7136
Langmuir 2005, 21, 7136-7142
Microstructure of Un-neutralized Hydrophobically Modified Alkali-Soluble Emulsion Latex in Different Surfactant Solutions Sheng Dai†,‡ and Kam Chiu Tam*,†,‡ Singapore-MIT Alliance, School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Republic of Singapore Received March 10, 2005. In Final Form: May 9, 2005 At low pH conditions and in the presence of anionic, cationic, and nonionic surfactants, hydrophobically modified alkali-soluble emulsions (HASE) exhibit pronounced interaction that results in the solubilization of the latex. The interaction between HASE latex and surfactant was studied using various techniques, such as light transmittance, isothermal titration calorimetry, laser light scattering, and electrophoresis. For anionic surfactant, noncooperative hydrophobic binding dominates the interaction at concentrations lower than the critical aggregation concentration (CAC) (C < CAC). However, cooperative hydrophobic binding controls the formation of mixed micelles at high surfactant concentrations (C g CAC), where the cloudy solution becomes clear. For cross-linked HASE latex, anionic surfactant binds only noncooperatively to the latex and causes it to swell. For cationic surfactant, electrostatic interaction occurs at very low surfactant concentrations, resulting in phase separation. With further increase in surfactant concentration, noncooperative hydrophobic and cooperative hydrophobic interactions dominate the binding at low and high surfactant concentrations, respectively. For anionic and cationic surfactant systems, the CAC is lower than the critical micelle concentration (CMC) of surfactants in water. In addition, counterion condensation plays an important role during the binding interaction between HASE latex and ionic surfactants. In the case of nonionic surfactants, free surfactant micelles are formed in solution due to their relatively low CMC values, and HASE latexes are directly solubilized into the micellar core of nonionic surfactants.
Introduction Fully neutralized hydrophobically modified alkalisoluble emulsion (HASE) is one type of associative thickener that is widely used in water-borne coating formulations.1 HASE is a comblike polymer with the chemical structure described in Scheme 1. On the methacrylic acid (MAA) and ethyl acrylate (EA) copolymer backbone, small amounts of hydrophobic moieties are grafted through a dimethyl m-isopropenylbenzyl isocyanate and a short poly(ethylene oxide) (PEO) linkage, where the grafted chains are referred to as the associative macromonomers (AM). At low pH, HASE is insoluble and possesses a compact structure. With the addition of a base, MAA segments are neutralized, leading to the swelling and dissolution of HASE polymers in water. After full neutralization, the HASE backbone possesses a polyelectrolyte character and solubilizes in water, while the hydrophobic groups associate to form aggregates in solution. For a semidilute HASE solution, these small aggregates become connected via bridging chains to form a gel-like network, which enhances the solution viscosity. The dissolution behaviors of dilute HASE solutions, the association behaviors of fully neutralized HASE dilute solutions, and the rheological properties of semidilute HASE solutions have been studied extensively by potentiometric and conductometric titrations,2,3 steady-state * To whom correspondence should be addressed. Fax: (65) 6791 1859. E-mail:
[email protected]. † Singapore-MIT Alliance. ‡ School of Mechanical and Aerospace Engineering. (1) Winnik, M. A.; Yekta, A. Curr. Opin. Colloid Interface Sci. 1997, 2, 424. (2) (a) Wang, C.; Tam, K. C.; Jenkins, R. D.; Bassett, D. R. Phys. Chem. Chem. Phys. 2000, 2, 1967. (b) Wang, C.; Tam, K. C.; Jenkins, R. D. J. Phys. Chem. B 2002, 106, 1195. (3) Siddiq, M.; Tam, K. C.; Jenkins, R. D. Colloid Polym. Sci. 1999, 277, 1172.
Scheme 1 Chemical Structures of HASE and Cross-Linked HASE Used in This Study
fluorescence spectroscopy and time-resolved fluorescence quenching,4-6 static and dynamic laser light scattering,7,8 pulsed-gradient NMR,9 and rheological10-15 techniques. When HASE polymers are fully neutralized, the solution properties resemble those of polyelectrolyte solutions that (4) Karunasena, A.; Brown, R. G.; Glass, J. E. In Polymers in Aqueous Media: Performance through Association; Glass, J. E., Ed.; American Chemical Society: Washington, DC, 1989; p 496. (5) Araujo, E.; Rharbi, Y.; Huang, X.; Winnik, M. A.; Bassett, D. R.; Jenkins, R. D. Langmuir 2000, 16, 8664. (6) Horiuchi, K.; Rharbi, Y.; Spiro, J. G.; Yekta, A.; Winnik, M. A.; Jenkins, R. D.; Bassett, D. R. Langmuir 1999, 15, 1644. (7) (a) Dai, S.; Tam, K. C.; Jenkins, R. D. Eur. Polym. J. 2000, 36, 2671. (b) Dai, S.; Tam, K. C.; Jenkins, R. D. Macromolecules 2000, 33, 404. (c) Dai, S.; Tam, K. C.; Jenkins, R. D.; Bassett, D. R. Macromolecules 2000, 33, 7021. (d) Dai, S.; Tam, K. C.; Jenkins, R. D. Macromol. Chem. Phys. 2001, 202, 335. (e) Dai, S.; Tam, K. C.; Jenkins, R. D. Macromol. Chem. Phys. 2002, 203, 2312. (8) Islam, M. F.; Jenkins, R. D.; Bassett, D. R.; Lau, W.; Ou-Yang, H. D. Macromolecules 2000, 33, 2480. (9) Nagashima, K.; Strashko, V.; Macdonald, P. M.; Jenkins, R. D.; Bassett, D. R. Macromolecules 2000, 33, 9329. (10) Tirtaatmadja, V.; Tam, K. C.; Jenkins, R. D. Macromolecules 1997, 30, 1426.
10.1021/la050651e CCC: $30.25 © 2005 American Chemical Society Published on Web 07/02/2005
Un-neutralized HASE Latex in Surfactant Solutions
possess an amphiphilic character. Hence, the presence of surfactant alters its association in aqueous media. The interaction between fully neutralized HASE and different types of surfactants has been studied.16-19 From the studies on HASE and nonionic surfactant (alkyl ethoxylate, CmEOn) or HASE and anionic surfactant (sodium dodecyl sulfate, SDS), it was obvious that the hydrophobic interaction between the hydrophobic tails of surfactants and the hydrophobic domains of HASE aggregates dominates the binding interaction. With further increase in surfactant concentration, the networklike cluster can be destroyed, as evident from the reduction in solution viscosity.20 The binding interaction between fully neutralized HASE and cationic surfactant (dodecyltrimethylammonium bromide, DoTab) shows different binding behaviors, which falls into the category of binding interactions between charged polymer and oppositely charged surfactants. Electrostatic interaction plays an important role during the binding process, which results in phase separation at low surfactant concentrations, and usually the precipitates can be resolubilized when excess amounts of surfactants are added. The electrostatic interaction can be minimized through the addition of electrolytes.18 At higher surfactant concentrations, hydrophobic interactions dominate and lead to the destruction of the HASE network structure. During our previous experiments, we noted that the addition of surfactant to un-neutralized HASE could lead to the solubilization of HASE latexes in aqueous solution, where the cloudy emulsion solution become clear. Although the solution behaviors of HASE and the binding interaction between fully neutralized HASE and different types of surfactant have been extensively studied, the interactions between un-neutralized HASE latexes and surfactants have not been reported. From a survey of the literature, a few studies on the interaction between surfactant and partially neutralized or un-neutralized polycarboxylic acids have been reported.21-27 The effect of charge density on the binding of cationic surfactant onto polyanions, such (11) Tirtaatmadja, V.; Tam, K. C.; Jenkins, R. D. Macromolecules 1997, 30, 3271. (12) (a) Tam, K. C.; Farmer, M. L.; Jenkins, R. D.; Bassett, D. R. J. Polym. Sci., Part B: Polym. Phys. 1998, 36, 2275. (b) Tam, K. C.; Guo, L.; Jenkins, R. D.; Bassett, D. R. Polymer 1999, 40, 6369. (13) (a) Tan, H.; Tam, K. C.; Jenkins, R. D. J. Colloid Interface Sci. 2000, 231, 52. (b) Tan, H.; Tam, K. C.; Jenkins, R. D. J. Appl. Polym. Sci. 2001, 79, 1486. (14) English, R. J.; Gulati, H. S.; Jenkins, R. D.; Khan, S. A. J. Rheol. 1997, 41, 427. (15) (a) Guo, L.; Tam, K. C.; Jenkins, R. D. Macromol. Chem. Phys. 1998, 199, 1175. (b) Ng, W. K.; Tam, K. C.; Jenkins, R. D. Eur. Polym. J. 1999, 35, 1245. (16) (a) Seng, W. P.; Tam, K. C.; Jenkins, R. D.; Bassett, D. R. Macromolecules 2000, 33, 1727. (b) Seng, W. P.; Tam, K. C.; Jenkins, R. D.; Bassett, D. R. Langmuir 2000, 16, 2151. (17) Tam, K. C.; Seng, W. P.; Jenkins, R. D. J. Polym. Sci., Part B: Polym. Phys. 2000, 38, 2019. (18) Wang, C.; Tam, K. C.; Jenkins, R. D.; Tan, C. B. J. Phys. Chem. B 2003, 107, 4667. (b) Wang, C.; Tam, K. C.; Tan, C. B. Langmuir 2004, 20, 7933. (19) Tan, H.; Tam, K. C.; Jenkins, R. D.; Tirtaatmadja, V.; Bassett, D. R. J. Non-Newtonian Fluid Mech. 2000, 92, 167. (20) (a) Seng, W. P.; Tam, K. C.; Jenkins, R. D. Colloids Surf., A: Physicochem. Eng. Aspects 1999, 154, 363. (b) Seng, W. P.; Tam, K. C.; Jenkins, R. D.; Bassett, D. R. In ACS Symposium Series 765, Associating Polymers in Aqueous Media; Glass, J. E., Ed.; American Chemical Society: New York, 2000; p 351. (c) Tan, H.; Tam, K. C.; Jenkins, R. D. Langmuir 2000, 16, 5600. (d) Tirtaatmadja, V.; Tam, K. C.; Jenkins, R. D. AIChE J. 1998, 44, 2756. (e) Tirtaatmadja, V.; Tam, K. C.; Jenkins, R. D. Langmuir 1999, 15, 7537. (21) Kogej, K.; Theunissen, E.; Reynaers, H. Langmuir 2002, 18, 8799. (22) Chandar, P.; Somasundaran, P.; Turro, N. J. Macromolecules 1988, 21, 950. (23) Katsuura, H.; Kawamura, H.; Manabe, M.; Kawasaki, H.; Maeda, H. Colloid Polym. Sci. 2002, 280, 30.
Langmuir, Vol. 21, No. 16, 2005 7137
as PAA, poly(methacrylic acid) (PMAA), poly(vinyl sulfate), and carboxymethyl cellulose, was reported. The effect of pH on the interactions between PAA and cationic surfactant is driven by electrostatic interaction at high pH but controlled by hydrogen bonding at low pH.27 However, some reported that the interactions between cationic surfactant and PAA or PMAA are driven by hydrophobic interaction at low pH and electrostatic interaction at high pH to produce highly organized cubic structures of polymer/surfactant aggregates.21,23 In this paper, the interaction between un-neutralized HASE latex and different types of surfactants, that is, anionic (SDS), cationic (DoTab), and nonionic (C12E9), in dilute aqueous solution was systematically studied by UV-vis spectroscopy, dynamic light scattering (DLS), ζ-potential, and isothermal titration calorimetric (ITC) techniques. The interactions between anionic surfactant (SDS) and unneutralized HASE latex with and without a cross-linker were compared, while a comparison for the DoTab system has recently been reported.18 The possible binding isotherm and binding mechanism were elucidated. Experimental Section Materials. The HASE copolymer, designated as HASEEO35C16, was synthesized by Dow Chemicals (formerly Union Carbide), which contains 50 mol % methacrylic acid (MAA), 49 mol % ethyl acrylate (EA), and 1 mol % associative macromonomer. The hydrophobic linear hexadecyl group (-C16H33) was ethoxylated with ∼35 repeat unit of ethylene oxide in the macromonomer (Scheme 1a). The cross-linked HASE latex (Scheme 1b) was also synthesized by Dow Chemicals (formerly Union Carbide), which contains MAA/EA in the molar ratio (X/ Y) ∼ 50:50, and the density of di-allyl phthalate (DAP) crosslinker is 4 wt %. The purification of HASE and cross-linked HASE has been described elsewhere.7 The pH of the cleaned HASE latex is ∼3.5. Anionic surfactant (SDS, CMC ∼ 8.3 mM28) was purchased from BDH Laboratory Supplies. Cationic surfactant (DoTab, CMC ∼ 15 mM28) and nonionic surfactant polyoxyethylene dodecyl ether (C12E9, CMC ∼ 0.08 mM28) were purchased from Sigma. The deionized water used for solution preparation was from an Alpha-Q Millipore (Millipore Corp. Bedford, MA) water-purifying system, and a 0.22-µm filter was used to filter the water. Prior to conducting the light-scattering measurement, the sample solutions were filtered through a 0.2-µm filter to remove dust particles. Visible Spectrophotometer. To detect the light transmittance of HASE latex in the presence of different concentrations of surfactants, an Agilent UV-visible spectrophotometer (HP 8453) was used under the standard mode at a fixed wavelength of 488 nm and a path length of 1 cm. Dynamic Light Scattering (DLS). A Brookhaven BI-200SM goniometer system equipped with a 522-channel BI9000AT digital multiple τ correlator was used to perform the lightscattering experiments. The light source is a power-adjustable vertical polarized argon ion laser with the wavelength of 488 nm. The measured temperature was controlled at 25 ( 0.1 °C using a Science/Electronic water bath. The REPES inverse Laplace transform (ILT) routine supplied with the GENDIST software package was used to analyze the intensity-intensity autocorrelation functions. The probability of reject was set to 0.5. At low surfactant concentrations where the HASE solution is cloudy, the polydispersity index as determined from the second cumulant is 0.005. However, the polydispersity index increases to 0.15-0.25 with increasing surfactant concentration when the HASE solution becomes clear. (24) Shimizu, T. Colloids Surf., A: Physicochem. Eng. Aspects 1994, 84, 239. (25) Kiefer, J. J.; Somasundaran, P.; Anaathapadmanabhan, K. P. Langmuir 1993, 9, 1187. (26) Hayakawa, K.; Santerre, J. P.; Kwak, J. C. Macromolecules 1983, 16, 1642. (27) Yoshida, K.; Dubin, P. L. Colloids Surf., A: Physicochem. Eng. Aspects 1999, 147, 161. (28) Rosen, M. J. Surfactants and Interfacial Phenomena; Wiley: New York, 1980.
7138
Langmuir, Vol. 21, No. 16, 2005
Dai and Tam
Figure 1. SDS concentration dependence of light transmittance (open circle) and ζ-potential (filled circle) for 0.1 wt % HASE-EO35C16 latex at 298 K, 1 atm. The insert is the SDS concentration dependence of the conductivity and EMF (Na+).
Figure 2. Isothermal titration thermograms for titration of 0.2 M SDS into water and 0.1 wt % HASE-EO35C16 at 298 K and 1 atm. The open square is the differential curve between SDS into HASE latex and that into water.
Electrochemical Analysis. A Brookhaven ZetaPlus ζ-potential analyzer with a 671-nm He-Ne laser source was used to measure the electrophoretic mobility and the ζ-potential of the latex solutions in the presence of different concentrations of surfactants. An electromotive force (EMF) study was conducted using a Metrohm Na+ selective electrode relative to a Metrohm Ag/AgCl reference electrode. An ABU93 tri-buret titrator with a modified Aliquot software was used to conduct the titration experiments. The EMF and conductivity values were measured at 25.0 ( 0.1 °C, where the temperature was controlled by a PolyScience water bath. Isothermal Titration Calorimetry (ITC). ITC monitors the differential enthalpy changes isothermally while titrating surfactant solutions into latex solutions. A Microcal isothermal titration calorimeter was used to perform the ITC experiments. It has a reference cell and a sample cell of 1.35 mL, both insulated by an adiabatic shield. The titration was carried out by a stepby-step injection of concentrated surfactant solutions from a 250-mL injection syringe into the sample cell filled with latex solutions. Using an interactive software, an injection schedule was automatically carried out for a predefined number of injections, the volume of each injection, and the time between each injection. The time interval between each injection is set at 4 min. Both titrant and titrate were degassed to remove dissolved gas.
curve for the titration of SDS into water and into HASEEO35C16. In the SDS dilution thermogram, the transition at an SDS concentration of 8.3 mM corresponds to the CMC of SDS in aqueous solution. The titration curve of SDS into HASE latex solution is exothermic with an obvious transition point at ∼6.8 mM, which differs from the endothermic curve for titrating SDS into water. It then merges with the SDS dilution curve at ∼15 mM. The difference between the two titration curves is attributed to the interaction between SDS and HASE latex dispersion. The obvious exothermic peak indicates the hydrophobic interaction between SDS and HASE latex.29 From the enthalpy-difference curve, the sharp transition occurs at ∼3.2 mM SDS, which corresponds to the critical aggregation concentration (CAC) of SDS and HASE latex. Upon reaching ∼15 mM SDS, the polymer-surfactant interaction ceases, and this critical point is related to the saturation concentration (C2). ITC results indicate the interaction between SDS and HASE latex is dominated by the solubilization of polymers into the SDS mixed micellar core. Since SDS is an anionic surfactant, the binding process will be accompanied by surface charge fluctuations. The ζ-potential was utilized to monitor the surface charge fluctuation during the solubilization process as shown in Figure 1. The HASE latex possesses a slightly negative charge (-3.0 mV) at 0 mM SDS. When anionic surfactant SDS was added, the surface of the latex becomes more negative, and this is related to the noncooperative binding of SDS monomers onto the HASE latex. The ζ-potential exhibits a minimum of -18 mV in the region of 3-5 mM SDS. When the SDS concentration exceeds ∼5 mM, the HASE containing SDS mixed micelles are cooperatively formed in solution, which solubilizes the HASE latex into the hydrophobic micellar core. During the latex-induced micellization process, the accompanying counterion condensation gives rise to a reduction in the total surface charge and a corresponding increase in the ζ-potential. In addition, it is well-known that increases in salt concentration not only decrease the thickness of the double layer but also decrease the absolute value of the ζ-potential,30 which also contributes to the increase in ζ-potential at higher SDS concentrations. Beyond a SDS concentration of 8 mM, both the double-layer and HASE latex particles
Results and Discussion 1. Interaction between HASE Latex and Anionic Surfactant. In this section, light transmittance, dynamic light scattering (DLS), ζ-potential, and isothermal titration calorimetry (ITC) were used to monitor the interaction between anionic surfactant (SDS) and HASE latex. The latex is in the form of an emulsion at low pH, but it becomes clear with the addition of SDS. The SDS concentration dependence of the turbidity of HASE latex was monitored by light transmittance at a wavelength of 488 nm. Figure 1 shows the light transmittance of 0.1 wt % HASEEO35C16 in different concentrations of SDS. It is evident that HASE latex becomes clear when the SDS concentration exceeds 5 mM, where the light transmittance is greater than 80%. This suggests that HASE latex interacts with the anionic surfactant to form mixed micelles with SDS monomers, driven by the system to minimize its free energy. The thermodynamics of interaction between SDS and HASE latex was studied by ITC. Figure 2 shows the ITC thermograms for titrating 0.2 M SDS into water and 0.1 wt % HASE-EO35C16 aqueous solution at 298 K and 1 atm, respectively. The interaction between SDS and HASE latex can be observed from the enthalpy difference
(29) Dai, S.; Tam, K. C. Langmuir 2004, 20, 2177. (30) Hunter, R. J. Foundations of colloid science; Oxford University Press: New York, 2001.
Un-neutralized HASE Latex in Surfactant Solutions
Langmuir, Vol. 21, No. 16, 2005 7139
Figure 3. The relationship between Rh and SDS concentrations for 0.1 wt % HASE-EO35C16 at 298 K.
have been fully destroyed, making it difficult to obtain reliable ζ-potential data, as indicated by the dash line. The fluctuation in the ionic environment during the binding process was also monitored by conductivity and Na+ ion selective electrode measurements (insert of Figure 1). It is evident that both conductivity and Na+ concentration increase with the addition of SDS, but a distinct transition in the slope of the conductivity curve is observed at 3-5 mM SDS, which indicates the onset of cooperative binding. The free Na+ ion concentration does not increase with the addition of SDS beyond this critical concentration, which suggests that counterion condensation of sodium ions on the surface of surfactant micelles must have occurred. The particle size of latexes during the binding process was monitored by DLS. The decay-time distribution functions of 0.1 wt % HASE-EO35C16 in different concentrations of SDS were measured at different scattering angles. It was found that the decay rates Γ are linearly proportional to the square of the scattering vector q, which confirmed that the decay mode is caused by the translational diffusion of HASE latex or HASE latex containing mixed micelles. The translational diffusion coefficient was determined from the slope of the Γ ∼ q2 plot. On the basis of the Stokes-Einstein relationship, the hydrodynamic radius Rh of the particle in solution can be calculated:
Rh )
kT 6πη0D
(1)
where k is the Boltzmann constant, T is the absolute temperature in Kelvin, η0 is the solvent viscosity, and D is the diffusion coefficient. In addition, it is also evident that the narrow distribution peak was detectable for SDS concentrations ranging from 1 to 5 mM, while a relatively broad distribution in the decay-time distribution functions was observed for SDS concentration exceeding 8 mM. At low SDS concentration, HASE latex is in the form of a compact coil, which exhibits strong scattering and narrow decay-time distribution. For higher SDS concentrations, the HASE latexes are solubilized by SDS molecules to produce mixed micelles together with solubilized HASE latexes. The solution becomes clear, and the scattering intensity decreases; this is accompanied by a broadening of the decay-time distribution functions and an incremental increase in the decay time as the concentration of SDS increases. This increase corresponds to the increase in the hydrodynamic radius of the particles as shown in Figure 3.
Figure 4. The decay-time distribution functions for 0.1 wt % cross-linked HASE and different concentrations of SDS at 298 K.
In the absence of SDS, HASE latex is insoluble, and it possesses a hydrodynamic radius of ∼80 nm. In the presence of small amounts of SDS (up to 3 mM SDS), the hydrodynamic radius increases slightly. Combining this trend with ITC results, we concluded that uncooperative binding of SDS monomers onto the HASE latex occurs, where the charge density is increased. Further increase in SDS concentration results in an increase in the hydrodynamic radius, reaching a maximum of ∼125 nm at 5 mM SDS. Thereafter, the hydrodynamic radius decreases to ∼110 nm and exhibits a marginal increase over the range of 8-20 mM SDS. The increase in the hydrodynamic radius over this range may be attributed to the cooperative binding of SDS to HASE latex, where the HASE latexes are solubilized into the core of SDS mixed micelles as evident from ITC thermograms and ζ-potential studies. For mixed micelles formed by ionic surfactant, counterion binding plays an important role in controlling the polymer-surfactant interaction and the resulting microstructure.31 The counterion binding is evident from ζ-potential measurements. With further increase in SDS concentration, the electrostatic repulsion is enhanced, and the balance between hydrophobic and electrostatic forces is altered, leading to a structural reorganization of mixed micelles into smaller stable microstructures, where the equilibrium of these two opposing forces is maintained. The transition at ∼8 mM SDS was also observed in the ITC thermogram, which may be related to this structural reorganization. The possible binding mechanism is given by the pictorial representation described in Figure 3. 2. Interaction between Cross-Linked HASE Latex and Anionic Surfactant. Due to the presence of crosslinkers, cross-linked HASE latex does not disintegrate with increasing pH. Instead, it swells as a result of the buildup of osmotic pressure exerted by bound surfactant molecules and counterions trapped within the cross-linked latex particle. Light-transmittance study of cross-linked HASE latex in the presence of different concentrations of SDS revealed that the cloudy solution does not become clear after the addition of SDS. DLS measurements of 0.1 wt % crosslinked HASE in different concentrations of SDS were taken at different scattering angles (Figure 4). Only one q2-dependent decay mode was evident, which is caused by the translational diffusion of cross-linked latex, and the (31) Goddard, E. D.; Ananthapadmanaban, K. P. Interactions of Surfactants with Polymer and Proteins; CRC Press: Boca Raton, 1993.
7140
Langmuir, Vol. 21, No. 16, 2005
Figure 5. SDS concentration dependence of Rh (open circle) and ζ-potential (closed circle) for 0.1 wt % cross-linked HASE at 298 K.
translational diffusion coefficients can be determined from the slope of the straight line between decay rates Γ and q2. The Rh values were determined on the basis of the Stokes-Einstein equation (eq 1) and shown in Figure 5. The hydrodynamic radius decreases slightly at low SDS concentrations (
