during the life cycle of the bubble, while the second term will remain constant. Figure 4-1 shows a typical life cycle of a bubble and the resulting pressure within the capillary. After a bubble breaks off from the capillary tip, the pressure is the lowest. As more gas flows into the capillary, the pressure builds up as the gas is pushed out of the capillary and the radius of curvature at the tip decreases. During this expansion process, surfactant is populating the new interface and acting to lower the surface tension. At the point of minimum interface radius of curvature, where a hemisphere of gas is formed at the capillary tip, the pressure is maximum. Both the minimum radius of curvature and the surface tension, as described by the Laplace equation, govem the maximum pressure experienced. On the incremental addition of gas, the bubble expands out of the capillary tip and the radius of curvature gradually increases. This results in a drop in the Laplace pressure, and a resulting rapid expansion of the

PAGE 63

48 bubble. At some point, the bubble breaks off from the capillary tip and the whole process starts again. Figure 4-1. Characteristic bubble pressure vs. time curve in the maximum bubble pressure method to determine dynamic surface tension. The maximum pressure is reached when the bubble is a perfect hemisphere at the tip of the capillary. Interface lifetime, A, is controlled by adjusting the bubble rate; B represents the dead time. 4.3 Dynamic Surface Tension and Micellar Stability Dynamic surface tension depends on several factors: monomer concentration (CMC), micellar stability, diffusion rate of the surfactant molecule to the interface, and the total surfactant concentration. During the formation of bubbles, surfactant monomers adsorb onto the freshly created interface from the bulk solution. The bubble dynamics are controlled mainly by diffusion below the CMC [Rillaerts and Joos, 1982]. Above CMC, however, the diffusion of monomers is augmented by the spontaneous breakdown of micelles. If the monomer is depleted by the adsorption process, micelles must breakup to provide additional monomers. If the micelles in solution are very stable, they cannot > Time m

PAGE 64

49 provide monomer fast enough and the dynamic surface tension remains higher. However, if the micelles are relatively unstable, their disintegration resupplies the depleted monomer and lower dynamic surface tensions are obtained. This is illustrated in Figure 42. Low D.S.T. High D.S.T. Figure 4-2. Effect of micellar stability on dynamic surface tension. In summary, for long bubble lifetimes, the equilibrium surface tension determines the interfacial tension at the air/water interface. However, when the bubble lifetime decreases, more and more monomer is depleted from the bulk solution and thus micelles must breakup in order to provide additional monomers. In that case, the breakup of micelles and thus the micellar stability determines the surface tension lowering. In order to show the importance of micellar breakup in the dynamic surface tension measurement, a dimensionless parameter 9 was introduced [Engels et al., 1998], y _y (4-2) f w I eq

PAGE 65

50 where Yd is the dynamic surface tension, Yeq the equilibrium surface tension as measured by the Wilhelmy plate method, and Yw the surface tension of pure water. This equation normalizes the surface tension with respect to the surface-activity of the solution. The denominator (Yw Yeq) can be considered as the effectiveness of the surfactant [Rosen, 1989]. When Yb = Yeq6 = 0, which indicates that the surfactant concentration at the surface of the bubble is the same as that under equilibrium conditions. However, when Yd = Yw, 6 = 1, indicating that no surfactant is present at the interface of the bubble. Values between 0 and 1 are a measure for the surfactant concentration at the surface and hence, the stability of micelles, assuming the diffusion time of monomers to be negligible. This is known to hold for ionic surfactants [Oh et al., 1992], however, the validity of this assumption for nonionic surfactants will be discussed in section 4.5. In this study, the dynamic surface tension behavior of three technical grade nonionic (Brij 35, Synperonic A7 and A50) and two pure nonionic surfactants was studied (Ci2(EO)5 and Ci2(EO)8). 4.4 Experimental Procedure 4.4.1 Materials Brij 35, Synperonic A7 and A50 were supplied by ICI Americas, Lie. (Wilmington, DE). Monodisperse penta-ethyleneglycol mono n-dodecyl ether (Ci2(EO)5) and octa-ethyleneglycol mono n-dodecyl ether (Ci2(EO)8) were purchased from Nikko Chemicals Co. (Tokyo, Japan). Deionized, distilled water was used in all experiments. Surfactant concentrations were calculated using their molecular weights disregarding the presence of any probable impurities. All experiments were performed at 22°C.

PAGE 66

51 4.4.2 Dynamic Surface Tension The maximum bubble pressure apparatus was constructed using a differential pressure transducer purchased from Omega Engineering, Inc. (Stanford, CT), with a sensitivity of 0 to 10 in. (25 cm) H2O (0 to 2500 Pa). A #23 steel needle was used as a capillary, with a nominal 0.025 in. (0.64 mm) external diameter, 0.013 in. (0.33 mm) internal diameter, and a flush cut tip. The capillary diameter was chosen so that the viscous resistance of water to bubble growth could be ignored. Such internal and viscous effects are a potential source of error in these measurements that need to be taken into consideration [Garrett and Ward, 1989]. All measurements were conducted with the capillary tip 1 cm beneath the liquid surface. Compressed air was used as the bubbling gas and an oscilloscope connected to the pressure transducer was used to determine the bubble frequency and the dynamic surface tension. A schematic diagram of the setup is shown in Figure 4-3. Rotameter Air flow Junction tube (enlarged scale) 1 Pressure transducer Surfactant solution (pressure) (voltage) Capillary tube (ID 0.013") 1 .0 cm depth Bubble formation Oscilloscope Figure 4-3. Setup for the measurement of dynamic surface tension by means of the maximum bubble pressure method.

PAGE 67

52 4.5 Results and Discussion Some properties used in the calculation of 9 are shown in Table 4-1. The three commercial (technical grade) surfactants have similar structures and CMCs (Table 4-1). Table 4-1. Equilibrium surface tension, CMC, and the slow micellar relaxation time, X2, as measured by the stopped-flow dilution technique. Surfactant Structure Yeq CMCoye 1:2 (mN/m) (mM) (s) Synperonic A7 C12-C15 Alkanol Ether (EO7) 29.0 0.050 150 Brij 35 Lauryl Alcohol Ether (EO23) 38.7 0.068 80 Synperonic A50 C12-C15 Alkanol Ether (EO50) 49.5 0.084 40 C,2(EO)5" Lauryl Alcohol Ether (EO5) 30.0 0.060 10 C,2(EO)8'' Lauryl Alcohol Ether (EOg) 34.7 0.072 4 " Pure (monodisperse) nonionic surfactant. The only difference between the molecules is an increase in degree of ethoxylation. Figure 4-4 shows the dimensionless parameter 0 versus the bubble lifetime for 2 mM solutions of Synperonic A7, Brij 35 and Synperonic A50. Figure 4-5 shows similar curves for the pure nonionics Ci2(EO)5 and Ci2(EO)8. The slow micellar relaxation constant, T2, did not show an appreciable change as function of surfactant concentration and therefore the relaxation kinetics (determined at the concentrations given in Table 3-1) are similar at 2 mM concentration [Eastoe, 1997].

PAGE 68

53 1 n 0.9 0 : ^ 1 0 0.5 1 1.5 2 Bubble Lifetime (s) Figure 4-4. Dimensionless dynamic surface tension vs. bubble lifetime for 2 mM solutions of Synperonic A7, Brij 35 and Synperonic A50. Figure 4-5. Dimensionless dynamic surface tension vs. bubble lifetime for 2 mM solutions of Ci2(EO)5 and Ci2(EO)8.

PAGE 69

54 It is clear that Synperonic A7 shows the slowest rate of adsorption of surfactant molecules due to the stability of micelles, resulting in 9 values close to 1. On the other hand, Synperonic A50 shows a faster adsorption of surfactant molecules, indicated by the lower 0 values. The same trend is observed for the pure nonionic surfactants. More stable micelles result in lower 9 values. Thus, increasing the degree of ethoxylation destabilizes micelles, resulting in lower dynamic surface tensions. A mathematical model which relates the slow micellar relaxation time (X2) to dynamic surface tension has been proposed by Fainerman and Makievski [1993], where Yeq is the equilibrium surface tension, R is the gas constant, T is the surface excess concentration, c the total surfactant concentration, t is the bubble lifetime and D the diffusion coefficient. It is clear that for long bubble lifetimes (t-»-oo), the dynamic surface tension approaches the equilibrium surface tension (Yeq). Furthermore, Fainerman [1992] found that the dynamic surface tension of micellar solutions at high surfactant concentration does not depend on the total surfactant concentration. Rather, it depends on the slow dissociation process of micelles. Equation 4.3 can be used to determine the slow micellar relaxation constant T2 from the dynamic surface tension measurements in the present study. Plotting the dynamic surface tension Yd vs. t ' results in curve with slope (d-^b/dt) proportional to (12)'^^ for very large t. The resulting values for X2, however, showed very poor agreement with the values obtained in this study. This can be attributed to the slope, which is almost zero at very long bubble lifetimes. Therefore, the error in the (4.3)

PAGE 70

55 calculation becomes very large. For example, Figure 4-6 shows the dynamic surface tension data for the two, monodisperse nonionic surfactants. The absolute value of T2 is hard to obtain from the curve at long bubble lifetimes, however, it is clear that (dyo/dt) for Ci2(EO)5 is larger than for Ci2(EO)8, and hence X2 of Ci2(EO)5 > Ci2(EO)8. 0.5 1 1.5 Bubble Lifetime (s) Figure 4-6. Dynamic surface tension vs. bubble lifetime for 2 mM solutions of C,2(EO)5andC,2(EO)8. A number of alternative theoretical approaches have also been proposed to account for the effects of adsorption barriers and micellar breakdown kinetics on dynamic surface tension [Filippov, 1994a,b; Rillaerts and Joos, 1982]. Although these treatments have met with some success for certain surfactants, they do not appear to be generally applicable. In general, dynamic surface tension is a useful tool to qualitatively confirm the

PAGE 71

56 slow micellar relaxation times, T2, of nonionic surfactants, obtained by the stopped-flow dilution or pressure-jump techniques with optical detection. 4.6 Importance of Diffusion Time of Nonionic Surfactants in DST Measurements It is desirable to compare the diffusion time of surfactant monomers from the bulk to the air/water interface with the relaxation time of micelles during the bubble process. In this section, a simple model will be presented and applied to the nonionic surfactants studied in the dynamic surface tension measurements (Synperonic A7, Brij 35 and Synperonic A50 and Ci2(EO)5 and Ci2(EO)8). The model assumes that micelles do not adsorb at the air/water interface [Horozov et al., 1997]. Consequently, the number of monomers (N/) needed to saturate the bubble surface of radius Rj can be calculated according to, N, = ^ (4.4) A where A is the area per molecule of the surfactant at the air/water interface. The number of molecules per mL solution {N2) at the CMC (in mM) is given by. AT CMC * N iv 2 ^iviK^ ly ^^^^ ^^^^ where Navos is the Avogadro number. The volume around the bubble, which contains Nj monomers, is given by N1/N2 mL. Therefore, a shell of radius R2 (containing Nj monomers) can be calculated by, ^=±;r(Rl-Rf) (4-6) 3

PAGE 72

57 The unknown quantity R2 can be determined by using the values of Ni, N2 and Rj calculated previously. The average characteristic diffusion length Ld is {R2-Ri)/2, which can be used to calculate the characteristic diffusion time tj, according to Overbeek [1977], (4.7) This is the diffusion time when the solution is assumed to be stationary. In fact, the diffusion time in real situations is much shorter than this value due to the convective motion of the solution induced by bubble motion. In order to calculate for the three commercial and the two pure nonionic surfactants (Synperonic A7, A50, Brij 35 and Ci2(EO)5 and Ci2(EO)8 respectively), the diffusion coefficient D and the molecular area need to be known. Consequently, the characteristic diffusion time can be calculated as function of bubble radius /?/. Table 4-2 shows the constants used in the calculation of td of the nonionic surfactants investigated in this study. Table 4-2. Constants used in the calculation of the characteristic diffusion time td. Surfactant Mw CMC DxlO^ Area/molecule I2 (g/mol) (mM) (cm^/s) (A) (s) Synperonic A7 522 0.050 3.67 50.3 150 Brij 35 1198 0.068 2.43 69.2 80 Synperonic A50 2414 0.084 1.71 210 40 C,2(EO)5" 406 0.058 4.17 51" 10 538 0.070 3.62 66" 4 " Pure (monodisperse) nonionic surfactant. " In good agreement with Rosen [1982].

PAGE 73

58 For example, for Synperonic A7 the following values are found assuming a bubble diameter of Rj = I mm: Ni = 2.5xl0'\ N2 = 3.0x10^^ R2 = 1.06x10 ^ m, = 3.11x10"^ m, and td= 1.31 s. Monomer diffusion coefficients, obtained for C8(EO)4 by Faucompre and Lindmann [1987], were used to approximate the other D values using the molecular weight (Mw), according to [Eastoe et al., 1996], ^ Mw 2 ^ Mw , (4.8) The areas per molecule were calculated from the surface excess concentration T using the Gibbs adsorption isotherm, equations (1.2) and (1.3). The results of the diffusion model are shown in Figures 4-7 and 4-8. Bubble Radius (mm) Figure 4-7. Characteristic diffusion time as function of bubble size in stationary solutions of Synperonic A7, Brij 35 and Synperonic A50. The characteristic diffusion time increases as the bubble radius increases and then flattens off. A similar trend was observed for the pure nonionic surfactants. Beyond a

PAGE 74

59 critical bubble radius Ri, tj does not further increase anymore. This can be attributed to the fact that AR (R2-R1) remains constant at larger bubble size. This constancy is determined by the CMC of the surfactant, and hence, higher CMC's result in constant td values at smaller bubble radii. The mathematical proof is given in Appendix B. Figures 47 and 4-8 assume a stationary solution, however, due to the convective motion induced by the bubbling of the gas, the real diffusion time is much shorter. It is clear that, for any bubble size, the characteristic diffusion times of the technical grade surfactants are much smaller than the slow micellar relaxation time Xj. This indicates that the micellar breakup time is a rate-limiting step in the supply of monomers to newly created interface. For the pure nonionic surfactants, however, the micellar breakup time is much smaller and hence, diffusion becomes a more competitive process. Table 4-3 lists the average characteristic diffusion times for bubble radii Ri > I

PAGE 75

60 mm. In case of the pure nonionic surfactants, diffusion accounts for approximately 10% of the adsorption process. Table 4-3. Comparison of the characteristic diffusion time, td, and slow micellar relaxation time, T2, in the dynamic surface tension measurement for Ri > 1 mm. Surfactant DxlO^ td X2 (cm^/s) (s) (s) Synperonic A7 3.67 1.3 150 Brij 35 2.43 0.55 80 Synperonic A50 1.71 0.05 40 Ci2(EO)5" 4.17 0.8 10 C,2(EO)8" 3.62 0.4 4 " Pure (monodisperse) nonionic surfactant. Another observation is that the size of the molecule does not seem to play an important role in the diffusion process. Earlier, it was expected that Synperonic A50, the largest molecule investigated in this study, would adsorb much slower to the interface due to its size. Still, a very small characteristic diffusion time was obtained. This can be attributed to; 1) higher CMC and 2) larger the area per molecule as compared to other nonionic surfactant molecules. Even though the diffusion coefficient of Synperonic A50 is small (Table 4-2), the CMC as well as the area per molecule are significantly larger than for the other surfactants. Thus, for Synperonic A50 relatively more monomers are available in the bulk solution (higher CMC) and, moreover, less molecules are required to cover the bubble surface (larger area per molecule).

PAGE 76

61 4.7 Conclusions 1. The dynamic surface tension behavior of three technical grade nonionic (Brij 35, Synperonic A7 and A50) and two pure nonionic surfactants (Ci2(EO)5 and Ci2(EO)8) was studied by means of the maximum bubble pressure method and related to the slow micellar relaxation time X2. A slow breakup of micelles (i.e. a long relaxation time T2), corresponds to high dynamic surface tensions, whereas very labile micelles result in lower dynamic surface tensions. 2. A dimensionless dynamic surface tension parameter, 6, was introduced indicating the importance of micellar stability in processes involving an increase in interfacial area. 6 values close to 0 indicate a very fast breakup of micelles, resulting in low surface tensions. 9 values close to 1 indicate a very slow breakup of micelles, resulting in relatively high surface tensions. 3. Increasing degree of ethoxylation leads to shorter relaxation times for nonionic surfactants and hence lower dynamic surface tensions. 4. The characteristic diffusion time, tj, of surfactant monomers is small as compared to the micellar relaxation time ii. However, for the pure nonionic surfactants, diffusion accounts for approximately 10% of the adsorption process, ignoring convective motion due to bubble formation. 5. The size of the surfactant molecule does not seem to influence the characteristic diffusion time of monomers. Synperonic A50, the largest molecule investigated in this study, showed much faster adsorption than the other nonionic surfactants. This can be attributed to its high CMC and large head group area.

PAGE 77

CHAPTER 5 EFFECT OF TAILORING MICELLAR STABILITY ON FOAMING PROPERTIES AND FOAMING METHODOLOGY 5.1 Introduction Foams are dispersions of gas in a liquid in which the volume of the dispersed phase is so high that the system can be regarded as a network of interconnected films [Void, 1983]. Foam is produced when air or some other gas is introduced beneath the surface of a liquid that expands to enclose gas with a film of liquid. Foam has a stable honeycomb structure of gas cells whose walls consist of thin liquid films with approximately plane parallel sides. Foams cannot be formed from pure liquids: a surfaceactive specie needs to be present in the system. This can be confirmed by a thermodynamic consideration. The Helmholtz free energy of a foam contains the surface area as an extensive variable, according to [Everett, 1988], where y is the surface tension of the liquid, A is the total surface area, and |j and n refer to the chemical potential and concentration of the components in the system, respectively. At constant temperature Tand concentration n, equation 5.1 yields after integration. dF = -pdV SdT +')dA+ ^^i^dn^ (5.1) (5.2) 62

PAGE 78

63 From equation (5.2) it is clear that a decrease in Helmholtz free energy results both from a loss of area and from the expansion of a gas. This can only occur in a coalescence process and hence, foam composed of gas in a pure liquid is thermodynamically unstable. Therefore, a third component is necessary to produce a stable foam. Foamability and foam stability primarily depend on chemical composition and properties of the adsorbed surfactant molecules. These, in turn, influence numerous factors, such as diffusion rate, micelle breakup time, rheology of the adsorbed layer, gaseous diffusion out of and into bubbles, size distribution of the bubbles, surface tension, bulk and surface viscosity and microstructure of the foam. Also, the presence of electrol)4e, temperature and pressure influences foam behavior [Bikerman, 1973; Rosen, 1989; Ross, 1958, 1980; Pugh, 1996]. As shown earlier by Oh and Shah [1991], the rate of adsorption of surfactant monomers onto the newly created surface during the foam generation process is primarily dependent upon the disintegration of micelles. Very stable micelles (for SDS at 200 mM concentration) cannot breakup fast enough to supply monomers necessary to stabilize the newly created area. Hence, higher dynamic surface tensions and thus less foam is generated. On the other hand, very labile micelles breakup quickly enough to provide additional monomers and thus more foam is generated. Therefore, by tailoring the micellar stability, one can control dynamic interfacial processes, such as foaming, wetting emulsification and solubilization. A schematic representation of the adsorption of surfactant monomers to the expanding interface due to the disintegration of micelles during foam generation is shown in Figure 5-1.

PAGE 79

64 Surfactant Solution Air Air Thin Liquid Film Figure 5-1. Schematic diagram showing the adsorption of surfactant molecules onto new surface area due to the disintegration of micelles. Very stable micelles result in a lower foamability. Another way of altering the micellar stability (other than changing the surfactant concentration) is the addition of long chain alcohols or oppositely charged surfactants. An example of tailoring SDS micellar stability by the addition of 1-dodecanol (C12OH) or dodecyltrimethylammonium bromide (C12TAB), a cationic surfactant, is given in Figure 5-2. It is clear that the micellar stability (T2) can be increased significantly by the addition of 1-dodecanol (for a 25 mM SDS solution from 1 ms to 230 ms). The introduction of ion-dipole interactions causes the micelle to pack closer and hence, longer relaxation times are observed. Even more significant is the addition of C12TAB. In that case, ion-ion electrostatic interactions are responsible for the three orders of magnitude increase in Xz (from 1 ms to 5000 ms). Lessner et al. [1981a,b] studied the influence of salt (NaC104) on the micellar stability of SDS. An increase in micellar stability was observed and a shift of the maximum relaxation time to lower surfactant concentrations.

PAGE 80

65 25 mM SDS, T2 = 1 ms. 25 mM SDS + 1.25 mM C12OH, T2 = 230 ms. 100 mM SDS + 10 mM C12TAB, X2 = 5000 ms (!) Figure 5-2. Tailoring SDS micellar stability by the addition of 1-dodecanol (C12OH) or dodecyltrimethylanmionium bromide (C12TAB). The study of the effect of alcohols on SDS micellar stability is not new. Earlier investigations by Leung and Shah [1986] and Yiv et al. [1981] showed that short chain alcohols (Ci to C5) labilize SDS micelles, which was explained on the basis of the Aniansson and Wall theory [1976]. A similar trend was found for the addition of glycerol [Huibers, 1996]. Glycerol increases the bulk viscosity, but decreases the slow micellar relaxation time X2. In summary, the addition of alcohols or alkyltrimethylammonium bromides allows for the tailoring of micelles with specific stability, which in turn determines the dynamic surface tension and hence dynamic interfacial processes, for example foaming. In this study the effect of micellar kinetics on foaming properties of mixtures of sodium dodecyl sulfate (SDS), alkyltrimethylammonium bromides (CnTAB for n = 8, 10, 12, 14 and 16) and long chain alcohols (CnOH, for n = 8, 10, 12, 14 and 16) was investigated. Varying the chain length of the cosurfactant gives rise to the concept of chain length compatibility. In addition, the effect of the foaming methodology on foamability was investigated. The

PAGE 81

66 rate at which foam is produced can result in opposite foaming behavior, which will be explained based upon micellar relaxation time and dynamic surface tension. 5.2 Experimental Procedure 5.2.1 Materials Sodium dodecyl sulfate (99% purity) was supplied by Sigma Chemical Co. (St. Louis, MO). The following chemicals were also used without further purification: alkyltrimethylammonium bromides, CgTAB (Lancaster Inc., Windham, NH), CioTAB (Acros Organics, Pittsburgh, PA), C12TAB, C14TAB (Sigma Chemical Co.) and CieTAB (Aldrich Chemical Company, Milwaukee, WI). CgOH was supplied by Fisher Scientific (Fair Lawn, NJ). CiqOH, and C12OH were supplied by Aldrich Chemical Company, Inc. (Milwaukee, WI). C14OH was supplied by Eastman Kodak Company (Rochester, NY) and CigOH was supplied by Sigma Chemical Co. All solutions were prepared using water that was both deionized and distilled. All experiments were carried out at 22°C. 5.2.2 Relaxation Time Measurement by PressureJump with Electrical Conductivity Detection The slow micellar relaxation time, T2. was measured using a pressure-jump apparatus from Dia-Log Corporation (DUsseldorf, Germany) by means of change in conductivity that results from micelle formation or disintegration. The surfactant solution was pressurized to 100-120 atmospheres and the solution was allowed to reach its new equilibrium state (at high CMC). Subsequently, the pressure was suddenly released to atmospheric (initial CMC) by rupture of a thin metal diaphragm. To reduce the monomer

PAGE 82

67 concentration after the pressure drop, some monomers will enter micelles already present, which can be seen as the fast relaxation process, commonly referred to as A much slower relaxation process is the formation of new micelles (referred to as X2) and can be observed as an exponential drop in electrical conductivity. The relaxation time T2 can then be calculated from the exponential decay in electrical conductivity [Huibers et al., 1996]. 5.2.3 Surface Tension Equilibrium surface tensions were measured from freshly prepared solutions by the Wilhelmy plate method. The platinum plate was always cleaned and heated to a red/orange color with a Bunsen burner before use. 5.2.4 Surface Viscosity A deep-channel surface viscometer [Wasan et al., 1971; Chattopadhyay et al., 1992] was used to measure the surface viscosity of each solution. In the deep-channel surface viscometer the channel walls are stationary concentric cylinders, while the floor moves with a constant angular velocity. To measure the centerline velocity of the air/water interface, a small Teflon particle was placed at the interface, and the time for that particle to make one complete revolution was recorded from visual observation. After measuring the centerline velocity of the air/water interface, the surface viscosity can be calculated using. Wo £ = 7t 8V, h TtVe -1 (5.3)

PAGE 83

68 where e is the surface viscosity, r\ the bulk viscosity of the solution, yo the channel width, Vb the plate rotational speed, V the centerline velocity of the air/water interface and D the ratio of depth to width of the liquid channel. 5.2.5 Foamability by Shaking A 100 mL graduated cylinder was used for the foamability measurements by vigorously shaking. The measurements were carried out by shaking 15 mL of the sample solution for 10 s. The foamability was recorded as the volume of foam produced immediately after shaking. Each solution was tested at least seven times. 5.2.6 Foamability by Air Bubbling A quartz column, 3.5 cm in diameter, was used to acquire the foamability by air bubbling measurements. At the base of the cylinder a single capillary, 2.5 mm in diameter, was used to generate the bubbles. Fifty mL of the sample solution was poured into the column using a long funnel that reached to the bottom to avoid any initial foam formation. The air was then turned on at a constant flow rate of 158.2 cc/min. The foam volume produced after 2 minutes was recorded. The measurement was repeated at least five times for each sample. 5.2.7 Foam Stability A quartz cylinder (3.5 cm diameter) was used for the foam stability measurements. The cylinder contained a single capillary (2.5 mm diameter) to generate

PAGE 84

69 the bubbles. 50 rtiL solution was poured into the cylinder using a long tube reaching the bottom, thereby avoiding contact with the walls, since any solution on the walls could act as an additional supply of surfactant molecules increasing the foam stability. After the foam height reached 50 cm, the airflow was stopped and the time recorded to collapse to half of its initial height. 5.2.8 Dynamic Surface Tension Dynamic surface tensions were measured for the SDS and SDS/C12OH mixtures using the setup described earlier in Chapter 4. All measurements were taken with the capillary tip 1 cm beneath the liquid surface. 5.3 Foaming Properties of SDS/CnTAB Mixtures Foaming properties, such as surface tension, surface viscosity, micellar relaxation time (X2), foam stability and foamability (by shaking) were measured for mixtures of SDS and CnTAB (for n = 8, 10, 12, 14 and 16). A low molecular ratio SDS/CnTAB (21/1) was used to avoid precipitation. The SDS concentration was chosen to be 100 mM, which is considerably higher than the CMC (CMC = 8.3 mM), because at SDS concentrations close to CMC the relaxation time is too small (on the order of milliseconds) to observe a significant change after the addition of CnTAB. Moreover, a high SDS concentration minimizes the effect of possible trace impurities. Figure 5-3 shows a master diagram of all the foaming properties of the SDS/CpTAB solutions. The T2 of pure SDS at 100 mM was determined to be 0.15 s, in perfect agreement with the value measured by Lessner et al. [1981b] and Oh and Shah [1991]. The T2 increases to 1.35 seconds when 5 mM

PAGE 85

70 CnTAB is added to 100 mM SDS solutions, which indicates the formation of more stable micelles as compared to pure SDS micelles. The maximum relaxation time was observed when the chain length of both the surfactant and cosurfactant were the same. The equilibrium surface tension (y) was lowest when the chain length of both surfactants was the same (SDS/C12TAB). Mixtures of anionic and cationic surfactants show a lower surface tension than pure anionic or cationic surfactant solutions alone, because of the electrostatic interaction between the ionic head groups. The lowest surface tension as observed for SDS/C12TAB, however, can be explained on the basis of chain length compatibility of both the surfactant molecules. Another parameter indicating the close packing of molecules at the air/water interface is surface viscosity. Measurements have shown that the maximum surface viscosity is obtained when the foaming agents possessed similar chain lengths. So, apparently the optimum packing of surfactant molecules does not only appear at the air/water interface but in micelles as well. A schematic diagram showing the proposed explanation for the molecular packing at the air/water interface is presented in Figure 5-4. In mixtures of anionic and cationic surfactants there are two factors contributing to the micellar stability. First is the electrostatic interaction between the ionic head groups. This will results in a closer packing as compared to pure SDS micelles. Second is the chain length of the surfactants. Since the electrostatic interaction between the head groups plays a role in all the solutions, only the difference in chain length can be observed.

PAGE 86

71 Figure 5-3. Effect of 5 mM CnTAB on foaming properties of 100 mM SDS solutions. The dashed lines represent 100 mM pure SDS solution.

PAGE 87

72 4 G 0
PAGE 88

73 In addition to the surface measurements described above, foam stability as well as foamability measurements were performed. Foamability was lowest for solutions containing SDS and C12TAB. This can be explained based upon rate of micellar breakup. Micelles must be broken up into monomers for adsorption onto newly created surface of bubbles. Without this process, foam can not be generated. If the micelles in solution are very stable, they can not rapidly provide surfactant monomers to the newly created surface. Hence, foamability is poor. However, if the micelles are relatively unstable, their disintegration provides the surfactant monomers, which can rapidly adsorb to the newly created surface. This enhances foamability of the micellar solutions. A relation between micellar stability and foam stability is less pronounced. Patel et al. [1996] found that stable SDS micelles form very stable foam films. Also, it is known that foam stability increases with bulk viscosity or surface viscosity [Davies and Rideal, 1963; Brown et al., 1953; Shah et al., 1978]. Normally, counterions decrease the repulsion between adjacent surfactant head groups, causing a more condensed film of higher surface viscosity [Chattopadhyay, 1992], thereby increasing foam stability. Another important factor influencing foam stability is the micellar structure inside the thin liquid film, which has been investigated by Nikolov and Wasan [1989a,b]. The stratification of thin liquid films can be explained as a layer by layer thinning of ordered structures of micelles inside the film. This structured phenomenon is affected by micellar effective volume fraction, their stability, interaction and polydispersity.

PAGE 89

74 5.4 Foaming Properties of SDS/CnOH Mixtures In this section, the influence of long chain alcohols (CnOH for n = 8, 10, 12, 14 and 16) on the SDS micellar stability will be discussed. Furthermore, the effect of ndodecanol (C12OH) on the micellar stability was related to foaming properties, such as foamability (by two different methods), dynamic and equilibrium surface tension and surface viscosity. The slow micellar relaxation time T2 of the SDS/long chain alcohol mixtures, which is directly related to the micellar stability, was determined using the pressure-jump technique with electrical conductivity detection. Figure 5-5 shows the slow relaxation time T2 as function of SDS concentration in the presence of 5 mol% CnOH for n = 8, 10, 12, 14 and 16 (higher concentrations of CnOH lead to solubility problems at 22°C). SDS Concentration (mM) Figure 5-5. The effect of 5 mol% CnOH on the slow micellar relaxation time, T2, in SDS solutions.

PAGE 90

75 It is clear that at lower SDS concentrations (< -150 mM) the micellar stability can be maximized by the addition of C,20H, where the alcohol chain length is equal to the surfactant chain length. The presence of shorter or longer chain alcohols also results in an increase in micellar stability. Apparently, the introduction of ion-dipole interactions between the hydroxyl group of the alcohol and the sulfate group of the SDS cause a tighter packing of the micelle, resulting in greater micellar stability. Beyond approximately 150 mM, depending on the alcohol chain length, long chain alcohols other than C12OH start destabilizing micelles, due to mismatching of the chains resulting in a disruption of the molecular packing causing the micelles to destabilize and hence, lower micellar relaxation times are obtained. The effect of 5 mol% CnOH on the micellar stability of 25 and 200 mM SDS solutions is presented in Figure 5-6. At a concentration of 25 mM of SDS or at approximately three times the CMC (8.2 mM), the stabilizing effect of C12OH is much more significant than it is at 200 mM as shown in Figure 5-5 and 5-6. This indicates that the micellar stability of relatively low concentration SDS solutions can be greatly enhanced by the addition of C12OH (at 25 mM almost 230 times). The stabilizing effect of alcohol may be due to the shielding of negative charges of SDS with hydroxyl groups of the alcohol molecules and a stabilizing effect of the hydrocarbon tails, resulting in tightly packed micelles. However, at 200 mM the enhanced micellar stability resulting from the C12OH addition is very small, showing that the contribution of C12OH to already stable micelles is not significant. The dependence of T2 on the concentration of pure SDS solutions was discussed earlier in Chapter 1 [also Oh and Shah, 1993a].

PAGE 91

76 1 \ } \ 1 CeOH CioOH C12OH CuOH CieOH Long Chain Alcohol Figure 5-6. The effect of 5 mol% CnOH on the slow micellar relaxation time, X2, in 25 and 200 mM SDS solutions. The effect of C12OH on the SDS micellar stability was related to the following interfacial properties: surface viscosity, equilibrium surface tension and foamability (by shaking and air bubbling through a single capillary). Figure 5-7B shows the results of the

PAGE 92

ft Figure 5-7. The effect of 5 mol% C12OH on foaming properties of SDS solutions. Different methods of foaming can result in opposite foaming behavior.

PAGE 93

78 surface viscosity measurements in the presence and absence of C12OH. From this graph it becomes clear that molecular packing is an important phenomenon in micelles as well as at the air/water interface (Figure 5-7 A). At low SDS concentration (25 mM) the effect of C12OH is much more pronounced than it is at 200 mM. The alcohol causes the molecules to pack tighter resulting in a high surface viscosity. However, at 200 mM the SDS molecules are already tightly packed, which does not allow the alcohol to increase the surface viscosity significantly. The equilibrium surface tension of 25-200 mM SDS solutions in the presence and absence of 5 mol% C12OH is shown in Figure 5-7C. It is clear that in the presence of C12OH the surface tension is approximately lowered by 7 mN/m due to closer packing of molecules at the air/water interface. The alcohol decreases the molecular area due to the ion-dipole interactions between the SDS head group and the hydroxyl group of the long chain alcohol and a maximum hydrophobic interaction between the carbon chains. Two different methods were applied for the foamability experiments: vigorous hand shaking and air bubbling through a single capillary. Both methods show different results. In the first case, larger foam volumes were obtained for pure SDS solutions than SDS/C12OH mixtures as shown in Figure 5-7D. Especially at low SDS concentrations (25mM) the SDS/C12OH mixtures produced significantly less foam than the pure SDS samples. This can be explained based on the ability of micelles to breakup in order to provide monomers to stabilize newly created interface. Very stable micelles cannot breakup fast enough to augment the flux of monomers necessary to stabilize the new air/water interface, resulting in higher interfacial tensions and hence, foaming ability is low (Figure 5-7D). So, apparently the breakup of micelles is a rate-limiting step in the

PAGE 94

79 supply of monomers to rapidly created air/water interface. At 200 mM the SDS and SDS/C12OH micelles are equally stable (see Figures 5-5 and 5-7 A) and hence, equal foam volumes are produced (Figure 5-7D). Foamability measurements performed by blowing air through a single capillary yielded different results. Figure 5-7E shows that the foam volume of the SDS/C12OH naixtures is consistently higher than pure SDS solutions, irrespective of the SDS concentration. Apparently, by using a single capillary enough time is given to the surfactant molecules to diffuse and stabilize the newly created air/water interface. Therefore a lower dynamic surface tension is obtained. The relation between the dynamic surface tension and the amount of foam created can be given by [Adamson and Gast, 1997], W = } (5.4) where W is the work done, y the interfacial tension at the air/water interface and AA the change in interfacial area. The same relation holds for emulsification processes [Walstra, 1983]. Obviously, when the same amount of work is applied, lower surface tension results in more interfacial area (either by decreasing the bubble size or by increasing foam volume), provided the presence of a surface-active specie. Since the surface tension of the SDS/C12OH mixtures is significantly lower than for pure SDS (Figure 5-7C), the former one will produce more foam using the single capillary foam column. The results of the foamability measurements were evaluated by a more quantitative experiment, namely dynamic surface tension by the maximum bubble pressure method (Chapter 4). The understanding of dynamic surface tension is important in any technological application where a new gas/liquid interface is rapidly being created

PAGE 95

80 in a surfactant solution. In most cases the equilibrium surface tension is never reached and the actual surface tension experienced at the air/water interface is much higher. In this study two solutions of 15 mM SDS and one containing SDS + 5 mol% C12OH were investigated. A concentration of 15 mM was chosen, since at too high surfactant concentrations, deviations from equilibrium surface tension are negligible at the bubble frequencies accessible with the current setup. The dynamic surface tension (DST) as a function of bubble lifetime (reciprocal value of bubble frequency) is given in Figure 5-8. The effect of 5 mol% C12OH is clearly visible. At higher bubble lifetimes, the SDS/C12OH mixture shows a significant lower surface tension than pure SDS, indicating that the equilibrium surface tension of the SDS/C12OH solution is much lower, just as observed earlier in Figure 5-7C. 45 42 39 CO 36 Q 33 30 -•-SDS -IH-SDS + C120H — ^ • . nl I 0.5 1 1.5 Bubble Lifetime (s) Figure 5-8. Dynamic surface tension (Yd) of SDS and SDS/C12OH solutions (15 mM SDS, 5 mol% C12OH).

PAGE 96

81 However, when the bubble lifetime decreases and thus the bubble frequency increases, the curves are getting closer to each other up to a bubble lifetime of approximately 0.15 s where they actually cross. This means that at bubble lifetimes smaller than 0.15 s (frequencies higher than ~7 s"'), the SDS/CnOH mixed micelles are not able to breakup quickly enough to augment the flux of monomers necessary to stabilize the newly formed bubbles as compared to pure SDS. In this region the micellar stability and thus the ability of micelles to breakup fast enough, determines surface tension reduction. In order to show the importance of micellar breakup in the dynamic surface tension measurement, a dimensionless parameter 0 was introduced, as discussed in Chapter 4, equation (4.2). Values between 0 and 1 are a measure for the surfactant concentration at the surface and hence, the stability of micelles, assuming the diffusion time of monomers to be negligible [Oh et al., 1992]. The more stable the micelles, the less monomer flux and hence 9 values closer to 1 will be obtained. Figure 5-9 shows the dimensionless parameter 6 versus the bubble lifetime for SDS and SDS/CnOH solutions of 15 mM SDS and 5 mol% C12OH. In this graph the 6 values are consistently higher for SDS/C12OH than for pure SDS over all bubble lifetimes. Apparently, when accounted for the surface-activity of SDS and SDS/C12OH, the breakup of SDS/C12OH mixed micelles is a rate-limiting step in highspeed dynamic processes. Therefore, it is very important to consider the time scale of generating newly created interfaces in industrial processes, since that determines whether the breakup of micelles and the dynamic surface tension are the dominant factors in foaming, emulsification, wetting and solubilization processes.

PAGE 97

82 0.5 1 Bubble Lifetime (s) 1.5 Figure 5-9. Dimensionless dynamic surface tension (0) of SDS and SDS/CnOH mixtures (15 mM SDS, 5 mol% C12OH). 5.5 Conclusions 1. The SDS micellar stability can be greatly enhanced by the introduction of ion-ion (SDS/CnTAB) or ion-dipole interactions (SDS/CnOH). 2. For mixed solutions of anionic and cationic surfactants or anionic surfactants and long chain alcohols, the foaming properties depend on the chain length of the individual molecules. Li general, the chain length of the surfactant and cosurfactant must be the same to maximize lateral molecular interactions, resulting in minimum surface tension, maximum surface viscosity, maximum micellar stability, minimum foamability (by the shaking method) and maximum foam stability. 3. Long chain alcohols (CnOH for n = 8, 10, 12, 14 and 16) stabilize SDS micelles, up to approximately 150 mM SDS (depending on the carbon chain length of the alcohol)

PAGE 98

83 due to the strong ion-dipole interaction between the negatively charged SDS head group and the hydroxy! group of the alcohol. Beyond this critical concentration the chain length compatibility starts playing a role. Therefore, only C12OH will cause a further increase in micellar stability, whereas the mismatch in chain length between the other alcohols and the SDS results in a disruption of the molecular packing in the micelle, thereby decreasing the stability. 4. The effect of adding C12OH is most pronounced when the stability of pure SDS micelles is very low, i.e., at low SDS concentrations (25 mM). At higher SDS concentrations, the micellar stability of SDS alone increases, which makes the effect of C12OH less pronounced. 5. The effect of micellar stability plays an important role in processes involving a rapid increase in surface area. If enough time is allowed for the interface to form, the dynamic surface tension approaches the equilibrium surface tension and thus more foam is generated (more in case of SDS/C12OH mixtures). However, in very highspeed processes, the micellar stability and thus the time it takes for micelles to breakup determines the rate of adsorption of surfactant molecules and therefore higher surface tensions will be attained for SDS/C12OH solutions. In that case less foam is generated, even though the equilibrium surface tension of the SDS/C12OH system is lower. In conclusion, different methods of foaming can produce opposite results as illustrated by the foamability measurements in this study.

PAGE 99

CHAPTER 6 IMPORTANCE OF MICELLAR STABILITY ON ANTIFOAMING ACTION 6.1 Introduction Foams are used for many different purposes, such as mineral flotation, food processing, purification (foam fractionation), processing of textiles, personal care, enhanced oil recovery and fire fighting [Bikerman, 1973; Prud'homme and Khan, 1996]. In some processes, however, the formation of foam is not desired and its presence can cause serious problems [Kroshwitz, 1993; Garrett, 1993]. Foaming properties, such as surface tension, surface viscosity, micellar stability and film elasticity can be greatly modified by the addition of organic materials. The antifoaming mechanisms by foam breakers can be summarized as follows [Rosen, 1989; Ross, 1958]: 1 . Removing surface-active materials from the air/water interface. Surfactant molecules are removed from the surface by adsorption onto or dissolution in the soil. Finely divided hydrophobic silica particles break the foam by adsorbing surfactant molecules from the bubble surface and carrying them into solution. The presence of certain types of soil in a surfactant solution shows decreased foaming due to this mechanism. 84

PAGE 100

85 2. Replacing surfactant molecules with other molecules at the surface. The surfactant molecules at the bubble surface are replaced by adding non-cohesive molecules of limited solubility in the solution. The tertiary acetylenic glycols, ethyl ethers and isoamyl alcohols break foam in this manner. 3. Converting the surfactant film into a solid brittle film with no elasticity. Calcium salts of long chain fatty acids break foams of SDS or sodium dodecylbenzene sulfate by this mechanism. 4. Reducing the surface viscosity of the film. Tributyl phosphate has a large cross sectional area at the air/water interface. This reduces the cohesive forces between the surfactant molecules and consequently reduces the surface viscosity, which leads to increased drainage of the liquid film. The nature of the electrolyte is of course very important. Electrolytes containing multivalent ions can be anticipated to cause larger effects as compared to those with monovalent ions, especially for ionic surfactants [Ross and Bramfitt, 1957]. The stabilization of foam films containing high surfactant concentrations caused by stratification of long range ordered microstructures in thin films, was shown by Ivan and Dimitrov [1988]. Nikolov and Wasan [1992] have theoretically and experimentally shown that the stepwise thinning of a foam film formed from a SDS micellar solution is governed by a long-range electrostatic repulsion by ionic micelles and a restricted volume effect in the film. Bergeron and Radke [1992] determined the disjoining pressure isotherms for single isolated foam films stabilized by SDS above the CMC. Studies of the oscillatory form of the disjoining pressure permitted quantitative interpretation of this

PAGE 101

86 stepwise thinning behavior. The stabilizing action of liquid crystals in foam systems has been established by Friberg and coworkers [1986]. Finally, Manev and Pugh [1997] showed that the presence of tetraalkylammonium counterions in aqueous foams and thin film lamella stabilized by SDS act either to increase or decrease the foam stability. As discussed in Chapter 5, the stability of micelles significantly influences foamability and foam stability. The stability of micelles, in turn, can be tuned by additives, such as long chain alcohols or oppositely charged surfactants. Also, the difference in chain length between the surfactant and cosurfactant was found to be an important factor in interfacial behavior (chain length compatibility effect). A typical list of antifoaming agents includes 2-ethylhexanol, tributyl phosphate, polydimethylsiloxane (PDMS), amides, mineral oil, fatty acids and their derivatives. Previous investigations by Blute et al. [1994] demonstrated that the foam destabilizing efficiency of tetraalkylammonium bromide salts is equivalent or superior to the traditional antifoaming agents tributyl phosphate and 2-ethylhexanol, which are used in many commercial foam inhibiting formulations. Furthermore, unlike most antifoaming agents, tetraalkylammonium salts are soluble in water. This is very important, since solubility in water allows the addition of foam inhibitor into the solution prior to foaming. With such a wide range of foam inhibiting chemicals it is not surprising that there are so many alternative theories to explain the antifoaming action [Garrett, 1993]. In the present study, an attempt has been made to correlate the antifoaming efficiency of tetraalkylammonium chloride (TCnAC), tributyl phosphate (TBP), and 2ethylhexanol (EH) with SDS micellar stability. The aim is to show that the influence of antifoaming agents on foam stability and foamability is a result of the effect of these

PAGE 102

87 compounds on the molecular packing at the air/water interface and in micelles. Hence, a correlation between (anti)foaming properties and micellar stability can be established. 6.2 Experimental Procedure 6.2.1 Materials Sodium dodecyl sulfate (99% purity) was supplied by Sigma Chemical Co. (St. Louis, MO). Tetramethylammonium chloride (97%) was supplied by Fisher Scientific (Fair Lawn, NJ). Tetraethylammonium chloride, tetrapropylammonium chloride, tetrabutylammonium chloride and 2-ethylhexanol were supplied by Eastman Fine Chemical (Rochester, NY) and used as received. Tributyl phosphate was obtained from Lancaster, Inc. (Windham, NH). The antifoaming agent concentration was varied from 0 to 50 mM in solutions of 150 mM SDS. Sodium perchlorate (99+% purity) was purchased from Acros (Fair Lawn, NJ). Deionized, distilled water was used for all solutions. All experiments were performed at 22°C. 6.2.2 Methods The slow micellar relaxation time Xi was determined by the pressure-jump technique with electrical conductivity detection (Chapter 5). Experimental procedures for the determination of the slow micellar relaxation constant and the measurement of surface tension, surface viscosity, foamability (by shaking) and foam stability were described earlier in Chapter 5.

PAGE 103

88 6.3 Results and Discussion The chemical structures of the antifoaming agents used in this study are shown in Figure 6-1. i. CH3-CH2-CH2-CH2-CH-CH2OH CH2CH3 O II H9C4O — P — OC4H9 u. OC4H9 CnH2n+l ill. H2n+lCn N CnH2n+l CI I CnH2n+l Figure 6-1. Structures of the antifoaming agents used in the present study, (i) 2ethylhexanol (EH); (ii) tributyl phosphate (TBP); (iii) tetraalkylammonium chloride (TCnAC). The effect of 2-ethylhexanol (EH), tributyl phosphate (TBP) and tetrabutylammonium chloride (TBAC) on the slow micellar relaxation time and foamability of 150 mM SDS solutions is shown in Figure 6-2. A surfactant concentration far above CMC (8.3 mM) was chosen, because at too low SDS concentrations, the slow micellar relaxation time is too small (on the order of milliseconds) to observe a significant change in the presence of antifoaming agents. Moreover, far above the CMC of SDS the presence of antifoaming agents results in transparent solutions and hence.

PAGE 104

89 other mechanisms of foam inhibiting action, such as spreading of insoluble droplets or emulsion droplets on the surface can be ignored. Figure 6-2. Effect of antifoaming agents on the slow micellar relaxation time, X2 and foamability of 150 mM SDS solutions. Hq, foam height in the absence of antifoaming agent; Ha, foam height in the presence of antifoaming agent.

PAGE 105

90 It is evident from Figure 6-2 that an increase in antifoaming agent concentration leads to in increase in micellar stability. However, beyond a critical concentration T2 decreases with increasing antifoaming agent concentration. The curves indicate that the maximum observed in X2 and the foamability relate to the antifoaming agent rather than a specific surfactant concentration. The effectiveness in increasing the SDS micellar stability follows the order: TBAC > TBP > EH. TBAC has a more significant stabilizing effect than TBP and EH, indicating the strong electrostatic attractions between the SDS headgroup and the tetrabutylammonium salt. On the other hand, the linking between TBP and SDS and between EH and SDS molecule is primarily due to hydrogen bonding or ion-dipole interactions, which shields the repulsion between the negatively charged SDS headgroups. The decrease in T2 beyond a critical concentration of antifoaming agent is attributed to the destabilization of micelles. In Figure 6-2, Ha/Ho represents the performance of the antifoaming agents at the particular dosage employed for the SDS micellar foaming solutions. Ho and Ha are the foam heights in the absence and presence of antifoaming agents, respectively. In mechanistic terms, HyHo represents the degree to which the foamability has been reduced by the addition of antifoaming agent. Thus, a Ha/Ho value equal to unity indicates the total ineffectiveness of the antifoaming agent and Ha/Ho less than unity indicates the inhibition power of foam formation. The curves show a minimum HJHo value at maximum micellar stability, which is in agreement with the conclusions drawn earlier in Chapter 5 that very stable micelles result in poor foamability. The foamability of micellar solutions depends on the ability of micelles to disintegrate into surfactant monomers, which in turn determines the efficiency of antifoaming agents. So, the decease in foamability of the SDS solutions in the presence of a particular dosage

PAGE 106

91 antifoaming agent can be attributed to the lower flux of monomers coming to the newly created interface. In order to confirm the importance of micellar stability on antifoaming action, the effect of tetrabutylammonium chloride on the SDS micellar stability was extended to tetraalkylammonium chlorides with shorter chain length (TCnAC for n = 1, 2 and 3). Figure 5-3 shows the slow relaxation time X2 of SDS/TCnAC mixtures as a function of TCnAC concentration. 3.0 TCnAC Concentration (mM) Figure 6-3. Effect of tetraalkylammonium chloride (TCnAC, for n = 1, 2, 3 and 4) on the slow micellar relaxation time, in 150 mM SDS solutions. All curves show the appearance of maximum micellar stability depending on the alkyl chain length of the antifoaming agent. Upon increasing the TCnAC concentration, the micellar stability increases. This indicates a greater stability of the micelles in this concentration range as compared to pure SDS, which can be explained by the electrostatic

PAGE 107

92 interaction between the ionic head groups. The TCnAC concentration is large enough to partially screen the charges between the head groups of the SDS molecules, reducing the area/molecule occupied by each head group, yet not sufficient to disrupt the molecular packing of the micelles. Besides ionic interactions (shielding of charges) the so-called salting out effect plays a role. When the monomeric form of a surfactant is salted out by the presence of an electrolyte, micellization is favored and the CMC is decreased [Rosen, 1989]. The effects discussed above are more pronounced when longer alkyl chain lengths are used and therefore the maxima in Figure 5-3 shift to lower concentrations as the alkyl chain length increases from methyl to butyl. Beyond a critical concentration the micellar stability decreases again, which is attributed to the bulkiness of the antifoaming agent molecules. At this point the concentration is large enough to obstruct the close packing of surfactant molecules, thereby decreasing the micellar stability. So, tetrabutylammonium chloride is the most effective antifoaming agent in this study, due to its size and shape. The slow relaxation time, as measured by the pressure-jump technique, was related to interfacial properties such as foamability, foam stability, surface viscosity and surface tension for the system SDS/TC2AC. The results are shown in Figure 6-4. At 5 mM TC2AC, minimum foamability, maximum foam stability, maximum surface viscosity and minimum surface tension were observed. A minimum foamability at 5 mM coincides with maximum micellar stability. This can again be explained based on the ability of micelles to breakup in order to provide monomers to stabilize newly created interface. At 5 mM the micelles are very stable so that they cannot breakup fast enough to augment

PAGE 108

Figure 6-4. Effect of tetraethylammonium chloride on foaming properties of 150 mM SDS solutions.

PAGE 109

94 the flux of monomers necessary to stabilize the new air/water interface. Hence, foaming ability is low. As explained earlier in Chapter 5, counterions decrease the repulsion between adjacent surfactant head groups, causing a more condensed film of higher surface viscosity [Chattopadhyay et al., 1992], thereby increasing foam stability. Another important factor influencing foam stability is the micellar structure inside the thin liquid film, which has been investigated by Nikolov and Wasan [1989a,b]. The stratification of thin liquid films can be explained as a layer by layer thinning of ordered structures of micelles inside the film. This structured phenomenon is affected by micellar effective volume fraction, their stability, interaction and polydispersity. The presence of TC2AC above 5 mM affects the micellar volume fraction by decreasing the thickness of the electrical double layer, which inhibits the formation of layers inside the film. Hence, there is a decrease in foam stability. So, at concentrations higher than 5 mM TC2AC a decrease in the thickness of the electrical double layer together with a decrease in surface viscosity leads to a reduction in foam stability. Finally, a minimum surface tension was observed at 5 mM TC2AC. At concentrations lower than 5 mM the ion-ion interactions between the SDS head groups and the tetraalkylammonium salt cause the surfactant molecules to pack closer, thereby lowering the surface tension. However, beyond 5 mM TCnAC the antifoaming agent starts disrupting the SDS molecular packing at the air/water interface resulting in a larger area per molecule and hence, a higher surface tension is obtained. The effect of antifoaming agents on the molecular packing at interfaces and in micelles is schematically shown in Figure 6-5.

PAGE 110

95 Labile Micelle Stable Micelle Labile Micelle _ Increasing Antifoaming Agent Concentration '^'^^ SDS ^1^^ Antifoaming Agent Figure 6-5. Schematic representation of the micro-structural changes in the SDS micellar packing upon addition of antifoaming agents. 6.4 Comparison between Electrolyte and Antifoaming Agent on Micellar Stability Since it is known that sodium counter ions can serve to increase micellar stability, it is interesting to compare the effect of Na^ and TCnA"^ ions for the system used in this study. Lessner et al. [1981b] measured the effect of NaC104 on the micellar stability in SDS solutions. Their results can be fit to the following equations [Huibers, 1996]: Cg= 0.325 CsDS + CNaci04 (6-1) 8 < Cg < 65 -log (T2) = -4.41 log (Cg) + 7.393 (6-2) 65 < Cg < 1 15 -log (X2) = 3.64 log (Cg) 7.201 ^^'^^

PAGE 111

96 where Cg is the total counterion concentration (niM) and the slow relaxation time (s). The data from Lessner et al. [1981b], the model described in equations 6.1, 2 and 3, and some additional points measured in the present study are plotted in Figure 6-6 for 150 mMSDS. 0) E ic g CO 2 X ^ o DC o 0 Model Lessner etal. [1981b] Present Study 1 \ — j \ 0 10 20 30 40 NaCI04 Concentration (mM) 50 Figure 6-6. Effect of Na"^ counterions on the micellar stability of 150 mM SDS solutions. It is clear that at concentrations higher than 10 mM (corresponding to the maximum T2 for TCiAC in Figure 6-3) , sodium ions still have a stabilizing effect on micelles, while all the TCnAC ions have a distinct destabilizing effect as observed in Figure 6-3. Eventually, at a concentration of about 17 mM, the sodium ions start destabilizing the micelles. Such a trend or behavior is analogous to the electrolyte effect for which beyond a critical value the ionic attraction becomes greater than the double layer repulsion, which leads to destabilization of micelles [Angarska et al., 1998]. It is clear, however, that the

PAGE 112

97 antifoaming agent is much more effective in disrupting the molecular packing at interfaces, due to their bulkiness, as compared to sodium ions. As pointed out by Koczo [1994], it is very important to distinguish between two ways of controlling foam: (1) antifoaming, where the chemical is added in the liquid prior to foam formation in order to prevent foaming or (2) defoaming, where the chemical is added from outside to eliminate an existing foam. In the first case we have shown that at low concentrations the antifoaming agent actually serves to stabilize foams and decreases foam ability, whereas at high concentrations the foamability increases and the stability of the foam decreases. When using TCnAC as defoamer, the foam will break as the defoamer dissolves into the foam. This also shows the importance of the amount of tetraalkylammonium salt added to a solution in order to produce the antifoaming effect. 6.5 Conclusions 1. The addition of antifoaming agents 2-ehtylhexanol (EH) and tributyl phosphate (TBP) to SDS solutions shows two opposing effects, depending on the concentration of antifoaming agent. The antifoaming agent can stabilize the SDS micelles at lower concentrations and in turn act as foam inhibiters. On the other hand, beyond a critical concentration, the antifoaming agent destabilizes the micelles (shorter relaxation time), which improves foamability of SDS solutions. 2. The addition of tetraalkylammonium salts to SDS solutions also shows two opposing effects depending on its concentration. The tetraalkylammonium counterions act to stabilize foams at lower concentrations. At concentrations lower than approximately 5 mM TC2AC (varying from 2 mM for TC4AC to 10 mM for TCiAC), the area per

PAGE 113

98 molecule of the SDS head group is decreased by shielding of charge repulsion, due to the addition of positive ions to the solution, together with a salting out effect. This results in more stable micelles, less foaming ability, higher foam stability, lower surface tension and higher surface viscosity. At higher concentrations (> 5 mM for TC2AC), this stabilizing effect is countered by the increasing substitution of sodium counterions (Na"^) by bulky tetraalkylammonium ions (TC2AC), causing an increase in the area per molecule as the charge shielding is overwhelmed by the steric effect resulting in less stable micelles. The same ion effect on the packing of the head groups occurs at the air/water interface, resulting in increased foamability and surface tension but decreased foam stability and surface viscosity. The maxima and minima in the (dynamic) interfacial properties shift to lower concentrations of TCnAC as the alkyl chain length is increased, due to the decrease in molecular packing at the interface and a stronger salting out effect.

PAGE 114

CHAPTER 7 TAILORING INTRAMICELLAR FORCES TO CONTROL DISPERSION STABILITY 7.1 Introduction Surfactant adsorption on mineral surfaces is a phenomenon of significant importance to many different industrial processes ranging from ore flotation, lubrication and paint technology to enhanced oil recovery [Rosen, 1989]. The process of surfactant aggregation or micellization is well understood in bulk solutions, leading to spherical or cylindrical micelles, bilayers and bicontinuous phases. At interfaces, however, the selfassembly process is influenced by additional surfactant-surface and solvent-surfactant interactions, including surface roughness, heterogeneity and charge behavior. The critical hemimicelle concentration and critical aggregation number have been calculated theoretically [Fuerstenau, 1964, 1975] but the structure and shape of the adsorbed aggregates still remains controversial due to limited direct experimental evidence. Traditionally, adsorption isotherms have been used for describing the adsorption characteristics of surfactants on substrates. Several experimental methods, such as NMR [Soderlind and Stilbs, 1991], calorimetry [Partyka et al., 1989], ellipsometry [Wangnerud and Olofsson, 1992], fluorescence decay [Chandar et al., 1987], neutron reflection [McDermott et al., 1994] and surface force measurements [Kekicheff et al., 1989; Pashley et al., 1988], have been employed to elucidate adsorption mechanisms. Recently, the , structure and shape of those adsorbed aggregates have been revealed by atomic force \ 99

PAGE 115

100 microscopy (AFM) [Manne et al., 1994; 1995]. For example, mica and silica have atomically smooth homogeneous surfaces, which become negatively charged in aqueous solution. For quaternary ammonium surfactants, the resulting structures are consistent with half-cylinders on crystalline hydrophobic silica, full cylinders on mica and spheres on amorphous silica [Manne et al., 1994; Ducker and Wanless, 1999; Pashley and Israelachvili, 1981]. The positioning of surfactant molecules at an interface between water and a substrate (e.g., mica, silica, or sapphire) is the result of an energy minimization for the surfactant, the water, and the surface of the substrate. Some important contributions to this energy are: 1. Interactions between the surfactant and solid substrate. The hydrophilic and/or hydrophobic part of the surfactant must be attracted to the interface. 2. Interactions between the water and substrate. A weak interaction promotes surfactant adsorption. 3. Interactions between the surfactant tail groups and water. This interaction is weaker than the interaction between water molecules, so it promotes surfactant aggregation both in bulk and at surfaces. Considering contributions 1 and 2 together, it is obvious that the surfactant must displace water from the substrate. Thus, a surface with higher energy drives the surfactant to form a more continuous film on the surface [Grant, 1988]. Although much has been published on the structure of surfactants adsorbed on minerals using AFM [Ducker and Wanless, 1999; Rutland and Parker, 1994; Israelachvili, 1994; Manne et al., 1994], very little has been reported on the film stability or technological implications. In the present study, an attempt has been made to correlate the stability of micelles in bulk with the stability of micelles adsorbed on mica and silica.

PAGE 116

101 The central hypothesis of this work is that the strength of the adsorbed surfactant film on particles is an important parameter in the stability of solid/liquid dispersions. Since the ion-electrostatic repulsion is negligible at high electrolyte concentrations, the repulsive force must come from steric stabilization by an adsorbed film. The use of surfactants for nanodispersion stability has the following advantages over conventional polymers [Hiemenz and Rajagopalan, 1997]. 1. Micelles are small as compared to polymers, which is important for nanoparticle stability; 2. Reversible adsorption (washable); 3. Control of adsorption (competitive adsorption). The main assumption in this work is that the micellar structures, as revealed by AFM on flat pieces of mica or silica, are also present on spherical particles. Different adsorption structures are possible as depicted in Figure 7-1. Surfactant molecules can form bilayers, semicylinders, full cylinders, semi spheres and full spheres adsorbed onto the surface. It is proposed that the maximum compressive force of the adsorbed surfactant aggregates is directly related to the stability of solid/liquid dispersions. hi the present study, the adsorption of alkyl trimethylammonium bromides (CnTAB for n = 10, 12, 14 and 16) on mica is studied by means of atomic force microscopy (AFM). hnages as well as force-distance curves are obtained. Consequently, from the force-distance curves, the yield stress and elasticity (Young's modulus) of the adsorbed micellar layer are derived. The stability of the adsorbed micellar layer is then

PAGE 117

102 ABC Figure 7-1. Schematic diagram showing the possible structures of surfactants adsorbed at the solid/liquid interface. A, bilayer formation; B, semi cylindrical micelles or semi-spheres; C, full cylinders or spheres. compared to the stability of micelles in bulk solution by means of the pressure-jump technique with electrical conductivity detection. The results are compared to preliminary dispersion stability studies of silica particles in the presence of salt, SDS and C12TAB to delineate the molecular mechanisms by which surfactants influence dispersion stability. 7.2 Experimental Procedure 7.2.1 Materials Alkyltrimethylammonium bromides CioTAB and C12TAB (both 99%) were supplied by Acros Organics (Fair Lawn, NJ). C14TAB, CieTAB (both 99%) and sodium dodecyl sulfate (SDS, >99%) were supplied by Sigma Chemical Co. (St. Louis, MO).

PAGE 118

103 Silica particles (200 nm) were obtained from Geltech Corporation (Orlando, FL). Deionized water was passed through a Milli-Q water purification system before use. All experiments were performed at 22°C. 7.2.2 Microscopy Images were captured using a Nanoscope HI AFM (Digital Intruments, Santa Barbara, CA) using silicon nitride cantilevers with a spring constant of 0.12 N/m. The image presented was obtained in Tapping Mode with low integral and proportional gain (1.0) and a scan rate of 1 Hz. Force-distance curves were obtained with the same Nanoscope HI AFM instrument in contact mode. A freshly cleaved mica surface was placed in the cell and allowed to equilibrate with the surfactant solution for ten minutes. Manne et al. [1995] found that mica immersed for several days showed no difference in the aggregate structure, indicating that the adsorbate reaches equilibrium configuration in a few minutes. Experiments were performed with at least three freshly prepared samples on different days. 7.2.3 Pressure-Conductivity Measurements The stability of micelles in bulk was measured by a pressure-jump apparatus with electrical conductivity detection from Dia-Log Corporation (Dusseldorf, Germany) as described earlier in Chapter 5. In this case, however, the change in electrical conductivity was measured as function of pressure for solutions of 2 CMC CioTAB and CieTAB instead of the slow micellar relaxation time, Xi-

PAGE 119

104 7.2.4 Turbidity Measurements The turbidity of silica dispersions (0.05 wt% solids at pH 4) was measured by a Hach turbidimeter, model 2100A (Hach Chemical Corporation). The intensity of scattered light at 90° was recorded after 60 minutes in NTU units. 7.3 Results and Discussion 7.3.1 AFM Images When muscovite mica is placed in solution, the ions from the basal surface desorb and exchange with other ions, resulting in a net negative charge. Electrokinetic [Scales et al., 1990] and surface force apparatus measurements [Kekicheff et al., 1989; Claesson et al., 1984] have been used to determine the surface potential of mica in a variety of electrolyte solutions. The concentration of CnTAB was chosen to be two times the CMC (Table 7-1), since at this concentration the adsorption has changed from bilayers to full cyliners on mica and semicylinders on silica [Gao et al, 1987; Gu et al., 1988; 1990; Rupprecht and Gu, 1991; Somasundaran, 1975; Sharma et al., 1996]. Table 71 . CMC values of alkyltrimethylammonium bromides [Rosen, 1989]. Surfactant Mw CMC (g/mol) (mM) CioTAB 280.3 68 C12TAB 308.3 16 CuTAB 336.4 3.6 C16TAB 364.5 0.92

PAGE 120

105 Figure 7-2 and 7-3 present two AFM images of mica in water (pH 6) and mica in the presence of CuTAB respectively. The images are very similar to the images obtained earlier by Manne and Gaub [1995] and more recently by Ducker and Wanless [1999]. Figure 7-2. Top view AFM image of mica immersed in water (pH 6). Size: 50 by 50 nm. The main driving force for adsorption of CnTAB on mica is the electrostatic force between the negatively charged surface and the ionic surfactant head group and the lateral hydrophobic attraction between the hydrophobic chains. The diagonal stripes in Figure 73 represent C14TAB cylindrical micelles in which one axis is approximately equal to one diameter of a micelle (5.2 nm) and the other is very long. Similar images were obtained for the other QTAB series (n = 10, 12 and 16).

PAGE 121

106 50.0 hI-O hm 0.5 nM 25.0 O.O nN Figure 7-3. Top view AFM image of mica immersed in a 2 CMC C14TAB solution (pH 6). Size: 50 by 50 nm. The diagonal stripes represent full cylindrical micelles, with an intermicellar distance of 5.2 nm. 7.3.2 Force-Distance Curves Force-distance curves of CnTAB samples were measured in the fluid cell using contact mode. An example of a force-distance curve for a 2 CMC C12TAB solution on mica is given in Figure 7-4. The initial repulsion at distances larger than about 8 nm is yet unclear. Electrical forces as predicted by the DLVO theory (van der Waals and double layer forces) are approximately 100-1000 times smaller than the forces measured here [Verwey and Overbeek, 1948; Derjaguin and Landau, 1941]. A possible explanation for the force experienced beyond 8 nm could be that of micelles adsorbed onto the AFM tip. As the tip continues to approach the surface, the micellar layer deforms due to compression. The micellar compression results in a steep increase in the interaction

PAGE 122

107 Separation Distance h (nm) Figure 7-4. Force-distance curve for a 2 CMC C^TAB solution showing the different stages as the tip approaches the mica surface. force up to approximately 4 nm where the micelle breaks and the tip jumps into contact with the mica surface. Force-distance curves for the entire series of CnTAB (for n = 10, 12, 14 and 16) are shown in Figure 7-5 and the maximum compressive force as function of alkyl chain length is plotted in Figure 7-6. It appears that the maximum compressive force (Fmax in Figure 7-4) increases linearly as function of alkyl chain length. Initially, octyltrimethylammonium bromide (CsTAB) was also investigated in this study. However, images as well as force-distance curves did not show adsorption of CgTAB on mica. This is more or less confirmed by extrapolating the maximum compressive force in Figure 7-6 to shorter alkyl chain lengths. The maximum compressive force approaches zero, indicating that CgTAB does not adsorb in the same manner as the other surfactants (n>8).

PAGE 123

108 Figure 7-6. Maximum compressive force as function of alkyl chain length for 2 CMC solutions of CnTAB (for n = 10, 12, 14 and 16) on mica.

PAGE 124

109 As discussed in Chapter 5, the stabiUty of sodium dodecyl sulfate (SDS) micelles in bulk solution can be greatly enhanced by the addition of C12TAB or long chain alcohols, due to the introduction of ion-ion or ion-dipole interactions. In order to show that the concept of tailoring micellar stability in bulk solution also holds for adsorbed micelles, a small amount of SDS was added to the 2 CMC CnTAB solutions. Figure 7-7 shows the force-distance curves for C12TAB cylindrical micelles with increasing SDS concentration. The C12TAB concentration was kept constant at twice the CMC. 13 1 h (nm) Figure 7-7. Force-distance curves for a 2 CMC C12TAB solution on mica containing increasing amounts of SDS (pH 6). The chain lengths between the two oppositely charged surfactants should be kept the same in order to maximize the lateral interactions, attributed to the chain length compatibility effect (see Chapter 5). Figure 7-8 presents the maximum compressive force as function of SDS concentration. The maximum compressive force increases linearly

PAGE 125

110 with SDS concentration and levels off beyond approximately 20 mol% of SDS. The addition of more than 20 mol% SDS to C12TAB solutions results in precipitation and hence could not be measured. 0 5 10 15 20 25 Mol% SDS Figure 7-8. Maximum compressive force as function of SDS concentration for a 2 CMC C12TAB solution on mica. It is interesting to note that the electrical force should decrease as the net charge on the micelle decreases due to the addition of SDS. Experimental results however, show an opposite effect, proving the steric origin of the repulsive force. Hence, the maximum compressive force measured in this study represents the stability of the micellar layer and thus the concept of tailoring micellar stability in bulk seems to also hold for micelles adsorbed onto solid/liquid interfaces. The effect of chain length on micellar stability in bulk was investigated by the pressure-jump technique with electrical conductivity detection, which was discussed

PAGE 126

Ill earlier in Chapter 5. However, in this case the change in electrical conductivity as function of pressure was measured instead of Xi. Figure 7-9 shows the effect of pressure on the electrical conductivity of 2 CMC solutions of C„TAB (for n = 10 and 16). 35 1 0 ' \ ^ ' ' 0 25 50 75 100 125 Pressure (atm) Figure 7-9 Change in electrical conductivity due to micelle breakup in the pressurejump experiment. The change in electrical conductivity is proportional to the signal represented on the y-axis. It is clear that CieTAB micelles are more stable as compared to CioTAB micelles, consistent with the force-distance curves in Figure 7-5. For CieTAB higher pressures are required to obtain the same change in conductivity as compared to CioTAB, meaning that the micelles are more stable. Aniansson and coworkers [1976] measured the stability of a series of sodium alkylsulfates and found an increase in micellar stability as the chain length was increased from Cio to Ci6, which is in agreement with the observations in this study. The longer the hydrocarbon chain length, the stronger the hydrophobic forces.

PAGE 127

112 resulting in more stable micelles. In conclusion, the results indicate that the stability of micelles adsorbed on a solid/liquid interface is consistent with the stability of micelles in bulk and hence, the stability of the adsorbed layer can be controlled by additives, such as oppositely charged surfactants, long chain alcohols or the presence of electrolyte. 7.3.3 Relation between Micellar Stability and Dispersion Stability As mentioned earlier, the central hypothesis of this work is that the strength of the adsorbed surfactant film on particles (i.e., the stability of the adsorbed micellar layer) is an important parameter in the stability of dispersions. At high ionic strength, ionelectrostatic forces don't play a role and steric repulsion of the adsorbed micelles remains the only factor that accounts for the stability of the dispersion. Therefore, dispersion stability experiments were performed with 0.05wt%, 200 nm, silica particles in solutions containing C12TAB in the presence and absence of (100 mM) NaCl or a varying amount of SDS at pH 4. A pH of 4 (close to the lEP between 2 and 3) was chosen to assure an unstable dispersion in the presence of NaCl. The pH was adjusted with HCl. Figure 7-10, shows the turbidity of silica dispersions after 60 minutes as well as the maximum compressive force as function of C12TAB concentration. A high turbidity indicates very stable dispersions, whereas very low values indicate unstable dispersions. Initially, the addition of C12TAB destabilizes the dispersion due to the adsorption of C12TAB onto the silica particles. At low concentrations of C12TAB (< 2 mM), adsorption is driven by electrostatic, attractive forces between the cationic surfactant and the oppositely charged silica surface. This results in monolayer coverage where the hydrophobic groups point towards the water [Sharma et al., 1996].

PAGE 128

113 Concentration of C12TAB (mM) Figure 7-10. Turbidity of silica dispersions and the maximum compressive force vs. concentration of C12TAB at pH 4 after 60 min. The attractive forces between the now hydrophobic silica results in a non-stable silica dispersion. With an increase in surfactant concentration, surface aggregation around the initial adsorption site occurs, which results in bilayer formation [Bijsterbosch, 1974]. Further increase in surfactant concentration results in the formation of semicylindrical micelles [Somasundaran, 1975; Manne and Gaub, 1995]. Above 5 mM of CnTAB, both the turbidity and the maximum compressive force increase. Beyond approximately 20 mM of C12TAB the silica dispersion becomes stable again. It is clear that the maximum compressive force follows the same trend as the stability of the dispersion and thus the strength of the adsorbed micellar layer seems to be directly related to the stability of the silica dispersion.

PAGE 129

114 Figure 7-11 shows the turbidity of the same system as in Figure 7-10 in the presence of 100 mM NaCl. Adding NaCl compresses the electrical double layer and therefore the dispersion becomes unstable. Interestingly, the presence of NaCl enhances the formation of surface aggregates, causing the plateau to shift to lower C12TAB concentrations (from approximately 20 to 10 mM C12TAB). 0 5 10 15 20 25 30 35 40 Concentration of C12TAB {mM) Figure 7-11. Turbidity of silica dispersions and the maximum compressive force vs. concentration of C12TAB in the presence of 100 mM NaCl at pH 4. This shift to the left is based upon the hydrophobic or salting out effect [Rosen, 1989]. The association of surfactant molecules to form aggregates on surfaces or in bulk is an entropically favored process due to the removal of hydrocarbons from the aqueous environment. This phenomenon, referred to as the hydrophobic effect, is analogous to micelle formation in solution [Rosen, 1989]. The presence of saU in the system reduces

PAGE 130

115 the entropy of the water molecules even further and hence, micellization is favored and surfactant adsorption enhanced, causing the dispersion to become more stable at lower C12TAB concentrations. Another factor causing the plateau in Figure 7-11 to shift to lower concentrations of C12TAB is the ability of NaCl to screen the charges between the ionic head groups of the surfactant. The influence of SDS on the stability of the dispersion in the presence of 5 mM C12TAB and 100 mM NaCl (pH 4) is presented in Figure 7-12. The addition of SDS stabilizes the mixed surfactant layer as shown in Figure 7-7 and 7-8. It appears that even a very small amount of SDS (0.007 mM) is enough to stabilize the silica dispersion again. Concentration of SDS (mM) Figure 71 2. Turbidity and the maximum compressive force vs. concentration of SDS in the presence of 5 mM C12TAB, 100 mM NaCl at pH 4.

PAGE 131

116 The ratio of C12TAB to SDS is approximately 700:1 for achieving maximum stability. More experimental work is required, however, it is remarkable that such as very low concentration strikingly enhances the stability of dispersions. In conclusion, the stability the silica dispersions appears to be directly related to the compressive strength of the adsorbed micellar layer, which can be tuned the same way as micellar stability in bulk solution, i.e., chain length or by additives, such as oppositely charged surfactants, electrolyte or long chain alcohols. 7.3.4 Mechanical Properties of the Adsorbed Micellar Laver It is interesting to calculate the mechanical strength of the adsorbed surfactant layer with that of polymers used as dispersants. Force-distance curves allow for the calculation of some important mechanical properties of the adsorbed CnTAB micelles, such as elasticity of the adsorbed layer and yield stress. However, first the dimensions of the AFM tip need to be known. The size of the tip and its exact profile can be determined by scanning electron microscopy (SEM), however, two problems arise in this procedure. First is the resolution. The resolution is usually worse than 1 nm, which is on the order of the size of the tip. Second is that the SEM cannot be run in-situ during AFM measurements. For these reasons another method was applied to determine the size of the AFM tip. The difference between the actual profile and the ideal profile is related to a finite, non-zero size of the tip. This is schematically illustrated in Figure 7-13. The radius of the tip can the be calculated from the geometry,

PAGE 132

117 where R^ic is the radius of cylindrical micelles (5.2/2 nm) and h is the difference between the cylindrical profile and the experimentally obtained profile. Using experimental values ofh = 0.3 nm and Rmic= 2.6 nm, equation (7.1) yields Rup = 8.8 nm. Figure 7-13. Calculation of the actual tip radius from the micellar profile obtained by the AFM. In order to calculate the elasticity of the adsorbed micelles, the micellar layer was considered to be flat as a first approximation. Figure 7-14 represents the standard geometry for a Hertzian calculation of the interaction between a hard sphere and a soft plate [Holland, 1964]. Figure 7-14. Interaction parameters between the AFM tip and the adsorbed layer of micelles for the calculation of the elasticity.

PAGE 133

118 In this study, the radius of the sphere is equal to the radius of the tip, Rup. The contact zone Rc can then be calculated according to, Rc=^J^np"-(R
PAGE 134

119 Wm\ Note that for low-density polyethylene Y = 6 20x10^ Wm\ i.e., the same order of magnitude as for the adsorbed micellar layer. 7.4 Conclusions 1. CnTAB molecules (for n = 10, 12, 14 and 16) form full cylindrical micelles on mica surfaces, even though the bulk concentration is relatively low (2 CMC). 2. The stability (i.e., max. compressive force) of CpTAB cylindrical micelles increases linearly with alkyl chain length. 3. The stability of CnTAB micelles can be greatly enhanced by the addition of SDS due to reduction of intramicellar ion-ion repsulsion, which suggests that the stability of adsorbed micelles shows a good correlation with micellar stability in bulk as shown by the pressure-electrical conductivity measurements. 4. The results suggest that ion-ion and hydrophobic interactions are possible molecular mechanisms by which surfactants influence dispersion stability. 5. Preliminary experiments have shown that unstable silica dispersions in the presence of 100 mM NaCl can be stabilized by the addition of C12TAB. The addition of SDS lowers the minimum concentration of C12TAB required to stabilize the silica dispersion. An extremely small amount of SDS (SDS/C,2TAB, 1/700) strikingly improves the stability of the adsorbed surfactant film and hence the stability of the dispersion. 6. The yield stress and elasticity of the adsorbed micellar layer appears to be of the same order of magnitude as that for low-density polyethylene.

PAGE 135

CHAPTER 8 SUMMARY AND RECOMMENDATIONS FOR FUTURE WORK H. 1 Measurement of CMC of Nonionic S urfactants The importance of the method to be employed for the measurement of critical micelle concentration (CMC) has not been brought out in the literature satisfactorily. A large difference is often observed between the CMC values of (technical grade) nonionic surfactants determined by different methods. Furthermore, for nonionic surfactants, a clear break in the surface tension (y) vs. concentration (logC) curve is usually not obtained. In this study the effect of the distribution pattern of ethoxylation on two commonly used methods for the determination of CMC was investigated, namely dye micellization and the surface tension methods. The surface tension method (Wilhelmy plate) for the determination of the CMC of commercial (technical grade) nonionic surfactants was found to be very sensitive to the presence of molecular species with higher surface-activity. In the presence of high surface-active impurities, the air/liquid interface gets saturated at concentrations much below the true CMC leading to a wrong interpretation of the break in the y vs. logC curve. The micellized dye method gives reliable results for micelle formation in bulk solution, even in the presence of such impurities. Foam fractionation of a solution of technical grade nonionic surfactant can remove the species with higher surface-activity preferentially. The CMC values, as 120

PAGE 136

121 measured by the surface tension and dye micellization method, are in close agreement with each other after foam fractionation. « 7 Micellar Relaxation Time of Nonionic S urfactants The association of many classes of surface active molecules into micellar aggregates is a well-known phenomenon. Micelles are often drawn as static structures of spherical aggregates of oriented molecules. However, micelles are in dynamic equilibrium with surfactant monomers in the bulk solution constantly being exchanged with the surfactant molecules in the micelles. Additionally, the micelles themselves are continuously disintegrating and reforming. The first process is a fast relaxation process typically referred to as Ti. The latter is a slow relaxation process with relaxation time Xi. Thus, T2 represents the entire process of the formation or disintegration of a micelle. The slow relaxation time is directly correlated with the average lifetime of a micelle, and hence the molecular packing in the micelle, which in turn relates to the stability of a micelle. It was shown earlier by Shah and coworkers that the stability of sodium dodecyl sulfate (SDS) micelles plays an important role in various technological processes involving an increase in interfacial area, such as foaming, wetting, emulsification, solubilization and detergency. The slow relaxation time of SDS micelles, as measured by pressure-jump and temperature-jump techniques was in the range of 10"^ to lO' seconds depending on the surfactant concentration. A maximum relaxation time and thus a maximum micellar stability was found at 200 mM SDS, corresponding to the least foaming, largest bubble size, longest wetting time of textile, largest emulsion droplet size

PAGE 137

122 and the most rapid solubilization of oil. These results are explained in terms of the flux of surfactant monomers from the bulk to the interface, which determines the dynamic surface tension. The more stable micelles lead to less monomer flux and hence to a higher dynamic surface tension. As the SDS concentration increases, the micelles become more rigid and stable due to the decrease in intermicellar distance. The smaller the intermicellar distance, the larger the Coulombic repulsive forces between the micelles leading to enhanced stability of micelles (presumably by increased counterion binding to the micelles). In this study a method has been developed using stopped-flow and pressure-jump with optical detection to determine the slow micellar relaxation time of nonionic surfactants. The results show relaxation times T2 in the range of seconds for Triton X-100 to minutes for polyoxyethylene alkyl ethers. The slow relaxation times are much longer for nonionic surfactants than for ionic surfactants, because of the absence of ionic repulsion between the head groups. The observed relaxation time T2 was related to dynamic surface tension and foaming experiments. It appears that micellar stability and thus the micellar breakup time of nonionic surfactants is a key factor in controlling technological processes involving a rapid increase in interfacial area, for example foaming. First, the available monomers adsorb onto the freshly created interface. Then, additional monomers must be provided by the breakup of micelles. Especially when the free monomer concentration is low, as indicated by a low CMC, the micellar breakup time is a rate-limiting step in the supply of monomers, which is the case for many nonionic surfactant solutions. A slow breakup of micelles (i.e. long relaxation time, X2)

PAGE 138

123 corresponds to high dynamic surface tensions (low foamability), whereas very labile micelles result in lower dynamic surface tensions (high foamability). S 3 Tailoring Micellar Stability to Control Foaming and Antifoaming Action Relaxation time data of ionic and nonionic surfactant solutions can predict the performance of a given surfactant solution. Moreover, the results suggest that appropriate micelles with specific stability or Xi can be designed by controlling the surfactant structure, concentration and physicochemical conditions, as well as by mixing anionic/cationic or ionic/nonionic surfactants for a desired technological application. The micellar stability of sodium dodecyl sulfate (SDS) micelles can be greatly enhanced by the addition of long chain alcohols (e.g. 1-dodecanol, C12OH) or cationic surfactants (e.g. alkyltrimethylammonium bromide). For mixed solutions of anionic and cationic surfactants or anionic surfactants and long chain alcohols, the foaming properties depend on the chain length of the individual molecules, hi general, the chain length of the surfactant and cosurfactant must be the same to maximize lateral molecular interactions, resulting in minimum surface tension, maximum surface viscosity, maximum micellar stability, minimum foamability and maximum foam stability. The effect of C12OH on SDS micellar stability is most pronounced when the stability of pure SDS micelles is very low, i.e. at low SDS concentrations (25 mM). At higher SDS concentrations, the micellar stability of SDS alone increases, which makes the effect of C12OH less pronounced. The effect of micellar stability plays an important role in processes involving a rapid increase in surface area. For the first time it has been shown that the foam volume generated by a surfactant solution depends on the method of producing foam. If enough time is allowed

PAGE 139

124 for the interface to form, the dynamic surface tension approaches the equilibrium surface tension and thus more foam is generated. However, in very high-speed foam generation processes, the micellar stability and thus the time it takes for micelles to breakup determines the rate of adsorption of surfactant molecules to the interface, resulting in higher surface tensions. In the latter case, less foam is generated, even though the equilibrium surface tension of the system is lower. Thus, different methods of foaming can produce opposite results, depending on the dynamic surface tension and micellar stability as demonstrated by the foamability measurements in this study. In the present study an attempt has been made to correlate the antifoaming efficiency of tetraalkylammonium chloride (TCAC), tributyl phosphate (TBP) and 2ethylhexanol (EH), with SDS micellar stability. The aim was to show that the influence of antifoaming agents on foam stability and foamability is a result of the effect of these compounds on the molecular packing at the air/water interface and in micelles. A direct correlation between (anti)foaming properties and micellar stability was demonstrated. The addition of EH, TBP and TCnAC to SDS solutions shows two opposing effects, depending on the concentration of antifoaming agent. The antifoaming agent can stabilize the SDS micelles at lower concentrations and in turn act as foam inhibiters. On the other hand, beyond a critical concentration, the antifoaming agent destabilizes the micelles (shorter relaxation time), which improves foamability of SDS solutions. The critical concentration shifts to lower concentrations of TCnAC as the alkyl chain length is increased, due to the decrease in molecular packing at the interface and a stronger salting out effect.

PAGE 140

125 S.4 Tailoring Tntramicellar Forces to C ontrol Dispersion Stability The understanding of the stability of niicelles in bulk solution can be used to design the strength of surfactant films adsorbed onto solid/liquid interfaces. In this study, one of the first applications of this concept, viz., dispersion stability is presented. The central hypothesis of this work is that the strength (or yield stress) of the adsorbed surfactant film on particles is an important parameter in the stability of dispersions. Alkyltrimethylammonium bromides adsorb as full or semicylinders on surfaces, such as mica or silica, even at relatively low bulk concentration (2 CMC). The stability of the adsorbed layer, which was measured by the maximum compressive force, increases linearly with alkyl chain length and can be greatly enhanced by the addition of oppositely charged surfactant, due to reduction of intramicellar repulsion. The results show a good correlation with micellar stability in bulk as shown by the pressure-electrical conductivity measurements. Preliminary dispersion stability measurements have shown that unstable silica dispersions can be stabilized by the addition of C12TAB. The addition of SDS lowers the minimum concentration of C12TAB required to stabilize the silica dispersion. We envision that tuning the compressive barrier required to stabilize solid/liquid dispersions might become a useful way of stabilizing high ionic strength dispersions. 8.5 Recommendations for Future Work The understanding of the effect of micellar relaxation time on various technological processes, such as foaming, antifoaming, wetting, emulsification, solubilization and detergency is not yet complete. Recommendations for further research and development are given below.

PAGE 141

126 The development of a pressure-jump apparatus with electrical and optical detection methods allows for the measurement of the slow relaxation time of mixed ionic/nonionic surfactant systems. The relaxation times can be measured simultaneously using both detection techniques. Polymer-surfactant systems are an upcoming area of interest. The stability of surfactant-polymer aggregates can be measured by the technique developed in this dissertation and related to dynamic interfacial processes, for example foaming. The long relaxation times obtained for the nonionic surfactants investigated in this study can be compared to the Aniansson and Wall theory of step-wise micellization using theoretical computer models (Appendix A). The dynamic surface tension setup could be improved in order to measure the maximum bubble pressure at higher bubble frequencies. This allows for the evaluation of the slow micellar relaxation time T2 at higher surfactant concentrations as well as the correlation between the dynamic surface tension data and available models on surfactant adsorption from bulk solution. The effect of cationic surfactants on the stability of nanoparticle dispersions needs further investigation. The preliminary results presented in this dissertation could be extended to, for example, high ionic strength suspensions used in Chemical Mechanical Polishing (CMP) processes.

PAGE 142

APPENDIX A HISTORICAL PERSPECTIVE ON MICELLAR KINETICS The study of the kinetics of micelUzation reached its peak in the 1970's. However, as early as 1965, Mijnlieff and Ditmarsch [1965] reported pressure-jump studies on sodium dodecyl sulfate and sodium tetradecyl sulfate. From that point on all attention was focused on the theoretical implementation of the step-wise formation and disintegration of micelles. The primary breakthrough was the discovery of the existence of two relaxation processes (fast and slow) [Lang et al., 1975] and the development of a model for the kinetic process of micelle formation and disintegration by Aniansson and coworkers [1974; 1975; 1976]. This model was supplemented by Lessner et al. [1981a,b] and Hall [1981]. The free surfactant monomers are assumed to be completely dissociated and the size distribution of the aggregates in a surfactant solution is assumed to have the shape schematically shown in Figure A-1, where C(AJ denotes the total concentration of aggregates containing n monomers. Furthermore, the size distribution curve is a function of temperature, pressure and concentration. The association and dissociation of micelles is considered to be a stepwise process involving the entry and departure of one monomer at a time from the micelle. 127

PAGE 143

128 >• Aggregation number n Figure A1 . Typical size distribution curve of aggregates in a micellar solution, with mean aggregation number, n Thus, there is a series of equilibria. Ai + K. =! A„ /i = 2,3,4,.... (A.1) where A„ denotes an aggregate containing n monomers, and k/ and k„" are the forward and reverse rate constants for a given step. Assuming the aggregation number n to be a continuous variable and applying a treatment analogous to heat conduction, Aniansson and coworkers found the following expression for the fast relaxation process Xi, 1 ' 1 + — a n J with a = C-CMC CMC (A.2) where a is the half-width of the distribution curve of micellar sizes (assumed to be Gaussian, Figure A-1), k' is the stepwise dissociation rate constant, which is assumed to be independent of n in the micellar region, C the total surfactant concentration and CMC

PAGE 144

129 the critical micelle concentration. Equation (A.2) predicts a linear relationship between 1/Ti and the total surfactant concentration, in agreement with experiments [Aniansson, 1976]. It is obvious that as the total surfactant concentration increases, the number of micelles increases, resulting in a decrease in intermicellar distance. Hence, the time required for a monomer to collide with a micelle is shorter at higher surfactant concentration. The magnitude of ii depends on the length of the hydrocarbon tail of the surfactant: the shorter the chain length, the faster the relaxation time (since micelles are more loosely packed structures for shorter chain surfactants). The expression for the slow relaxation time T2 can be simplified to [Aniansson, 1976] 1 n T2 CMC*R 2 f ^2 1 + — a V « J (A.3) where is a term which may be visualized as the resistance to flow through the critical region (i.e. the narrow passage in Figure A-1 going from monomers to micelles) and is given by "2 1 n=n,+l (A.4) where n is the aggregation number of some aggregate and A„ is the equilibrium concentration of aggregates of order n. The dependence of I/T2 upon ionic strength, concentration and temperature has been interpreted in terms of their effect upon R. Interestingly, the two relaxation times can be used to calculate two important parameters of a micellar solution; (1) the residence time of a surfactant molecule in a micelle and (2) the average Ufetime or stability of micelles [Attwood, 1983; Muller, 1979; Gormally,

PAGE 145

130 1980; Lang, 1987]. The residence time of a surfactant monomer in micelles is equal to nAwhere n is the mean aggregation number (n in Figure A-1) and k the dissociation rate constant of a monomer from a micelle. The average micellar lifetime Tm is given by [Aniansson, 1985], T^ = ^2—^-^^2 (A.5) 1 + — a n When the concentration of surfactant is much greater than the CMC, the micellar lifetime is approximately equal to nXiAlthough derived for nonionic surfactants, the results by Aniansson and coworkers were mainly applied to ionic systems. The agreement between theory and experiment was, in general, satisfactory. Equation (A.5) predicts that X2 should increase with concentration of a surfactant. However, it has been reported that for some ionic surfactant systems, Xi first increases, then passes through a maximum and then decreases again [Lang, 1975; Lessner 1981a,b; Innoue, 1980]. Kahlweit concluded that in ionic systems at high concentration, the reaction path for the formation of micelles must be different than at low concentration. Therefore, the following model was proposed explaining the occurrence of a maximum in Xj. Ionic micelles, including submicellar aggregates, can be considered as charged particles. At low counterion concentration, these particles are stable with respect to coagulation due to the repulsive electrostatic forces. Consequently they can grow by stepwise incorporation of monomers according to, N,,i + Ni .i==t N,. (A.6) where / is the aggregation number. With increasing counterion concentration, however, the electric double layer around each particle becomes increasingly compressed, so that

PAGE 146

131 the attractive dispersion forces (van der Waals forces) lead to a reversible coagulation according to, N, + J^=^ N, k^l = i (A.7) where it and / are classes of submicellar aggregates. Kahlweit and coworkers [1982] then represented the reaction path of the formation of micelles by two parallel resistors Ri and R2. At low counterion concentration, R2 is very high due to electrostatic repulsion between submicellar aggregates. Therefore Ri determines the rate of micelle formation, according to equation (A.6). By increasing the counterion concentration, Ri becomes very high so that R2 determines the rate of micelle formation and the reaction follows equation (A.7). The model also predicts a shift of the maximum T2 to lower surfactant concentrations, by the addition of electrolyte. For nonionic systems both reaction paths compete right from the CMC on. Kahlweit then compared the results with predictions of the DLVO theory, which were in good agreement with the experiments. Although many other kinetic treatments have been proposed, that of Aniansson and coworkers is possibly the most comprehensive today.

PAGE 147

APPENDIX B MATHEMATICAL PROOF OF CONSTANT CHARACTERISTIC DIFFUSION TIME FOR LARGE BUBBLE RADH As shown in Chapter 4, the number of monomers (Ni) needed to saturate the bubble surface of radius Ri can be calculated according to, (B.l) where A is the area per molecule of the surfactant at the air/water interface. The number of molecules per mL solution (N2) at the CMC (in mM) is given by, where Naws is the Avogadro number. The volume around the bubble, which contains Ni monomers, is given by N1/N2 mL. Therefore, a shell of radius R2 (containing Ni monomers) can be calculated by. Substituting for R2 = Ri + AR in (B.3) yields. 3 132

PAGE 148

133 equals, 3 (B.5) Binomial expansion of (1+AR/Rif yields. AR ^AR^ AR^ 1 +3 — + 3 (B.6) For large Ri, the last two terms in (B.6) are negligible, and hence (B.5) can be written as, —^ = 47lR^AR (B.7) Substituting (B.l) and (B.2) in (B.7) yields, AR = 1 (B.8) A * CMC * N Avog Thus, AR (= R2-R1), and hence Lj (Ld=AR/2) and td (L//2D) are constant for large bubble radii. So, tj only depends on the area per molecule and the CMC. For surfactants with higher CMC, AR is smaller and hence td reaches a plateau at smaller bubble radii, as shown in Figures 4-7 and 4-8.

PAGE 149

REFERENCES Adamson, A.W. and Gast, A.P., "Physical Chemistry of Surfaces," Wiley (6'' ed.). New York, 1997. Alexandridis, P., Athanassiou, V., Fukuda, S. and Hatton, T.A., Langmuir 10, 2604 (1994). Angarska, J.K., Tachev, K.D., Kralchevsky, P.A., Mehreteab, A. and Broze, G., T Colloid Interface Sci . 200, 3 1998. Aniansson, E.A.G., Progr. Colloid Polvm. Sci. 70, 2 (1985). Aniansson, E.A.G. and Wall, S.N., J. Phvs. Chem. 78, 1024 (1974). Aniansson, E.A.G. and Wall, S.N., J. Phvs. Chem. 79, 857 (1975). Aniansson, E.A.G., Wall, S.N., Almgren, M., Hoffmann,, H., Kielmann, I., Ulbricht, W., Zana, R., Lang, J. and Tondre, C, J. Phvs. Chem. 80, 905 (1976). Ashby, M.F. and Jones, D.R.H., "Engineering Materials 1, Introduction to their Properties and Materials," Pergamon Press, Oxford, England, 1980. Attwood, D. and Florence, A.T., "Surfactant Systems," Chapman and Hall, Ltd, London, 1983, Chapter 3. Bergeron, V. and Radke, C.J., Langmuir 8, 3020 (1992). Bergethon, P.R. and Simons, E.R., "Biophysical Chemistry: Molecules to Membranes," SpringerVerlag, New York, 1990. Bijsterbosch, B.H., T. Colloid Interface Sci . 47, 186 (1974). Bikerman, J.J., "Foams," SpringerVerlag, New York, 1973. Blute, I., Jansson, M., Oh, S.G. and Shah, D.O., J. Am. Oil. Chem. Soc. 71, 41 (1994). Borchardt, J.K. and Yates, C.W., J. Am. Oil. Chem. Soc . 70, 47 (1993). 134

PAGE 150

135 Brown, A.G., Thuman, W.C. and McBain, J.W., T. Colloid Interface Sci. 8, 491 (1953). Chandar, P., Somasundaran, P. and Turro, N.J., J. Colloid Interface Sci. 117, 31 (1987). Chang, C.H., Wang, N.H.L. and Franses, E.I., Colloids Surf. 62, 321 (1992). Chattopadhyay, A.K., Ghaicha, L., Oh, S.G. and Shah, D.O., J. Phys. Chem. 96, 6509 (1992). Chiu, H.L. and Huang, S.D., Se p. Sci. Technol . 26, 73 (1991). Chiu, Y.C. and Wang, S.J., Colloids Surf. 48, 297 (1990). Claesson, P., Horn, R.G. and Pashley, R.M., J Colloid hiterface Sci. 100, 250 (1984). Davies, J.T. and Rideal, E.K., "Interfacial Phenomena," Academic Press (2"*^ ed.), New York, 1963. Derjaguin, B.V. and Landau, L.D., J. Exp. Theory Phvsics of USSR U, 802 (1941). Ducker, W.A. and Wanless, E.J.. Langmuir 15, 160 (1999). Eastoe, J., Dalton, J., Roguega, P., Crooks, E., Pitt, A.R. and Simister, E. J., J. Colloid hiterface Sci . 188, 423 (1997). Eastoe, J., Dalton, J., Roguega, P., Sharpe, D., Dong, J. and Webster, J.R.P., Langmuir 12, 2706(1996). van Ee, J.H., Misset, O. and Baas, E.J. "Enzymes in Detergency," Marcel Dekker, New York, 1997. Ekwall, P., In "Chemistry, Physics and Applications of Surface Active Substances," J.Th.G. Overbeek (Ed.) Gordon and Breach Science Publisher, New York, 1967. Elving, P.J., "Treatise on Analytical Chemistry," Part I, Vol. 5, Wiley (2"'' ed.). New York, 1982. Engels, Th., von Rybinski, W. and Schmiedel, P., Proer. Colloi d Polvm. Sci. iU, 1 17 (1998). Everett, D.H., "Basic Principles of Colloid Science," Royal Society of Chemistry, London, 1988. Fainerman, V.B., Colloids and Surf. 62, 333 (1992).

PAGE 151

136 Fainerman, V.B. and Makievski, A.V., Colloids Surf. 69, 249 (1993). Fainerman, V.B., Miller, R. and Joos, P., Colloid Polym. Sci. 272, 731 (1994). Faucompre, B. and Lindmann, B., .1. Phvs. Chem. 91, 383 (1987). Fendler, J., "Catalysis in Micellar and Macromolecular Systems," Academic, New York, 1975. Filippov, L.K., T Cnllnid Interface Sci. 163, 49 (1994a). Filippov, L.K., T Colloid Interface Sci. 164, 471 (1994b). Friberg, S.E., Blute, I., Kuineda, H. and Stenius, P. Langmuir 2, 659 (1986). Frindi, M., Michels, B. and Zana, R., J. Phvs. Chem . 98, 6607 (1994). Fruhner, H. and Czichocki, G., Tenside. Surf actants. Detergents 33, 310 (1996). Fuerstenau, D.W., Healy, T.W. and Somasundaran, P., Trans. AME 229, 321 (1964). Fuerstenau, D.W. and Wakamatsu, T., Faradav Disc uss. Chem. Soc. 59, 157 (1975). Gao, Y., Du, J. and Gu, T T Chem. Soc. Faradav Trans. 1 83, 2671 (1987). Garrett, P.R. and Ward, D.R., J. Colloid Interface Sci. 132, 475 (1989). Garrett, P.R. in "Defoaming Theory and Industrial Applications," Garrett, P.R. (Ed.) Marcel Dekker, New York, 1993. Gibbs, J.W., "Collected Works 1", Yale University Press, New Haven, CT, 1948. Gloxhuber, C. and Kunstler, K., "Anionic Surfactants: Biochemistry, Toxicology, Dermatology," Marcel Dekker (2"" ed.). New York, 1992. Gormally, J., Gettings, W.J. and WynJones, E., in "Molecular Interactions," H. Ratajczak and W.J. Orville-Thomas (Eds.), Vol. 2, Wiley, New York, 1980, p. 143. Goebel, A. and Lunkenheimer, K., Langmuir 13, 369 (1997). Grant, L.M., Tiberg, F. and Ducker, W.A, J. Phvs. Chem. 102, 4288 (1988). Gratzel, M. and Kalyanasundaram, K., "Kinetics and Catalysis in Microheterogeneous Systems," Marcel Dekker, New York, 1991.

PAGE 152

137 Greenshields, J., personal communication, ICI Surfactants, Wilmington, DE, 1998. Gu, T., Gao, Y. and He, L., T Chem. Soc. F aradav Trans 1 84, 4471 (1988). Gu, T. and Rupprecht, H., rolloid Polvm. Sci. 268, 1 148 (1990). Hall, D.G., J. Chem. Soc Faradav Trans. 1 77, 1973 (1981). Hartley, G.S., "Aqueous Solutions of Paraffm-Chain Salts," Hermann, Paris, 1936. Hartley, G.S., Collie, B. and Samis, C.S., Trans. Faradav Soc. 32, 795 (1936). Hiemenz, P.C. and Rajagopalan, R., "Principles of Colloid and Surface Chemistry," Marcel Dekker (3"* ed.). New York, 1997, Chapter 7. Hoffmann, H., Nagel, R., Platz, G. and Ulbricht, W.J.. Colloid Polym. Sci. 254, 812(1976). Holland, L. "The Properties of Glass Surfaces," Chapman & Hall, London, 1964. Holmberg, K., "Novel Surfactants: Preparation, Applications, and Biodegradability," Marcel Dekker, New York, 1998. Horozov, T.S., Kralchevsky, P.A., Danov, K.D. and Ivanov, I.E., J. Disp. Sci. Technol. 18, 593 (1997). Hua, X.Y. and Rosen, M.J., T. Colloid Interface Sci . 141, 180 (1991). Huibers, P.D.T., Ph.D. Thesis, University of Florida, Gainesville, FL, 1996. Huibers, P.D.T., Oh, S.G. and Shah., D.O. in "Surfactants in Solution," A.K. Chattopadhyay and K.L. Mittal (Eds.), Marcel Dekker, New York, 1996. Hunter, R.J., "Foundations of Colloid Science," Vol.1, Oxford University Press, 1987. Innoue, T., Tashiro, R., Shibuya, Y. and Shimozawa, R., J. Colloid Interface Sci. 73, 105(1980). Israelachvili, J.N., "Intermolecular and Surface Forces," (2"'' ed.), Academic Press, London, 1991. Israelachvili, J.N.. Langmuir 10, 3774 (1994). Israelachvili, J.N., Mitchell, D. J., Ninham, B.W., J. Chem. Soc. Faradav Trans. 1 72, 1525 (1976).

PAGE 153

138 Ivan. LB. and Dimitrov, D.S. in 'Thin Liquid Films" Vol. 29, Marcel Dekker, New York, 1988. James, A.D., Robinson, B.H. and White. N.C.. T rolloid Interface Sci. 59, 328 (1977). Jho. C. and Burke, R., T roUnid Interface Sci. 95. 61 (1983). Kahlweit. M.. T. Colloid Interface Sci. 90, 92 (1982). Kahlweit, M., Pure & AppI. Chem. 53, 2069 (1981). Kaneshina, S., Shibata, O., Nakamura, M. and Tanaka, M., Colloids and Surf. 6, 73 (1983). Kato, S., Harada, S. and Sahara, H., J. Phvs. Chem . 99, 12570 (1995). Kekicheff, P., Christenson, H.K. and Ninham. B.W., Colloids Surf. 40. 31 (1989). Kelvin, L., Phil. Mag . 42, 368 (1871). Knoche, W. and Wiese, G., Rev. Sci. Instrum . 47, 220 (1976). Koczo. K., Koczone, J.K. and Wasan, D.T., J. Colloid Interface Sci. 166, 225 (1994). Kroshwitz, J. I. and Howe-Grant, M. (Eds.) in "Kirk-Othmer Encyclopedia of Chemical Technology," Wiley-hiterscience. New York, 1993; Vol. 2, p 430. Lang. J. and Eyring, E.M.. J. Polvm. Sci . 10. 89 (1972). Lang, J., Tondre, C, Zana, R.. Bauer, R., Hoffmann, H. and Ulbricht, W., J.Phys. Chem. 79, 276(1975). Lang, J. and Zana, R. in "Chemical Relaxation Methods in Surfactant in Solutions -New Methods of hivestigation," R. Zana (Ed.), Marcel Dekker. New York. 1987. Lasic, D., "Liposomes: From Physics to Applications." Elsevier Scientific. Amsterdam. Netherlands. 1993. Lessner, E.. Teubner, M. and Kahlweit. M.. J. Phvs. Chem . 85, 1529 (1981a). Lessner, E., Teubner, M. and Kahlweit. M.. J. Phvs. Chem . 85,.3167 (1981b). Leung, R. and Shah, D.O.. J. Colloid hiterface Sci. 113, 484 (1986). MacLeod, C.A. and Radke. C.J., J. Colloid Interface Sci. 160. 435 (1993).

PAGE 154

139 Manev, E. and Pugh, R.J., T Cnlloid Interface Sci . 186, 493 (1997). Manne, S., Cleveland, J.P., Gaub, H.E., Stucky, G.D. and Hansma, P.K., Langmuir 10, 4409 (1994). Manne, S. and Gaub, H.E., Science 270, 1480 (1995). McBain, J.W., Trans. Faraday Soc . 9, 99 (1913). McBain, J.W. and Salmon, C.S^ .1. Amer. Chem. Soc . 42, 426 (1920). McDermott, D.C., McCamey, J. Thomas, R.K. and Rennie, A.R., J. Colloid Interface Sci. 162, 304(1994). Michels, B., Waton, G. and Zana, R., Langmuir 13, 31 1 1 (1997). Mijnlieff, P.P. and Ditmarsch, R., Nature 208, 889 (1965). Miller, R., Dukhin, S.S. and Kretzschmar, G., "Dynamics of Adsorption at Liquid Interfaces," Elsevier, Amsterdam, 1995. Miller, C.A. and Neogi, P., "Interfacial Phenomena, Equilibrium and Dynamic Effects," Surfactant Science Series, Vol. 17, Marcel Dekker, New York, 1985. Miller, R., Sedev, R., Schano, K.H., Ng, C. and Neumann, A.W. Colloids Surf. 69, 209 (1993). Mitchell, D. J. and Ninham, B.W., J. Chem. Soc. Faradav Trans. 2 . 77, 601 (1981). Mukerjee, P. and Mysels, K.J. "CMCs of Aqueous Surfactant Systems," NSRDS-NBS 36, U.S. Dept. of Commerce, Washington, DC, 1971. MuUer, N. in "Solution Chemistry of Surfactants," K.L. Mittal (Ed.), Vol. 1, Plenum Press, New York and London, 1979. Myers, D. "Surfaces, Interfaces and Colloids: Principles and Applications," VCH Publishers, New York, 1991. Mysels, K.J., Colloids Surf. 43, 241 (1990). Mysels, K.J., Langmuir 2, 428 (1986). Mysels, K.J. and Stafford, R.E., Colloids Surf. 51, 105 (1990).

PAGE 155

140 Nikolov, A.D., Kralchevsky, P.A., Ivanov, I.B. and Wasan, D.T., J. Colloid Interface Sd. 133, 13 (1989a). Nikolov, A.D. and Wasan, D.T., JColloid Int erface Sci. 133, 1 (1989b). Nikolov, A.D. and Wasan, D.T., Langmuir 8, 2985 (1992). Oh, S.G. and Shah, P.O.. Langmuir 7, 1316 (1991). Oh, S.G., Klein, S.P. and Shah, D.O., AIChE Journal 38, 149 (1992). Oh, S.G. and Shah, D.O., Langmuir 8, 1232 (1992). Oh, S.G. and Shah, D.O., I. Am. Oil Chem. Soc , 70, 673 (1993a). Oh, S.G. and Shah, D.O., J. Phvs. Chem. 97, 284 (1993b). Oh, S.G., Jobalia, M. and Shah, D.O., J. Colloid Interface Sci . 156, 51 1 (1993). Overbeek, J. Th. G., T. Colloid Interface Sci. 58, 408 (1977). Partyka, S.E., Lindenheimer, M. and Groszek, A., Colloids Surf . 37, 309 (1989). Pasley, R.M. and Israelachvili, J.N., Colloids Surf. 2, 169 (1981). Pashley, R.M., McGuiggan, P.M., Horn, R.G. and Ninham, B.W., J. Colloid Interface Sci. 126, 569(1988). Patel, S.S., Kumar, K., Shah, D.O. and Delfmo, J.J., J. Colloid Interface Sci. 183, 603 (1996). Preston, W.C.. I. Phvs. Colloid Chem . 52, 84 (1948). Prud'homme, R.K. and Khan, S.A. "Foams: Theory, Measurement and Applications," Surfactant Science Series, Vol. 57, Marcel Dekker, New York, 1996. Pugh, R.J., Adv. Coll. Interface Sci. 64, 67 (1996). Reiss-Husson, F. and Luzzati, V., J. Phvs. Chem . 68, 3504 (1964). Rillaerts, E. and Joos, P., J. Phvs. Chem. 86, 3471 (1982). Rosen, M.J., "Surfactants and Interfacial Phenomena," (2"^* ed.), John Wiley & Sons, New York, 1989.

PAGE 156

141 Rosen, M.J., Cohen, A.W., Dahanayake, M. and Hua, X.Y., J.Phys. Chem. 86, 541 (1982). Ross, J.L., Bruce, W.D. and Janna, W.S., Langmuir 8, 2644 (1992). Ross, S., "Encyclopedia of Chemical Technology," John Wiley & Sons (3^"* ed.). Vol. 1 1 (1980), pp 127-145. Ross, S. and Bramfitt, J.H., J. Phvs. Chem . 61, 1261 (1957). Ross, S. and Hauk, R.M., .T. Phvs. Chem. 62, 1260 (1958). Rupprecht, H. and Gu, T., Colloid Polvm. Sci. 269, 506 (1991). Rutland, M.W. and Parker, J.L., Langmuir 10, 1 1 10 (1994). Rybinski, W. and Stoll, G., "Alkyl Polyglycosides: Technology, Properties, and Applications," VCH, New York, 1997. Scales, P.J., Grieser, F. and Healy, T.W., Langmuir 6, 582 (1990). Shah, D.O., in "Micelles, Microemulsions and Monolayers," D.O. Shah (Ed.), Marcel Dekker, New York, 1998, Chapter 1. Shah, D.O., Djabbarab, N.F. and Wasan, D.T., J. Colloid and Polym. Sci. 276, 1002 (1978). Shah, D.O. and Schulman, J.H., J. Lipis Res . 8, 227 (1967). Sharma, E.G., Basu, S. and Sharma, M.M.. Langmuir 12, 6506 (1996). Shiao, S.Y., Ph.D. Thesis, University of Florida, Gainesville, FL, 1976. Shiao, S.Y., Chhabra, V., Patist, A., Free, M.L., Huibers, P.D.T., Gregory, A., Patel, S., and Shah, D.O., Adv. Colloid Interface Sci . 74, 1 (1998). Shinoda, K. and Nakagawa, T., "Colloidal Surfactants: Some Physicochemical Properties," Academic Press, New York, 1963. Soderlind, E. and Stilbs, P., J. Colloid Interface Sci . 143, 586 (1991). Somasundaran, P., AIChE Symposium Series . No 150, 71, 1 (1975). Strey, R. and Pakusch, A. in "Surfactants in Solution," K. Mittal and P. Bothorel (Eds.), Vol. 4, Plenum Press, New York and London, 1986, p 465.

PAGE 157

142 Tamura, T., Kaneko,Y. and Ohyama, M., T rnllnid Interface Sci. 173, 493 (1995). Tanford, C, "The Hydrophobic Effect. The Formation of Micelles and Biological Meinbranes," Wiley, (2"" ed.). New York, 1980. Tharapiwattananon, N., Scamehom, J.F., Osuwan, S., Harwell, J.H. and Haller, K.J., Sep, Sci. Technol. 31, 1233 (1996). Thomas, W.D.E. and Hall, D.J. T r.nllnid Interface Sci. 51, 328 (1975). Tondre, C, Lang, J. and Zana, R., T Tnlloid Interface Sci. 52, 372 (1975). Verwey, E. and Overbeek, J. Th. G., "Theory of the Stability of Lyophobic Colloid," Amsterdam, Elsevier, 1948. Void, R.D. and Void, M.J., "Colloid and Surface Chemistry," Addison-Wesley, London 1983. Walstra, P. in "Encyclopedia of Emulsion Technology," P. Becher (Ed.), Vol. 1, Dekker, New York and Basel, 1983. Wangnerud, P. and Olofsson, G., J. Colloid Interface Sci. 153, 392 (1992). Wanka, G., Hoffmann, H. and Ulbricht, W., Colloid Polvm. Sci. 268, 101 (1990). Wasan, D.T., Gupta, L. and Vora, M.K., AIChE Journal 17, 1287 (1971). Wyn-Jones, E., "Chemical and Biological Aspects of Relaxation Spectroscopy," D. Reidel, Dordrecht and Boston, 1975. Yiv, S., Zana, R., Ulbricht, W., and Hoffmann, H., T Colloid Interface Sci. 8, 224 (1981).

PAGE 158

BIOGRAPHICAL SKETCH Alexander Patist was born on May 21, 1971, in Jutphaas, The Netherlands. He completed high school (HAVO) at the Oosterlicht College in Nieuwegein (1988) and went on to the Hogeschool van Utrecht (location Hilversum) to complete his bachelor's degree in chemical engineering in 1992. He joined the chemical engineering department of Eindhoven, University of Technology and received his master's degree in polymer science and technology from the group of Professor Piet Lemstra in 1995 on a new process for the production of water-blown polystyrene foam. Through this project, which was conducted at the Shell Research and Technology Centre in Amsterdam, he came in contact with Professor Dinesh Shah, director of the Center for Surface Science and Engineering at the University of Florida. He joined Professor Shah's group in March 1995 for a post-master's research project on adsorption and desorption of surface-active polymers on rigid gas-permeable contact lenses. He decided to continue to work toward his Ph.D. degree in the department of chemical engineering at the University of Florida in January 1996 and completed his degree under the supervision of Professor Shah in 1999. 143

PAGE 159

I certify that I have read this study and that in my opinion it confirms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Dinesh O. Shah, Chairman Professor of Chemical Engineering I certify that I have read this study and that in my opinion it confirms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Ben Koopman ^^^^ Professor of Environmental Engineering Sciences I certify that I have read this study and that in my opinion it confirms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Brij M. Moudgil ^ Professor of Materials Science and Engineering I certify that I have read this study and that in my opinion it confirms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Engineering I certify that I have read this study and that in my opinion it confirms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Spyros Svoronos Professor of Chemical Engineering

PAGE 160

This dissertation was submitted to the Graduate Faculty of the College of Engineering and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August 1999 M. Jack Ohanian Dean, College of Engineering Winfred M. Phillips Dean, Graduate School