Adhesive Interaction between Polyelectrolyte Multilayers of

Feb 4, 2009 - DiVision of Fibre Technology, School of Chemical Science and Engineering, Royal Institute of Technology,. SE - 100 44 Stockholm, Sweden,...
0 downloads 8 Views 277KB Size
Langmuir 2009, 25, 2887-2894

2887

Adhesive Interaction between Polyelectrolyte Multilayers of Polyallylamine Hydrochloride and Polyacrylic Acid Studied Using Atomic Force Microscopy and Surface Force Apparatus Erik Johansson,*,† Eva Blomberg,‡,§ Rikard Lingstro¨m,† and Lars Wågberg† DiVision of Fibre Technology, School of Chemical Science and Engineering, Royal Institute of Technology, SE - 100 44 Stockholm, Sweden, DiVision of Surface Chemistry, School of Chemical Science and Engineering, Royal Institute of Technology, SE - 100 44 Stockholm, Sweden, and Institute of Surface Chemistry, Box 5607, SE - 114 86 Stockholm, Sweden ReceiVed October 31, 2008. ReVised Manuscript ReceiVed December 23, 2008 In the present work, the adhesion between substrates treated with identical polyelectrolyte multilayers (PEM) from polyallylamine hydrochloride (PAH) and poly(acrylic acid) (PAA) was studied using atomic force microscopy (AFM) and the surface force apparatus (SFA). The AFM measurements, conducted under wet conditions for PEMs formed at pH 7.5, showed a higher adhesion (pull-off force) when PAH was adsorbed in the outermost layers. There was also a difference depending on the molecular mass of the polymers, demonstrating a greater adhesion for the low molecular mass combination of polyelectrolytes. Furthermore, the time in contact showed to be of importance, with increasing pull-off forces with contact time at maximum load. The SFA measurements were conducted under dry conditions, at 100% RH, and under wet conditions for PEMs adsorbed at pH 7.5/3.5. The SFA adhesion measurements showed that under dry conditions, the adhesive forces between two high energetic mica substrates were lowered when they were covered by PEMs before the measurements. The thickness of the adsorbed layers was also measured using SFA. This showed that there was a significant swelling when the dry layers were exposed to 100% RH or to wet conditions. The swelling was higher, indicating a less rigid layer, when PAH was adsorbed in the outermost layer than when the PEM was capped with PAA.

Introduction Polyelectrolyte multilayers (PEMs) were first introduced by Decher in the early 1990s, and over the past 15 years the technique has developed as a general and simple substrate treatment method for achieving various desirable properties.1 The technique shows a great potential in wide areas of application, for example paper making,2-5 and is already in use in applications such as sensor technology6 and surface treatment of contact lenses.1 PEMs are formed by consecutively treating a charged substrate with oppositely charged polyelectrolytes; PEMs showing very different properties can be created by using different combinations of polyelectrolytes and can also be formed from combinations of polyelectrolytes and colloids.1 The properties of the layers can also be influenced by changing fundamental parameters such as salt concentration, type of salt, temperature, and charge of the polyelectrolytes that are added. This has also been extensively reviewed elsewhere.1,7 The first decade of PEM studies was to a large extent focused on studies of thickness and adsorbed amount of different * Corresponding author. Phone: +46-(0)8-7908311. E-mail: ejoha@ polymer.kth.se. † Division of Fibre Technology, Royal Institute of Technology. ‡ Division of Surface Chemistry, Royal Institute of Technology. § Institute of Surface Chemistry. (1) Decher, G.; Schlenoff, J. B. Multilayer Thin Films: Sequential Assembley of Nanocomposite Materials; Wiley-VCH: Weinheim, Germany, 2003. (2) Wågberg, L.; Forsberg, S.; Johansson, A.; Juntti, P. J. Pulp Pap. Sci. 2002, 28, 222. (3) Eriksson, M.; Notley, S. M.; Wågberg, L. J. Colloid Interface Sci. 2005, 292, 38. (4) Eriksson, M.; Pettersson, G.; Wågberg, L. Nord. Pulp Paper Res. J. 2005, 20, 270. (5) Lingstro¨m, R.; Wågberg, L.; Larsson, P. T. J. Colloid Interface Sci. 2006, 296, 396. (6) Sun, Y.; Zhang, X.; Sun, C.; Wang, B.; Shen, J. Macromol. Chem. Phys. 1996, 197, 147. (7) Klitzing, R. v. Phys. Chem. Chem. Phys. 2006, 8, 5012.

combinations of polyelectrolytes and conditions during layerby-layer (LbL) formation, whereas only a few studies investigated the internal structures of PEMs. Since a PEM coating changes the properties of substrates, it can also be used as a way of changing the adhesion between the interacting substrates. Forces between substrates treated by individual layers of weak polyelectrolytes have been extensively studied during the last decades,8-11 but to our knowledge there are still only a few studies of the forces between substrates treated by PEMs. Force measurements using atomic force microscopy12 and the colloidal probe technique13 have recently been performed under wet conditions for multilayers formed from polyallylamine hydrochloride (PAH) and poly(acrylic acid) (PAA)14,15 by using different pH strategies; both polymers adsorbed at pH 7.5,14 at which both PAH and PAA are relatively highly charged, and adsorption of PAH at pH 7.5 and PAA at pH 3.5,15 at which PAH is relatively highly charged and PAA is weakly charged.16 Both investigations showed a higher pull-off force when PAH was adsorbed in the outermost layer. To date, a large proportion of the existing investigations regarding force measurements between PEMs have been conducted using PEMs formed from PAH and polystyrene sulfonate (PSS), that is, layers formed from one strong and one (8) Marra, J.; Hair, M. L. J. Phys. Chem. 1988, 92, 6044. (9) Claesson, P. M.; Dahlgren, M. A. G.; Eriksson, L. Colloids Surf., A 1994, 93, 293. (10) Biggs, S.; Proud, A. D. Langmuir 1997, 13, 7202. (11) Notley, S. M.; Biggs, S.; Craig, V. S. J. Macromolecules 2003, 36, 2903. (12) Senden, T. J. Curr. Opin. Colloid Interface Sci. 2001, 6, 95. (13) Ducker, W. A.; Senden, T. J.; Pashley, R. M. Nature 1991, 353, 239. (14) Notley, S. M.; Eriksson, M.; Wågberg, L. J. Colloid Interface Sci. 2005, 292, 29. (15) Lingstro¨m, R.; Notley, S. M.; Wågberg, L. J. Colloid Interface Sci. 2007, 314, 1. (16) Shiratori, S. S.; Rubner, M. F. Macromolecules 2000, 33, 4213.

10.1021/la803628w CCC: $40.75  2009 American Chemical Society Published on Web 02/04/2009

2888 Langmuir, Vol. 25, No. 5, 2009

weak polyelectrolyte. Force measurements using the surface force apparatus (SFA) for the evaluation of forces between substrates coated by a maximum of two17 or four18 layers have been conducted. One monolayer of PAH showed adhesion18 that was attributed to bridging between the substrates. This phenomenon is, however, not a general explanation for the increased adhesion between PEMs consisting of a larger number of layers, because it is known that only the properties of the first adsorbed layers of the PEM are influenced by the substrate.19 In a recent investigation AFM force measurements were used to study the interaction, in asymmetric systems, between a bare glass sphere and a substrate coated by PEM formed from PAH/ PSS and PAH/DNA.20 This showed higher pull-off forces when the cationic polymer was adsorbed in the outermost layer, compared to when the anionic polymer was adsorbed in the outermost layer. This effect was most significant for the PAH/ DNA multilayer, and for this system it also increased with the number of layers adsorbed. For these asymmetrical systems, the higher adhesive force detected with the cationic polymer in the outermost layer was attributed to bridging between the negatively charged bare glass sphere and the positively charged polymer on the substrate of the PEM. PAH/PSS showed no adhesive force when PSS was adsorbed in the outermost layer, and PAH/DNA showed only a small adhesion when DNA was adsorbed outermost. For dry systems, Creton et al.21 showed that A-B diblock polymers added to the interface between blocks of A and B polymers could dramatically improve the adhesion by mechanically entangling on both sides of the interface with the respective homopolymers (A and B).21 The fracture toughness of the interface is then influenced by both the number of interacting chains and the degree of polymerization of the blocks. This model, translated into the interaction between two PEM-treated substrates, implies that the number and length of entanglements would strongly influence the strength of the interaction. Entanglement between substrates has also been discussed by de Gennes,22 who proposed a model for describing the strong adhesion properties of weakly cross-linked rubber. Another relevant model is that proposed by Aradian et al.23 for describing the interdiffusion of polymer chains between two polymer particles and the way in which a cross-linking agent can dramatically slow down this diffusion process. Chen et al.24 used SFA measurements of polystyrene and poly(vinylbenzyl chloride) to show that the number of interacting chain ends, the mobility of the chains, and the rate of interpenetration are all of great importance for achieving a strong adhesion between polymer surfaces. From this, it is not obvious that a thick PEM formed from high molecular mass polymers would give an adhesion higher than that for a thinner layer formed from low molecular mass polyelectrolytes, because the mobility of the polyelectrolytes and the time of interaction between the interacting substrates are also very important for the development of adhesion. Once the anchoring of the polymers within the layers is good, the mechanical properties of the surface layers (17) Lowack, K.; Helm, C. A. Macromolecules 1998, 31, 823. (18) Blomberg, E.; Poptoshev, E.; Claesson, P. M.; Caruso, F. Langmuir 2004, 20, 5432. (19) Kulcsar, A.; Voegel, J. C.; Schaaf, P.; Kekicheff, P. Langmuir 2005, 21, 1166. (20) Gong, H. F.; Garcia-Turiel, J.; Vasilev, K.; Vinogradova, O. I. Langmuir 2005, 21, 7545. (21) Creton, C.; Kramer, E. J.; Hui, C. Y.; Brown, H. R. Macromolecules 1992, 25, 3075. (22) de Gennes, P. G. Langmuir 1996, 12, 4497. (23) Aradian, A.; Raphae¨l, E.; de Gennes, P. G. Macromolecules 2000, 33, 9444. (24) Chen, N.; Maeda, N.; Tirrell, M.; Israelachvili, J. Macromolecules 2005, 38, 3491.

Johansson et al.

will also play an important role for the fracture toughness of the formed contact zone.21 With this short summary as a background, it is obvious that the mechanisms behind the increase in adhesion between PEMtreated substrates are not fully understood; it is unclear whether it is due to an increase in the thermodynamic work of adhesion, or if it is also influenced by nonequilibrium mechanisms such as polymer interdiffusion across the interface as proposed by Creton et al.,21 de Gennes et al.,22,23 and Chen et al.24 The present work was aimed at investigating pull-off forces when separating substrates treated by PAH/PAA PEMs on silicon oxide (using AFM) and on mica (using the SFA). By comparing SFA measurements of the pull-off forces between dry PEMs and AFM measurements of the pull-off forces between PEMs interacting in wet conditions, it may be possible to separate out the influence of the change in thermodynamic work of adhesion from the influence of polymer entangling. SFA was also used to study the change in thickness of the PEMs when the conditions were changed from dry to wet or 100% humidity, in order to get a better fundamental understanding of the swelling behavior of the layers.

Materials and Methods Substrates. The silica used as model substrates for the AFM measurements was delivered as silicon wafers, supplied by MEMC, Electronic Materials SpA, Novara, Italy, with a natural silicon oxide layer of about 4 nm. The wafers were rinsed consecutively with ethanol and milli-Q water and were then blown dry with nitrogen. Finally the substrates were hydroxylated and cleaned with the aid of a plasma cleaner (PDC-002, Harrick Plasma, Inc., Ithaca, NY) for 3 min at 30 W. The muscovite mica used in the SFA measurements was obtained from Axim Enterprises, Inc., New Hyde Park, NY. The mica was freshly cleaved before the measurements. Polyelectrolytes. Polyallylamine hydrochloride (PAH) was used as the cationic polymer, at two different molecular weights; 15 000 and 70 000 (according to the supplier). Polyacrylic acid (PAA) was used as the anionic polymer, again at two molecular weights, 8000 and 240 000 (according to the supplier). All polyelectrolytes were delivered from Sigma Aldrich and were used as supplied. Aqueous NaCl (Kebo, Sweden) solution, 0.01 M, was used to dissolve the PAH and the PAA to a polyelectrolyte concentration of 30 mg/L. For the AFM measurements, both the PAH and PAA solutions were adjusted to pH 7.5. At this pH, PAA shows a high degree of dissociation while PAH has reached about 50% of its maximum charge.25 For the SFA measurements, the PAA solutions were adjusted to pH 3.5 before adsorption, thus decreasing the degree of dissociation of the PAA. Milli-Q grade water was used to prepare the solutions. Water with a resistivity of 18.2 MΩ cm and total organic carbon content less than 10 ppb was obtained from a Millipore system comprising RiOs-10 and Milli-Q+ 185 units. The solutions were filtered through a 0.2 µm filter. Atomic Force Microscopy. Surface forces were measured using a Nanoscope IIIa AFM with a Picoforce scanner (Veeco, Ltd., Santa Barbara, CA). The general, principle of force measurements using AFM12 is described in detail elsewhere and will not be further discussed here. The colloidal probe technique13 was used to measure the wet adhesive forces between two silica substrates covered with PEM. Borosilicate glass spheres (Duke Scientific, Inc.) with a diameter of 10 µm were glued to standard V-shaped Si3N4 cantilevers (Veeco, Ltd.) with a nominal spring constant of 0.12 N/m. The borosilicate glass spheres were glued to the cantilevers using a two-component (25) Horn, D. In Polymeric Amines and Ammonium Salts: InVited Lectures and Contributed Papers Presented at the International Symposium on Polymeric Amines and Ammonium Salts, Ghent, Belgium, 24-26 September, 1979; Goethals, E. J., Ed.; Oxford: New York, 1980.

AdhesiVe Interactions of Polyelectrolyte Multilayers

Langmuir, Vol. 25, No. 5, 2009 2889

Figure 1. Schematic picture of a symmetrical system of PEMs formed from PAH and PAA onto negatively charged substrates in a liquid cell. In the AFM force measurements, PEMs were formed onto one flat and one spherical silica substrate while in the SFA measurements the PEMs were formed on two crossed mica cylinders.

epoxy resin (Casco Adhesives AB, Stockholm, Sweden). A very small amount of glue was picked up by dipping an etched tungsten wire attached to a XYZ translation stage into a glue sample. The glue was then transferred to the edge of the cantilever. A second tungsten wire was used to pick up a glass sphere and deposit it on top of the glue spot. The gluing and sphere attachment were done under a microscope. The epoxy resin was allowed to cure for at least 24 h before the cantilevers were cleaned, examined under a microscope, and used in the AFM force measurements. For each individual measurement, the borosilicate probe diameter was measured using a light microscope (Nikon) and the normal spring constant was determined by the thermal noise method.26 PEMs were formed in situ in the AFM liquid cell; that is, both the flat silica substrate and the silica sphere were covered by PEMs, and the substrates were never allowed to dry. Polyelectrolyte solution was injected into the liquid cell through a syringe filter with a 0.45 µm polyethersulfone membrane and was allowed to adsorb for 10 min, followed by rinsing with 0.01 M NaCl solution. The forces were measured for each layer in the PEM at the end of the rinsing cycle, when there was no free polyelectrolyte present in the solution. Figure 1 shows a schematic picture of identical PEMs formed from PAH and PAA onto two interacting substrates in a liquid cell. The interacting charged substrates could be the flat and spherical silica substrates in the AFM experiments or the crossed mica cylinders in the SFA experiments. The surface delay feature in the Nanoscope software was used to adjust the time in contact at maximum load (0 s, 1 s, and 5 s) before separating the substrates. The maximum load was 20 nN, the scan size was 2 µm, and the scan rate was 0.447 Hz, resulting in a load/unload rate of 1.79 µm/s. Ten force-distance curves were measured for each setting. Representative force curves are shown, and the adhesive pull-off forces, i.e., the maximum force on separation of the substrates, are presented as average pull-off forces. To allow comparison between experiments, the pull-off forces were normalized with the probe radius. Surface Force Apparatus. The thickness of the multilayers and the adhesion between identical multilayer coatings were measured with a Mark IV interferometric SFA.27 The substrates were mounted inside the measuring chamber of the SFA in a crossed cylindrical geometry. The separation between the substrates was controlled by a motor or by applying voltage to a piezoelectric crystal to which the upper substrate was attached. The layer thickness was determined interferometrically by using fringes of equal chromatic order (FECO),28 from which the absolute surface separation can be determined to within an interval of 0.1-0.2 nm. At the beginning of each experiment, the bare mica substrates were brought into contact in dry air in order to measure the zero contact separation in the absence of the adsorbed layers. The thickness of the multilayer films was then obtained from the shift of the contact position toward larger separation. The adhesion (pull-off) force acting on separation of the substrates was calculated from the jump distance.29 The (26) Hutter, J. L.; Bechhoefer, J. ReV. Sci. Instrum. 1993, 64, 1868. (27) Parker, J. L.; Christenson, H. K.; Ninham, B. W. ReV. Sci. Instrum. 1989, 60, 3135. (28) Israelachvili, J. N. J. Colloid Interface Sci. 1973, 44, 259. (29) Horn, R. G.; Israelachvili, J. J. Chem. Phys. 1981, 75, 1400.

Figure 2. Normalized force versus apparent separation upon retraction for PEM-covered silica substrates. (a) Low molecular weight PAH and PAA and (b) high molecular weight PAH and PAA. Both were adsorbed at pH 7.5/7.5 in a background electrolyte concentration of 0.01 M NaCl. The force curves shown correspond to PAH in the outermost layer, i.e., layer 1, 3, 5, 7, 9, 11, and 0 s of contact time at maximum load. Layer 1 (0), layer 3 (∆), layer 5 (]), layer 7 (9), layer 9 (2), layer 11 ([).

measurements of the adhesion force were done with a rather stiff spring (spring constant 1386 N/m) in order to reduce errors due to shearing and sliding. In the results presented below, the pull-off force is normalized with the geometric mean radius of the mica cylinders. Experiments in liquid were performed by filling the gap between the substrates with a drop of electrolyte solution.

Results Force Measurements Using AFM. Figure 2a and 2b show representative force-distance curves upon retraction for silica substrates covered by PEMs constructed from PAH and PAA of low molecular weight (LMw) (Figure 2a) and high molecular weight (HMw) (Figure 2b), respectively. Only the forcedistance curves with PAH in the outermost layer are shown. For these curves, there was no time delay at maximum load; that is, the contact time at maximum load was 0 s. The multilayers were constructed in situ in the AFM liquid cell from polyelectrolyte solutions at pH 7.5 and 0.01 M NaCl. Between each adsorption step, the cell was rinsed with electrolyte solution with the same ionic strength and pH in order to wash away weakly attached polyelectrolytes. The force measurements were made at the end of the rinsing cycle. For both LMw and HMw PAH/PAA, the pull-off force, and range of interaction both increased with the number of adsorbed layers. The shapes of the curves were similar for LMw and HMw PAH/PAA, but the pull-off forces were higher for the LMw system than for the HMw combination. PEMs formed from LMw PAH/

2890 Langmuir, Vol. 25, No. 5, 2009

Johansson et al.

Figure 3. Normalized pull-off force as a function of layer number and contact time at maximum load for PEM-covered silica substrates. (a) Low molecular weight PAH and PAA and (b) high molecular weight PAH and PAA. Both were adsorbed at pH 7.5/7.5 in a background electrolyte concentration of 0.01 M NaCl. Contact time at maximum load: 0 s (9), 1 s (∆), 5 s ([).

Figure 4. Thickness of PEMs formed on mica, determined using SFA in dry air, at 100% RH, and under wet conditions (drop) for four and five adsorbed layers, respectively. Figure 4a shows the total thickness for two approaching PEMs, that is 8 and 10 layers, respectively, and Figure 4b shows the thickness per individual layer.

PAA also showed interaction over larger distances in comparison to the HMw system. Figure 3a shows the average normalized pull-off forces as a function of layer number and contact time at maximum load for silica substrates covered by LMw PAH/PAA multilayers. Odd layer numbers mean that PAH is in the outermost layer, and even numbers mean that PAA is in the outermost layer. The pull-off forces were significantly higher when PAH was adsorbed in the outermost layer as compared to when PAA was in the outermost layer. The increase in pull-off force with increasing layer number was also higher for PAH-covered PEMs. These trends, and the level of the pull-off forces at 0 s of contact time at maximum load, are in accordance with the results of earlier colloidal probe measurements on multilayers constructed from LMw PAH/ PAA.14,15 It has also been shown15 that very similar trends were found when the PEM was formed at pH 7.5 (PAH) and pH 3.5 (PAA) showing the relevance of using the pH strategy in the present work for the formation of the PEM. For each layer, force measurements were first carried out with 0 s of contact at maximum load, and then with prolonged contact times of 1 and 5 s. The pull-off force increased with increasing contact time at maximum load, for all numbers of layers. Repeating the 0-s measurement after the 5-s measurement produced about the same value as for the first 0-s measurement. Typically the 0-s values before and after the 5-s measurements were within 10-20% and with no general trend of increasing or decreasing values before and after. Figure 3b shows the average normalized pull-off forces as a function of layer number and contact time at maximum load for

silica substrates covered by HMw PAH/PAA multilayers. The effects on pull-off force of layer number, the polyelectrolyte that was adsorbed in the outermost layer, and contact time showed similar patterns to those seen for LMw PAH/PAA. However, the absolute values of the pull-off forces were lower for the HMw multilayer, with a maximum pull-off force of 3.5 mN/m at layer number 11 and 5 s in contact at maximum load, compared to 6.0 mN/m for the same conditions in the LMw multilayer. Thickness of the Adsorbed Layers Measured Using SFA. The thickness measurements of the adsorbed layers, performed with the SFA, were conducted in dry air, at 100% relative humidity (RH), and in wet conditions. For the experiment conducted under wet conditions, 0.01 M NaCl was added, as during the PEM formation, and the pH of the added solution was the same as for the preceding adsorption step: 7.5 when PAH was adsorbed in the outermost layer and 3.5 when PAA was adsorbed in the outermost layer. From Figure 4 it is obvious that the measured thickness was significantly influenced by the ambient conditions during the measurements, and also that this effect was significantly dependent on which of the polyelectrolytes, PAH or PAA, was adsorbed in the outermost layer. It should be noted that the results presented in Figure 4a show the total thickness of four and five layers on each mica substrate that is eight and ten layers in total, respectively. In the dry state (0% RH), the total thickness was about 15 nm for the five-layer PEM and about 12 nm for the four-layer PEM. This means that the thickness per deposited individual layer was determined to about 1.4 nm for both four deposited polyelectrolyte layers (PAH-

AdhesiVe Interactions of Polyelectrolyte Multilayers

Langmuir, Vol. 25, No. 5, 2009 2891

water meniscus could not have been the only explanation for the increase in adhesion. When a drop of the solution was placed between the PEMcoated substrates, and the SFA measurement was conducted under wet conditions, a small pull-off force was detected. This adhesion, 42.6 mN/m when PAA was adsorbed in the outermost layer and 1.9 mN/m when PAH was adsorbed in the outermost layer, must have originated from interaction between the PEMs, because the pull-off force between two substrates of bare mica in 0.01 M NaCl is zero due to the short-range repulsive hydration force originating from adsorption of hydrated sodium ions at the mica substrate.31 Figure 5. The pull-off force measured using SFA and normalized with the geometric mean radius of the mica-covered cylinders. The force was measured for bare mica and for mica treated by PEMs formed from four and five layers of PAH/PAA, respectively. The experiments were conducted in dry air (0% RH), humid air (100% RH), and wet conditions (drop).

PAA)2 and five deposited polyelectrolyte layers (PAH-PAA)2PAH. In humid air (100% RH), the layers were significantly swollen, and when PAA was adsorbed as the outermost layer, the total thickness was about 14 or 1.8 nm per layer, an increase of about 23% in comparison to the thickness in dry air. This thickness increase was significantly lower compared to when the PAH-covered PEM was exposed to 100% humid air, showing a total thickness of 26 or 2.6 nm per layer, which was 79% higher compared to the thickness of the PEM in dry air. A similar dependence on the polyelectrolyte that was adsorbed as the outermost layer was found when the layers were exposed to a 0.01 M NaCl solution before drying the substrates. In this case the swelling was even larger. With PAA outermost, the thickness per layer was about 2.1 nm, which corresponds to an increase of 49% compared to the thickness per layer at 0% RH and an increase of 21% compared to that at 100% RH. The swelling of the layers was larger when PAH was adsorbed as the outermost layer; the thickness per layer was 3.5 nm in 0.01 M NaCl and was increased by 138% and 32%, respectively, in comparison to the thickness in dry and humid air. Force Measurements Using SFA. The adhesion forces between the high energetic mica substrates were very high; normalized with the geometric mean radius of the mica cylinders, the pull-off force in dry air (0% RH) was about 1500 mN/m (Figure 5). When PEMs were deposited onto the mica, this adhesion force decreased to about 300 mN/m when PAA was adsorbed as the outermost layer and about 100 mN/m when PAH was adsorbed as the outermost layer. As can be seen from Figure 5, the adhesion force in humid air (100% RH) was significantly higher, around 500 mN/m. It is likely that the main reason for this adhesion was a capillary force caused by the condensate of water between the substrates. Condensed water could be observed visually, under the microscope, as small water droplets on the substrates. This observation was further supported by an observed discontinuity in the refractive index of the FECO interference pattern in the spectrometer during the separation of the substrates. In fact, this discontinuity was positioned to the left of the main fringe, which shows that this refractive index was smaller.30 When the substrates had been in contact and then separated, the condensed water droplets in the contact zone disappeared, which is consistent with a lower refractive index, indicating that air was nevertheless the main component in the contact zone and so a continuous (30) Christenson, H. K. J. Colloid Interface Sci. 1985, 104, 234.

Discussion The strength of adhesive joints between polymer-coated substrates may be discussed in terms of interaction between the chains when the joint is formed and how this interaction is then broken when the substrates are separated. The strength of the joint is determined by parameters such as the properties of the polymer layers (thickness, viscoelastic properties, and molecular mobility), whether the joint is formed and broken in dry or wet conditions, and the time of interaction when the joint is formed. For example, fiber joints in paper, reinforced by PEM treatment, are formed under wet conditions and then dried, and finally the joint strength is utilized under dry conditions.15 Several studies have shown that the most important controlling factor for the adhesive forces between polymer-covered substrates is the entanglement of polymer chains across the interface.21,22,24,32-34 Entanglements across the interface increase the molecular contact area; when the substrates are separated, the polymer chains must be disentangled, leading to increased adhesion with increased degree of entanglements across the interface. The polymers in these previous studies were thermoplastics, and they were studied in air.21,24,32,33 In the present study, we suggest that the same line of argument used for these systems can also be applied when discussing adhesive forces between wet substrates covered by PEMs. AFM. The AFM colloidal probe technique was used to study the interactions between PEM-covered silica substrates under wet conditions, that is, conditions typical of those under which joint formation takes place in many biological systems and, for example, during formation of strong fiber/fiber joints in papermaking. Effect of Polyelectrolyte Adsorbed in the Outermost Layer on the Wet Adhesion. As shown in Figure 3a and 3b, a common feature for LMw and HMw PAH/PAA PEMs was that the adhesive forces were higher when cationic PAH was adsorbed in the outermost layer than when anionic PAA was adsorbed in the outermost layer. This was in agreement with recent AFM pulloff force measurements, which were also compared to energy dissipation measurements using a quartz crystal microbalance (QCM) for PEMs formed from LMw PAH/PAA.14 The energy dissipation during the build up of PEM, a measure of the rigidity of the adsorbed film, was higher when PAH was adsorbed in the outermost layer, indicating that the film was less rigid when PAH was adsorbed in the outermost layer than when PAA was adsorbed in the outermost layer.14 The frequency signal, however, showed a steady build-up of the multilayer, indicating no changes in bound solvent during the change from PAH to PAA.14 It was (31) Pashley, R. M. J. Colloid Interface Sci. 1981, 83, 531. (32) Maeda, N.; Chen, N.; Tirrell, M.; Israelachvili, J. N. Science 2002, 297, 379. (33) Luengo, G.; Pan, J.; Heuberger, M.; Israelachvili, J. N. Langmuir 1998, 14, 3873. (34) McKenzie, A. W. Appita 1984, 37, 580.

2892 Langmuir, Vol. 25, No. 5, 2009

therefore interpreted that the change in dissipation was caused by a higher mobility for the PEM when covered by PAH which allowed the interfacial chains to interdiffuse across the interface, between the two PEMs, to a greater extent than for the lessmobile PEM with PAA as outermost layer. Therefore, it seems that it is the difference in degree of entanglement across the interface that causes the difference in adhesive force between PEMs with PAH and PAA, respectively, as the outermost layer. The exact reason for the large difference in mobility when changing from PAH to PAA is not known. Because both PAH and PAA are weak polyelectrolytes, it could be anticipated that the degree of dissociation should be similarly affected for both polymers when changing the sign of the polyelectrolyte in the capping layer. This is not the situation and merits further investigations. Earlier investigations of multilayers formed from PAH and PSS,35 that is, one weak and one strong polyelectrolyte, have shown differences in the solvent volume fraction of the multilayer depending on which polyelectrolyte was adsorbed in the outermost layer. These differences were only present when the multilayers were deposited at a high ionic strength of 1 M NaCl and not when they were deposited at a comparatively lower ionic strength of 0.25 M NaCl, and it was suggested that uncompensated polyion charges were responsible for the oscillations in the solvent volume fraction.35 In the present investigation, the ionic strengths were much lower, 0.01 M NaCl, and thus the difference in solvent volume fraction depending on the polyelectrolyte in the outermost layer should be small. Similar effects of the charge balance in the PEMs have been found for the thermal behavior of polyelectrolyte multilayer microcapsules formed from strong polyelectrolytes,36 PDADMAC and PSS, but the present results are more probably linked to changes in the degree of dissociation of the weak polyelectrolytes. Effect of Molecular Weight on Wet Adhesion. The adhesive forces between PEM-covered substrates were higher for the LMw PAH/PAA system than for the HMw equivalent (Figure 3a and 3b), and the interactions were also of longer range (Figure 2a and 2b). These results are consistent with the results of earlier studies of the effect of molecular weight on adhesion hysteresis and friction between polystyrene and poly(vinylbenzyl chloride) surfaces.24,32 The current results suggest that polymer chain ends are much more important than the polymer loops, in terms of producing a strong adhesion. Chain ends are more mobile than polymer loops and interdiffuse faster into an opposite layer to form entanglements, and hence it can be suggested that the level of adhesion between polymer-coated substrates is to a large extent determined by the number of polymer chain ends. Since substrates treated by LMw polylelectrolytes will have a higher surface density of chain ends that will diffuse across the interface, the adhesive forces will be higher for substrates treated by LMw PAH/PAA than for substrates treated by HMw PAH/PAA, provided that the hypothesis about the importance of chain ends is correct. Another possible effect of molecular weight is that the low molecular mass polymers form PEMs with a lower elastic modulus, meaning that these PEMs are more easily deformable than PEMs formed from HMw polyelectrolytes. At a given load, the contact area might therefore increase more due to compression of the PEM for LMw PEMs than for HMw PEMs. An increased contact area would then increase the adhesion, in addition to the entanglements across the interface. More work regarding the mechanical properties of wet PEMs is needed to test this hypothesis, and to the knowledge of the authors this information is still missing today. (35) Carrie`re, D.; Krastev, R.; Scho¨nhoff, M. Langmuir 2004, 20, 11465. (36) Ko¨hler, K.; Shchukin, D. G.; Mo¨hwald, H.; Sukhorukov, G. B. J. Phys. Chem. B 2005, 109, 18250.

Johansson et al.

In contrast to our results, Creton et al.21 found that increased molecular weight led to increased adhesion. In that study21 the interdiffusion process was allowed to continue for several hours at an elevated temperature and pressure so the polymers could fully interpenetrate across the interface. Higher molecular weight gave a larger amount of entanglements per polymer chain and thereby stronger adhesion. However, in a dynamic system where the entanglement time is limited, as in the present study, equilibrium is not reached and so the rate of interdiffusion of the polymers becomes very important. It therefore seems logical that the low molecular mass polyelectrolytes with a larger number of chain ends and a higher mobility would lead to the highest adhesion for the experimental systems tested in the present work. Luengo et al.33 have, however, shown that adhesion peaks at an intermediate polymer mobility, where the entanglement rate is high enough for creating extensive interdigitation across the interface but the disentanglement rate is low enough to give a high resistance upon separation. Thus, if the mobility of the polyelectrolytes in a PEM is too high, this might not be beneficial for adhesion. Effect of Contact Time on the Wet Adhesion. Our results show that increased time in contact at maximum load increased the adhesive forces, for all layer numbers and for both LMw and HMw PAH/PAA. These results suggest that the polyelectrolytes at the external parts of the PEMs diffuse across the interface to form entanglements. With longer contact time, and hence more time to interdiffuse, both the number and the length of entanglements will increase, resulting in increased adhesive forces between the PEM-covered substrates. Similar results and links between contact time, interdiffusion, and the effect on adhesion between polymer surfaces have been seen for other polymer systems.24,32 The strength-enhancing mechanism for PEM systems hence seems to be similar to the mechanisms in these thermoplastic polymer systems24,32 though on a different time scale. Considering the results from the colloidal probe measurements regarding the influence of the polyelectrolyte in the outermost layer, the molecular weight of the polyelectrolytes forming the PEM, and the effect of contact time, it is obvious that the entanglements across the interface are important in forming strong adhesive joints between PEM-covered substrates. For the AFM measurements conducted in this work, the substrates were also separated and the joint was broken under wet conditions. In a real applied system, the joint-forming time might be different, the breaking might take place under other conditions (e.g., preceded by drying), and the rate of separation might also be different. The conditions used during the forming and the breaking of the adhesive joint will of course influence the degree of entanglement, the ease or difficulty of disentanglement, and thereby the adhesive forces between PEM-covered substrates. Stability of the PEM. It might be argued that the PEMs could be damaged by repeatedly bringing the substrates together, for example by transfer of molecules from one substrate to another, and that the interaction properties would thereby be changed, thus affecting our results. In order to investigate this possibility, the pull-off force for the 0-s maximum load contact time was measured again after the first set of measurements (0 s, 1 s, and 5 s of contact time) were taken, for each layer adsorbed. The pull-off forces with 0 s of contact time at maximum load, before and after the measurements made with extended contact times, are shown in Figure 6a (LMw PAH/PAA) and 6b (HMw PAH/ PAA). As can be seen in the figures, the pull-off forces were at about the same level before and after the measurements carried out with prolonged contact times. From this it can be concluded

AdhesiVe Interactions of Polyelectrolyte Multilayers

Figure 6. Normalized pull-off force as a function of layer number for PEM-covered silica substrates. (a) Low molecular weight PAH and PAA and (b) high molecular weight PAH and PAA. Both were adsorbed at pH 7.5/7.5 in a background electrolyte concentration of 0.01 M NaCl. Contact time at maximum load was 0 s. Measurements were taken both before (0) and after (9) the measurements with prolonged contact times.

that the PEMs were not damaged or changed by the colloidal probe measurements, at least not severely enough to change the interactions, as detected with AFM, between the PEM-covered substrates. SFA. As has been shown for PAH/PAA PEMs formed at pH 7.5/3.5, there was a significant difference in thickness depending on the ambient conditions of the substrates. The thickness of the PEMs was also highly dependent on the polymer that was adsorbed in the outermost layer. This difference in thickness was in accordance with recent energy dissipation measurements made under wet conditions using QCM-D;14 the PEMs in this QCM study, formed from LMw PAH/PAA under the same conditions as in the present study, showed a higher energy dissipation when PAH was adsorbed in the outermost layer,14 indicating a less compact layer than when PAA was adsorbed in the outermost layer. From Figure 4a and 4b it is also clear that the thickness of the PEM was higher at 100% RH than under dry conditions, and the thickness per layer also increased with moisture content. This means that the layers become softer and in turn this means that the molecular contact area increases as the moisture increases. This increase in molecular contact area can thus be suggested to be an important reason for the detected increase in adhesion when changing from dry conditions to 100% RH, together with the earlier-mentioned partial condensing of water at the higher humidity. The higher thickness/softness of the PAH-terminated layer may also explain the difference in adhesion between the PAA-terminated and the PAH-terminated layers.

Langmuir, Vol. 25, No. 5, 2009 2893

When changing from 100% RH to a wet system, there was a rapid drop in adhesion from levels around 500 mN/m down to levels around 1-2 mN/m. Compared to the level of adhesion demonstrated using AFM, the experiments using SFA were in good agreement with other recent results regarding PEMs formed from PAH/PAA using the same pH sequence, that is, PAH adsorbed at pH 7.5 and PAA adsorbed at pH 3.5.14 When five layers were adsorbed these experiments demonstrated a pull-off force of about 2 mN/m, which can be compared to 1.9 mN/m for the corresponding SFA-measurements in the present work. However, in the case of four layers there was a large difference; 0.3 mN/m for the previously conducted AFM force measurements and 42.9 mN/m for the SFA force measurements. In comparison to the AFM measurements in the present work, too, the results for five layers were almost similar, but a large difference could be detected for four layers. There might, naturally, be different explanations for this phenomenon, but considering the differences in geometry and maximum load before separation for the different methods, a direct comparison should only be made with great care. Another factor that must be considered when comparing the SFA and the AFM data is that the AFM measurements were conducted under totally wet conditions whereas the wet SFA measurements were conducted by placing a drop of water in the contact zone between the two mica cylinders. This means that the wetting of the cylinders might have had an impact if the contact angles were very different for the PAA-terminated and PAH-terminated PEMs. As shown earlier, the PAH-terminated substrates will have had a contact angle against water of around 105°, whereas the PAA-terminated substrates will have had a contact angle of around 45°.5,15 This means that the PAAterminated substrates will have shown a larger pull-off force purely due to wetting reasons. This might also explain the large differences between the AFM and the SFA measurements for this polymer system. Because the contact angle of the PAHterminated substrate, 105°, was relatively close to 90°, the influence of the wetting will be very limited for this polymer system. A comparison of Figures 3 and 5 reveals another interesting phenomenon. Figure 5 clearly demonstrates that the deposition of PAH/PAA multilayers on mica substrates lowers the work of adhesion between these substrates. Considering that the mica substrate is a high energy substrate this might not be so surprising, but it is also clear that the effect of a PAH/PAA treatment in other systems, such as for silicon oxide/silicon oxide substrates and cellulose fibers,2,5,15 cannot be ascribed to an improvement in the thermodynamic work of adhesion between the substrates. Instead, the improvement in the adhesion between these substrates must be due to the way in which the joints are formed under wet conditions. As discussed earlier, a strong adhesive contact is formed if: (a) a large molecular contact area is formed as the substrates come into contact, which is naturally promoted if the substrates are soft; and (b) the molecules can migrate across the interface to form a large number of entanglements. If these substrates then are dried together, they will form a three-dimensional contact zone that will be more efficient in distributing stresses when loaded, compared with the situation when two molecularly smooth substrates are brought together. The results of the present work can be related to recent studies of interactions between PEM-treated wood fibers when a sheet is formed. Sheets made from fibers treated by PEMs formed from the same polymer system and under the same conditions showed an increase in tensile index, that is, stress at break

2894 Langmuir, Vol. 25, No. 5, 2009

normalized with the amount of material tested, from about 20 kNm/kg for nontreated fibers to about 55 kNm/kg when treated with eight layers of LMw PAH/PAA adsorbed at pH 7.5/3.5.3 This result was in agreement with the results described in the present paper, showing a more significant improvement in tensile index when PAH was adsorbed in the outermost layer than when PAA was adsorbed in the outermost layer. Analysis of the contact area of individual fiber-fiber joints37 for fibers treated by PAH/PAA using light microscopy demonstrated an increase of the degree of contact from about 18% to 32% for fibers treated with five layers adsorbed at pH 7.5/3.5. For layers 3-5 the degree of contact was higher when PAH was adsorbed outermost than when PAA was adsorbed outermost indicating softer layers when the multilayer is capped by PAH which is in agreement with the results in the present study. This shows upon a very interesting chain of events ranging all the way from viscoelastic properties of formed layers, to wet adhesion forces as evaluated with AFM, to mechanical properties of single fiber/fiber joints and to fiber network properties. It also means that by molecular engineering and optimization of PEM properties it will be possible to tailor interactions between macroscopic substrates. What is still missing are the mechanical properties of thin PEM films, but recent developments have shown large promise38 at least for evaluation of the elastic properties of the films.

Conclusions In this work, adhesion forces between substrates treated with PEMs formed from PAH/PAA have been measured using AFM and SFA. From the AFM measurements with PEMs formed at pH 7.5, it can be concluded that the adhesion was increased when more layers were adsorbed and that the adhesion forces were greater when PAH was adsorbed in the outermost layer. (37) Eriksson, M.; Torgnysdotter, A.; Wågberg, L. Ind. Eng. Chem. Res. 2006, 45, 5279. (38) Nolte, A. J.; Cohen, R. E.; Rubner, M. F. Macromolecules 2006, 39, 4841.

Johansson et al.

Also, it was found that a low molecular mass combination (15k/ 8k) showed a greater influence on the adhesion than a high molecular mass combination (70k/240k). This was explained as being due to a higher number of interacting chain ends which were suggested to be essential to achieve a strong adhesion due to entanglement of the chains across the interface. The number of chain ends is lower for high molecular mass polyelectrolytes than for low molecular mass polyelectrolytes, and the mobility of the chains is also greater for the low molecular mass polyelectrolytes. From the AFM measurements it was also concluded that the adhesive forces increased with contact time at maximum load because of the longer time to form entanglements across the interface. SFA adhesion measurements were conducted for PAH/PAA adsorbed at pH 7.5/3.5 under dry conditions, at 100% RH, and under wet conditions. The SFA adhesion measurements showed that under dry conditions, the adhesive forces between two high energetic mica substrates were lowered when they were covered by PAH/PAA PEMs before the measurements. At 100% RH, the adhesive forces increased for the PEM-covered surfaces, in comparison to the adhesion under dry conditions, due to partial condensing of water and increased softness of the PEMs. From 100% RH to wet conditions, there was a rapid drop in adhesion. The SFA measurements also showed that there was a swelling of the layers when exposed to 100% RH or wet condition. This swelling was greatest when PAH was in the outermost layer, which is also in agreement to what has been shown by measuring the energy dissipation using QCM-D of PEMs formed from PAH/ PAA, showing higher values, i.e., less rigid structure when PAH was adsorbed outermost. Acknowledgment. E.J. acknowledges the Swedish Center for Biomimetic Fiber Engineering (Biomime) and the Lyckeby Research Foundation for financial support. E.B. acknowledges the Swedish Research Council for financial support. R.L. thanks the Bio fibre Materials Centre (BiMaC) for financial support. LA803628W