Adsorption On Polystyrene Latex Particles

Langmuir 2000, 16, 7905-7913

7905

Competitive Adsorption of the Anionic Surfactant SLS and the Nonionic Surfactant Triton X-405 on Polystyrene Latex Particles Damien Colombie´,† Katharina Landfester,‡ E. David Sudol,† and Mohamed S. El-Aasser*,† Emulsion Polymers Institute and Department of Chemical Engineering, Lehigh University, Iacocca Hall, 111 Research Drive, Bethlehem, Pennsylvania 18015, and Max Planck Institute for Colloid and Interface Science, Am Muehlenberg, 14424 Golm, Germany Received December 3, 1999. In Final Form: July 13, 2000 The competitive adsorption of an anionic surfactant (SLS) and a nonionic surfactant (Triton X-405) on monodisperse polystyrene particles (92 nm) was studied and quantified using a variety of experimental techniques (filtration, desorption via serum replacement, 1H NMR spectroscopy). All experiments were performed at 25 °C using a cleaned polystyrene latex (2% solids). In the competitive adsorption of a 1/1 molar ratio of SLS to Triton X-405 on the polystyrene particles, Triton X-405 adsorbed preferentially at total surfactant concentrations in the aqueous phase below 2.5 × 10-3 M, due to its low free energy of adsorption. At higher surfactant concentrations, the particle surface became saturated with Triton X-405, and cooperative interaction between the two surfactants took place. An excess amount of the two surfactants was noted on the surface. Triton X-405 was observed to adsorb on polystyrene particles precovered with SLS. Approximately 20% of the SLS was removed from the surface below 7.0 × 10-4 M Triton X-405 aqueous concentration. However, cooperative adsorption occurred at higher concentrations, and a large excess of the two surfactants was noted on the surface at saturation. SLS was also shown to adsorb on polystyrene particles precovered with Triton X-405. The adsorption was small at SLS aqueous concentrations below 2.5 × 10-3 M. At higher concentrations, more of the SLS was adsorbed, and an excess surfactant content was present at the surface. No significant effect on the adsorption of Triton X-405 was noted for increasing SLS concentrations in the system. Competitive desorption experiments were performed with the two surfactants using serum replacement. SLS was found to desorb more readily than Triton X-405. The particle surface composition was richer in Triton X-405 as the total surfactant concentration in the system decreased.

Introduction Various types of surfactants have been used in the synthesis and stabilization of polymer latexes. Anionic surfactants provide electrostatic stability as described by the DLVO theory.1,2 The efficiency of such surfactants is highly dependent on many parameters such as ionic strength and pH. This dependency can be a major drawback in terms of the stability of the latex. Nonionic or polymeric surfactants provide steric stabilization. The repulsion between particles is provided by the thermodynamically favored steric repulsion of the adsorbed materials.3,4 It is therefore practical to use mixtures of anionic and nonionic surfactants in emulsion polymerization to combine different stabilization mechanisms. Most industrial emulsion polymerization processes use mixtures of anionic and nonionic emulsifiers. A common process can be described as follows: the reaction is started with the anionic surfactant, and the nonionic is added at a higher conversion or as a poststabilizer. In emulsion polymerization, the interaction between the surfactants and the polymer particles is fundamental in the control of the process and the final properties of the latex. The †

Lehigh University. Max Planck Institute for Colloid and Interface Science. * To whom correspondence should be addressed.

‡

(1) Derjaguin, B. V.; Landau, L. Acta Physicochim. USSR 1941, 14, 633. (2) Verwey, E. J. W.; Overbeek, J. G. In Theory of Stabilization of Lyophobic Colloids; Elsevier: Amsterdam, 1948. (3) Ottewill, R. H. J. Colloid Interface Sci. 1977, 58, 357. (4) Napper, D. H. In Polymeric Stablilization of Colloidal Dispersions; Academic Press: New York, 1983.

surfactant molecules distribute between various locations (surface of the particles, aqueous phase, micelles, monomer droplets) by a dynamic equilibrium. This behavior determines phenomena such as the overall stability of the latex, the nucleation of new particles, and limited aggregation. Therefore, it is very important to study and quantify the interaction between the surfactant molecules and the polymer particles. Adsorption isotherms of a surfactant on a latex particle surface are usually determined by following the change of the aqueous bulk surfactant concentration with increasing amounts of emulsifier in the system. Various techniques have been used to determine adsorption isotherms of surfactants on polymer particles.5,6 Although the adsorption of various nonionic surfactants on latex particles has been widely studied,7-12 only a small amount of work has been published on the competitive adsorption of mixtures of anionic and nonionic surfactants on latex particles. Kronberg and his team13,14 applied a (5) Maron, S. H.; Elder, M. E.; Moore, C. J. Colloid Sci. 1954, 9, 104. (6) Ahmed, S. M.; El-Aasser, M. S.; Pauli, G. H.; Poehlein, G. W.; Vanderhoff, J. W. J. Colloid Interface Sci. 1980, 73, 388. (7) Kronberg, B.; Kall, L.; Stenius, P. J. Dispersion Sci. Technol. 1981, 2, 215. (8) Kronberg, B.; Stenius, P. J. Colloid Interface Sci. 1984, 102, 410. (9) Zhao, J.; Brown, W. J. Phys. Chem. 1996, 100, 3775. (10) Zhao, J.; Brown, W. J. Phys. Chem. 1996, 100, 5908. (11) Haggerty, J. F.; Roberts, J. E. J. Appl. Polym. Sci. 1995, 58, 271. (12) Colombie´, D.; Landfester, K.; Sudol, E. D.; El-Aasser, M. S. J. Colloid Interface Sci. 1998, 202, 554. (13) Hulden, M.; Kronberg, B. J. Coat. Technol. 1994, 66, 67. (14) Kronberg, B.; Linstrom, M.; Stenius, P. In Phenomena in Mixed Surfactant Systems; Scamehorn, J. H., Ed.; ACS Symposium Series 311; American Chemical Society: Washington, DC, 1986.

10.1021/la9915825 CCC: $19.00 © 2000 American Chemical Society Published on Web 09/19/2000

7906

Langmuir, Vol. 16, No. 21, 2000

thermodynamic model to the adsorption of such a mixture by extending an existing model for mixed micelles. The surface composition was governed in a manner similar to the composition of the mixed micelles. The model was correlated with experimental values obtained by solvent depletion. It showed both theoretically and experimentally that when a small amount of Triton X-100 was added to a surface precovered with SLS, the nonionic surfactant adsorbed readily and displaced the anionic surfactant. Thus, there was a strong preferential adsorption in favor of the nonionic surfactant. The results of Kronberg et al. were corroborated by Bolze et al.15 using small-angle X-ray scattering (SAXS) as a technique to probe the adsorption of mixtures of SLS and Triton X-405 on the surface of polystyrene latex particles. The latex particles were covered with SLS and increasing amounts of Triton X-405 were added. Most of the Triton adsorbed and replaced the SLS on the surface. When the particles were saturated with Triton X-405 and increasing amounts of SLS were added, the SLS also was able to remove some of the Triton X-405 from the surface. The objective of the current study is to obtain a better qualitative and quantitative understanding of the mixed adsorption of SLS and Triton X-405 on polystyrene particles, as a prelude to conducting emulsion polymerizations with this pair of stabilizers. First, the adsorption isotherms of SLS and Triton X-405 used as sole emulsifiers are determined. Then various techniques are used to determine the competitive adsorption of a 1/1 molar mixture of SLS and Triton X-405. The adsorption of Triton X-405 on PS particles precoated with SLS is next reported. Similarly, the adsorption of SLS on PS particles precoated with Triton X-405 is performed. As a comparison, the adsorption isotherm of SLS on PS particles on which PEO segments are grafted is determined. All the adsorption results are correlated with the competitive desorption of the two emulsifiers (using “serum replacement”). The surface packing is estimated in all experiments. Experimental Section Materials. The anionic surfactant used in this study was Ultrapure Bioregeant sodium lauryl sulfate (99+% pure, J.T. Baker, Inc.). The cmc of this surfactant in water at 25 °C was experimentally determined to be 6.0 mM. No further purification was performed. The nonionic surfactant is a 70% aqueous solution of Triton X-405 (octylphenoxy poly(ethylene oxide) with an average number of ethylene oxide (EO) units of 40, Aldrich). The nonionic surfactant is polydisperse, the number of EO units ranging from 20 to 60. The nonionic emulsifier was dried at 70 °C for 24 h prior to use. The cmc of this dried surfactant in water at 25 °C was experimentally determined to be 0.8 mM. The monodisperse polystyrene (PS) latex used (LS-1039E, The Dow Chemical Co.) has a number-average diameter of 92 nm (determined by electron microscopy16). In one experiment, this monodisperse polystyrene latex was modified by grafting PEO segments on the surface via a seeded copolymerization. For this synthesis, styrene monomer (Aldrich) was washed with a 10 wt % NaOH solution to remove the inhibitor and distilled no more than 2 weeks before the polymerization was run. The comonomer, poly(ethylene glycol) methyl ether methacrylate (PEGMA, 50% solution in water, Mn ) 2080 g/mol Aldrich), was used as received. Prior to use, the latexes (5.7% solids) used in these adsorption studies were cleaned for 3 weeks by serum replacement to remove any surfactant and electrolyte remaining from their preparation. The surface charge density of this cleaned latex was 2.5 µC/cm2. The conductivity of the effluent was monitored to determine the extent of cleaning. The water used was deionized (DI). D2O (Cambridge Chemical Laboratories) and chloroform (99%+, (15) Bolze, J.; Horner, K. D.; Ballauf, M. Colloid Polym. Sci. 1996, 274, 1099. (16) Miller, C. M. Ph.D. Dissertation, Lehigh University, 1995.

Colombie´ et al. Aldrich) were used as received. Diimidium bromide (Aldrich), sodium chloride (Aldrich), and the citrate buffer solution (Aldrich) were used as received. Synthesis and Characterization of the PEO Grafted Polystyrene Latex. The seeded emulsion polymerization was carried out as follows. The seed latex (100 g of uncleaned latex, 20% solids) was filtered to remove any large aggregates and poured into a glass bottle. The surfactant (0.07 g of sodium lauryl sulfate) and buffer (0.05 g of sodium bicarbonate) were added, and the mixture was stirred at 300 rpm for 2 h using a magnetic bar. The monomers (0.80 g of styrene and 1.60 g of PEGMA) were then added to swell the latex particles with overnight mixing. The initiator (0.22 g of potassium persulfate) was then added, and the bottle was capped and placed in a bottle polymerizer at 60 °C for 24 h. The resulting latex was cleaned for 3 weeks by serum replacement to remove the surfactant and impurities in the latex. The serum was clear during the whole course of the cleaning process, suggesting the absence of any secondary nucleation during the synthesis. This was confirmed by transmission electron microscopy (Phillips 400 transmission electron microscope). A sample of the resulting latex was dried in an oven at 70 °C for 24 h. The sample was redissolved in deuterated chloroform for 1H NMR analysis (details given in the following section). Resonances for the protons of the polystyrene backbone chain (-CH(Ph)-CH2-) were visible between 1.2 and 2.2 ppm. Aromatic protons resonances were observed between 6.5 and 7.5 ppm. The solvent peak (chloroform) was located at 7.2 ppm. A peak at 3.6 ppm representing the resonance of ethylene oxide protons (-O-CH2-CH2-) was noted. These data confirmed that the grafting of some PEGMA onto the PS particles was successful during this reaction. From the NMR spectrum, it was possible to evaluate the fraction of grafted PEGMA by integration of the phenyl protons and the ethylene oxide protons. From the ratio of the areas of these peaks, the fraction of grafted PEGMA was estimated to be 45% of the initial amount (i.e., 1.74 × 10-4 mol). This amount corresponds to a density of 8.5 × 1016 molecules of PEGMA per m2 of particle surface. Determination of the Adsorption Isotherms. Filtration. This technique is based on the extraction a small volume of the aqueous phase from a latex for surfactant analysis. To reduce the perturbations created by the extraction of the samples containing varying amounts of surfactant, the following procedure was applied. The desired amount of surfactant was added to approximately 25 g of cleaned monodisperse polystyrene latex diluted to 1.9% solids. The sample was allowed to equilibrate by stirring for 24 h before the extraction was performed. A 50 cm3 serum replacement cell (Advantec MFS) with a polycarbonate membrane (pore size 100 nm, Osmotics) was rinsed thoroughly with DI water and allowed to dry for 24 h. The latex sample was poured in the cell, and approximately 3 cm3 of the aqueous phase was extracted and analyzed. Desorption by Serum Replacement. In some cases, the technique developed by Ahmed et al.6 was used. The experimental procedure is as follows. In experiment DC-SR1, SLS (0.23 g) was added to a 100 g sample of cleaned monodisperse polystyrene latex diluted to 1.9% solids. In experiment DC-SR2, SLS and Triton X-405 were added to the latex in combination (0.38 and 0.55 g, respectively). The samples were stirred for 24 h prior to the experiment. A 400 cm3 serum replacement cell (Advantec MFS) with a polycarbonate membrane (pore size 100 nm, Osmotics) was rinsed thoroughly with DI water and allowed to dry for 24 h. The latex was then charged into the serum replacement cell. DI water was passed through the cell, and the surfactant content in the exit stream was monitored over approximately 8 h. The amounts of water extracted from the cell were carefully weighed to perform an accurate mass balance on the system. Surfactant Concentrations. The SLS concentration in the aqueous phase samples was monitored by two techniques. The conductivity of the aqueous solutions was determined at 25 °C using a YSI model 32 conductance meter, and the SLS concentration was determined via a calibration curve. However, the conductivity of mixed surfactant solutions can be affected by the presence of micelles, which cannot be readily calibrated. As a consequence, colorimetry was also used to determine the SLS concentration.

SLS and Triton X-405 on Polystyrene Latex The colorimetry technique used here, first reported by Orthgieb and Dobias,17 allows one to measure the concentration of sulfate or sulfonate groups and was successfully applied by Urquiola et al.18 to monitor the concentration of the reactive surfactant, sodium dodecyl allyl sulfosuccinate, in an emulsion polymerization. A 200 mL solution of diimidium bromide dye was prepared as follows: 0.1 g of diimidium bromide and 5.0 g of sodium chloride were mixed in 100 mL of citrate buffer solution (pH ) 4) and diluted to 200 mL with DI water. Surfactant solution samples were diluted 10, 15, or 40 times so that their concentration ranged from 2 × 10-5 to 3 × 10-4 M. A 2.0 g sample of dye solution was added to 2.0 g of the diluted surfactant solution. The sample was shaken vigorously for 2 min to allow the formation of the dyeSLS complex. A 3.0 g sample of chloroform was added. The mixture was stirred for 3 min to allow the extraction of the dyeSLS complex into the oil phase. The chloroform phase (bottom of the tube, purple color) was separated from the aqueous phase. The optical absorbance of the chloroform phase at a wavelength of 525 nm was measured against a reference solution of pure chloroform using a Spectronics Genesys 2 UV/vis spectrometer. The Triton X-405 aqueous concentration was monitored by measuring the absorbance of the phenyl ring (π f π* transitions) in the UV range at 223 and 274 nm using the UV/vis spectrometer. NMR Technique. This technique was described in detail in a previous article.12 Due to a large water signal in the proton NMR spectrum of the latex itself, the polystyrene latex had to be prepared carefully by replacing the water with D2O. An 8.73 g amount of cleaned monodisperse latex (5.74% solids) was placed in a serum replacement cell with 16.32 g of D2O. Three successive additions of 35 g of D2O were performed, and 35 g of the serum was recovered after each addition so that most of the water was replaced by D2O (>99%). At the end of the procedure, 24.42 g of polystyrene latex (1.9% solids) in deuterated water was recovered. This procedure was repeated four times so that approximately 100 g of latex was available for sample preparation. Three stock solutions of Triton X-405 in D2O (9.70 × 10-2, 9.22 × 10-3, and 1.06 × 10-3 M) and two stock solutions of SLS in D2O (9.35 × 10-3 and 9.90 × 10-2 M) were subsequently prepared. The different samples were prepared from these solutions to adjust the initial surfactant concentrations. All samples had an approximate volume of 1 cm3. The evaluation is based on the ratios of the peaks to each other. As a consequence, the precise volume was not important. All NMR spectra were obtained using a Bruker 500 MHz spectrometer. The pulse length was 3.0 µs. To minimize the water proton contribution in the spectra, they were presaturated with the frequency of the water peak. To obtain a good signal-to-noise ratio in the spectra, 200-1500 scans were accumulated. The repetition delay was chosen to be 6 s to ensure full relaxation of all protons. Tetramethylsilane (TMS) was used as an external standard.

Langmuir, Vol. 16, No. 21, 2000 7907

polymer particles,19 one of which is the Langmuir approach, which has been used successfully in many cases. This approach is based on the following assumptions. The surface of the particles is energetically uniform. No interaction exists between the adsorbed molecules. All mechanisms for adsorption are identical. The adsorption is up to (and including) a complete monolayer on the surface. Although all of these assumptions are not typically met by surfactant adsorption, the Langmuir model has nonetheless been accepted to provide an adequate description of the process. In the Langmuir adsorption model, eqs 1 and 2 can be derived:

Γ)

asc 1 + bc

(1)

where Γ is the moles of solute adsorbed per unit surface area, c is the aqueous surfactant concentration, and as and b are Langmuir constants. b is the ratio of the rate constant for adsorption over the rate constant for desorption. as is the area occupied by one molecule of surfactant at saturation for a monolayer of emulsifier on the surface of the polymer particles. Equation 1 can be transformed into eq 2, the linear Langmuir form, to allow the experimental determination of as and b.

1 1 ) as + Na bc

(2)

where Na is the number of molecules of SLS adsorbed per unit area when the aqueous concentration of the surfactant is c. The standard molar free energy for adsorption (∆Gads, J/mol) can be calculated from eq 3.

∆Gads ) -RT ln(bω)

(3)

Adsorption Behavior of Each Individual Surfactant. Adsorption of SLS on PS Particles. The adsorption isotherm of SLS on polystyrene particles was determined using the serum replacement technique (experiment DCSR1). The objective of this determination was to characterize the adsorption of SLS by itself for comparison to its adsorption in the presence of Triton X-405. The adsorption isotherm at 25 °C is shown in the Supporting Information (Figure a), where the number of molecules of SLS adsorbed (for a particle surface of 1 m2) is plotted as a function of the aqueous concentration of the surfactant. The adsorption behavior is as expected; at low SLS concentrations most of the SLS is adsorbed with the proportion (SLS adsorbed/SLS aqueous phase) decreasing as the surface approaches saturation. Several models have been used to describe qualitatively and quantitatively the adsorption of surfactants on

where T is the absolute temperature (K) and ω is the number of water molecules per dm3 of water (55.6 mol/ dm3). From the Langmuir adsorption isotherm, the surface area per molecule of SLS is 44 Å2, the adsorption constant is 550 dm3/mol, and the Gibbs adsorption free energy is -25.6 kJ/mol. The surface area per molecule found here is the same as that reported by Ahmed20 for the same system (44 Å2). A similar molar energy for adsorption of SLS on PBMA particles has been reported by Kronberg et al.14 (-23.7 kJ/mol). Adsorption Isotherm of Triton X-405 on PS Particles via Proton NMR. In a recent study,12 the adsorption isotherm of Triton X-405 molecules on polystyrene particles was determined using 1H NMR spectroscopy (Figure b in the Supporting Information). The corresponding Langmuir parameters can be calculated from the data up to and including the plateau region. Above this concentration, the Langmuir approximation is no longer valid due to the greater than monolayer coverage of the particles. From the adsorption isotherm, the surface area per molecule of Triton X-405, the adsorption constant, and the Gibbs free energy for adsorption can be determined. The estimated surface area per molecule (172 Å2) is similar to that found by Kronberg et al.7 for a similar poly(ethylene oxide) type surfactant (200 Å2 for an emulsifier having an average chain length of 50 EO units). The molar energy for adsorption (-36.2 kJ/mol) is similar to that of PEO on PBMA particles as reported by Kronberg et al.14 (-34.1

(17) Orthgieb, E.; Dobias, B. Poster presented at the 7th Symposium on Surfactants in Solution, Ottawa, 1988. (18) Urquiola, M. B. Ph.D. Dissertation, Lehigh University, 1994.

(19) Adamson, W. A. Physical Chemistry of Surfaces; 5th ed.; Wiley: New York, 1990. (20) Ahmed, S. M. Ph.D. Dissertation, Lehigh University, 1979.

Results and Discussion

7908

Langmuir, Vol. 16, No. 21, 2000

Colombie´ et al.

Figure 1. Adsorption isotherm of SLS and Triton X-405 on polystyrene particles (1 m2 surface area) at 25 °C from a 1/1 molar SLS/Triton X-405 mixture obtained via filtration/ conductivity/colorimetry (SLS) and proton NMR (Triton X-405).

kJ/mol). This energy is very low compared to the free energy of adsorption for SLS (-25.6 kJ/mol). Competitive Adsorption of SLS and Triton X-405 on PS Particles. As shown in the previous section, SLS and Triton X-405 have very different adsorption behaviors on the PS particles, when used alone. Triton X-405 has a significantly lower free energy for adsorption than SLS. This energy as well as the possible interaction between the two types of surfactants will be determining parameters in the competitive adsorption of the two emulsifiers. Even though the NMR technique was very powerful in the determination of the adsorption behavior of nonionic surfactants, it was not possible to use it for the determination of the SLS adsorption behavior. The SLS molecules adsorbed on the surface are not mobile enough to be detected within the 5 MHz window of the proton NMR spectrum. At high SLS concentrations, a signal for the aqueous SLS protons was detected, but the correlation to the actual concentration was not successful. As a consequence, a variety of techniques are used in combination for the determination of the competitive adsorption of SLS and Triton X-405 (filtration/colorimetry and filtration/conductivity for SLS; proton NMR and filtration/ UV spectroscopy for Triton X-405). To simplify the experimental determination of the adsorption isotherms, all experiments were performed at 25 °C using a monodisperse polystyrene latex (1.9% solids, 92 nm). All adsorption results are reported for a sample of latex having 1 m2 of surface available for adsorption. Adsorption of a 1/1 Molar Ratio of SLS and Triton X-405. Figure 1 depicts the adsorption isotherms of SLS and Triton X-405 on polystyrene particles at 25 °C from 1/1 molar SLS/Triton X-405 solutions. The first three data points for SLS were determined by colorimetric titration only since the conductivity technique was not sensitive enough to detect the aqueous SLS at extremely low concentrations. However, both techniques are consistent for the determination of the adsorption of SLS at the higher concentrations. At very low surfactant concentrations, both surfactants adsorb strongly on the surface of the polystyrene particles. The slope of the number of adsorbed surfactant molecules as a function of total aqueous surfactant concentration is high. However, this slope decreases significantly above 3 × 10-4 M total aqueous surfactant concentration, indicating less adsorption of both surfactants with further increases in their concentrations. Up to 2 × 10-3 M total aqueous surfactant concentration, Triton X-405 adsorbs preferentially on the surface of the polystyrene particles as more Triton X-405 molecules occupy the surface of the particles than SLS molecules. Between 2 × 10-3 and 3 × 10-3 M total aqueous surfactant

Figure 2. (a) Molar ratio of SLS to Triton X-405 in the overall system (solid line), in the aqueous phase, and on the polystyrene particle surface and (b) percentage of adsorbed SLS and Triton X-405 as a function of the total surfactant concentration in the aqueous phase.

concentration, the adsorption of the two surfactants becomes more competitive. Eventually there is more SLS on the surface than Triton X-405. In Figure 2a, the SLS to Triton X-405 molar ratio in the aqueous phase and on the particle surface is plotted as a function of the total aqueous surfactant concentration. The same trend described above is observed: at low concentrations, the aqueous phase is richer in SLS. As reported in a previous section and in the literature,8 nonionic surfactants such as Triton X-405 have been shown to have a strong affinity for the surface of the particles (∆Gads ) -36 kJ/mol). This has been attributed to an exclusion of the surfactant from the water phase (to prevent hydrophobe-water interactions). Lateral interactions on the particle surface also play an important role in decreasing the total free energy in the system. With increasing concentration, the ratio of the two surfactants in the aqueous phase becomes close to the total ratio of the two surfactants in the system. At intermediate concentrations, the particle surface may be saturated with the nonionic surfactant. Some of the Triton X-405 remains in the water phase and associates with SLS as described in the literature.21 Finally, the presence of mixed micelles affects the partitioning of both surfactants, and more of the Triton X-405 will be present in the aqueous phase. The cmc of a 1/1 molar solution of SLS and Triton X-405 was measured to be 1.3 × 10-3 mol/dm3 total surfactant concentration at 25 °C (determined by surface tension measurements). Of course, this was measured in the absence of particles and may not be the same ratio as when micelles appear in the presence of the particles. Figure 2b depicts the partition of both SLS and Triton X-405 between the particle surface and the aqueous phase as a function of the total aqueous surfactant concentration. (21) Cabane, B. J. Phys. Chem. 1977, 81, 1639.

SLS and Triton X-405 on Polystyrene Latex

Langmuir, Vol. 16, No. 21, 2000 7909

Table 1. Comparison between the Theoretical and Experimental Polystyrene Particle Surface Coverage (for a Particle Surface of 1 m2) at Saturation in the Competitive Adsorption Experiments

experiment

1/1 molar SLS/Triton X-405 adsorption on bare PS particles

Triton X-405 adsorption on SLS covered PS particles

SLS adsorption on Triton X-405 covered PS particles

SLS/Triton X-405 competitive desorption from PS particles

5.2 × 1017 1.0 × 1018 192 4.5 × 1017

7.2 × 1017 1.6 × 1018 222 4.0 × 1017

1.5 × 1018 2.3 × 1018 159 2.1 × 1017

8.0 × 1017 1.5 × 1018 187 3.8 × 1017

8.8 × 1017

9.8 × 1017

3.3 × 1017

7.1 × 1017

194 1.1

244 1.6

155 7.0

187 2.1

nSLS sat theoretical (molecules) nSLS sat experimental (molecules) excess SLS (%) nTriton X-405 sat theoretical (molecules) nTriton X-405 sat experimental (molecules) excess Triton X-405 (%) SLS/Triton X-405 molar ratio at saturation

At low surfactant concentrations, the proportion of Triton X-405 adsorbed is almost constant (around 88%, over a narrow range however), whereas a smaller and increasing proportion of SLS is adsorbed (from 60 to 70%). As the concentration of the surfactants increases, a smaller proportion of both surfactants is adsorbed as might be expected as the surfaces become more crowded. Above 1.5 × 10-3 M, however, the proportion of SLS adsorbed increases again, this being attributed to a cooperative adsorption of the two surfactants. Eventually, above 3.0 × 10-3 M, the proportion of both adsorbed surfactants decreases as the surface becomes fully saturated. To determine the surface packing at saturation, the following approach was used. The theoretical number of molecules of the two surfactants packed on the polystyrene particles (having 1 m2 of available surface area) is calculated, knowing the experimental molar ratio of the two surfactants. This calculation is based on the monolayer unit surface coverage for each surfactant (determined in the previous section). No interaction between the two surfactants is taken into account in the calculation. The number obtained for each surfactant is subsequently compared with the actual number determined experimentally in the adsorption isotherm (in which SLS/Triton X-405 interactions may occur). The excess surfactant on the surface of the particles is then calculated. For a particle surface of 1 m2, the theoretical number of molecules of the two surfactants (nSLS th and nTriton X-405 th, respectively) at saturation in a monolayer without interaction should be described by

nSLS thas0 SLS + nTriton X-405 thas0 Triton X-405 ) 1 (4) where as0 SLS is the surface area occupied by one molecule of SLS on polystyrene particles in the absence of Triton X-405 and as0 Triton X-405 is the surface area occupied by one molecule of Triton X-405 on polystyrene particles in the absence of SLS. At saturation, the molar ratio of the two surfactants (r) can be determined experimentally. By assuming the theoretical ratio would be equal to the experimental ratio at saturation:

r)

nSLS th nTriton X-405 th

(5)

Equations 4 and 5 can be solved, and the result is given in the following equations:

nSLS th )

r ras0 SLS + as0 Triton X-405

nTriton X-405 th )

1 ras0 SLS + as0 Triton X-405

(6) (7)

The excess surfactant (e) on the particle surface is defined as

e)

nsurf exp × 100 nsurf th

(8)

The theoretical and experimental values for the number of molecules of both emulsifiers on the surface of the polystyrene particles (for 1 m2 of surface) are given in Table 1. The results indicate a large excess of both of the surfactants on the particles at saturation when they adsorb simultaneously (>190%). Several possible explanations can be proposed for this cooperative adsorption. When more Triton X-405 is available in the aqueous phase, the association with SLS21 may lead to a greater adsorption of the hydrophobic complex. Just as in the formation of mixed micelles, the exclusion of the SLS/ Triton X-405 complex from the aqueous phase is energetically favored. Furthermore, some interactions on the particle surface can decrease the free energy for adsorption. The electrostatic repulsion between the sulfate end groups of the SLS molecules could be shielded by the presence of the nonionic surfactant, allowing a closer packing on the surface. The Triton X-405 may also reorganize its structure by being in a more extended conformation. In this case, the loss of entropy due to the extended conformation of the polymeric chain can be balanced by energetically favored side chain interactions (limiting water-hydrophobe interactions). Eventually, it is possible that SLS adsorbs in the hairy layer created by the nonionic surfactant in the vicinity of the particle. Its adsorption may be favored by side-chain interactions. Adsorption of Triton X-405 on PS Particles Precovered with SLS. In most industrial emulsion polymerization processes, the surfactants are added at various process times in the reaction. Very often, the latex is synthesized using an anionic surfactant. At the end of the reaction, a nonionic surfactant can then be added to enhance stability of the system. As a consequence, it can be helpful to know what is the adsorption behavior of a nonionic surfactant on latex particles already stabilized with an adsorbed anionic surfactant. In this experiment, 0.3269 g of SLS was first added to 249.52 g of the cleaned monodisperse polystyrene latex (2% solids, 92 nm). The SLS was allowed to adsorb for 24 h to ensure equilibrium. The amount of SLS added to the system is that required to cover 100% of the surface of the particles if all the SLS was to adsorb on the latex particles with a 44 Å2 surface area per molecule. The Triton X-405 was then added to the latex. The mixture was stirred for 24 h before any measurement was performed. Under these conditions, equilibrium was assumed to be established between all the components of the system.

7910

Langmuir, Vol. 16, No. 21, 2000

Colombie´ et al.

Figure 3. Number of molecules of SLS and Triton X-405 on 1 m2 of polystyrene particles (precovered with an amount of SLS sufficient to theoretically cover 100% of the surface) at 25 °C for increasing Triton X-405 concentrations obtained via filtration/colorimetry (SLS) and proton NMR (Triton X-405).

The adsorption isotherm (25 °C) of Triton X-405 on the polystyrene particles precovered with SLS is depicted in Figure 3. At low concentrations, Triton X-405 adsorbs readily on the surface of the particles, despite the presence of SLS at the surface. As a consequence, some of the SLS initially present on the surface is displaced by the Triton X-405 (between 0 and 8 × 10-4 M aqueous Triton X-405) as seen in Figure 3. This displacement reaches a minimum at 2.5 × 10-4 M aqueous Triton X-405. After this point, the SLS begins to readsorb, reaching (and exceeding) its original level of adsorption at about 8 × 10-4 M aqueous Triton X-405. As reported in the previous section, some cooperative adsorption probably occurs as SLS and Triton X-405 increasingly associate in the aqueous phase and on the particle surface. The adsorption of Triton X-405 has been reported to be extremely strong, and the removal of SLS from the surface has already been observed by Bolze et al.15 using SAXS (small-angle X-ray scattering). In Figure 4a, the molar ratios of SLS to Triton X-405 in the whole system, on the surface of the particles, and in the aqueous phase are plotted as a function of the aqueous phase concentration of Triton X-405. At low nonionic surfactant concentrations (below 8.0 × 10-4 M aqueous Triton X-405), the molar ratio of SLS to Triton X-405 in the aqueous phase is significantly higher than the ratio of the two surfactants in the entire system. Figure 4b describes the partition of both SLS and Triton X-405 between the particle surface and the aqueous phase as a function of the Triton X-405 aqueous surfactant concentration. The desorption and readsorption of SLS at low Triton X-405 concentrations is also reflected in this curve; the proportion of adsorbed SLS decreases (from 62% to 53%) and then increases (from 53% to 70%) with increasing concentration of Triton X-405. This indicates that more SLS is finally adsorbed on the particles in the presence of the nonionic surfactant than in its absence. Equations 4-7 were also applied to this adsorption data, and the results are given in Table 1. Once again, a strong excess of both surfactants (more than 200%) is noted. The molar ratios of SLS to Triton X-405 at different locations (aqueous phase, particle surface, whole system) are very close at the high Triton X-405 concentration. The total ratio of the two emulsifiers is close to 1.9, as well as the ratio on the surface (with a slight predominance of SLS over Triton X-405) and in the aqueous phase. Adsorption of SLS on PS Particles Precovered with Triton X-405. The experiment described in the above section can be reversed by fully covering the polystyrene particles with Triton X-405. The adsorption isotherm of SLS is determined subsequently. In this experiment,

Figure 4. (a) Molar ratio of SLS to Triton X-405 in the whole system (total), in the aqueous phase, and on the polystyrene particles (precovered with SLS) at 25 °C and (b) percentage of SLS and Triton X-405 adsorbed for increasing Triton X-405 concentrations obtained via filtration/colorimetry (SLS) and proton NMR (Triton X-405).

Figure 5. Number of molecules of SLS and Triton X-405 on 1 m2 of polystyrene particles (precovered with an amount of Triton X-405 sufficient to cover 100% of the surface) at 25 °C for increasing SLS concentrations obtained via filtration/ colorimetry (SLS) and proton NMR (Triton X-405).

0.4832 g of Triton X-405 was added to 250.02 g of a cleaned monodisperse polystyrene latex (2% solids, 92 nm). This amount of Triton X-405 added to the system was enough to cover 100% of the surface of the particles if all the Triton X-405 was to adsorb on the latex particles with a 170 Å2 area per molecule. The Triton X-405 was allowed to adsorb for 24 h. The SLS was then added to the latex. The mixture was stirred for another 24 h before any measurement was performed. As in the previous section, a dynamic equilibrium is assumed to be established between all the components of the system. The adsorption isotherm (25 °C) of SLS on the polystyrene particles precovered with Triton X-405 is given in Figure 5. No Triton X-405 was removed from the surface of the polystyrene particles when SLS was added to the system even though SLS is shown to adsorb. The results given by NMR and UV spectroscopy are consistent, even

SLS and Triton X-405 on Polystyrene Latex

Langmuir, Vol. 16, No. 21, 2000 7911

Figure 7. Number of molecules of SLS on 1 m2 of polystyrene particles (precovered with Triton X-405 or functionalized with PEO segments) at 25 °C for increasing SLS concentrations obtained via filtration/colorimetry.

Figure 6. (a) Molar ratio of SLS to Triton X-405 in the whole system (total), in the aqueous phase, and on the polystyrene particles (precovered with Triton X-405) and (b) percentage of SLS and Triton X-405 adsorbed at 25 °C for increasing SLS concentrations obtained via filtration/colorimetry (SLS) and filtration/UV spectroscopy (Triton X-405).

though some systematic discrepancy is noted between the two techniques. These results confirm that the Triton X-405 is bound tightly to the polymer surface, once it has been adsorbed. The addition of another adsorbing species does not seem to affect its affinity for the surface. This should be important with respect to the resulting latex properties particularly in terms of stability, especially when the latex is formulated with a lot of additives. However, it may be a problem when the latex must be coagulated by chemical means or when the surfactant must be removed for other studies. The adsorption of the SLS is not very strong as shown in Figure 5. The SLS continues to adsorb on the particles up to the highest concentration used here ([SLS]aq ) 5 × 10-3 M). This is evidence for cooperative adsorption. As expected, the aqueous phase also becomes richer in SLS molecules. In Figure 6a, the molar ratios of SLS to Triton X-405 in the whole system, on the surface of the particles, and in the aqueous phase are plotted as a function of the aqueous phase concentration of Triton X-405. Since the amount of Triton X-405 on the surface is not varying, the main factor determining the SLS to Triton X-405 ratio in the aqueous phase and on the particle surface is the adsorption behavior of the SLS. Figure 6b describes the partitioning of both SLS and Triton X-405 between the particle surface and the aqueous phase as a function of the SLS aqueous surfactant concentration. No variation in the Triton X-405 partitioning is observed over the range of concentrations studied. The SLS partitioning also reflects the cooperative adsorption that takes place in the system. Once again, eqs 4-7 were applied to these data, and the results are given in Table 1. As before, a strong excess of adsorbed surfactants is noted when mixtures of surfactant are used. The surface packing is greatly increased

when both surfactants are used in combination. The proposed mechanism for this cooperative adsorption is similar to the one described in the previous sections. Adsorption of SLS on a PS Latex with Grafted PEO Segments. The previous section showed that SLS adsorbed on a PS latex precoated with Triton X-405. The nonionic surfactant was physically bound to the surface during this experiment as no desorption of Triton X-405 was noted over the range of concentrations studied. A comparative experiment was performed by determining the adsorption isotherm of SLS on a PS latex with chemically bound PEO segments (Figure 7). The chemical nature of the grafted PEO segments is close to the Triton X-405 based on the average number of ethylene oxide units (45 for the PEGMA, as opposed to 40 for the Triton X-405). However, the end group is a methyl group, as opposed to the hydroxyl at the end of the chain in Triton X-405. For a sample having 1 m2 of particle surface area, there are 8.5 × 1016 molecules of grafted PEO segments. This is lower than the density of Triton X-405 precoating the latex particles in the previous section (4.4 × 1017 molecules). As noted in Figure 7, SLS adsorbed on the surface of the polystyrene particles, regardless of whether the PEO segments were physically or chemically bound. When the nonionic emulsifier was used, the adsorption of the SLS was lower than in the case of the latex functionalized with PEO segments, where a somewhat stronger initial adsorption was noted. However, this difference may reside in the number of PEO chains at the surface of the particles (lower density). The initial adsorption on the bare surface is higher for similar reasons. At higher concentrations, the adsorption on the bare surface approaches saturation. When chemically or physically bound PEO are used, an increased adsorption of the SLS is noted. These results suggest that the adsorption of the SLS may not occur directly at the surface of the particles in the presence of PEO segments. Instead, SLS/PEO complexes21 may form at the interface and a multilayer adsorption phenomenon is likely to occur. Competitive Desorption. An increased understanding of the adsorption behavior of surfactants can be obtained by desorbing them from the surface of the latex particles (92 nm, 1.9% solids). In experiment DC-SR2, the initial total concentrations were 13 and 3 mM for SLS and Triton X-405, respectively. To perform the competitive desorption experiment, two separate methods of monitoring the surfactant concentration must be used: conductivity for SLS and UV spectroscopy for Triton X-405. The concentration profile in the exit stream from the serum replacement cell for experiment DC-SR2 is plotted in Figure c of the Supporting Information as a function of the number

7912

Langmuir, Vol. 16, No. 21, 2000

Colombie´ et al.

(ranging from 2.1 to 0.2). In terms of the surface composition, the trend observed in this desorption experiment is very similar to that observed in the adsorption experiments. The surface coverages were calculated as before, and the results are given in Table 1. Once again, a strong excess of adsorption (around 185%) is noted when mixtures of surfactant are used. The results are consistent with the data obtained from the adsorption isotherms. Conclusions

Figure 8. Adsorption isotherms of SLS and Triton X-405 on PS particles obtained by the desorption technique at 25 °C.

Figure 9. Ratio of SLS to Triton X-405 in the overall system (total), in the aqueous phase, and on the particle surface as a function of the total surfactant concentration in the aqueous phase obtained by desorption at 25 °C.

of residence volumes (q/v, where q is the volume of serum recovered and v is the volume of the aqueous phase contained in the cell). The desorption rates are different for the two surfactants. SLS has a high rate of desorption compared to Triton X-405. As a consequence, the ratios and amounts of the two surfactants will vary simultaneously. Even though, it is not possible to control this variation, the partition of the two surfactants can still be determined by mass balance for the different ratios and amounts sampled during the experiment. The number of molecules of both surfactants adsorbed on the surface of the polystyrene particles at 25 °C as a function of the total aqueous surfactant concentration is given in Figure 8. The data obtained from the desorption experiment are very similar to the data obtained through the adsorption studies. At low surfactant concentrations, there is a preferential adsorption of Triton X-405 on the surface. Little desorption of the nonionic surfactant is observed during the experiment. Above 5 × 10-3 M, SLS is in the majority on the surface of the particles. This preferential adsorption can be characterized by comparing the molar ratio of the two surfactants at the particle surface and in the water phase to the total ratio of the two surfactants in the system, since there is a continuous variation of the total composition. These ratios are plotted in Figure 9. The initial total molar ratio of SLS to Triton X-405 (before desorption) is 4.5. The total molar ratio of SLS to Triton X-405 decreases with decreasing surfactant concentration (from 4.0 to 1.3), since the anionic surfactant desorbs more readily (as mentioned above). As a consequence, the aqueous phase is rich in SLS (SLS/Triton X-405 molar ratio ranging from 8 to 12) as compared to the whole system. The strong preferential adsorption of the nonionic surfactant is shown by the low SLS to Triton X-405 ratio at the surface as compared to the total ratio in the system

In this work, the competitive adsorption of mixtures of SLS and Triton X-405 on PS particles was studied at 25 °C. First, the individual adsorption behaviors of SLS and Triton X-405 were determined. The adsorption isotherm of SLS was determined using the desorption technique. The adsorption data were fitted to the Langmuir equation. The area occupied by one molecule of SLS on the PS particles was determined to be 44 Å2, and the Gibbs free energy for adsorption was -26 kJ/mol. The adsorption isotherm of Triton X-405 was monitored by a quantitative nondisruptive technique based on proton NMR. At low surfactant concentrations, the adsorption data were fitted to the Langmuir equation. The area occupied by one molecule of Triton X-405 on the PS particle surface was 172 Å2, and the Gibbs free energy for adsorption was -36 kJ/mol, showing a lower adsorption energy than SLS. Deviations from the Langmuir model were observed at higher concentrations. This was possibly due to incipient phase separation between the surfactant molecules in the aqueous phase (containing micelles) and those at the surface of the particles. Other possible reasons are the adsorption of a second layer or an effect of the polydispersity of the nonionic surfactant. In the adsorption from 1/1 molar ratio solutions of SLS and Triton X-405, the following features were noted. At low surfactant content, there was a strong adsorption of Triton X-405 and SLS, the nonionic surfactant being preferentially adsorbed at the surface of the particles. At higher concentrations, the surface composition approached the initial 1/1 molar ratio. A strong excess of adsorbed surfactant was noted when both surfactants were used together. When Triton X-405 was adsorbed on PS particles precovered with SLS, the anionic emulsifier was removed from the surface at low Triton X-405 contents, due to its higher energy of adsorption. With further increases in the Triton X-405 concentration, the anionic surfactant readsorbed, and a large excess of both adsorbed surfactants was noted at the surface. When SLS was adsorbed on particles precovered with Triton X-405, the amount of adsorbed anionic emulsifier was small at first. It increased with increasing SLS concentration as the surface packing increased and cooperative adsorption took place. The Triton X-405 was not directly affected by the adsorption of SLS. Similarly, SLS was shown to adsorb on a PS latex surface containing grafted PEO segments. In the competitive desorption of SLS and Triton X-405, the same trends were noted. At high emulsifier contents, there was a cooperative adsorption of the two surfactants. The anionic surfactant desorbed at a higher rate than the nonionic surfactant and the surface of the particles was richer in Triton X-405 at low emulsifier concentrations. From these experiments, a qualitative mechanism for the competitive adsorption of SLS and Triton X-405 on PS particles is proposed as illustrated schematically in Figure 10. At low surfactant concentrations, both SLS and Triton X-405 exist as individual molecules, and the surface of the particles is largely free of surfactants. As

SLS and Triton X-405 on Polystyrene Latex

Figure 10. Schematic of the proposed mechanisms of competitive adsorption of SLS and Triton X-405 on PS particles.

a consequence, the surfactant with the lowest energy for adsorption (i.e., Triton X-405) adsorbs strongly and preferentially. There is little Triton X-405 remaining in the aqueous phase, in which SLS is predominant. As the surfactant concentration increases, the surface becomes saturated with Triton X-405. More of the Triton X-405 is available in the aqueous phase, and it can associate with SLS to form hydrophobic complexes as described in the literature.21 In the absence of latex particles, the favored configuration for these complexes would be to form aggregates of mixed composition in order to lower the free energy of the system. The formation of mixed micelles has been shown to be a nonideal process, which occurs at emulsifier concentrations lower than what would be expected considering the cmc of each individual surfactant.22 In the present study, an additional effect should be added. The surface of the particles is a favorable location for the surfactant complexes in terms of decreasing free (22) Rubingh, D. N. In Solution Chemistry of Surfactants; Mittal, K., Ed.; Plenum Press: New York, 1979.

Langmuir, Vol. 16, No. 21, 2000 7913

energy. As the surface is already crowded with Triton X-405, a reorganization must occur to allow the cooperative adsorption to take place. The packing of the surfactant molecules can be increased according to the following mechanisms. The charge on the anionic surfactant may be shielded by the presence of the nonionic surfactant. The limiting factor in terms of packing of an anionic surfactant on a polymer particle is the electrostatic repulsion between the charged end groups of two adjacent molecules. In the presence of a nonionic surfactant, this repulsion effect may decrease, allowing for better packing. Another possible explanation is a reorganization of the nonionic surfactant chains. At low concentrations, nonionic surfactants have been shown to adsorb in a relatively flat conformation on the surface of polystyrene. At higher concentrations, a more extended conformation has been observed.6 This occurrence is not favored in terms of entropy. However, it allows better hydration of the poly(ethylene oxide) segments and side chain interactions. Eventually, the adsorption of the two surfactants may lead to a complex multilayer structure. The SLS may adsorb in the adsorbed Triton X-405 layer, either by interacting directly with the hydrophobe of the nonionic emulsifier or by lateral side chain interactions. In terms of stability during the course of an emulsion polymerization, the combination of the two surfactants may lead to either enhanced or decreased stability. The combination may be powerful if the anionic surfactant content at the particle surface is initially high and the nonionic surfactant is added during or after the polymerization. However, the addition of a little nonionic surfactant on particles covered with a little anionic surfactant may be detrimental, since the Triton X-405 was shown to remove some of the SLS at low surfactant contents. Acknowledgment. Financial support from Elf Aquitaine and Elf Atochem and the research assistance of undergraduate students Keisha Antoine and Vincent Colombie´ are greatly appreciated. Supporting Information Available: Adsorption isotherms of SLS on PS particles and profile of aqueous concentrations of SLS and Triton X-405. This material is available free of charge via the Internet at http://pubs.acs.org. LA9915825