Langmuir 1996, 12, 3221-3224
3221
An Atomic Force Microscopy Study on the Transition from Mushrooms to Octopus Surface “Micelles” by Changing the Solvent Quality Amalia Stamouli, Eric Pelletier, Vasileios Koutsos, Eric van der Vegte, and Georges Hadziioannou* Department of Polymer Chemistry and Materials Science Centre, University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands Received December 19, 1995. In Final Form: April 17, 1996X Atomic force microscopy (AFM) is used to study the behavior of a diblock copolymer onto a solid surface while the solvent quality is changed. In a first step, the copolymer poly(2-vinylpyridine)/polystyrene (P2VP/PS) is adsorbed onto mica from a selective solvent (the PS block is well solvated and the P2VP is not solvated). In such a case, the P2VP block adsorbs preferentially on the substrate and anchors the PS block to the surface. In a second step, the solvent quality is reduced to such an extent that the PS block is no longer solvated. This procedure allows the formation of octopus surface “micelles”. The parameters characterizing this regime are measured using AFM. In addition, a direct determination of the adsorbance is performed. It reveals that the PS blocks when immersed in good solvent do not overlap. A mushroom surface regime instead of a brush one is thus created at the surface.
Introduction Adsorption phenomena of polymers at solid-liquid interfaces play an important role in steric stabilization of colloid particles suspended in solution,1,2 chromatography, adhesion,3 and biocompatibility of artificial organs in medicine.4 Thus, the structure of the adsorbed polymer is of fundamental importance. Different experimental techniques have been used to study the conformations of molecules adsorbed onto surfaces and the kinetics of adsorption: hydrodynamic methods,5,6 small angle neutron scattering (SANS),7 fluorescence excited by evanescent waves (EWIF),8 radio labeling,9 ellipsometric studies,10 surface plasmon oscillations (SPO),11,12 neutron reflectometry,13 total internal reflection fluorescence (TIRF),14 nuclear magnetic resonance (NMR),8 and surface forces apparatus (SFA).15-17 Atomic force microscopy (AFM)18 has been proven a very successful technique for the imaging and characterization of solid surfaces from micrometer scale down to the atomic level. Only few data X
Abstract published in Advance ACS Abstracts, June 1, 1996.
(1) Napper, D. Polymeric Stabilization of Colloidal Dispersions; Academic: London, 1983. (2) Gast, A.; Leibler, L. Macromolecules 1986, 19, 686. (3) Lee, L. H. Adhesion and Adsorption of Polymers; Plenum Press: New York, 1980. (4) Ruckenstein, E.; Chang, D. B. J. Colloid Interface Sci. 1988, 123, 170. (5) Koopal, L. K.; Hlady, V.; Lyklema, J. J. Colloid Interface Sci. 1988, 121, 49. (6) Priel, Z.; Silerberg, A. J. Polym. Sci., Polym. Phys. Ed. 1978, 16, 1917. (7) Cosgrove, T. J. Chem. Soc., Faraday Trans. 1 1990, 86, 1323. (8) Rondalez, F.; Ausserre, D.; Hervet, H. Annu. Rev. Phys. Chem. 1987, 38, 317. (9) Pefferkorn, E.; Carroy, S.; Varoqui, R. J. Polym. Sci., Polym. Phys. Ed. 1985, 23, 1997. (10) Takahashi, A.; Kawaguchi, M.; Hirota, H.; Kato, T. Macromolecules 1980, 13, 884. (11) Tassin, F. J.; Siemens, R. L.; Tang, W. T.; Hadziioannou, G.; Sxalen, J. D.; Smith, B. A. J. Phys. Chem. 1989, 93, 2106. (12) Munch, M. R.; Gast, A. P. Macromolecules 1990, 23, 2313. (13) Field, J. B.; Toprakcioglu, C.; Dai, L.; Hadziioannou, G.; Smith, G.; Hamilton, W. J. Phys. II 1992, 2, 2221. (14) Lok, B. R.; Cheng, Y. I.; Robertson, C. R. J. Colloid Interface Sci. 1983, 91, 104. (15) Tauton, H. J.; Toprakcioglu, C.; Klein, J. Macromolecules 1988, 21, 3333. (16) Hadziioannou, G.; Patel, S.; Granick, S.; Tirrell, M. J. Am. Chem. Soc. 1986, 108, 2869. (17) Belder, G. F. Ph.D. Thesis, University of Groningen, 1995.
S0743-7463(95)01557-5 CCC: $12.00
have been reported so far about the structure of endgrafted polymer films by using AFM. The adsorption of two-dimensional micelles of P2VP/PS diblock copolymer on flat substrates has been imaged.19 A highly regular microdomain structure covering the scanned domain has been shown. Zhao et al.20 studied the lateral structure of end-grafted PS chains by a small reactive end group on Si substrates, whereas in another study21 the PS was grafted via a short block, PEO, much shorter than the PS block. Both studies confirmed22 the transition occurring, in poor solvent, from dimples to isolated islands when the grafting density is decreased. Williams23 has theoretically studied the formation of octopus “micelles” of grafted polymers that were immersed in a bath of solvent of which the quality is changed. The transition from a good solvent to one that is worse than θ solvent, for polymer chains which are irreversibly grafted to a surface, has been studied. Depending on the grafting density, three different regimes can be found. At high grafting densities a uniform collapsed layer occurs;22 at low grafting densities each chain forms its own globule and in the intermediate regime octopus surface “micelles” are formed. See Figure 1. Two opposite effects play a role in their formation, on the one hand they want to minimize the contact area with the solvent, but this is opposed by the stretching energy of the chains which form the micelles. This kind of micelle have been also seen in computer simulations.24,25 But it has to be noted that they have nothing to do with the micelles observed in bulk copolymer systems.26,27 (18) Binnig, G.; Quate, C. F.; Gerber, Ch. Phys. Rev. Lett. 1986, 56, 930. (19) Meiners, J. C.; Ritzi, A.; Rafailovich, M. N.; Sokolov, J.; Mlynek, J.; Krausch, G. Appl. Phys. A 1995, 61, 519. (20) Zhao, W.; Krausch, G.; Rafailovich, M. N.; Sokolov, J. Macromolecules 1994, 27, 2933. (21) O’Shea, S. J.; Welland, M. E.; Rayment, T. Langmuir 1993, 9, 1826. (22) Yeung, C.; Balazs, A. C.; Jansnow, D. Macromolecules 1993, 26, 1914. (23) Williams, D. R. M. J. Phys. II 1993, 3, 1313. (24) Lai, P.-Y.; Binder, K. J. Chem. Phys. 1992, 97, 586. (25) Grest, G. S.; Murat, M. Macromolecules 1993, 26, 3108. (26) Tang, W. T. Ph.D. Thesis, Stanford University, 1987. (27) Sikora, A.; Tuzar, Z. Makromol. Chem. 1983, 184, 2049.
© 1996 American Chemical Society
3222 Langmuir, Vol. 12, No. 13, 1996
Stamouli et al.
Table 1. Presentation of the Different Parameters Measured from AFM Pictures time (min)
xa
Dhalfb (nm)
Γd (mg/m2)
σc (10-16 m-2)
σ* e (10-16 m-2)
Nf (10-12 mol cm-2)
9 76 224 5760
92 98
27.67 24.85 25.90 20.20
1.525 1.624
0.519 0.553
0.219 0.234
0.862 0.918
2.486
0.846
0.358
1.405
150
x, number of features within a scanned area of 500 × 500 nm2. b Dhalf, half the distance between the centers of two neighboring features. c σ, the number of chains per unit surface area. d Γ, the mass coverage. e σ*, the grafting density. f N, the number of adsorbed molecules per unit surface area. a
In this paper we investigate the conformations of PS chains adsorbed via a P2VP block, when the solvent quality is decreased. The images obtained with AFM show that each polymer chain does not collapse individually but that the PS blocks fuse to form octopus surface “micelles”. Although the change of the solvent quality alters the conformation of the PS segments, it does not change the adsorbance. The number of the P2VP blocks that are adsorbed onto the substrate is expected to remain the same. Thus, information about the chain conformation in good solvent can be obtained. The system and the procedure used are described in the first section. The second part is devoted to a theoretical background in which the characteristic parameters of the system are defined. In the remaining last part, we discuss our experimental results. Experimental Section Poly(2-vinylpyridine)/polystyrene diblock copolymer (Mw,P2VP ) 102 000 g mol-1, Mw,PS ) 75 000 g mol-1 with an index of polydispersity of 1.12) is used. The monomeric size of a polystyrene segment is comparable with the dimension of a poly(2-vinylpyridine) segment: aPS ) 5.46 Å and aP2VP ) 5.32 Å, these values were obtained from their bulk densities, FPS ) 1.06 g/mL and FP2VP ) 1.17 g/mL.28 The block copolymer is dissolved in toluene (Merck) at a concentration of 0.05 mg/mL. Data for the radius of gyration of PS in toluene have been reported.29 RPS ) 1.86NPS0.595 ) 8.9 nm. In such a solvent, the P2VP/PS copolymer forms micelles with the P2VP block as the core. Tang et al.,26 using light scattering, determined a critical micelle concentration (cmc) of 62 µg/mL for the 76/577 diblock and 65 µg/mL for the 8/60 and 68/60 diblocks, in toluene. Sikora et al.27 used block copolymers with larger PVP blocks and determined a cmc of the order of 3 mg mL-1. We expect therefore that our solutions are below the cmc for the concentration that we have used. It is known19 that the presence of micelles results in the formation of a micellar structure above the homogeneous brush layer, affecting the kinetics of adsorption.11,12 Mica sheets (clear ruby, grade 2) are cleaved in a laminar flow hood and immediately immersed in the polymer solution. Toluene is a selective solvent for the P2VP/PS diblock copolymer; the P2VP block is not solvated and adsorbs on the surface while at the same time the PS block is well solvated and stretches away from the surface. Hadziioannou et al.16 were the first to show that this copolymer could be used to mimic a brush by using a surface forces apparatus. After an appropriate incubation time, presented in Table 1, the mica sheets are taken out, rinsed with toluene, and dried under a stream of argon. A commercially available AFM (Topometrix, Explorer Model) is used with microfabricated Si3N4 cantilevers, having a spring constant of 0.036 N m-1. Because of the softness of the PS chains the scanning is performed in water so as to avoid the deformation or even destruction of polymer chains that can occur while imaging the samples. Several samples were made for each incubation time and each sample was imaged at different areas. Similar pictures were obtained, demonstrating the reproducibility of the results. (28) Webber, R. M.; Anderson, J. L. Langmuir 1994, 10, 3156. (29) Higo, Y.; Ueno, N.; Noda, I. Polym. J. 1983, 15, 367.
Theoretical Background Marques et al.30 have theoretically studied the adsorption of A-B diblock copolymers onto a solid plane immersed in a highly selective solvent. The A part is in a poor solvent and forms a molten layer on the solid wall where the solvent does not penetrate. The B part is in a good solvent and forms a brush grafted layer on this molten layer. Various adsorption regimes are found depending on the asymmetry between the two parts of the copolymer which is measured by β. This parameter is defined as the ratio of the respective molecular dimensions of the diblock copolymer, β ) RFB/RGA. The B partspolymerization index NBshas an unperturbed radius of gyration of RFB ) NB3/5aB, where aB is the segment length. The A partspolymerization index of NAsis strongly attracted to the wall with a characteristic radius of RGA ) NA1/2aA, where aA is the segment length. Assuming that the segment lengths of the A and B blocks are equal, β ) NB3/5/NA1/2. Three areas have been found as a function of the value of β.31 Copolymers with β > NA1/2 can be identified as highly asymmetric, those with 1 < β < NA1/2 as moderately asymmetric, and those with β < 1 as symmetric. Parsonage et al.32 measured the adsorption of a series of P2VP/ PS diblock copolymers from toluene solutions onto mica sheets and silicon wafers. A fit to Parsonage’s data has been made by means of the following expression28
σP ) σ0
0.06866β1.0577 1 + 0.004672β1.2258
(1)
where σP is the coverage corresponding to their experimental data and σ0 is the overlap coverage of the PS blocks onto the P2VP layer adsorbed onto the solid surface. σ0 is defined by
σ0 ) (πRPS2)-1
(2)
Results and Discussion A typical image of an AFM scanning performed on a moderately asymmetric, β = 25, diblock copolymer P2VP/ PS adsorbed on mica is presented in Figure 1. An array of close-packed spherical features are observed on the substrate. The gray scale indicates the height of the features, ranging from 0 (black) to 14 nm (white). From the AFM images we are able to measure the following parameters: x, number of spherical features within a scanned area, and Dhalf, half the distance between the centers of two neighboring features. These parameters are presented in Table 1. In good solvent the PS blocks are swollen, Figure 2a, with an unperturbed radius of gyration29 equal to RPS ) (30) Marques, C.; Joanny, J. F.; Leibler, L. Macromolecules 1988, 21, 1051. (31) Guzonas, D. A.; Boils, D.; Tripp, C. D.; Hair, M. L. Macromolecules 1992, 25, 2434. (32) Parsonage, E.; Tirrell, M.; Watanabe, H.; Nuzzo, R. G. Macromolecules 1991, 24, 1987. Tirrell, M.; Parsonage, E.; Watanabe, H.; Dhoot, S. Polym. J. 1991, 23, 641.
Micelle Transition with Solvent
Langmuir, Vol. 12, No. 13, 1996 3223
of the individual collapse of PS blocks is not considered since Dhalf is 5 times as large as the size of one individual collapsed PS block. The comparison of the theoretical value of Rc with the values of Dhalf ) 24 ( 4 nm (Table 1) shows a better agreement. This demonstrates that the PS blocks have fused to form surface octopus micelles at the mica surface. Parameters such as the number of chains per unit surface area, σ, and the mass coverage, Γ, can be determined using the AFM. The number, x, of features within surface area (S ) 500 × 500 nm2) is counted. As n diblock copolymers are involved in the formation of one feature, σ is given by the relation
σ ) nx/S
(4)
From σ, the mass coverage Γ of polymer onto the mica surface is directly calculated Figure 1. AFM picture of an adsorbed layer of P2VP/PS diblock taken in water after immersion in toluene. The mica sheet was immersed in the polymer solution for 4 days. The blackto-white height difference equals 14 nm. The scan area is (1000 × 1000) nm2.
Γ)
(Mw,PS + Mw,P2VP) σ Na
(5)
where Na is the Avogadro number. An additional parameter, which will allow a comparison with results obtained with different techniques by other groups, which have not considered the P2VP blocks in their evaluation of the mass coverage, is the grafting density σ*, which can be calculated as:
σ* ) σ Figure 2. Scheme of the diblock copolymer as a function of the solvent quality. (a) Conformation of an adsorbed layer of P2VP/ PS immersed in a good solvent (toluene) and adsorbed onto a mica surface. The P2VP block is not solvated by toluene and anchors the PS block to the mica surface. The PS block does not adsorb onto mica and dangles in the solution. (b) The solvent quality has been reduced from good (toluene) to a worse than θ solvent (water) for the PS block. The PS segments have fused to form octopus “micelles”. Rn is the globule size and Rc is the corona size of a micelle.
8.9 nm. If some overlapping occurs, they can be slightly stretched away from the surface depending on the lateral confinement. When the solvent quality is reduced slowly from good to a worse than θ solvent, each PS block can individually collapse to form its own globule. Its size is given by ∼aPSNPS1/3 ) 4.77 nm. If the PS blocks in good solvent are slightly interpenetrated, n PS blocks can fuse together to form a surface octopus micelle, when the solvent quality is reduced to below θ, Figure 2b. Its characteristic radius should be Rn ∼ aPS(nNPS)1/3, which can be determined, using scaling laws, from the minimization of the free energy.23 Let all the chains in a circle of radius Rc collapse to form a globule of radius Rn, then
Rc ≈ aPSNPS8/15
(3a)
n ≈ NPS2/5
(3b)
Rn/Rc ≈ NPS-1/15
(3c)
Theoretically, therefore, 14 PS blocks have fused to form one micelle with dimension Rc ∼ 18.64 nm and Rn ∼ 12 nm. The numerical prefactors in the above expressions are of order unity, so a semiquantitative comparison between theory and experimental results can be done. According to the model, Dhalf can be equated to Rc. The hypothesis
(
)
Mw,PS MW,PS + MW,P2VP
(6)
The results are presented in Table 1 as a function of the incubation time of the mica surface within the toluene solution. The mass coverage values are consistent with published data determined by surface forces apparatus33 showing the validity of the technique of measurement. Equation 2 gives the grafting density value at which the PS blocks just start to overlap, σ0 ) 0.3654 × 10-16 m-2. It is about equal to the value that we have obtained for the highest incubation time (4 days). Webber’s expression provides a much higher grafting density value (eq 1), σP ) 0.65865 × 10-16 m-2. This shows that the PS blocks in good solvent are closer to a mushroom regime than to a brush regime. However, the PS blocks have to overlap slightly to allow the formation of octopus surface “micelles” when the solvent is changed. Such a process is supported by the results of Field et al.13 obtained with neutron reflectivity. These authors found that the polymer density profile for the block copolymers with large P2VP blocks was better described by the mushroom-type profile rather than by a brushtype profile. By reducing the size of the adsorbing block, they found a soft crossover from mushrooms to brushes. All measurements dealing with the adsorption of block copolymers reveal that the adsorbed amount of polymer increases with time until it reaches a first plateau value. Figure 3 shows the number of adsorbed molecules, N (mol/ cm2), calculated from the values of σ (see table), as a function of the incubation time of the mica sheets in the polymer solution. The log scale is selected, instead of a linear one, because a thermodynamic equilibrium requires a long time to establish. Assuming that the adsorption energy depends only on the thermodynamic characteristics of the diblock layer, we can use the Arrhenius relation for describing the adsorption rate (33) Patel, S. S.; Tirrell, M. Annu. Rev. Phys. Chem. 1989, 40, 597.
3224 Langmuir, Vol. 12, No. 13, 1996
Stamouli et al.
Conclusions
Figure 3. Plot of the number of adsorbed molecules, N, as a function of the incubation time of the P2VP/PS layer within the toluene solution. It shows that N varies linearly with log t, having a slope of 2.032 × 10-13 mol/h.
dN k ) dt t
(7)
where k denotes the average number of adsorbed molecules per hour. From the slope of Figure 3 we obtain k ) 2.032 × 10-13 mol/h. A detailed analysis of the adsorption kinetics cannot be performed since only a few timedependent measurements are available (Figure 3). But the value of k is in agreement with measurements performed on P2VP/PS layers on silica substrates, adsorbed from toluene solutions.34 (34) Huguenard, C.; Varoqui, R.; Pefferkorn, E. Macromolecules 1991, 24, 2226.
We have presented an AFM study of an adsorbed P2VP/ PS copolymer layer. The polymer was adsorbed from a selective solvent in such a way that the P2VP was the anchor block and the PS formed the tails extending in the solution far from the surface. The reduction of the solvent qualitystoluene to watershas changed the conformation of the PS chains in water. The PS chains have fused to create octopus surface micelles. The AFM images furnish a direct way to measure their sizes and time-dependent adsorption parameters such as the number of chains per unit surface area, the mass coverage, or the grafting density. The mass coverage data are in agreement with independent measurements performed with SFA. In addition, the results confirm previously reported data on similar systems showing that a mushroom regime can be expected, in good solvent, rather than a brush one. To obtain a more complete understanding on the kinetics of adsorption above and below the cmc, more measurements have to be performed. However the salient features have been described. Moreover the imaging of single, isolated chains should be possible by decreasing the concentration of the polymer solution. Acknowledgment. We thank Richard Gill for providing us with the block copolymer and Paul van Hutten for his careful reading of the manuscript. This research was supported by the Netherlands Foundation of Chemistry (SON), the Fundamental Research on Matter (FOM), and the HCM-Network “Functional Materials Organized at Supermolecular level”. LA9515579
