Adsorption Properties of Polyelectrolyte−Surfactant Complexes on

Jul 19, 2006 - Department of Chemistry, Surface Chemistry, The Royal Institute of Technology, SE-100 44 Stockholm, Sweden, and Institute for Surface C...
0 downloads 8 Views 85KB Size
Langmuir 2006, 22, 7639-7645

7639

Adsorption Properties of Polyelectrolyte-Surfactant Complexes on Hydrophobic Surfaces Studied by QCM-D Ali Naderi*,† and Per M. Claesson†,‡ Department of Chemistry, Surface Chemistry, The Royal Institute of Technology, SE-100 44 Stockholm, Sweden, and Institute for Surface Chemistry, P.O. Box 5607, SE-114 86 Stockholm, Sweden ReceiVed April 25, 2006. In Final Form: June 21, 2006 Adsorption and deposition from turbid solutions are common in many industrial processes but notoriously difficult to investigate using standard optical techniques such as ellipsometry and reflectometry. In this report, we have addressed this problem by employing a quartz crystal microbalance with dissipation monitoring ability, QCM-D. The system under investigation consisted of a cationic polyelectrolyte, poly(vinylamine), PVAm, and an anionic surfactant, sodium dodecyl sulfate, SDS, which were mixed together in 10 mM NaCl solution. The polyelectrolyte and the surfactant readily associate in bulk solution, resulting in increased solution turbidity once large aggregates are formed. The solutions were placed in contact with a polystyrene surface, and the adsorption process was monitored by following the changes in the resonance frequency and dissipation factor. The results obtained can in most cases be evaluated using the Sauerbrey relation, but in some cases a more elaborate analysis is necessary. It is found that PVAm adsorbs to polystyrene in the absence of SDS. In the turbid region, deposition is observed, and the sensed mass exceeds the sum of that obtained for each of the components alone. On the other hand, at high SDS concentrations, the surfactant dominates in the adsorbed layer. Adsorption equilibrium is in most cases established within 1-2 h, the exception being found around the solution composition that results in the formation of charge-neutralized aggregates. In this case, a slow deposition of aggregates persists over prolonged times.

1. Introduction Mixed polymer-surfactant systems attract much research interest due to their intriguing properties and wide commercial usage. Among application areas, one can point out health and personal care and pharmaceutical and mining applications, where simultaneous introduction of polymers and surfactants provide an effective tool for altering processing conditions. The area has seen considerable progress in recent years due to new theoretical and modeling approaches1-12 suitable for unravelling the mechanisms behind the interactions between polymers and surfactants and the influence of structural and solution properties. However, still much work remains to be done in this area, not only due to the ever-growing application areas but also because of the introduction of new surfactant and polymer classes and the complexities that thereby follow. A system class that has gained particular attention is that consisting of polyelectrolytes mixed with oppositely charged surfactants. Most efforts have been focused on the interaction of polyelectrolyte-(oppositely charged) ionic surfactant in bulk solution where these components associate through electrostatic and hydrophobic forces.13-16 The association between one * To whom correspondence should be addressed. † The Royal Institute of Technology. ‡ Institute for Surface Chemistry. (1) Wallin, T.; Linse, P. Langmuir 1996, 12, 305-314. (2) Wallin, T.; Linse, P. J. Phys. Chem. 1996, 100, 17873-17880. (3) Wallin, T.; Linse, P. J. Phys. Chem. 1997, 101, 5506-5513. (4) Wallin, T.; Linse, P. Langmuir 1998, 14, 2940. (5) Wallin, T.; Linse, P. J. Chem. Phys. 1998, 109, 5089-5100. (6) Akinchina, A.; Linse, P. Macromolecules 2002, 35, 508-519. (7) Akinchina, A.; Linse, P. J. Phys. Chem. B 2003, 107, 8011-8021. (8) Jonsson, M.; Linse, P. J. Chem. Phys. 2001, 115, 10975-10985. (9) Jonsson, M.; Linse, P. J. Chem. Phys. 2001, 115, 3406-3418. (10) Nguyen, T. T.; Shklovski, B. I. J. Chem. Phys. 2001, 114, 5905-5916. (11) Schiessel, H.; Bruinsma, R. F.; Gelbart, W. M. J. Chem. Phys. 2001, 115, 7245-7252. (12) Skepo¨, M.; Linse, P. Macromolecules 2003, 36, 508-519. (13) Hayakawa, K.; Kwak, J. C. T. J. Phys. Chem. 1982, 86, 3866-3870. (14) Hayakawa, K.; Kwak, J. C. T. J. Phys. Chem. 1983, 87, 506-509.

polyelectrolyte chain and surfactants results in the formation of a complex, while the entity formed by any number of polyelectrolyte chains and surfactants is usually referred to as an aggregate. A cooperative association process normally starts at a well-defined surfactant concentration, termed the critical aggregation concentration, CAC, that typically is one or several orders of magnitude lower than the critical micelle concentration (CMC) of the surfactant. Monte Carlo simulations and mean field calculations have shown that parameters such as the polyelectrolyte concentration, its charge density, flexibility, and hydrophobicity are variables that are of importance for the CAC of the system.1-5 It was concluded that an increased polyelectrolyte charge density and flexibility lowers the CAC, and that the association process is both energetically and entropically favorable.1 Simulations8,9 and theoretical predictions10,11 have also shown that spherical macroions (models for charged micelles) bind to polyelectrolyte chains in such amount that they in the plateau region strongly overcompensate the charge of the polyelectrolyte chain. It has been proposed that this is due to electrostatic correlation forces.10 Modeling of mixtures of polyelectrolytes and macroions has also suggested that the sequence of events that occurs as more and more polyelectrolytes are added to a macroion solution is as follows: formation of stable clusters, phase separation, and redispersion.12 Experiments are largely in line with the predictions mentioned above. For instance, when the charge of the polyelectrolytesurfactant complex becomes reduced, flocculation into larger aggregates occurs. At this stage, the solution becomes noticeably turbid to the naked eye. In a large excess of surfactant (or polyelectrolyte), a clear solution is often, but not always, formed. However, an experimental difficulty which is not captured by (15) Hayakawa, K.; Santerre, J. P.; Kwak, C. T. Macromolecules 1983, 16, 1642-1645. (16) Lindman, B.; Thalberg, K. Interactions of Surfactants with Polymers and Proteins, Goddard, E. D., Ananthapadmanabhan, K. P., Eds. CRC Press: Boca Raton, FL, 1993.

10.1021/la061118h CCC: $33.50 © 2006 American Chemical Society Published on Web 07/19/2006

7640 Langmuir, Vol. 22, No. 18, 2006

modeling is that mixtures of polyelectrolytes and surfactants are prone to be trapped in long-lived nonequilibrium states, and once large aggregates are formed, they are difficult to redisperse.17-19 It is now well established that overcharging of the polyelectrolyte occurs at high surfactant concentrations due to incorporation of excess surfactant, a process that, besides electrostatic correlation forces, is driven by hydrophobic interactions. The large aggregates formed in the presence of high charge density polyelectrolytes most often have an ordered internal structure where surfactant aggregates are packed in hexagonal, lamellar, or cubic structures connected by polyelectrolyte chains.20-24 This demonstrates that the shape of the surfactant aggregates associated with the polyelectrolyte in many cases, but not all, is different to that of spherical micelles, and this aspect has as yet not been investigated by simulation methods. The excess surfactant has been suggested to be accommodated in an outer layer, a suggestion that is based on the observation that the electrophoretic mobility of the aggregates changes as more surfactants are incorporated, whereas the internal structure remains unaffected as judged from the unaltered position of Bragglike small angle scattering peaks.21,25 In contrast to the abundant work done on polyelectrolytesurfactant association in bulk, the literature on the adsorption properties of preformed aggregates at solid-liquid interfaces is scarce, but some experimental reports can be found.19,26-33 Here one has to consider both how the polyelectrolyte and the surfactant interact with the surface and how they interact with each other. Tilton et al.34 have suggested a simple and useful classification scheme to distinguish between the major different interaction possibilities. In their scheme, a class I system is defined as a system where the surfactant has an affinity for the polymer and the surface is selective, i.e., only the surfactant or the polymer, but not both, has an affinity for the surface. A class II system is defined as one where the surfactant binds to the polymer and the surface is nonselective. In a class III system, the surfactant has no affinity for the polymer and the surface is selective; and a class IV system constitutes a system where the surfactant does not bind to the polymer and the surface is nonselective. Ellipsometry and reflectometry are powerful techniques for evaluating adsorption properties of polymers and surfactants. However, these instruments are limited to systems where the turbidity of the solution is not an issue. In this paper, we report (17) Naderi, A.; Claesson, P. M.; Bergstro¨m, M.; Dedinaite, A. Colloids Surf. A 2005, 253, 83-93. (18) Naderi, A.; Claesson, P. M. J. Disp. Sci. Technol. 2005, 26, 329-340. (19) Dedinaite, A.; Claesson, P. M. Langmuir 2000, 16, 1951-1959. (20) Dedinaite, A.; Claesson, P. M.; Bergstro¨m, M. Langmuir 2000, 16, 52575266. (21) Claesson, P. M.; Bergstro¨m, M.; Dedinaite, A.; Kjellin, M.; Legrand, J. F.; Grillo, I. J. Phys. Chem. B 2000, 104, 11689-11694. (22) Bergstro¨m, M.; Kjellin, U. R. M.; Claesson, P. M.; Pedersen, J. S.; Nielsen, M. M. J. Phys. Chem. B 2002, 106, 11412-11419. (23) Bergstro¨m, L.-M.; Kjellin, U. R. M.; Claesson, P. M.; Grillo, I. J. Phys. Chem. B 2004, 108, 1874-1881. (24) Bastardo, L.; Garamus, V. M.; Bergstro¨m, M.; Claesson, P. M. J. Phys. Chem. B 2005, 109, 167-174. (25) Dedinaite, A.; Meszaros, R.; Claesson, P. M. J. Phys. Chem. B 2004, 108, 11645-11653. (26) Esumi, K.; Masuda, A.; Otsuka, H. Langmuir 1993, 9, 284-287. (27) Yamanaka, Y.; Esumi, K. Colloids Surf. A 1997, 122, 121-133. (28) Bury, R.; Desmazieres, B.; Treiner, C. Colloids Surf. A 1997, 127, 113124. (29) Claesson, P. M.; Fielden, M.; Dedinaite, A.; Brown, W.; Fundin, J. J. Phys. Chem. B 1998, 102, 1270-1278. (30) Rojas, O. J.; Claesson, P. M.; Berglund, K. D.; Tilton, R. D. Langmuir 2004, 20, 3221-3230. (31) Pagac, E. S.; Prieve, D. C.; Tilton, R. D. Langmuir 1998, 14, 2333-2342. (32) Terada, E.; Samoshina, Y.; Nylander, T.; Lindman, B. Langmuir 2004, 20, 1753-1762. (33) Terada, E.; Samoshina, Y.; Nylander, T.; Lindman, B. Langmuir 2004, 20, 6692-6701. (34) Braem, A. D.; Prieve, D. C.; Tilton, R. D. Langmuir 2001, 17, 883-890.

Naderi and Claesson

on adsorption/deposition from aqueous solutions containing polyelectrolytes and oppositely charged surfactants. Since these solutions are highly turbid over a large composition range, we have employed the QCM-D technique which works equally well in transparent and turbid solutions. The bulk properties of the same system, poly(vinylamine) mixed with SDS, have been described in a previous publication.18 2. Materials and Methods 2.1. Materials. The polyelectrolyte used (obtained as a gift from BASF) was poly(vinylamine), PVAm, with a molecular weight of 90 000 g/mol (supplier specifications). The polydispersity index (obtained by dynamic light scattering measurements) of the PVAm sample in 10 mM NaCl was found to be 0.25. At pH 6-7 and low ionic strength, which are the conditions in our solutions, about 5060% of the segments are charged.35 The surface tension of 20 ppm PVAm (in 10 mM NaCl) was >72 mN/m, indicating the absence of surface active compounds in the sample. The anionic surfactant sodium dodecyl sulfate (SDS), with >99% purity, was obtained from BDH. It was recrystallized twice from ethanol before being used; no minimum in the surface tension vs surfactant concentration isotherm around the CMC of SDS was observed. This indicates high purity of the sample. Sodium chloride, with >99.5% purity, was obtained from Merck and used as received. The water was first pretreated with a Milli-RO 10 Plus system and purified further with a Milli-Q PLUS 185 system. QCM-D sensors with a polystyrene layer, spin-coated on gold, were purchased from Q-Sense AB (Gothenburg, Sweden). The thickness of the polystyrene layer is 30-40 nm (supplier specifications). The sensors are AT-cut crystals with a thickness of 0.3 mm (fundamental frequency ≈ 5 MHz). The equilibrium water contact angle on the polystyrene surface was measured to be 88-90°. 2.2. Mixing Procedure. Stock solutions of PVAm (ca 1000 ppm) and SDS (8.3 and 16.4 mM) in 10 mM NaCl solutions were first made. A premix of polyelectrolyte and 10 mM NaCl was then prepared by adding the polyelectrolyte stock solution to 10 mM NaCl and blending by turning the sample tube upside down a few times. Finally, the surfactant stock solution was added to premix solutions containing the polyelectrolyte. The addition was done slowly in a dropwise fashion, taking ca. 5 s, in such a way that it in the end gave the desired sample concentrations of polyelectrolyte and surfactant. This method of mixing was referred to as STP, surfactant added to the polyelectrolyte, in our previous work focusing on the effect of the mixing process.17,18 Each sample investigated contained 20 ppm PVAm, whereas the amount of SDS was varied between the different samples. We stress that each sample was prepared separately. 2.3. Methods of Investigation 2.3.1. Quartz Crystal Microbalance with Dissipation monitoring (QCM-D). The adsorption properties of the samples were investigated 15 min after their preparation using a commercially available QCM-D apparatus (Q-Sense AB, Gothenburg, Sweden). It allows the simultaneous determination of changes in the resonance frequency (∆f) and dissipation (∆D). During the adsorption process the resonance frequency of the crystal, f0, decreases to a lower value f. If the adsorbed mass is evenly distributed, rigidly attached and small compared to the mass of the crystal, ∆f ) (f0 - f) can be related to the adsorbed mass per unit area (∆m) by the Sauerbrey equation36 ∆m )

C∆f n

(1)

where n is the vibrational resonance overtone number (1, 3, 5, 7) and C is a constant (C ) 0.177 mg m-2 Hz-1 in our device) that describes the sensitivity of the device to changes in mass; n ) 1 (35) van Treslong, B. C. J.; Staverman, A. J. R. Netherlands Chem. Soc. 1974, 93, 171. (36) Saurbrey, Z. Z. Phys. 1959, 155, 206-222.

Polyelectrolyte-Surfactant Complexes

Langmuir, Vol. 22, No. 18, 2006 7641

represents the fundamental frequency which in our case is ≈ 5 MHz. In the measurements reported in this article, we used the 3rd overtone (≈ 15 MHz), if not otherwise stated. The Sauerbrey equation is derived for uniform ultrathin rigid films in a vacuum with material properties indistinguishable from those of the crystal resonator. It has, however, been shown that eq 1 is applicable also for measurements in liquids, provided the adsorbed layer is homogeneous, thin, and rigid.37 Energy losses occur in the crystal and due to the adsorbed material, which, when the driving voltage is turned off, leads to a damping of the oscillation with an amplitude decay rate that is affected by all energy dissipative mechanisms in the system. The dissipation factor D is defined as D)

Edis 2πEst

(2)

where Edis is the dissipated energy and Est is the stored energy during one oscillation cycle. The dissipation change is defined as, ∆D ) D - D0, where D0 denotes the dissipation when the sensor is immersed in solution prior to the start of the adsorption process and D is the dissipation at any given time during the experiment. A large value of ∆D signifies a large energy dissipative power of the adsorbed layer, and this is most often observed for thick and nonrigid layers. The QCM-D sensors were first cleaned in a Deconex11 solution (a cleaning agent, supplied by Go¨teborgs Termometerfabrik) for 5 min before being rinsed with excess Milli-Q water and then placed in a container filled with Milli-Q water for at least 2 h. Next, the sensors were cleaned with ethanol and blow-dried with N2 before being mounted inside the measuring chamber. The set temperature for the measurements was 22 °C. The experiments were initiated by first establishing a stable baseline using the background solution (10 mM NaCl). When this was achieved, the PVAm-SDS sample, in 10 mM NaCl, was injected (about 0.5 mL was flushed through the measuring cell that has a volume of 80 µL). This was done 15 min after mixing of the components. The adsorption process was monitored for at least 1 h, and the measuring chamber was then rinsed twice with 5 mL of 10 mM NaCl; each rinsing step was followed during 10 min. 2.3.2. Dynamic Light Scattering. Dynamic light scattering measurements were performed using a BI-200SM goniometer system connected to a BI-9000AT digital correlator from Brookhaven Instruments and a water-cooled Lexel 95-2 laser with maximum power of 2 W. The intensity of the scattered light was measured at an angle of 90° relative to the incident beam (λ ) 514 nm). The samples were investigated at room temperature (20-23 °C), and the measurements were carried out 15 min after the blending of the final ingredient. The data were analyzed as described previously.17,18 2.3.3. XPS. The adsorbed amount of PVAm on mica and polystyrene surfaces was determined by employing a Kratos AXISHS X-ray photoelectron spectrometer (XPS, also known as ESCA) equipped with a hemispherical analyzer. Photoelectron emission was induced by nonmonochromatic X-rays (Al KR, 1486.6 eV) emitted from a dual anode. Electrostatic lenses were used to collect the photoelectrons that were emitted normal to the surface plane. The quantification on mica was carried out as described previously38 but also including the baseline correction recently suggested,39 using the sensitivity factors for K2p (emanating from the mica crystal) and N1s (from the polyelectrolyte) provided by Kratos. The quantification of the adsorbed amount of polystyrene was achieved by comparing the N1s signal from the adsorbed layer on mica and on polystyrene.38 The samples used for XPS analysis were prepared in the following way. Freshly cleaved mica pieces were immersed in 10-mL beakers containing 20 ppm PVAm and 10 mM NaCl for 1 h. Next the beakers with the surfaces were immersed in 2.5 L of Milli-Q water for 20 (37) Kanazawa, K. K.; Gordon, J. G. Anal. Chim. Acta 1985, 175, 99-105. (38) Rojas, O.; Ernstsson, M.; Neuman, R. D.; Claesson, P. M. J. Phys. Chem. B 2000, 104, 10032-10042. (39) Dedinaite, A.; Ernstsson, M. J. Phys. Chem. B 2003, 107, 8181-8188.

Table 1. Adsorbed Mass and Sensed Mass Evaluated from XPS and QCM-D, Respectively surface

nitrogen atoms/ m2

adsorbed mass (mg/m2)a

sensed mass (mg/m2)

mica polystyrene

5.4 ( 0.3 × 1018 3.4 ( 0.1 × 1018

0.39 ( 0.02 0.24 ( 0.007

1.45 ( 0.15

a

Evaluated from XPS measurements using the molecular weight for the uncharged segment. No chlorine signal was observed on polystyrene whereas up to 10% Cl- ions, counted per adsorbed poly(vinylamine) segment was observed on mica. Each sample was studied in at least triplicates and the values given in the table is the mean of these experiments with the uncertainties showing the maximum and minimum values obtained in repeated experiments.

min. The surfaces were then withdrawn and dried by a nitrogen jet. The adsorption of PVAm on polystyrene surfaces was done inside the QCM-D chamber. After adsorption of the polyelectrolyte for 1 h the surfaces were rinsed 2 times with 10 mM NaCl and 2 times more with Milli-Q water. They were then dried by a nitrogen jet. All surface preparation steps were carried out inside a dust-free laminar flow cabinet.

3. Results We will first consider the adsorption properties of PVAm alone and then that of SDS. Next, we will briefly recapitulate some of the bulk association properties between these components and finally the adsorption/deposition from mixtures of PVAm and SDS will be described. 3.1. Adsorption of PVAm. Adsorption of PVAm on mica was evaluated by means of XPS, whereas PVAm adsorption on polystyrene was evaluated using both XPS and QCM-D. The results are summarized in Table 1. We note that the mica basal plane contains 2.1 × 1018 negative surface sites per square meter, and that the number of charged segments in the plateau region of the adsorption isotherm normally is close to this value when highly charged polyelectrolytes are adsorbed from low ionic strength solutions.38,39 Since each segment in PVAm contains one nitrogen atom, one would expect about 2.1 × 1018 charged segments to be adsorbed. Clearly, more segments than this are present on the surface, which suggests that a considerable fraction of the segments in the adsorbed layer is in the uncharged form. This is supported by the ESCA spectrum that shows that the main peak is located at a binding energy of 400 eV, corresponding to uncharged amines (the peak from charged ammonium groups is located at 402-403 eV).40 The resolution of the spectra obtained with the electrostatic lens was not sufficient to allow a meaningful deconvolution of the N1s peak to separately determine the number of charged and uncharged segments, but this requires high quality spectra obtained by monochromatic X-rays.39 The sensed mass and dissipation change that accompany adsorption of PVAm on a polystyrene surface are illustrated in Figure 1. The adsorption was allowed to proceed from a 20 ppm PVAm solution in 10 mM NaCl. Adsorption equilibrium is reached after about 80 min, giving a sensed mass of ca. 1.45 ( 0.15 mg/m2. The sensed mass is considerably higher than the adsorbed mass, see Table 1. In fact, in our case, water and chloride ions associated with the layer in wet state constitute 80-85% of the sensed mass. Nevertheless, the dissipation change amounts to only 0.4-0.5 × 10-6, a figure that is in good agreement with the values observed for highly charged polyelectrolytes adsorbed on gold crystals.41 (40) Wagner, C. D. Anal. Chem. 1977, 49, 1282. (41) Plunkett, M. A.; Claesson, P. M.; Ernstsson, M.; Rutland, M. W. Langmuir 2003, 19, 4673-4681.

7642 Langmuir, Vol. 22, No. 18, 2006

Naderi and Claesson

Figure 1. Sensed mass and dissipation change as a function of time for adsorption from a 20 ppm PVAm solution in 10 mM NaCl onto polystyrene surfaces. The baseline was obtained in 10 mM NaCl. PVAm was introduced at t ≈ 15 min. The chamber was thoroughly rinsed with 10 mM NaCl two times (at t ≈ 120 and 140 min). Sensed mass (9), dissipation change (0).

These results allow us to conclude that the adsorbed mass of PVAm on polystyrene is low, and the low dissipation value indicates that the adsorbed layer is rather flat. This conclusion is qualitatively consistent with predictions based on the Scheutjens-Fleer theory for adsorption of polyelectrolytes to an uncharged (or weakly charged) surface driven by nonelectrostatic forces.42 In passing, we note that we have measured the force between our polystyrene surface and a negatively charged silica surface using the AFM colloidal probe technique. The results demonstrate that the polystyrene surface is weakly negatively charged, presumably due to some surface oxidation. This conclusion is supported by the presence of a small oxygen peak in the XPS spectrum of the polystyrene surface. The equilibrium water contact angle on the polystyrene surface decreased from close to 90° to ca. 65° after PVAm adsorption. Thus, the adsorption of the polyelectrolyte increased the hydrophilicity considerably, and we suggest that the reduction in interfacial energy between polystyrene and water is the main driving force for the adsorption of PVAm to polystyrene since the small surface charge (normally present on polystyrene43) would require a significantly smaller amount of PVAm for its neutralization. 3.2. Adsorption of SDS. The adsorption isotherm of SDS on polystyrene surfaces from a 10 mM NaCl solution, in terms of sensed mass and dissipation, is shown in Figure 2. At the lowest SDS concentrations, adsorption equilibrium was reached after 10-15 min, whereas not more than a couple of minutes were required at the higher SDS concentrations. The sensed mass and dissipation increase with SDS concentration up to the CMC of SDS, which is 6 mM in 10 mM NaCl.44 As expected, the adsorption is noncooperative. The maximum sensed mass obtained above the CMC was found to be about 1.45 mg/m2. The dissipation remains low at all SDS concentrations. Rinsing with 10 mM NaCl removes essentially all surfactants. 3.3. Bulk Properties of PVAm and SDS Mixtures. The bulk properties of mixtures of PVAm and SDS as a function of mixing protocol, mixing ratio, salt concentration, and time have been reported in detail in a previous publication.18 Here we recapitulate some information that is essential for understanding the adsorption behavior of the mixtures. It was noted that the PVAm-SDS (42) Fleer, G. J.; Cohen Stuart, M. A.; Scheutjens, J. M. H. M.; Cosgrove, T.; Vincent, B. Polymers at Interfaces; Chapman & Hall: London, 1993. (43) Schmitt, F.-J.; Ederth, T.; Weidenhammer, P.; Claesson, P. M.; Jacobasch, H.-J. J. Adh. Sci. Technol. 1999, 13 (1), 79-96. (44) Gunnarsson, G.; Jo¨nsson, B.; Wennerstro¨m, H. J. Phys. Chem. 1980, 84, 3114-3121.

Figure 2. Sensed mass (top) and dissipation change (bottom) as a function of SDS concentration. The adsorption occurs onto polystyrene from a 10 mM NaCl solution. The adsorption time was 1 h. The results in the absence of polyelectrolyte (0) are compared with the ones in the presence of 20 ppm PVAm (9). The results for pure PVAm (20 ppm) are for convenience shown at 0.01 mM SDS (2).

aggregates formed were positively charged below an SDS concentration of 0.33 mM and negatively charged above this value. The internal organization of the large aggregates formed around the charge neutralization concentration consists of SDS bilayers organized in a lamellar structure, whereas the polyelectrolyte is located between the surfactant bilayers.22 The excess SDS present in the aggregates above the charge neutralization concentration is suggested to be located at the outside of the aggregates, a structure that has been proposed also for other polyelectrolyte-surfactant systems.20,45 The initial turbidity reaches a maximum at the charge neutralization concentration (0.33 mM SDS), and the size of the aggregates is also largest around this concentration, see Figure 3. The aggregates flocculate and sediment with time in the SDS concentration range 0.17-2 mM, but stable aggregates are formed at higher (g4 mM SDS) and lower surfactant concentrations e0.12 mM SDS. The stability of the aggregates was determined by monitoring the solution turbidity over a time period of more than 1000 h.18 3.4. Adsorption Properties of PVAm and SDS Mixtures. The sensed mass of the adsorption layer formed from solutions containing 20 ppm PVAm and various amounts of SDS (in 10 mM NaCl) on polystyrene, after 1 h of adsorption, is shown in Figure 2. We note that the presence of SDS does not significantly affect the sensed mass up to a concentration of 0.17 mM, whereas at higher surfactant concentrations (0.33 and 0.83 mM SDS), the (45) Me´szaros, R.; Thompson, L.; Bos, M.; Varga, I.; Gilanyi, T. Langmuir 2003, 19, 609.

Polyelectrolyte-Surfactant Complexes

Figure 3. Hydrodynamic radius (Rh) of PVAm-SDS aggregates in 10 mM NaCl. The values shown are registered 15 min after the preparation of the samples. The dashed line indicates the Rh of pure PVAm in 10 mM NaCl.

Langmuir, Vol. 22, No. 18, 2006 7643

Figure 5. Fraction of sensed mass left on the polystyrene surface after 2 times rinsing with 10 mM NaCl at the end of the adsorption process. The results for pure PVAm (20 ppm) are for convenience shown at 0.01 mM SDS (2).

Figure 4. Sensed mass as a function of time for mixtures of 20 ppm PVAm and different SDS concentrations in 10 mM NaCl. The substrate used for adsorption was polystyrene. The adsorption process was initiated at t ≈ 5 min, and the chamber was thoroughly rinsed with 10 mM NaCl at t ≈ 65 and 75 min. The data shown are for SDS concentrations of 0.08 mM (×), 0.33 mM (0), and 4 mM (4).

Figure 6. Change in dissipation as a function of change in frequency for adsorption of 20 ppm PVAm (in 10 mM NaCl) in the presence and absence of SDS on a polystyrene surface. In absence of SDS (0); in the presence of SDS: 0.17 mM (O), 0.33 mM (4), and 8.3 mM (3).

sensed mass is significantly higher than for any of the components alone. At even higher SDS concentrations, the sensed mass approaches that determined for SDS alone at the same concentrations. As will be shown below, under these conditions, SDS dominates in the adsorbed layer. Most of the investigated mixtures reach adsorption equilibrium within the course of the experimental time, 1 h. The only notable exceptions are the systems closest to the charge neutralization point (0.33 mM). At these concentrations, a slow increase in adsorbed mass was obtained due to a continued deposition of polyelectrolyte-surfactant aggregates on the surface. In Figure 4, typical adsorption curves for some of the investigated systems are illustrated. In Figure 5, we display the fraction of sensed mass remaining on the surface for the investigated PVAm-SDS systems, after rinsing with 10 mM NaCl. Clearly, at low surfactant concentrations, most of the material remains on the surface, whereas at high surfactant concentrations, all material is removed by rinsing, indicating predominance of SDS in the formed layer. The dissipation changes resulting from the adsorption event are illustrated in Figure 2. At low surfactant concentrations, up to 0.08 mM, the dissipation values obtained for the mixtures are the same as that obtained for PVAm alone, whereas a noticeably higher value is obtained at 0.17 mM SDS. We note that at 0.33 and 0.83 mM SDS the dissipation change is large, considerably larger than in the presence of either the surfactant or the polyelectrolyte alone. This leads to the conclusion that the layers formed from these mixtures are thick and likely inhomogeneous

as previously observed by AFM for layers formed by deposition from mixtures of the cationic polyelectrolyte poly([2-(propionyloxy)ethyl]trimethylammonium chloride) and SDS.20 This finding is consistent with a slow deposition of large polyelectrolyte-surfactant aggregates as suggested by the time evolution of the sensed mass. At high surfactant concentrations, g4 mM, the dissipation changes are, within experimental error, the same as for the surfactant alone. Some information on the build-up of the layers can be obtained by plotting the change in dissipation factor as a function of the change in frequency, see Figure 6. We note that for the sample containing only 20 ppm PVAm in 10 mM NaCl, no change in dissipation is observed until a change in frequency of ca. 20 Hz has been reached. From this point, the dissipation increases as more polyelectrolyte adsorbs, which suggests that the chains that attach to the surface last adopt conformations that extend further away from the surface. This result is consistent with previous observations on polyelectrolyte adsorption on gold surfaces.41 In contrast, the sample containing 20 ppm PVAm and 8.3 mM SDS shows no change in dissipation with frequency, but steady values for both these quantities are reached very quickly, which is another result that demonstrates the predominance of SDS in the adsorbed layer under this condition. At a concentration of 0.17 mM SDS, positively charged aggregates are present in solution. Just like in the absence of SDS, the initial adsorption leads to a very limited change in dissipation, indicating a flat adsorption structure. However, as the adsorption proceeds, the dissipation increases to values well above that for PVAm alone,

7644 Langmuir, Vol. 22, No. 18, 2006

suggesting adsorption of aggregates that result in a thicker layer than that produced by the polyelectrolyte alone. Finally, the results obtained at 0.33 mM SDS, where charge neutralized aggregates slowly deposit on the surface, show a significantly larger increase in dissipation for a given change in frequency than under any other condition. This shows that an inhomogeneous and thick layer is formed from the very beginning of the deposition process.

4. Discussion 4.1. Sensed Mass Compared to Adsorbed Mass. The sensed mass calculated from the frequency change of the quartz resonator includes both the mass of the adsorbing species and any change in the mass of solvent that oscillates with the crystal due to formation of the adsorption layer. Thus, in general one expects that the sensed mass detected by QCM-D should be different to, and most often larger than, that detected by e.g. ellipsometry and reflectometry that are sensitive to the adsorbing species only. The sensed mass determined at the adsorption plateau of SDS, about 1.45 mg/m2, is rather large. If it would correspond to the adsorbed mass of only SDS, it would mean that the surfactant packs with an area/molecule in the layer that is 0.33 nm2, which seems small. Indeed, by comparing with the adsorbed amount of SDS determined on polystyrene employing other techniques, one finds lower values. For instance, Turner46 et al. report that the surface excess of SDS (determined by neutron reflection and ATR-FTIR) on polystyrene at and above CMC in water is 1.15 mg/m2, corresponding to an area of 0.42 nm2 per molecule. They also concluded that the adsorbed layer of SDS contains about 40% water. A similar value for the surface excess of SDS, 1 mg/m2, was reported by Dedinaite et al. for adsorption to hydrophobized silica as evaluated by relectometry.47 If we assume that the SDS adsorption is similar on our polystyrene surface and the one used by Turner et al., we have to conclude that the QCM-D senses a change in mass of water coupled to the oscillator amounting to 0.3 mg/m2, which means that 20% of the sensed mass is due to water associated with the adsorbed layer. We also note that the sensed mass determined by QCM-D has been compared with the adsorbed mass determined by ellipsometry for hexa(ethylene glycol) mono n-tetradecyl ether, C14E6, adsorbed on silica surfaces.48 The values obtained were 2.7 and 1.6 mg/m2, respectively. Thus, in this case, 40% of the sensed mass was due to water associated with the layer, which is a consequence of the significant hydration of the oligomeric ethylene oxide headgroup.49 In this study, we have found that about 80% of the sensed mass for PVAm on polystyrene is due to associated water and chloride counterions. There are no other similar studies using hydrophobic surfaces to compare with, but Plunkett et al. have investigated the adsorption of a series of cationic acrylamide copolyelectrolytes on gold surfaces employing QCM-D and XPS.41 They found that adsorption of high charge density polyelectrolytes resulted in low dissipation values and similar values for the adsorbed mass (determined by XPS) and the sensed mass (determined by QCM-D). On the other hand, high dissipation values were found for low charge density polyelectrolytes, and for these polyelectrolytes, the sensed mass was much larger than the adsorbed mass. 4.2. Evaluation beyond the Sauerbrey Equation. We have performed the analysis of the measured data using the Sauerbrey (46) Turner, S. F.; Clarke, S. M.; Rennie, A. R.; Thirtle, P. N.; Cooke, D. J.; Li, X. Z.; Thomas, R. K. Langmuir 1999, 15, 1017-1023. (47) Dedinaite, A.; Bastardo, L. Langmuir 2002, 18, 9383-9392. (48) Stålgren, J. J. R.; Eriksson, J.; Boschkova, K. J. Colloid Interface Sci. 2002, 253, 190-195. (49) Tyrode, E.; Johnson, M. C.; Kumpulainen, A.; Rutland, M. W.; Claesson, P. M. J. Am. Chem. Soc. 2005, 127, 16848-16859.

Naderi and Claesson Table 2. Sensed Mass (Using the Third Overtone) and the True Sensed Mass for PVAm-SDS Aggregates Formed in 10 mM NaCl and Adsorbed on Polystyrene Surfacesa

PVAm (20 ppm) 0.04 mM SDS 0.08 mM SDS 0.17 mM SDS 0.33 mM SDS 0.83 mM SDS 4.00 mM SDS 8.33 mM SDS

sensed mass (mg/m2)

true sensed mass (mg/m2)

1.5 ( 0.2 1.5 ( 0.2 1.6 ( 0.2 1.6 ( 0.2 3.1 ( 0.2* 3.5 ( 0.6* 1.4 ( 0.2 1.5 ( 0.1

1.5 ( 0.1 1.6 ( 0.2 1.6 ( 0.1 2.0 ( 0.2 3.5 ( 0.3* 4.5 ( 0.7* 1.6 ( 0.4 1.7 ( 0.2

a The values were obtained after 1h of measurement. The values indicated by “*” are not equilibrium values, as a slow deposition of aggregates continue beyond the 1 h measurement period.

equation, which is a good approximation provided the adsorbed layer is rigid. To check the validity of the Sauerbrey equation in our systems, we analyzed the data using the model of Johannsmann et al.50 They derived eq 3 below, which allows the true sensed mass to be calculated for viscoelastic layers.

(

)

f 3F2d 3 1 δfˆ ≈ -f0 fFd + Jˆ ( f ) 3 πxFq µq

(3)

where δfˆ is the shift in the complex frequency, f0 is the fundamental resonance frequency of the quartz crystal in air, f is the resonance frequency of the crystal in contact with solution, d is the thickness of the film, and Jˆ ( f ) is the complex shear compliance. The quantities Fq and µq are the specific density and elastic shear modulus of quartz, respectively, and F is the density of the fluid. Equation 3 can be transformed into a more convenient form by using the equivalent mass (m ˆ *)50 defined by

m ˆ*)-

xFq µq δfÄ 2f0

(4)

f

One thus obtains

(

)

Ff 2d 2 3

m ˆ * ) m0 1 + Jˆ ( f)

(5)

The true sensed mass, m0, was calculated under the assumption that Jˆ ( f ) is independent of the frequency in the accessible frequency range. A plot of the equivalent mass against the square of the resonance frequency f 2, gives the true sensed mass as the intercept. In Table 2, we present the true sensed mass (m0) obtained from eq 5 and the sensed mass calculated for the third overtone, using the Sauerbrey eq 1, for the PVAm-SDS mixtures. We notice that the Sauerbrey equation gives a very good estimate of the true sensed mass, except for the data obtained for PVAm mixed with SDS close to the charge neutralization concentration, where the adsorbed layer is built from deposited large aggregates. Note that the true sensed mass includes water associated with the adsorption layer, and it is thus not equal to the adsorbed mass. 4.3. Adsorption, Deposition, and Desorption from Polyelectrolyte-Surfactant Mixtures. In the present case, both the polyelectrolyte and the surfactant have affinity for the surface, and they associate with each other in solution. In the Tilton classification of polymer-surfactant systems, this corresponds to a class II system.34 The polystyrene surface used in this (50) Johannsmann, D.; Mathauer, K.; Wegner, G.; Knoll, W. Phys. ReV. B 1992, 46, 7808.

Polyelectrolyte-Surfactant Complexes

investigation is hydrophobic with a water contact angle of 8890°, and it is also weakly negatively charged. The main driving force for adsorption is hydrophobic interactions whereas electrostatic factors counteract surfactant adsorption and also polyelectrolyte adsorption once the polystyrene surface charge has been compensated by the initially adsorbing polyelectrolytes. In the mixed PVAm-SDS systems, the surfactant is present in aggregates and as free surfactants in solution, but the free surfactant concentration is not known. A rough estimate can be obtained at the charge neutralization concentration that occurs at a SDS concentration of 0.33 mM. The corresponding segment concentration of polyvinylamine is 0.46 mM. Approximately 60% of the segments are charged at the pH used and assuming that these segments are neutralized by SDS, we obtain a free surfactant concentration of 0.05 mM. In fact, the free surfactant concentration is likely somewhat lower since the degree of protonation of the polyelectrolyte segments is expected to increase due to association with the surfactant.45 At low surfactant concentrations, below the charge neutralization concentration of the bulk aggregates, the amount of free surfactants is thus low, and from the adsorption isotherm of SDS on polystyrene, we can estimate that if only the surfactant adsorbed a sensed mass of e0.3 mg/m2 would be expected. However, the results obtained show a sensed mass that is higher by a factor of about 5, and independent of surfactant concentration. Clearly, the sensed mass is dominated by the polyelectrolyte and the amount of surfactants incorporated in the adsorbed layer is low in this regime. However, some surfactant is incorporated as evidenced from the change in adsorption kinetics that is much faster in the presence of surfactant; compare the data in Figure 1 with the curve in the presence of 0.08 mM SDS in Figure 4. It thus seems clear that the presence of anionic surfactants limits the barrier for adsorption by reducing the net charge of the polyelectrolyte aggregates in solution and the similarly charged adsorbed layer. We note that the adsorbed layer formed below the charge stoichiometry point cannot readily be removed by dilution with 10 mM NaCl, see Figure 5, which is identical to the situation with PVAm alone. At the charge neutralization point of the bulk aggregates, at a SDS concentration of 0.33 mM, large aggregates are formed in solution. They deposit slowly on the surface, and their attachment does not set up any electrostatic barrier for further deposition. Thus, the sensed mass increases slowly with time, see Figure 4. During the rinsing process, due to the insolubility of these aggregates, they initially remain on the surface. However, with time, some desorption occurs, see Figure 4. Above the charge neutralization point of the bulk aggregates, the free surfactant concentration in solution increases rapidly. Now both the aggregates and the free surfactant are negatively charged, and an electrostatic repulsion exists between the adsorbed layer and all components present in solution, which limits the adsorption. At an SDS concentration of 0.83 mM aggregates still deposit slowly onto the surface. Thus, the data demonstrate that aggregates with a slight excess of surfactant (0.83 mM SDS)

Langmuir, Vol. 22, No. 18, 2006 7645

deposit more readily than aggregates with a slight excess of polyelectrolyte charges (0.17 mM SDS). We suggest that this is due to desorption of the excess surfactant during the deposition event. Also at this concentration, the deposited aggregates remain at the surface after rinsing, which is due to their insolubility. However, at high enough surfactant concentrations, the free surfactant dominates in the adsorbed layer. This is evidenced by the similar adsorbed mass obtained in the presence and absence of PVAm and in the observation that all sensed mass is removed by rinsing.

5. Conclusions SDS adsorption to polystyrene has been followed by QCM-D. It was shown that the sensed mass is considerably larger than the adsorbed mass and that the difference between these two quantities can be rationalized by assuming that 20% of the sensed mass is due to water associated with the adsorption layer. PVAm alone does also adsorb to polystyrene, with an adsorbed mass of 0.24 mg/m2, where the sensed mass is much higher, about 1.45 mg/m2. Thus, in this case, water and chloride ions associated with the layer contribute more than the adsorbed polyelectrolyte to the QCM-D response. The adsorption properties of the PVAm-SDS mixtures depend strongly on the composition of the solution, and three main regimes can be distinguished. At low surfactant concentrations, below the charge neutralization concentration in bulk solution, the adsorbed layer is dominated by the polyelectrolyte even though the presence of surfactant increases the adsorption rate. In this regime, the adsorbed layer is not removed by rinsing, which is rationalized by the predominance of PVAm. Close to the charge neutralization point, a slow deposition of large and uncharged aggregates persists for prolonged times. This also occurs for weakly negatively charged aggregates but not for weakly positively charged ones. This is rationalized by desorption of SDS from the negatively charged aggregates during deposition. Finally, well above the charge neutralization concentration, the free SDS concentration in solution increases and the surfactant dominates at the interface. At sufficiently large surfactant concentrations, the results for the PVAm-SDS mixtures become indistinguishable from that obtained with the surfactant alone. The rinsing process does in this case essentially remove all of the adsorbed material and we conclude that surfactants strongly predominate in the adsorbed layer. Finally, we have shown that the Sauerbrey approximation works surprisingly well, except when large aggregates are deposited on the surface. In this situation, more advanced theories have to be applied. Acknowledgment. We thank Marie Ernstsson (at the Institute for Surface Chemistry, Stockholm, Sweden) for her help with the ESCA measurements. Financial support from the Swedish Research Council, VR, is acknowledged. LA061118H