Surface Rheology of PEO−PPO−PEO Triblock Copolymers at the Air

Jun 10, 2005 - The dilatational rheological properties of monolayers of poly(ethylene oxide)-poly(propylene oxide)- poly(ethylene oxide)-type block ...
0 downloads 0 Views 274KB Size
Download PDF
Langmuir 2005, 21, 6373-6384

6373

Surface Rheology of PEO-PPO-PEO Triblock Copolymers at the Air-Water Interface: Comparison of Spread and Adsorbed Layers B. Rippner Blomqvist,*,†,‡ T. Wa¨rnheim,†,§ and P. M. Claesson†,‡ YKI, Institute for Surface Chemistry, Box 5607, 114 86 Stockholm, Sweden, and Department of Chemistry, Surface Chemistry, Royal Institute of Technology, Drottning Kristinas va¨ g 51, 100 44 Stockholm, Sweden Received December 30, 2004. In Final Form: April 29, 2005 The dilatational rheological properties of monolayers of poly(ethylene oxide)-poly(propylene oxide)poly(ethylene oxide)-type block copolymers at the air-water interface have been investigated by employing an oscillating ring trough method. The properties of adsorbed monolayers were compared to spread layers over a range of surface concentrations. The studied polymers were PEO26-PPO39-PEO26 (P85), PEO103PPO40-PEO103 (F88), and PEO99-PPO65-PEO99 (F127). Thus, two of the polymers have similar PPO block size and two of them have similar PEO block size, which allows us to draw conclusions about the relationship between molecular structure and surface dilatational rheology. The dilatational properties of adsorbed monolayers were investigated as a function of time and bulk solution concentration. The time dependence was found to be rather complex, reflecting structural changes in the layer. When the dilatational modulus measured at different concentrations was replotted as a function of surface pressure, one unique master curve was obtained for each polymer. It was found that the dilatational behavior of spread (Langmuir) and adsorbed (Gibbs) monolayers of the same polymer is close to identical up to surface concentrations of ≈0.7 mg/m2. At higher coverage, the properties are qualitatively alike with respect to dilatational modulus, although some differences are noticeable. Relaxation processes take place mainly within the interfacial layers by a redistribution of polymer segments. Several conformational transitions were shown to occur as the area per molecule decreased. PEO desorbs significantly from the interface at segmental areas below 20 Å2, while at higher surface coverage, we propose that segments of PPO are forced to leave the interface to form a mixed sublayer in the aqueous region.

Introduction Nonionic triblock copolymers of poly(ethylene oxide) and poly(propylene oxide) (PEOx-PPOy-PEOx), due to their properties as surface-active agents, are used in a wide range of industrial processes, such as dispersion stabilization, emulsification, emulsion stabilization, and foaming.1 These polymers may act as steric barriers at solid surfaces in medical devices to avoid undesired adhesion of proteins2 and stabilize the liquid interfaces of foams3 and emulsions4 to prevent coalescence and flocculation. Their rich phase behavior and other solution bulk properties, such as micellar structure, have been studied extensively; for example, see refs 5-8. On the other hand, until recently, experimental information on the structure * Corresponding author: e-mail [email protected]; telephone +46-8-5010 60 79; fax +46-8-20 89 98. † YKI, Institute for Surface Chemistry. ‡ Royal Institute of Technology. § Present address: ACO Hud AB, Box 622, 194 26 Upplands Va¨sby, Sweden. (1) Edens, M. W. Nonionic Surfactants, Polyoxyalkylene Block Copolymers; Marcel Dekker: New York, 1996; Vol. 60, Chapt. 4, pp 185-210. (2) Green, R. J.; Davies, M. C.; Roberts, C. J.; Tendler, S. J. B. J. Biomed. Mater. Res. 1998, 42, 165-171. (3) Sedev, R.; Jachimska, B.; Khristov, K.; Malysa, K.; Exerowa, D. J. Dispersion Sci. Technol. 1999, 20, 1759-1776. (4) Alexandridis, P.; Lindman, B. Amphiphilic Block Copolymers: Self-Assembly and Applications; Elsevier Science: Amsterdam, 2000. (5) Svensson, M.; Alexandridis, P.; Linse, P. Macromolecules 1999, 32, 637-645. (6) Wanka, G.; Hoffman, H.; Ulbricht, W. Macromolecules 1994, 27, 4145-4159. (7) Desai, P. R.; Jain, N. J.; Bahadur, P. Colloids Surf., A: Physicochem. Eng. Aspects 2002, 197, 19-26. (8) Malmsten, M.; Lindman, B. Macromolecules 1992, 25, 54405445.

and properties of adsorbed triblock copolymer layers at liquid interfaces, which is the focus of the present work, has been limited. The structure of flexible linear polymers at interfaces was treated theoretically in works by de Gennes and Alexander9-11 and by Scheutjens and Fleer.12,13 De Gennes and Alexander constructed scaling laws that describe the relation between the number of monomers (segments) and the structure of grafted polymer chains on a solid surface.9-11,14 For adsorbing chains, the theory predicted structural transitions due to excluded volume effects (steric forces). The layer structure changes from a close to two-dimensional flat structure (“pancake”), via swollen coils protruding into the solution (“mushroom”) to a regime of stretched chains (“brush”), as the surface coverage increases. Diblock copolymers composed of one insoluble part (“anchor”) and one soluble part (“buoy”) were expected to form a compact precipitate of the anchor chains while the soluble chains formed a brush.10 The scaling laws are frequently used to evaluate experimental data, for example, ref 15, but have the disadvantage that all numerical coefficients are not given by the theoretical modeling. The mean field lattice theory of Scheutjens and Fleer12,13 takes into account short-range interactions between segment, solvent and surface. The model describes the segment (9) de Gennes, P. G. Scaling Concepts in Polymer Physics; Cornell University Press: London, 1979. (10) de Gennes, P. G. Adv. Colloid Interface Sci. 1987, 27, 189-209. (11) Alexander, S. J. Phys. Colloq. 1977, 38, 983-987. (12) Scheutjens, J.; Fleer, G. J. J. Phys. Chem. 1979, 83, 1619-1635. (13) Scheutjens, J.; Fleer, G. J. J. Phys. Chem. 1980, 84, 178-190. (14) de Gennes, P. G. Macromolecules 1980, 13, 1069-1075. (15) daSilva, A. M. G.; Filipe, E. J. M.; d’Oliveira, J. M. R.; Martinho, J. M. G. Langmuir 1996, 12, 6547-6553.

10.1021/la0467584 CCC: $30.25 © 2005 American Chemical Society Published on Web 06/10/2005

6374

Langmuir, Vol. 21, No. 14, 2005

concentration profile of adsorbed layers and allows an evaluation in terms of trains (sequences of segments in contact with the surface), loops (segments in the solution with both ends at the surface) and tails (a chain fixed to the surface at one end). The model predicts that the conformation at low surface concentration is flat, with most segments in trains, some loops, and very few tails, while at higher concentration, the number of loops and tails become more significant. It was predicted that the segment density of adsorbed homopolymer layers at high surface concentration decays roughly exponentially with distance from the surface, though at large distances the decay is slower, due to long dangling tails that protrude into the solution. The early theories of de Gennes, Alexander, and Scheutjens and Fleer have been used and extended in different directions by several groups.16-20 Marques et al.17 discussed diblock copolymer adsorption in selective solvents, and Ligoure20 treated polymers weakly attracted to the surface. Linse and Hatton19 applied the theory of Fleer and Scheutjens to elucidate transitions in surface tension isotherms of PEO-PPO-PEO polymers and also to predict segment distribution profiles at the air-water interface. Aguie-Beghin et al.18 used scaling arguments to describe the relationship between surface pressure and surface concentration. Surface force measurements have confirmed that the thickness of, for example, adsorbed PEO homopolymer layers on solid surfaces (mica)21 and layers of poly(butylene oxide)-PEO block copolymers22 and poly(vinyl alcohol)23 in thin liquid films, can be much larger than the solution dimensions of the chains (good solvent conditions). Neutron reflectivity measurements of adsorbed PEO homopolymer24 and PEO-PPO-PEO block copolymers25-27 have added experimental information on segment density profiles perpendicular to the air-water interface. The results of Vieira et al.26 suggested that the adsorbed layer structure could be described as follows. The uppermost layer, next to air, contains only propylene oxide (PO) residues. Below this layer there is a substantial mixing of PO and ethylene oxide (EO) residues in water, while EO tails and the solvent are present farthest from the interface. There is also one published study on the structure and molecular orientation of PEO-PPO-PEO polymers as measured by sum frequency generation vibrational sprectroscopy (SFG),28 in which it is concluded that the methyl group of PPO at high surface coverage is oriented close to normal to the interface. Compression studies in Langmuir troughs of spread polymer layers have provided evidence for various struc(16) Baranowski, R.; Whitmore, M. D. J. Chem. Phys. 1995, 103, 2343-2353. (17) Marques, C.; Joanny, J. F.; Leibler, L. Macromolecules 1988, 21, 1051-1059. (18) Aguie-Beghin, V.; Leclerc, E.; Daoud, M.; Douillard, R. J. Colloid Interface Sci. 1999, 214, 143-155. (19) Linse, P.; Hatton, T. A. Langmuir 1997, 13, 4066-4078. (20) Ligoure, C. J. Phys. II 1993, 3, 1607-1617. (21) Klein, J.; Luckham, P. Nature 1982, 300, 429-431. (22) Rippner, B.; Boschkova, K.; Claesson, P. M.; Arnebrant, T. Langmuir 2002, 18, 5213-5221. (23) Lyklema, J.; Van Vliet, T. Faraday Discuss. Chem. Soc 1978, 65, 25-32. (24) Lu, J. R.; Su, T. J.; Thomas, R. K.; Penfold, J.; Richards, R. W. Polymer 1996, 37, 109-114. (25) Vieira, J. B.; Li, Z. X.; Thomas, R. K. J. Phys. Chem. B 2002, 106, 5400-5407. (26) Vieira, J. B.; Li, Z. X.; Thomas, R. K.; Penfold, J. J. Phys. Chem. B 2002, 106, 10641-10648. (27) Sedev, R.; Steitz, R.; Findenegg, G. H. Physica B 2002, 315, 267-272. (28) Chen, C.; Even, M. A.; Chen, Z. Macromolecules 2003, 36, 44784484.

Blomqvist et al.

tural transitions. Surface pressure-area (π-A) isotherms for PEO homopolymers29-31 have indicated that a PEO layer collapses into the aqueous subphase at surface pressures between 5 and 10 mN/m, depending on the PEO molecular weight and temperature. It was also claimed that structural changes occur below a segmental area of 20 Å2, while the polymer remains at the interface.29 π-A isotherms for different ethylene oxide-containing block copolymers15,32-35 have pointed to transitions or gradual conformational changes at π ≈ 5 mN/m and/or π ≈ 10-11 mN/m. The change occurring at π ≈ 10 mN/m is apparently universal for PEO-based block copolymers of a wide variety of anchoring hydrophobic blocks, and it has been associated with the PEO chains desorbing into the water subphase.15,33,34 Others interpret the plateau at π ≈ 5 mN/m as the pressure when PEO segments linked to a more hydrophobic part begin to form loops and tails.32 Of the mentioned reports, only the work of Kim and Yu35 investigates a PEO-PPO-PEO block copolymer (up to π ) 11 mN/m). Munoz et al.36,37 have presented a series of studies on PEO76PPO29PEO76 (F68) in which the interfacial structure was discussed on the basis of results from π-A isotherms of spread monolayers and amounts adsorbed from solution at equilibrium. The viscoelasticity of polymer monolayers at the airwater interface is intimately connected with the molecular conformation. Surface dilatational rheology methods38 are therefore very sensitive and useful for monolayer characterization. The dilatational viscoelasticity of PEO homopolymer at the air-water interface has been thoroughly investigated over wide frequency and concentration intervals by Noskov et al.39-41 using several dilatational methods. They found that the main relaxation process of the PEO system is rearrangements within the surface layer.41 The objective of this investigation is to obtain an understanding of the relationship between block copolymer structure and changes in the dilatational rheological properties of spread and adsorbed PEO-PPO-PEO block copolymer layers. Three polymers with different PEO and PPO block sizes, PEO99-PPO65-PEO99 (F127), PEO103PPO40-PEO103 (F88), and PEO26-PPO39-PEO26 (P85), have been compared by compression of spread layers in a Langmuir trough and by use of a low-frequency oscillating ring rheometer42 to determine dilatational surface rheological properties of adsorption layers. Dynamic viscoelastic data during adsorption from bulk (29) Shuler, R. L.; Zisman, W. A. J. Phys. Chem. 1970, 74, 15231534. (30) Kuzmenka, D. J.; Granick, S. Macromolecules 1988, 21, 779782. (31) Sauer, B. B.; Yu, H. Macromolecules 1989, 22, 786-791. (32) Holland, N. B.; Xu, Z.; Vacheethasanee, K.; Marchant, R. E. Macromolecules 2001, 34, 6424-6430. (33) Bijsterbosch, H. D.; de Haan, V. O.; de Graaf, A. W.; Mellema, M.; Leermakers, F. A. M.; Cohen Stuart, M. A.; van Well, A. A. Langmuir 1995, 11, 4467-4473. (34) Baekmark, T. R.; Elender, G.; Lasic, D. D.; Sackman, E. Langmuir 1995, 11, 3975-3987. (35) Kim, C.; Yu, H. Langmuir 2003, 19, 4460-4464. (36) Munoz, M. G.; Monroy, F.; Ortega, F.; Rubio, R. G.; Langevin, D. Langmuir 2000, 16, 1083-1093. (37) Munoz, M. G.; Monroy, F.; Ortega, F.; Rubio, R. G.; Langevin, D. Langmuir 2000, 16, 1094-1101. (38) Miller, R.; Wustneck, R.; Kragel, J.; Kretzschmar, G. Colloids Surf., A: Physicochem. Eng. Aspects 1996, 111, 75-118. (39) Noskov, B. A.; Alexandrov, D. A.; Loglio, G.; Miller, R. Colloids Surf., A: Physicochem. Eng. Aspects 1999, 156, 307-313. (40) Noskov, B. A.; Akentiev, A. V.; Loglio, G.; Miller, R. J. Phys. Chem. B 2000, 104, 7923-7931. (41) Noskov, B. A.; Akentiev, A. V.; Bilibin, A. Y.; Zorin, I. M.; Miller, R. Adv. Colloid Interface Sci. 2003, 104, 245-271. (42) Kokelaar, J. J.; Prins, A.; Degee, M. J. Colloid Interface Sci. 1991, 146, 507-511.

PEO-PPO-PEO Copolymers at the Air-Water Interface

solutions of various concentrations ranging from 0.02 to 1600 µM are presented. One interesting finding is a surface concentration-dependent redistribution of the relatively hydrophobic PPO block from the air-water interface to sublayers further down in the surface film. During the course of our work, three studies on the dilatational rheology of PEO-PPO-PEO monolayers, to which we will relate our results, have been reported.35,43,44

Langmuir, Vol. 21, No. 14, 2005 6375 π-ln A isotherms; that is, according to the general formula E ) -dπ/d ln A. The E versus π curves were smoothed by using mean values of five consecutive surface pressure values when the dilatational modulus was calculated. Theory: Surface Dilatational Rheology. The surface dilatational modulus E is defined as the resulting change in surface tension, dγ, caused by a change of the air-water interfacial area A, expressed as the relative area variation, dA/A:45-47

dγ dπ )(dA ) ) d dγ ln A d ln A

E)A

Experimental Section Materials. The triblock copolymers Synperonic PE/P85 (PEO26-PPO39-PEO26, M h w ) 4600 g/mol, batch 1005BD0428), Synperonic PE/F88 (PEO103-PPO40-PEO103, M h w ) 11 400 g/mol, batch A3), and Synperonic PE/F127 (PEO99-PPO65-PEO99, M hw ) 12 500 g/mol, batch 2804BK0518) were kindly provided from ICI Surfactants, Uniqema, The Netherlands. PEO represents a -CH2-CH2-O- unit, while PPO represents a -CH2-CH(CH3)O- unit. The block copolymers were used as received. Ultrapure deionized water, purified in an Elga Elgastat LabWater system (18.2 MΩ‚cm resistivity), was used. All glassware and the Wilhelmy plates were washed with Micro-90 (no. 9050, International Products Corp.) and rinsed with water, followed by 2-propanol (AnalaR grade), and then water. Stock solutions of the block copolymers were prepared by dissolving the polymer in water by gentle stirring overnight at least 1 day before the measurement and were stored (at most 3 days) under nitrogen at room temperature until use. Diluted solutions were prepared by weighing of appropriate amounts of stock solution. Methods: Surface Dilatational Rheology. An oscillating ring trough method as described by Kokelaar et al.42 was used for determining the dilatational rheological properties. The apparatus consisted of a ground glass cylinder (diameter 10 cm), which was dipped and raised in the vertical direction by a sinusoidal drive, in a glass trough (diameter 19 cm) containing the solution. The surface tension was measured by using a ground glass Wilhelmy plate placed inside the cylinder. The periodic expansion and compression caused an area change (dA/A) of 5%, which corresponds to an average area change of 0.5 cm2/s over each cycle. The oscillation frequency range was 0.001-0.5 Hz. Measurements were performed at a fixed frequency of 0.13 Hz (corresponding to an angular frequency of ω ) 0.82 rad s-1). Sample solution (250 mL) was poured into the trough at t ) 0, and the first data point could be recorded at a surface age of t ≈ 30 s. Layers obtained by adsorption, in separate experiments for each solution concentration, were investigated at room temperature (20-21°C) for 1 h. The investigated range of block copolymer concentrations was 0.02-1600 µM. The surface tension γ, the complex surface dilatational modulus E (sometimes called the dynamic elasticity35,41), the storage modulus Ed, and the loss modulus ηdω were determined as a function of time. Langmuir Film Measurements. The surface pressure-area isotherms were obtained by use of a Teflon Langmuir trough with the dimensions 10 × 5 × 60 cm, equipped with one moveable barrier. The barrier was positioned so that the initial (maximum) surface area was 514 cm2. The surface pressure π was determined by the Wilhelmy plate technique, with a thin ground glass plate, and was recorded against the total area accessible for the film. Data were collected every 1.2 s. The trough, which was housed in a Plexiglas chamber, was cleaned thoroughly with water and the bare surface was compressed to check the cleanliness before an experiment was started. The trough was filled with 1 L of water. Next, concentrated aqueous solutions of block copolymers were spread by use of a glass microliter syringe with a steel needle. After 10 min of equilibration, the film was compressed at a constant barrier speed of 1 mm/s (corresponding to an area change rate of 1 cm2/s) to the minimum surface area of 60 cm2, immediately followed by decompression of the layer at the same speed. The experiments were performed at room temperature (20-21 °C). The surface dilatational modulus (also called static elasticity in this case35,41) was calculated as the derivative of the (43) Hambardzumyan, A.; Aguie-Beghin, V.; Daoud, M.; Douillard, R. Langmuir 2004, 20, 756-763. (44) Mun˜oz, M. G.; Monroy, F.; Herna`ndez, P.; Ortega, F.; Rubio, R. B.; Langevin, D. Langmuir 2003, 19, 2147-2214.

(1)

When the surface layer is in slow equilibrium with the bulk solution or when it is composed of spread insoluble molecules, eq 1 can be expressed as E ) dπ/d ln Γ, where Γ is the surface excess. When the interfacial area is subjected to periodic compressions and expansions at a given frequency, relaxation processes such as diffusional exchange between the surface layer and the bulk solution or molecular rearrangements within the layer may cause a phase difference (measured by the phase angle θ) between the applied area variation and the surface tension response. In that case E is a complex number (E ) Er + iEi), and there are two contributions to the dilatational modulus: the elastic real part, Er, reflecting the energy stored in the layer, and the viscous imaginary part, Ei, accounting for the energy lost through relaxation processes. The nomenclature of surface rheological parameters found in the literature is not consistent. We use Ed and ηdω to refer to the storage and loss part, respectively, of the complex dilatational modulus. For sinusoidal variations of the surface area, the storage and the loss modulus are given by

Er ) Ed ) |E| cos θ

(2)

Ei ) ηdω ) |E| sin θ

(3)

and

where ηd is the surface dilatational viscosity and ω is the angular frequency of the oscillations. In the absence of relaxation processes affecting the surface dilatational modulus, the phase angle θ is equal to 0 and the surface layer behaves as a purely elastic body. Surface Shear Rheology. Measurements were carried out by using an automatic in-plane oscillatory ring apparatus, Camtel CIR 100 (Camtel Ltd., Royston, U.K.) The technique has been described previously.48 Briefly, a platinum Du Nou¨y ring (diameter 13 mm) was placed in the plane of the air-liquid interface. The oscillation resonance frequency was 3 Hz and the strain amplitude was set to 5 × 10-3 rad. The shear elastic G′ and shear viscous G′′ moduli were calculated from the applied forces required to maintain resonance frequency and amplitude (normalized resonance mode). Water was used as the reference interface, and a reference measurement was taken once a day. Solutions were poured into a glass vessel (diameter ) 4.6 cm) and the surface shear rheology of the adsorbed layer was measured for 1 h at room temperature (20-21 °C). Adsorbed layers (solution concentration range 0.02-1600 µM) were investigated. The relation between the surface dilatational and surface shear parameters has been discussed by Miller et al.38 and Freer et al.49

Results Some molecular characteristics of the investigated block copolymers P85, F88, and F127 are shown in Table 1. In quoting the cmc values, we have adopted the view that the critical micelle concentration (cmc) corresponds to the (45) Lucassen, J.; van den Tempel, M. J. Colloid Interface Sci. 1972, 41, 491-498. (46) Lucassen-Reynders, E. H. Food Struct. 1993, 12, 1-12. (47) Lucassen, J.; Giles, D. J. Chem. Soc., Faraday Trans. 1975, 71, 217-232. (48) Warburton, B. In Technologies in rheological measurement; Collyer, A. A., Ed.; Chapman & Hall: London, 1993; pp 55-97. (49) Freer, E. M.; Yim, K. S.; Fuller, G. G.; Radke, C. J. J. Phys. Chem. B 2004, 108, 3835-3844.

6376

Langmuir, Vol. 21, No. 14, 2005

Blomqvist et al.

Table 1. Molecular Characteristics of Investigated PEOx-PPOy-PEOx Triblock Copolymers polymer

mol. wt.,a g/mol

PEO, wt %

PPO, wt, g/mol

x-y-x

P85

4600

50

2300

26-39-26

F88

11 400

80

2300

103-40-103

F127

12 500

70

3800

99-65-99

cloud point, °C (wt %)

cmc, µM (°C)

85 (10)b 85 (1)c >100(10)b >100 (1)c >100(1)d >100(15)e

8700 (25)f 1500 (35)f 800 (25)g 600 (25)f

a Average value. b Reference 50. c Reference 51. d Reference 52. e Reference 8. f Reference 53. g Reference 54. Note that the cmc of PEOPPO-PEO block copolymers decreases with increasing temperature.

high-concentration breakpoint of surface tension isotherms. Alternatively it can be determined by dyesolubilization methods.19,55 For a discussion on this topic the reader is also referred to refs 22, 25, and 56. All measurements were carried out well below the cmc and the cloud point, except at the highest concentration of the adsorption experiments (1600 µM), which could be close to the cmc according to some reports. These particular polymers were chosen to form two pairs, one pair with similar-sized hydrophobic PPO blocks (P85 and F88) and one pair with similar-sized, relatively hydrophilic PEO blocks (F88 and F127), for the purpose of comparing the relative importance of the respective blocks for the rheological behavior of interfacial layers. The surface tension as a function of time for some selected samples is illustrated in Figure 1a. At the lowest concentration (0.02 µM) there was no measurable effect on the surface tension due to adsorption, except for the largest polymer F127. In this case a decrease of the surface tension was observed when the adsorption had been allowed to proceed for about 45 min. The reason is that the amount of polymer at the surface, due to slow diffusion to the surface region in combination with low bulk concentration, initially was too low to provide a measurable surface tension decrease. The solution diffusion coefficient of F127 is on the order of 10-11 m2/s.8 The equilibrium situation at the air-water interface is reached very slowly, which is consistent with data for other block copolymer systems.22,57,58 In addition to the relatively slow bulk diffusion and long relaxation times at the interface,37 this is partly due to the polydispersity of the samples. As the adsorption time increases, polymers that are more hydrophobic than average are expected to accumulate at the interface.59 The true equilibrium state was not attained at any concentration for any of the polymers within 1 h of adsorption. The surface tension continued to decrease very slowly with time, although almost constant surface tension values were reached at the highest bulk concentrations (rate of surface tension decrease is ≈0.1 mN/m every 10 min at t ) 60 min for the (50) Prasad, K. N.; Luong, T. T.; Florence, A. T.; Paris, J.; Vaution, C.; Seiller, M.; Puisieux, F. J. Colloid Interface Sci. 1979, 69, 225-232. (51) Desai, M.; Jain, N. J.; Sharma, R.; Bahadur, P. J. Surfactants Deterg. 2000, 3, 193-199. (52) Desai, P. R.; Jain, N. J.; Sharma, R. K.; Bahadur, P. Colloid Surf., A: Physicochem. Eng. Aspects 2001, 178, 57-69. (53) Alexandridis, P.; Holzwarth, J. F.; Hatton, T. A. Macromolecules 1994, 27, 2414-2425. (54) Lopes, J. R.; Loh, W. Langmuir 1998, 14, 750-756. (55) Alexandridis, P.; Athanassiou, V.; Fukuda, S.; Hatton, T. A. Langmuir 1994, 10, 2604-2612. (56) Kabanov, A. V.; Nazarova, I. R.; Astafieva, I. V.; Batrakova, E. V.; Alakhov, V. Y.; Yaroslavov, A. A.; Kabanov, V. A. Macromolecules 1995, 28, 2303-2314. (57) Cohen Stuart, M. A.; Cosgrove, T.; Vincent, B. Adv. Colloid Interface Sci. 1986, 24, 143-239. (58) Jiang, Q.; Chiew, Y. C.; Valentini, J. E. Colloids Surf., A 1996, 113, 127-134. (59) Schille´n, K.; Claesson, P. M.; Malmsten, M.; Linse, P.; Booth, C. J. Phys. Chem. B 1997, 101, 4238-4252.

Figure 1. Surface tension of P85 (PEO26PPO39PEO26), F88 (PEO103PPO40PEO103), and F127 (PEO99PPO65PEO99) as measured with a Wilhelmy plate at T ) 20-21 °C. (a) Variation with time of the surface tension for some selected samples. Solution concentrations were 0.02 µM P85 (O), 0.2 µM P85 (0), 2 µM P85 (]), 0.5 µM F88 ([), 1600 µM F88 (2), 0.02 µM F127 (+), 0.2 µM F127 (b), and 2 µM F127 (9). (b) Surface tension as a function of block copolymer solution concentration. The presented values represent the surface tension at the adsorption time t ) 1 h, which are not the equilibrium values.

200 and 1600 µM solutions). The shapes of the curves vary with the solution concentration, which reflects that the surface layer structure and dynamics change with surface coverage. A bimodal type of adsorption kinetics of PEO-PPO-PEO copolymers, including initial diffusion followed by an internal reorganization of the structure, was discussed recently.37 The surface tension values after 1 h of adsorption decrease in the order P85 > F88 > F127 (Figure 1b). A comparison of the results for F127 and F88 shows that a larger PPO block results in a larger surface tension decrease. This is as expected since PPO is more surfaceactive than PEO (ref 19 and references therein). Similarly, a larger PEO block (compare F88 and P85) results in a larger surface tension decrease, which we attribute to an

PEO-PPO-PEO Copolymers at the Air-Water Interface

Langmuir, Vol. 21, No. 14, 2005 6377

increased repulsion between the PEO chains that increases the surface pressure. The surface tension values after 1 h are reported, despite the fact that they are not equilibrium values, since this is the time span of the rheological measurements reported below. Surface Dilatational Rheology of Adsorbed Layers at the Air-Water Interface. The dilatational modulus E was measured at an oscillating frequency of 0.13 Hz during the period 30 s-1 h of adsorption from solutions of P85 (shown in Figure 2a,b), F88 (Figure 2c,d), and F127 (Figure 2e,f), at concentrations ranging from 0.02 to 1600 µM. Measurements were performed in separate experiments for each bulk concentration to collect data over a broad range of surface concentrations within a practical experimental time scale. At the lowest concentration, no response to dilatation of the surface was observed, consist with the surface tension measurements. The change of the modulus with time was rather complex. In some cases it increased initially, reached a maximum, and then decreased with time; for example, for F88 at 0.2 and 5 µM and F127 at 0.02-2 µM. It could also pass through a minimum, for example, for P85 at 1 µM and F88 at 0.5 µM. The highest values of the F88 and F127 moduli were 18-19 mN/m. A maximum of E ) 9-10 mN/m was also observed after 8 min at the concentration 0.2 µM for both F88 and F127. In the concentration range 20-1600 µM the modulus remained essentially stable, at least from t ) 30 min and throughout the experiments. It decreased with increasing solution concentration. The smallest molecule P85 differed in that the highest attained value of E was only 11 mN/m (see Figure 2a,b), obtained at the bulk concentration 1 µM. The real component of the modulus, Ed, was larger than the imaginary component, ηdω, for all three polymers over the whole concentration range (Figure 3). Thus, the elastic contribution clearly dominates the surface dilatational behavior. The data presented in Figure 3 represent the dilatational parameters after 1 h at the various solution concentrations. As noted already, they do not represent the true equilibrium situation, and only the parameters for the highest concentrations are likely to be close to equilibrium. The time evolution shown in Figure 2 explains, for example, why the elasticity at 0.02 µM F127 is higher than at 0.2 µM. Nevertheless, it is worth noting that the highest values of both elasticity and viscosity were obtained at solution concentrations close to 1 µM. The viscosity of the interfacial layers was less than ≈4 mN/m (ηdω ≈3 mN/m) even at the peak around 1 µM. At higher concentrations, the complex dilatational modulus as well as its elastic part decreased continuously with polymer concentration. The concentration dependence of the viscosity appears to be more complicated. ηdω of P85 and F88 was constant or decreasing in the interval 2-1600 µM, while for F127 the measured values were constant or slightly increasing at concentrations above 20 µM. The observed differences of ηdω are very small and close to the spreading of experimental data, but we chose to report the values since the trends were reproducible. The spreading of the dilatational parameters (E, Ed, and ηdω) and of the phase angle were 30 mN/m was exactly reproducible when the same layer was compressed once more directly after the first cycle, which also is not compatible with an irreversible dissolution into the bulk solution. (62) Barentin, C.; Muller, P.; Joanny, J. F. Macromolecules 1998, 31, 2198-2211. (63) Tamada, K.; Minamikawa, H.; Hato, M.; Miyano, K. Langmuir 1996, 12, 1666-1674. (64) Berg, J. M.; Eriksson, L. G. T.; Claesson, P. M.; Borve, K. G. N. Langmuir 1994, 10, 1225-1234. (65) Pezron, E.; Claesson, P. M.; Berg, J. M.; Vollhardt, D. J. Colloid Interface Sci. 1990, 138, 245-254.

Spread Monolayers versus Adsorbed Monolayers. Layers that assemble at the air-water interface through adsorption from bulk solution are common for all surfaceactive materials. The exchange of small soluble surfactants between the adsorbed layer and bulk solution is rapid. Under such conditions, classical Langmuir trough experiments cannot be carried out. On the other hand, for molecules with low aqueous solubility, like phospholipids, Langmuir trough experiments can be performed whereas classical surface tension measurements are impossible or impractical due to slow adsorption from a very dilute aqueous phase. Hence, most substances are only possible to investigate either as spread or as adsorbed monolayers. The block copolymers investigated here are an exception by being sufficiently hydrophobic to allow a spread film to be compressed significantly yet being water-soluble at room temperature and surface-active enough to adsorb to the interface. Even the PEO homopolymer is known to possess this property, which was first shown by Shuler and Zisman.29 The PEO-PPO-PEO block coplymers will naturally be even more strongly anchored to hydrophobic interfaces than PEO on its own, because of the methylene groups in the PPO block. Thus, we may make use of the unusual situation that surface rheological properties of adsorbed monolayers can be compared to those of spread layers, where the surface concentration is easily controlled. Before making this comparison, we would like to emphasize two issues. The first is concerned with the molecular composition of the layer of a polydisperse sample. The composition of the spread monolayer will, ignoring the possibility of dissolution of the most hydrophilic fraction of the sample into the subphase, reflect the composition of the block copolymer sample. On the other hand, for an adsorbed layer there will be a preferential adsorption of the most hydrophobic molecules in the layer. Thus, in general the exact composition of molecules at the interface may well be different for spread and adsorbed monolayers even though the same sample is used. The second issue is related to the measurement procedure. The spread monolayers are continuously compressed, which may lead to desorption into the subphase, particularly at high compressions. On the other hand, the adsorbed layers are exposed to sinusoidal oscillating changes of the surface area, which can in principle lead to desorption upon compression and readsorption during the expansion phase. Thus, the interfacial dynamics is in general not expected to be identical. However, if the adsorption and desorption kinetics are slow compared to the time scale of the measurement, the difference in measuring method will not play an important role.

PEO-PPO-PEO Copolymers at the Air-Water Interface

Langmuir, Vol. 21, No. 14, 2005 6381

Figure 6. (a-c) Dilatational modulus of block copolymer layers adsorbed at the air-water interface as a function of surface pressure: (a) P85; (b) F88; (c) F127. (d-f) Static dilatational modulus of spread layers, as calculated from surface pressure-area isotherms obtained by compression in a Langmuir trough. The surface pressure, π, is given by π ) γ0 - γ, where γ0 is the surface tension of water. (d) P85; (e) F88; (f) F127.

Variation of Dilatational Modulus with Surface Pressure. The surface rheological properties of the spread and adsorbed layers are compared by plotting the dilatational modulus as a function of surface pressure. This is possible since the surface tension and the dilatational modulus were recorded simultaneously for the adsorbed layers. In the dilatational modulus vs surface pressure plots shown in Figure 6a-c, the data originating from separate adsorption experiments at different bulk concentrations overlap to form one master curve for each block copolymer. This demonstrates that the modulus is a unique function of surface pressure, without large variations depending on adsorption time or solution concentration. This is somewhat surprising considering that the samples are polydisperse. Five regions (I-V, indicated in Figure 6c) are distinguished. At low surface pressures (0 to ≈5 mN/m) the modulus increases with π (region I), followed by a decrease of the modulus between π ≈ 5 and 10 mN/m (region II). After the minimum at π ≈ 10 mN/m, the modulus increases once again, initially in an apparently linear manner, to reach a second maximum (region III), and at even higher surface pressures the dilatational modulus again decreases (region IV). Region V, where

Table 2. Transitions in Adsorbed PEO-PPO-PEO Layers As Measured by Surface Dilatational Ring Trough Rheology first peak

minimum

second peak

block π E π E π E copolymer (mN/m) (mN/m) (mN/m) (mN/m) (mN/m) (mN/m) P85 F88 F127

6 5 5

9 9 9

10.5 11 10.5

8 2 3

16 22.5 21

11 19 19

lower values of the modulus are found, starts at π ≈ 3133 mN/m for F88 and F127 and at π ≈ 20 mN/m in the case of P85. The transition surface pressures taken at the centers of local maxima and the first minimum and the corresponding dilatational moduli are listed in Table 2. Transitions occur at π ≈ 5 and π ≈ 10 mN/m independently of block copolymer size, consist with previous results from surface-pressure isotherms of various other ethylene oxide-based polymers.32,34 The main difference between the polymers in region II is that the P85 modulus is higher at the minimum between the peaks, suggesting that the length of the PEO chains determines the behavior. The reason is that the shorter EO chains of P85 occupy a

6382

Langmuir, Vol. 21, No. 14, 2005

Blomqvist et al.

Table 3. Transitions in Spread PEO-PPO-PEO Layers upon Compression in a Langmuir Trough and Mass of Spread Polymer and Segmental Areas (A) Transitions in Spread PEO-PPO-PEO Layers first peak

minimum Aa

second peak Aa

polymer

π (mN/m)

E (mN/m)

(Å2)

π (mN/m)

E (mN/m)

(Å2)

π (mN/m)

E (mN/m)

Aa (Å2)

P85 F88 F127

6 5 5

11 10.5 10.5

1700 5000 5100

10 9.5 9.5

9 3 4

1200 2500 2700

15 23 20

14 27 19

800 400 600

(B) Mass of Spread Polymer and Segmental Areas first peak

minimum

second peak

polymer

Γb (mg/m2)

ac (Å2)

Γb (mg/m2)

ac (Å2)

Γb (mg/m2)

ad (Å2)

P85 F88 F127

0.45 0.38 0.41

19 20 19

0.64 0.76 0.77

13 10 10

0.95 4.7 3.5

9/21/15 2/10/2 2/9/3

a Area per molecule. Values were taken from plots of E versus the area per molecule constructed using the data in Figures 6d-f and 4a-c. The area at the center of the broad local maxima and minima is presented. b The surface coverage Γ ) (Mw × 103)(A × 10-20)-1NA-1, where A is the area per molecule and NA is Avogadro’s constant. c Area per segment ) area per molecule divided by (NPEO + NPPO), where N is the number of segments. d Area per segment ) area per molecule divided by (NPEO + NPPO)/NPPO/NPEO.

relatively smaller surface area and will accordingly give rise to a smaller variation of surface tension upon desorption. Above π ≈ 10 mN/m, there are clear differences between the polymers. The peak value of the modulus for P85 is only 11 mN/m, whereas it reaches 19 mN/m for the block copolymers with longer poly(ethylene oxide) chains. The larger repulsion between the longer PEO chains may explain this. However, we note that the results presented in ref 43 show that this is not the complete explanation because, according to their data, the dilatational modulus of a block copolymer with shorter EO chains could reach a higher value than that for one with longer EO chains, at constant length of the PPO block. Concentration-dependent exchange of small-molecule surfactants between the surface and the bulk phase is well-known to affect the dilatational rheology of adsorbed layers.45,46,66,67 However, in the case of polymers, effects of diffusional exchange are usually unimportant at frequencies >1 Hz.40 In region IV, the solution concentrations are in the range ≈1-5 µM, and the frequency used was only 0.13 Hz, which makes it tempting to consider an exchange relaxation process to explain the decrease of the modulus in this region. However, our experiments suggest a more complex picture. Note that the modulus at some concentrations (see, for example, P85 at 2 µM, F88 at 2 and 5 µM, and F127 at 1 and 2 µM) starts at a surface pressure around 10 mN/m, then increases to reach the second peak, and then decreases, while the bulk concentration is constant during the whole experiment. This is clearly inconsistent with an explanation based on increasing exchange between the bulk and the surface layer. Instead the second peak must be due to the onset of a conformational change. This view is supported by the results for spread monolayers discussed below. In region V, F88 and F127 show similar properties; the modulus remains fairly constant with time (and π) at each concentration but decreases with increasing bulk concentration. In contrast to region IV, this behavior may be consistent with an exchange with the bulk phase. The modulus is, however, not equal to 0, meaning that any relaxation due to exchange is not complete at the chosen frequency. The P85 modulus increased with time and π (66) Stubenrauch, C.; Miller, R. J. Phys. Chem. B 2004, 108, 64126421. (67) Langevin, D. Adv. Colloid Interface Sci. 2000, 88, 209-222.

at the highest solution concentrations. This is interpreted as being due to compositional changes of the adsorbed layer. We now turn to the E-π curves for spread (Langmuir) layers (Figure 6d-f). The similarities of the adsorbed and spread layers are striking, and the agreement at least up to surface pressures of ≈10 mN/m is remarkable. Identical viscoelastic behavior in spread and adsorbed films of PEO homopolymer and PEO-PPO-PEO copolymers up to the PEO collapse pressure of 10 mN/m has been noticed by others.31,35 The adsorbed and spread polymer layers of F127 behave nearly identically up to surface pressures of 30 mN/m, the only difference being that region III is shifted slightly toward higher π for the adsorbed layer. The P85 and F88 Langmuir layers show comparatively larger differences compared to the adsorption layers in regions IV and V, though the behaviors agree qualitatively (see peak values in Table 3A). These polymer layers are apparently more sensitive than F127 to the mode of surface deformation. In summary, the overall similarities between spread layers under compression and adsorbed layers confirm that the surface structure is determined by the surface concentration. Rearrangements within the layer explain most of the features of the surface dilatational properties. Exchange of molecules with the bulk is not important at a frequency of 0.13 Hz, except possibly at concentrations above ≈20 µM, where relaxation due to exchange could contribute to the rheological behavior. In one recent surface rheological study, the block copolymer F68 (PEO76-PPO29-PEO76) was investigated by the capillary wave technique at frequencies of 800 Hz and 5 kHz, a measurement principle that is completely different44 from the oscillating ring dilatation method used in our study. Nevertheless, the E-π curves for F88 and F127, which are similar polymers to F68 regarding the PEO/PPO balance, are very close to that of F68 over the whole pressure range. They also agree with the dependence of E on surface pressure for F68, as measured by the oscillating bubble technique at a frequency of 0.1 Hz43 (up to π ) 15 mN/m). As we have observed above, the E-π master curve for each polymer seems to be a result of the surface concentration only. The dilatational elasticity of this type of block copolymer appears to be insensitive to the type of method and frequency used, which again demonstrates that exchange processes between surface

PEO-PPO-PEO Copolymers at the Air-Water Interface

Langmuir, Vol. 21, No. 14, 2005 6383

layer and bulk solution are unimportant. We note, however, that in some cases clear frequency dependence has been reported.44 The transition areas per molecule and the corresponding mass of polymer (see Table 3) indicate that both the lowpressure maximum and the transition at the minimum at π ≈ 10 mN/m are determined by the mass of block copolymer present at the air-water interface. Furthermore, the transition molecular areas scale with the sum of the numbers of EO and PO monomers per molecule, as calculated by dividing the area by NEO+PO. This is consistent only with a mainly flat confomation where the area is simply the sum of the areas occupied by the PEO and PPO block. In contrast, the high-pressure maximum occurs at a surface coverage not related to the mass of polymer (even though values of Γ for F88 and F127 are more similar compared to that for P85); neither does the area per molecule scale with the number of EO, PO, or the sum of EO and PO units. Instead the molecular areas at the second peak decrease with increasing ratio between the polymerization degree of PEO and PPO blocks (NPEO/ NPPO), which is 1.3, 5.2, and 3.0 for P85, F88, and F127, respectively. Our data and those presented in ref 43, taken together, indicate that the surface pressure of the second maximum and the peak value of E are determined by the balance between PPO and PEO (and not only by the PEO chain length, as the data presented in this paper alone would indicate). However, it was suggested in ref 43 that the maximum dilatational modulus increased approximately linearly with the number percentage of PPO, but the reported “maximum” was not the real maximum in all cases, which makes this conclusion misleading, as also shown by our data. The area per molecule at the onset of the second local minimum (in the strongly compressed layers, region V) is in the range 200-300 Å2 for all three of the polymers and scales best with the number of PPO segments. The area per PPO segment at this point was 5-6 Å2 (not shown in Table 3B). Assuming thermodynamic equilibrium conditions, the exponent y, which was mentioned already in connection with Figure 4, is related to the dilatational modulus by the expression E ) yπ.18 Even though the system can hardly be regarded as being in true thermodynamic equilibrium, both the static and dynamic elasticity can be described by the scaling law at low surface pressures, with an exponent y ) 2-2.7, that is, close to that expected in a two-dimensional system (Figure 6). Values of y in this regime for similar PEO-PPO-PEO block copolymers were found to be 2.5-2.9,43,44 that is, consistent with our finding, and the same values apparently hold for both PEO and PPO coils.43 It has also been confirmed that y ) 2.85 for PEO homopolymers under good solvent conditions.68 A scaling relationship with E seems to hold also in the linear regimes above π ≈ 10-12 mN/m, where y is close to 1 for P85 but 2.5-2.7 for F88 and F127. Upon comparison with Figure 4d, one would expect a lower value for F88 and F127 in this regime (corresponding to a threedimensional structure). The reason for this discrepancy for F88 and F127 is suggested to be because the theory is strictly valid for dilute and semidilute equilibrium regime, and may thus not be applicable for the high surface concentrations encountered at high surface pressures. In Figure 7, both the compression and decompression curves of the static dilatational modulus are shown. (The data for the compression curve are the same as in Figure 6f, and the decompression curve was measured for the

same layer.) The data for the two curves agree with each other at π ) 0-5 mN/m and in the transition region around π ≈ 10 mN/m. Between 5 and 10 mN/m the decompression curve is slightly shifted, compared to the compression curve, toward higher pressures. At surface pressures above 15 mN/m the elasticity is different on decompression as compared to the compression values. We will not discuss this further, but we note that the dilatational elasticity of adsorbed monolayers at high surface pressures is considerably more similar to the elasticity of spread monolayes on compression than it is to the elasticity measured on decompression. An implication is that the compression curve reflects the equilibrium condition within the monolayer better than the decompression curve. The corresponding behavior has been observed for spread PEO homopolymers and was discussed in terms of conformational changes.29 The effects on the dilatational properties of polymer layers far from equilibrium were also discussed, for example, in ref 69. Conformational Changes at the Air-Water Interface. In this section the word “desorption” refers to the removal of a segment from direct attachment to the air-water interface. With this terminology a desorbed segment is still part of the adsorbed or spread layer but not directly in contact with the air-water interface. Before proceeding we note that the segment, of course, is a part of a polymer chain, which means that desorption of a segment is likely to lead to conformational changes along a region of the chain. A consequence is that the relaxation rate due to desorption of a segment will be longer than that associated with desorption of a free monomer. To fully understand the relaxation process, the rheological properties should be investigated over a very broad frequency range. This would mean employing a range of different techniques probing different frequency ranges. So far no such study has been reported. The ease with which a segment that is directly attached to the surface desorbs from the air-water interface during compression depends on the free energy barrier for moving the segment to the region just below the air-water interface. With increasing intralayer repulsion, that is, increasing surface pressure, the free energy of a segment adsorbed at the air-water interface increases, which in turn increases the tendency for desorption. Since a PO segment is anchored more strongly than an EO segment to the

(68) Sauer, B. B.; Yu, H.; Tien, C. F.; Hager, D. F. Macromolecules 1987, 20, 393-400.

(69) Boury, F.; Ivanova, T. Z.; Panaiotov, I.; Proust, J. E.; Bois, A.; Richou, J. J. Colloid Interface Sci. 1995, 169, 380-392.

Figure 7. Static dilatational modulus as a function of surface pressure during compression ([) immediately followed by expansion (]) of spread F127 block copolymer at the air-water interface at T ) 20-21 °C.

6384

Langmuir, Vol. 21, No. 14, 2005

interface, the energy barrier for desorption will be larger for the PO segment. We propose the following view of the structural changes of PEO-PPO-PEO block copolymers at the air-water interface. In the surface pressure range of π ) 0-5 mN/ m, both the PEO and PPO blocks are in contact with the air-water interface in a flat two-dimensional conformation. In this region the film is purely elastic, and as the available area decreases, the elasticity increases due to increasing overlap repulsion. The monolayer becomes saturated in the flat expanded conformation at a surface coverage of 0.4 mg/m2. This corresponds to an average area per segment of 19-20 Å2, which agrees with what was found previously for the PEO homopolymer.29 At the maximum around π ) 5-6 mN/m, PEO segments start to form loops and tails. Consequently the dilatational modulus falls, as the relaxation mechanism involving desorption of EO segments is increasingly facilitated. This process continues until π ) 9-10 mN/m, and then the majority of PEO segments leave the air-water interface to form a swollen three-dimensional structure in the aqueous phase. The transition occurs when the mass of polymer present at the interface is about 0.7 mg/m2 (0.8 mg/m2 for F88 and F127 and 0.6 mg/m2 for P85), corresponding to average segmental areas of 10 and 13 Å2, respectively. The mean area for P85 is somewhat larger because of the higher relative amount of the larger PO segments. From this point toward higher surface pressures, the sizes of the respective PEO and PPO blocks begin to determine the behavior. Upon further increase of π, the PPO blocks remain at the interface while the PEO chains gradually extend deeper into the aqueous phase. The modulus rises once again, due to increasing repulsion between PPO segments in the top layer and PEO segments in the sublayer. Both elasticity and viscosity contribute to the dilatational modulus of the extended layer. We propose that, at the second maximum, the repulsion within the layer has become significantly strong to facilitate desorption of the more hydrophobic PO segments, and the relaxation mechanism involving desorption of PO segments becomes progressively more important at higher surface pressures. Once more an increase of loops and tails leads to a faster redistribution between the upper and lower regions of the layer. This is the explanation for the decrease of the dilatational modulus in region IV. The desorption of PO segments is supported by the fact that the area per molecule in region V is significantly smaller (5-6 Å2) than expected for a close-packed PO segment in two-dimensional conformation. Munoz et al.36,44 reach the conclusion that PPO remains anchored to the interface at high pressures and forms some ordered structure of folded PPO chains that extend into the air phase. The reason for the decrease of the layer elasticity above π ) 20 mN/m was not discussed by them. Hambardzumyan et al.43 also discuss the possibility of ejection of PPO toward the air. Our explanation to the results is a partial solubilization/redistribution of the relatively hydrophobic PPO chains in the sublayer, which does not exclude that some PPO is situated above the air-water interface. Neutron reflectivity studies26 of the polymer PEO23PPO52PEO23 at two concentrations, one below and one above the cmc, showed that the top layer consisted of only PO segments and that the thickness of this layer was fairly constant over the concentration range. Additional PO segments were mainly accommodated in the layers below the top layer, that is, toward the aqueous

Blomqvist et al.

phase, and the PO profile extended deeper at the higher concentration. Theoretical calculations of volume fraction profiles for one concentration far below and one above the cmc19 have also shown that there is a segregation between EO and PO at the air-water interface but that PO penetrates both into the air phase and into a mixed sublayer of PO, EO, and water. Only EO contributes to the longest tails penetrating into the water phase. Thus, taken together, experimental and theoretical data do not support the simple stratified polymer brush model with a compact layer of anchoring PO blocks completely segregated from the soluble EO chains. The conformational variations seen in the present study are observed at low bulk solution concentrations. In condensed surface phases of PEO-PPO-PEO block copolymers corresponding to much higher solution concentrations, yet other types of structural transitions have been proposed.25 Conclusions The PEO-PPO-PEO block copolymers adopt different conformations as the available area per molecule becomes smaller due to increasing adsorption or decreasing airwater interfacial area, in the case of adsorbed or spread layers, respectively. Differences in molecular structure lead to distinctions in the surface relaxation time and consequently to differences in dilatational surface elasticity and viscosity. The relaxation processes are mainly rearrangements within the surface layer, except possibly at the highest solution concentrations. The block copolymer layers at the air-water interface are viscoelastic, though dominated by elastic behavior. Structural transitions occur gradually but in unique narrow intervals of surface pressure, independent of the path to reach this pressure. We propose the following conformational changes at the air-water interface in the interval of π ) 0-30 mN/m: (i) A flat conformation with most segments in contact with the interface up to π ≈ 5 mN/m. Above 5 mN/m, corresponding to a surface coverage Γ of 0.4 mg/m2 (≈20 Å2/segment), PEO segments start to form loops and tails. (ii) The main part of the PEO segments has desorbed from the interface at π ≈ 10 mN/m when Γ is ≈0.7 mg/m2 (≈10-13 Å2/segment) and a three-dimensional structure forms. (iii) A regime with PPO segments in a mainly flat conformation and PEO chains extending in the aqueous phase follows. An extended structure consisting mainly of PEO tails forms gradually with increasing pressure. (iv) At a limiting area and surface coverage (second maximum in Figure 6), which is determined by the PEO/ PPO ratio, the relatively hydrophobic PPO segments start to form loops and tails in the aqueous PEO layer below. This process continues until around π ≈ 20 and π ≈ 30 mN/m for P85 and F88/F127, respectively (corresponding to an average area per PPO segment of about 5-6 Å2). Acknowledgment. B.R.B. is grateful to Peter Wilde, Alan Mackie, and Mike Ridout for very valuable help and discussions at the Institute of Food Research in Norwich. The Foundation for Strategic Research, Colloid and Interface Technology Program, Sweden, and the European Union via the Marie Curie Training Site fellowship program (Contract QLK1-CT-2000-60030) at the Institute of Food Research, U.K., are acknowledged for funding. LA0467584