Langmuir 2001, 17, 6547-6552
6547
Controlled Adsorption of End-Functionalized Polystyrene to Silicon-Supported Tris(trimethylsiloxy)silyl Monolayers Christopher M. Stafford, Alexander Y. Fadeev,† Thomas P. Russell,* and Thomas J. McCarthy* Polymer Science and Engineering Department, University of Massachusetts, Amherst, Massachusetts 01003 Received April 3, 2001. In Final Form: July 9, 2001 Chemically grafted tris(trimethylsiloxy)silyl (sub)monolayers [tris(TMS)] were prepared on the native oxide of silicon by vapor-phase reaction of the corresponding monochlorosilane at elevated temperatures. By exploiting the inherently sluggish kinetics of the silanization reaction, the grafting density of the tris(TMS) (sub)monolayer can be tuned. Water contact angle, ellipsometry, and X-ray photoelectron spectroscopy were used to determine the surface coverage. Unreacted silanol groups on the substrate surface were used as adsorption sites for carboxylic acid end-functionalized polystyrenes (PS-COOH). The thickness of the adsorbed layer could be controlled by the tris(TMS) surface coverage, adsorbing solvent, and polymer molecular weight. Amino-tris(TMS) mixed monolayers were prepared by the reaction of (4-aminobutyl)dimethylmethoxysilane (ABDMS) with tris(TMS) submonolayers. Polymer adsorption to tris(TMS)/ABDMS mixed monolayers did not occur as readily as it did to the tris(TMS)/silanol surfaces. Aggregates of adsorbed PS were observed on the tris(TMS)-modified surfaces, with an average separation distance that increased with increasing tris(TMS) coverage. The effectiveness of these adsorbed layers in suppressing dewetting of a thin PS film was examined.
Introduction There has been considerable theoretical and experimental interest2-4in the adsorption of functionalized polymers to solid/liquid interfaces, which makes use of one or multiple functional groups on the polymer to anchor the polymer to the solid surface. This concept has been used for colloid stabilization,5,6enhanced wetting and adhesion,7,8fabrication of chemical microsensors,9 and biocompatibility.10,11 Covalent attachment of monofunctional organosilanes has proven to be a simple and versatile method for tuning the properties of solid surfaces such as wettability, adhesion, and surface activity. Recently, the preparation and properties of tris(trimethylsiloxy)silyl [tris(TMS)] monolayers and their use as patterns for the synthesis of uniformly mixed binary monolayers of organosilanes on oxidized silicon wafers was described.12 Contact angle * Corresponding authors: e-mail
[email protected] or
[email protected]. † Present address: Department of Chemistry, Seton Hall University, S. Orange, NJ 07079. E-mail
[email protected]. (1) Deleted in press. (2) Cohen Stuart, M. A.; Cosgrove, T.; Vincent, B. Adv. Colloid Interface Sci. 1986, 24, 143. (3) Takahashi, A.; Kawaguchi, M. Adv. Polym. Sci. 1982, 46, 3. (4) Fleer, G. J.; Lyklema, J. In Adsorptions from Solution at the Solid/ Liquid Interface; Parfitt, G. D., Rochester, C. H., Eds.; Academic Press: New York, 1983. (5) Napper, D. Polymeric Stabilization of Colloidal Dispersions; Academic: London, 1983. (6) Vincent, B. Adv. Colloid Interface Sci. 1974, 4, 193. (7) Yoo, D.; Shiratori, S. S.; Rubner, M. F. Macromolecules 1998, 31, 4309. (8) Mittal, K. L., Ed. Adhesion Aspects of Polymeric Coatings; Plenum: New York, 1983. (9) Yang, X.; Johnson, S.; Shi, J.; Holesinger, T.; Swanson, B. Sensors Actuators B: Chem. 1997, B45, 87. (10) Kumakura, M.; Yoshida, M.; Asano, M. J. Appl. Polym. Sci. 1990, 41, 177. (11) Freij-Larsson, C.; Jannasch, P.; Wesslen, B. Biomaterials 2000, 21, 307.
studies using probe fluids of different sizes showed that even closely packed monolayers of tris(TMS) have interstitial holes (nanopores) that can be used to modify the surface further. In this work, control over the surface coverage of tris(TMS) was used to manipulate the distribution and size of these nanopores. Such surfaces are unique for the study of polymer adsorption onto model chemically heterogeneous surfaces, providing a route to control lateral correlations of surface interactions. Experimental Section General. All reagents were used as received unless noted otherwise. Ethanol (HPLC grade) was purchased from VWR Scientific. Toluene (anhydrous), ethyldiisopropylamine (EDIPA), benzene, N,N,N′,N′-tetramethylethylenediamine (TMEDA), styrene, and sec-butyllithium were obtained from Aldrich. Benzene was dried by passing it through activated alumina to remove polar impurities and through a copper catalyst (Q-5, Engelhardt) to remove traces of oxygen.14 TMEDA was purified by vacuum distillation from calcium hydride and stored under nitrogen. Carbon dioxide (Coleman Grade) was purchased from Merriam Graves and dried in the same manner as benzene. Methanol (HPLC grade), hydrogen peroxide (30%), sulfuric acid, and sodium dichromate were purchased from Fisher. Tris(trimethylsiloxy)chlorosilane was purchased from Gelest and (4-aminobutyl)dimethylmethoxysilane was obtained from United Chemical. House-purified water (reverse osmosis) was used without further purification. Silicon wafers (4 in.) were obtained from International Wafer Service and subsequently cut into 1.5 cm × 1.5 cm squares. Synthesis of PS-COOH. Carboxylic acid end-functionalized polystyrene was synthesized by living anionic polymerization as reported by Quirk.15 All reactions were performed with Schlenk techniques. Styrene was purified by distilling first from CaH2 and then from Bu2Mg. To carry out the polymerizations, diluted (12) Fadeev, A. Y.; McCarthy, T. J. Langmuir 1999, 15, 7238. (13) Fadeev, A. Y.; McCarthy, T. J. Langmuir 1999, 15, 3759. (14) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.; Timmers, F. J. Organometallics 1996, 15, 1518. (15) Quirk, R. P.; Yin, J.; Fetters, L. J. Macromolecules 1989, 22, 85.
10.1021/la010498y CCC: $20.00 © 2001 American Chemical Society Published on Web 09/15/2001
6548
Langmuir, Vol. 17, No. 21, 2001
sec-butyllithium was added to stirred benzene, followed by the addition of styrene. The reaction was allowed to continue overnight. A 2-fold excess of TMEDA was then added to the poly(styryl)lithium solution and the solution was freeze-dried for 24 h to remove the benzene. Carbon dioxide was introduced and the flask was kept under CO2 overnight to ensure complete carboxylation of the chain ends. The carboxylate salt was redissolved in THF containing 10% concentrated aqueous HCl to yield the carboxylic acid end-functionalized polystyrene. The polymer was precipitated in methanol, recovered, and vacuum-dried at 6070 °C for several days. These polymers were stored in the dark and under constant N2 atmosphere to inhibit possible oxidation of the polymer that could lead to an increase in the number of adsorbing sites on the polymer chain. Thin-layer chromatography was carried out to ensure that the polymers adsorb to silica surfaces.16 Chemical Modification of Silicon Wafers. Silicon wafers were pretreated as described elsewhere.12,13 The samples were submerged in a mixture of H2SO4/H2O2/Na2Cr2O7 overnight, rinsed with a copious amount of water, and then dried in a clean oven at 130 °C for 90 min. The wafers were immediately placed in reaction flasks containing the organosilane of choice. Samples were placed in a custom-made wafer holder and suspended in a reaction tube containing 0.5 mL of silane. There was no direct contact between the silane and the silicon substrates. The reaction tube was placed in an oil bath and heated to 68 ( 1 °C. All silanization reactions were carried out in the vapor phase for varying reaction times. After silanization, the wafers were rinsed in the order 2 × 10 mL of toluene, 3 × 10 mL of ethanol, 2 × 10 mL of ethanol/water mixture (1:1), 2 × 10 mL of water, 2 × 10 mL of ethanol, and 2 × 10 mL of water and then dried in a clean oven at 130 °C for 10-15 min. One sample out of each batch was reserved for characterization by ellipsometry, contact angle, and X-ray photoelectron spectroscopy (XPS). Independent studies were performed to ensure all samples in the reaction tube were identical with respect to tris(TMS) coverage. The remaining wafers were kept for either adsorption or subsequent reaction with ABDMS and then adsorption. Adsorptions. Solutions for adsorption were prepared by dissolving PS-COOH in toluene at a concentration of 1 mg/mL. The solutions were filtered through a 0.45 µm Acrodisc filter just prior to adsorption. The surfaces prepared by vapor-phase reaction of tris(TMS) and/or ABDMS were placed in scintillation vials containing 5 mL of the PS-COOH solution. Adsorptions were carried out at 23 ( 1 °C for 24 h. The solutions were then decanted from the vials and the wafers were rinsed with 2 × 5 mL of toluene for 2 min. This should remove any PS that is not chemisorbed to the surface. The samples were dried under vacuum for 1 day and characterized by ellipsometry, contact angle, atomic force microscopy (AFM), and XPS. Adsorptions from cyclohexane (θ solvent) were carried out as described above with the exception that the temperature was kept at 35 ( 0.1 °C. This ensures that the polymer is truly adsorbing and not precipitating onto the surface. Dewetting Studies. One sample from each adsorption study was retained to study the dewetting of a polystyrene film from these surfaces. Dewetting studies were performed on tris(TMS) surfaces before and after adsorption of 11K PS-COOH from toluene and from cyclohexane. A 0.75% solution of 49.9K polystyrene was prepared in toluene and spin-coated onto the samples at 4000 rpm. This resulted in a 230 ( 10 Å film as determined by ellipsometry. The films were then annealed at 160 °C for 24 h, and then quenched rapidly to room temperature to freeze in the dewetted structure. The dewetted structures were examined by both optical microscopy and AFM. Characterization. Contact angle measurements were made with a Rame´-Hart telescopic goniometer equipped with a Gilmont syringe and a 24-gauge flat-tipped needle. Probe fluids used were water and n-hexadecane. Advancing (θA) and receding (θR) contact angles were recorded while the probe fluid was added to and withdrawn from the drop, respectively. All samples exhibited contact angles within ( 1° of the average value reported herein. pH-dependent contact angles were measured with buffer solu(16) Iyengar, D. R.; McCarthy, T. J. Macromolecules 1990, 23, 4344.
Stafford et al.
Figure 1. Kinetics of tris(TMS)-Cl vapor-phase reaction as monitored by advancing (b) and receding (O) water contact angles. It is assumed that an advancing CA of 108° is complete coverage. Table 1. Contact Angle Data for 11K PS-COOH Adsorption from Toluene onto Tris(TMS) Surfaces θA/θR (deg) before adsorption
after adsorption
sample
H2O
C16H34
H2O
clean Si tris(TMS) 1 h tris(TMS) 4 h tris(TMS) 24 h tris(TMS) 51 h tris(TMS) 72 h tris(TMS) 93 h
spreads 62/51 67/53 81/70 86/75 94/82 103/90
spreads 24/9 24/11 31/20 30/20 34/22 36/33
89/49 90/50 89/54 90/67 90/68 94/75 103/89
C16H34 8/0 8/0 10/0 20/7 19/10 22/10 36/32
tions prepared as described elsewhere.17 X-ray photoelectron spectra (XPS) were obtained on a Perkin-Elmer Physical Electrons 5100 with Mg KR excitation (400 W). Spectra were taken at two takeoff angles, 15° and 75° (between the plane of the surface and the entrance lens of the detector optics). The attenuation of the Si0 peak was used to determine layer thickness. Film thickness was also measured with a Rudolph Research AutoEL-II ellipsometer equipped with a helium-neon laser (λ ) 6328 Å) at an incidence angle of 70°. AFM images were obtained from a Digital Instruments Dimension 3100 scanning probe microscope operated in tapping mode.
Results and Discussion Synthesis of Covalently Attached Monolayers of Mixed Functionalities from Tris(trimethylsiloxy)chlorosilane. Vapor-phase reactions of tris(TMS) and ABDMS were chosen for both their simplicity and the ability to form more densely packed monolayers than the corresponding solution reactions.12 To create a range of surface coverages of tris(TMS), the kinetics of the vaporphase reaction was studied. The degree of coverage was determined by contact angle measurements on surfaces treated with tris(TMS) for varying times. The results of the kinetics study are shown in Figure 1. Initially there is a rapid rise in contact angle to values of 70° within 1 h, followed by a gradual increase over reaction times of several days. Consequently, the surface coverage can be controlled by simply varying reaction time. This is critical for controlling the adsorption of PS-COOH to these surfaces. The contact angle data can be analyzed in a manner proposed by Israelachvili and Gee18 for molecularly mixed heterogeneous surfaces. The observed contact angle, θobs, can be described in terms of the mole fractions of each (17) Lide, D. R., Ed. CRC Handbook of Chemistry and Physics, 76th ed.; CRC Press: Boca Raton, FL. (18) Israelachvili, J. N.; Gee, M. L. Langmuir 1989, 5, 288.
Adsorption of Polystyrene to Monolayers
Langmuir, Vol. 17, No. 21, 2001 6549
Table 2. XPS and Contact Angle Data for Tris(TMS)/ABDMS Mixed Surfaces Prepared by Consecutive Vapor-Phase Reactions XPS atomic concentration (%)a
reaction conditions (vapor phase)
C
Si
O
N
C/N
ABDMS 68 h tris(TMS) 1 h, ABDMS 68 h tris(TMS) 4 h, ABDMS 68 h tris(TMS) 24 h, ABDMS 68 h tris(TMS) 72 h, ABDMS 68 h tris(TMS) 93 h
14.6 29.3 30.4 26.3 26.1 16.7
39.0 31.8 32.5 34.2 34.3 41.1
45.2 36.8 35.1 37.9 38.5 42.2
1.24 2.09 1.97 1.52 1.13 0
11.75/1 14.04/1 15.45/1 17.32/1 23.10/1
a
ratiob tris(TMS)/ ABDMS
θA/θR (deg) H2O
0:1 1:3.9 1:2.4 1:1.6 1:0.8 1:0
73/39 73/54 77/57 85/69 94/73 103/90
Numbers are 75° takeoff angle data. bReference 38.
component, f1 and f2, as well as the contact angles for the pure surface of each component, θ1 and θ2, by
[1 + cos (θobs)]2 ) f1[1 + cos (θ1)]2 + f2[1 + cos (θ2)]2 (1) f 1 + f2 ) 1
(2)
In this study the surface was treated as a mixture of trimethylsilyl groups (TMS) (θ1 ) 108°)13 and silanols (θ2 ) 0°). It should be noted that the advancing contact angle was used for θobs in all of these studies (see Table 1). By this method, the percentage of tris(TMS) covering the surface was calculated by measuring the contact angle. The residual silanols, not blocked by tris(TMS), are sites available for adsorption or chemical reaction. Mixed surfaces containing tris(TMS) and amino groups were prepared by subsequent reaction of the tris(TMS) surfaces with (4-aminobutyl)dimethylmethoxysilane in the vapor phase. Mixed monolayers with different compositions were prepared by controlling the initial reaction time of tris(TMS). The chemical composition of the mixed monolayers was determined by XPS. The XPS data along with the contact angle results are shown in Table 2. The carbon-to-nitrogen ratio increases as the amount of tris(TMS) increases, showing that there are fewer sites available for subsequent reaction with ABDMS, as expected. From the carbon-to-nitrogen ratio of the pure ABDMS and tris(TMS) surfaces, the relative ratio of tris(TMS) to ABDMS for each of the surfaces was determined. Tris(TMS)/ABDMS mixed surfaces were also characterized by measurement of the contact angle by using water solutions having different pHs. This approach, referred to as contact angle titration, is widely used for the characterization of acid-base surface properties of polymers,19,20oxides,21 and supported monolayers.22-24 Figure 2 shows advancing contact angles for different tris(TMS)/amino mixed surfaces plotted as a function of the probe fluid pH. With decreasing pH, a transition in the contact angle from a more hydrophobic to a less hydrophobic surface is observed at pH ∼ 3.5. This transition is most evident for the pure amine surface and gradually decreases in magnitude for mixed surfaces containing less amine functionality. It is worth noting that this transition can still be detected for the pure tris(TMS) surface, which obviously does not contain any amine groups. This behavior can be explained as illustrated. (19) Holmes-Farley, S. R.; Reamey, R. H.; McCarthy, T. J.; Deutch, J.; Whitesides, G. M. Langmuir 1985, 1, 725. (20) Holmes-Farley, S. R.; Bain, C. D.; Whitesides, G. M. Langmuir 1988, 4, 921. (21) McCafferty, E.; Wightman, J. P. J. Colloid Interface Sci. 1997, 194, 344. (22) Bain, C. D.; Whitesides, G. M. Langmuir 1989, 5, 1370. (23) Creager, S. E.; Clarke, J. Langmuir 1994, 10, 3675. (24) Zhang, H. L.; Zhang, H.; Zhang, J.; Liu, Z. F.; Li, H. L. J. Colloid Interface Sci. 1999, 214, 46.
Figure 2. pH-dependent contact angles for pure tris(TMS) (b), mixed tris(TMS)/ABDMS (0), and pure ABDMS (2) surfaces.
For pH > 4, the amine groups are protonated by the residual surface silanols, exposing methylene groups from the butyl segment, and thus producing a more hydrophobic surface. Such protonation has been observed directly by a variety of spectroscopic techniques.25-27Near pH ∼ 3.5 the acidic buffer solution competes with the surface silanols (pKa ∼ 4)28 to protonate the amine. This frees the amino group to interact with the probe fluid and produces a less hydrophobic surface due to the presence of the charged species. Adsorption of Carboxylic Acid End-Functionalized Polystyrene on Mixed Surfaces. The adsorption of carboxylic acid end-functionalized polystyrene on the mixed surfaces is shown schematically in Figure 3. The adsorption of PS-COOH onto these surfaces is rapid and irreversible. Contact angle data for the tris(TMS) surfaces before and after adsorption are given in Table 1. The water contact angle for a smooth film of PS was measured to be 95°/79°. The differences observed could arise from either chemical heterogeneities on the surface, surface restructuring, or surface roughness. It should be noted that although the water contact angle remains constant for the first four samples, the hexadecane contact angle increases. This indicates that while water senses only a PS surface, hexadecane penetrates the PS layer and sees the original tris(TMS) surface. The kinetics of adsorption of 11K PS-COOH was determined on a series of surfaces having different tris(TMS) coverages. The thickness of the PS-COOH layer as a function of adsorption time is shown in Figure 4. As seen, the adsorption of PS-COOH occurs quickly, even for (25) Leyden, D. E.; Kendall, D. S.; Waddell, T. G. Anal. Chim. Acta 1981, 126, 207. (26) Culler, S. R.; Ishida, H.; Koenig, J. L. J. Colloid Interface Sci. 1985, 106, 334. (27) Mingalyov, P. G.; Fadeev, A. Y.; Staroverov, S. M.; Lisichkin, G. V.; Lunina, E. V. J. Chromatogr. 1993, 646, 267. (28) Iler, R. K. The Chemistry of Silica; John Wiley & Sons: New York, 1979.
6550
Langmuir, Vol. 17, No. 21, 2001
Stafford et al.
Figure 5. Adsorbed thicknesses, as determined by ellipsometry, of PS-COOH adsorption to tris(TMS) surfaces: 11K adsorbed from cyclohexane (2) and toluene (b); 41K adsorbed from toluene (O); and 11K adsorbed from toluene to mixed tris(TMS)/ABDMS surfaces (0).
Figure 3. Schematic of adsorption of solvated PS-COOH onto a tris(TMS)-modified surface.
Figure 4. Adsorption kinetics of 11K PS-COOH from toluene onto surfaces having increasing tris(TMS) surface coverage: 0% tris(TMS) (clean Si) (b); 55% tris(TMS) (O); 86% tris(TMS) (2); 92% tris(TMS) (4); and 97% tris(TMS) (9).
surfaces that contain high amounts of tris(TMS). In all cases the adsorption of PS-COOH saturates within 24 h. Therefore, all remaining experiments were conducted for this amount of time. Adsorbed thickness was determined by XPS and ellipsometry. The adsorption data for 11K and 41K PS-COOH to tris(TMS) surfaces from toluene, as well as 11K PS-COOH from cyclohexane (35 °C), are shown in Figure 5. As expected, the thickness of the adsorbed polymer layer decreases with increasing tris(TMS) coverage. Also, the 41K PS-COOH shows a decrease in the total adsorbed thickness. For a given number of chains anchored to the surface, an increase in the total adsorbed thickness would be expected as the molecular weight of the polymer increases. The observed decrease can be understood by a fine balance between the energetic gain of pulling the chain into solution (buoyancy effect) and the energy associated with anchoring the chain to the surface.16 Adsorption from cyclohexane (θ solvent) offers versatility in the process by allowing thicker layers to be obtained. This is due to the polymer adopting a more globular conformation, which results in a higher packing density
Figure 6. AFM images (height/phase) before (clean silicon) and after modification with tris(TMS).
on the surface in comparison to adsorption from a good solvent (toluene). The adsorption results for 11K PSCOOH to mixed tris(TMS)/ABDMS surfaces are also shown in Figure 5. To ensure that the ABDMS was not protonated by external acidic groups, the adsorption protocol was altered to include a rinse with EDIPA just prior to adsorption. Adsorption to these mixed monolayers resulted in thinner adsorbed layers, which can be attributed to the lower surface energy of the amine surface compared to the silanol surface. Finally, AFM was used to study in detail the surface topography for the 11K PS-COOH adsorbed to tris(TMS)modified surfaces. Figure 6 shows AFM images (both height and phase) for a clean silica surface and one modified with tris(TMS) prior to adsorption. In comparison to a clean substrate, the modified surface exhibits texture
Adsorption of Polystyrene to Monolayers
Langmuir, Vol. 17, No. 21, 2001 6551
Figure 7. AFM images (height/phase) of 11K PS-COOH adsorbed from toluene to tris(TMS)-modified surfaces: (A) 0% tris(TMS); (B) 52% tris(TMS); (C) 59% tris(TMS); (D) 76% tris(TMS); (E) 81% tris(TMS); and (F) 86% tris(TMS).
arising from the incomplete coverage of the surface with tris(TMS). Figure 7 shows images obtained for 11K PSCOOH adsorbed to surfaces having tris(TMS) coverages ranging from 0% (clean silicon) to 86%. Analysis of the AFM images for adsorption of 11K PSCOOH onto clean silicon (Figure 7A) shows aggregates of polystyrene having an average height of 1.2 nm and diameter of 22 nm.30Contact angle data, however, show that the entire surface is covered with polystyrene. If there is no lateral broadening due to the AFM tip shape, then the average number of chains within one of these aggregates can be calculated. By use of the bulk density (29) The refractive index for tris(TMS)-Cl could not be found and was assumed to be 1.386. This value is for tris(TMS)-H and should be an accurate assumption provided the thickness of the tris(TMS) layer is small. (30) These numbers were obtained by performing particle analysis on the height image with the software provided by Digital Instruments. This analysis yields average height, diameter, and number of particles as well as other information not listed here.
(1.05 g/cm3) of PS, the volume occupied by one PS chain (Vmol) can be written as
Vmol )
( )( ) M 1 F NA
(3)
where M is the molecular weight, F is the bulk density, and NA is Avogadro’s number. For an 11K PS chain, Vmol ) 1.7 × 104 Å3. The volume of a cylindrical disk is given by
V)
πd2 h 4
(4)
From the height and diameter data from the AFM image, the average volume of the observed clusters is 4.6 × 105 Å3. This corresponds to ∼26 polystyrene chains in each cluster. This represents an upper limit since we are assuming the AFM tip to be infinitely sharp. Similar
6552
Langmuir, Vol. 17, No. 21, 2001
Stafford et al.
Table 3. Number of Chains Per Aggregate as Calculated by Average Dimensions from AFM Image Analysis dimensions of aggregates corrected for a in AFM image tip diameter of 5 nm
clean Si tris(TMS) 1 h tris(TMS) 4 h tris(TMS) 24 h tris(TMS) 51 h tris(TMS) 93 h
D/H (nm)
no. of chains
D*/H (nm)
no. of chains
22/1.2 36/1.8 32/1.2 19/1.0 12/0.8 12/0.9
26 105 55 16 5 6
17/1.2 31/1.8 27/1.2 14/1.0 7/0.8 7/0.9
15 78 39 9 1 1
calculations were performed for the remaining samples and the results are shown in Table 3. If we assume that the diameter of the AFM tip is 5 nm, a first-order correction for the smearing of the data by the tip shape can be made since the observed image is a convolution of the real image and the tip shape. The corrected data are also shown in the last column of Table 3. The data in Table 3 strongly suggest that single PS molecules are being observed for high tris(TMS) coverages. Several studies have utilized end-grafted polymer brushes to aid in preventing the dewetting of polymer thin films. These approaches typically involve spin-casting a mixture of the functionalized and nonfunctionalized polymers onto a substrate, followed by annealing the film to allow for diffusion of the functionalized polymer to the polymer/substrate interface.31-34The resulting monolayer modifies interfacial interactions and promotes entanglements, thereby suppressing dewetting. If, on the other hand, the polymer brush is formed prior to spin-coating the nonfunctionalized film35,36(in contrast to forming the brush in situ as described above), a phenomenon termed autophobic dewetting occurs. This has been attributed to the formation of a densely packed polymer brush monolayer, and it becomes entropically unfavorable for the nongrafted film to diffuse into the brush layer. These films were shown to be highly unstable and dewetting occurred rapidly. PS films, 230 Å in thickness, were spin-coated onto surfaces modified with tris(TMS) and onto surfaces to which 11K PS-COOH was adsorbed onto tris(TMS) surfaces from either toluene or cyclohexane. The results of the dewetting studies are shown in Table 4. For the surfaces treated only with tris(TMS), dewetting was found in all cases with little change in the contact angle of the dewetted droplets as a function of coverage. It should be noted that the contact angle described here is of the static, dewetted droplet of PS on the surface as measured by AFM37 and should not be confused with the dynamic contact angles of probe fluids described earlier. For the surfaces to which PS-COOH was anchored, the dewetted (31) Henn, G.; Bucknall, D. G.; Stamm, M.; Vanhoorne, P.; Je´roˆme, R. Macromolecules 1996, 29, 4305. (32) Ferreira, P. G.; Ajdari, A.; Leibler, L. Macromolecules 1998, 32, 3994. (33) Leibler, L.; Ajdari, A.; Mourran, A.; Coulon, G.; Chatenay, D. In Ordering in Macromolecular Systems; Teramoto, A., Kobayashi, M., Norisuje, T., Eds.; Springer-Verlag: Berlin, Heidelberg, 1994. (34) Shull, K. R. Macromolecules 1996, 29, 2659. (35) Reiter, G.; Auroy, P.; Auvray, L. Macromolecules 1996, 29, 2150. (36) Shull, K. R. Faraday Discuss. 1994, 98, 203. (37) Vitt, E.; Shull, K. R. Macromolecules 1995, 28, 6349. (38) The ratio of tris(TMS)to ABDMS was calculated as follows: the C/N for pure ABDMS was adjusted to equal 6:1 to exclude contamination. All other surfaces were adjusted by the same factor. Subtracting the C/N of the pure ABDMS surface from the mixed surfaces and dividing by 9 leaves the number of tris(TMS) molecules per ABDMS molecule.
Table 4. Contact Angles of Dewetted PS on Various Tris(TMS) Surfaces as Determined by AFM Analysis coverage (%)
adsorbed thickness (Å)
R/H (µm)
θE (deg)
1.17/0.48 1.45/0.66 1.43/0.63 1.09/0.57 0.94/0.44
42 46 45 52 47
tris(TMS) 0 52 59 76 81
0 0 0 0 0
0 52 59 76 81
tris(TMS), 11K PS-COOH, toluene 19 1.88/0.24 13 2.09/0.57 12 1.88/0.68 8 0.98/0.47 5 1.21/0.56
14 28 37 48 46
0 30 42 75 80
tris(TMS), 11K PS-COOH, cyclohexane 36 2.87/0.20 24 14.5/0.90 23 4.65/0.42 10 1.15/0.48 8 0.82/0.36
7 7 10 42 45
contact angles were highly dependent on the grafting density. For tris(TMS) coverages greater than 60%, no substantial improvement in the wetting characteristics was found. However, as the thickness of the adsorbed PS-COOH layer increases, a considerable reduction in the contact angle and consequently, the interfacial energy is observed. PS does not wet the clean silicon surface covered with PS-COOH due to an autophobic dewetting process described by Leibler et al.33 and subsequently Shull.36 What is lacking in this system (PS-COOH/PS) is an enthalpic gain that could offset the entropic loss of stretching the surface-grafted chains, allowing for a net gain in total energy in the system and wetting to occur. Other polymer systems that do exhibit strong enthalpic interactions (e.g., PS/PVME) are currently being studied. Conclusions Chemically grafted (sub)monolayers of tris(TMS) were prepared by vapor-phase reaction of the corresponding monochlorosilane. By controlling reaction kinetics, a series of surfaces were prepared having increasing coverages of tris(TMS). The inherently poor packing efficiency of such a bulky silane allowed for preparation of a unique class of surfaces that can be used as templates for polymer adsorption. The adsorption behavior of a model polymer system, carboxylic acid end-functionalized polystyrene, was studied extensively as a function of tris(TMS) surface coverage. The adsorbed thickness of PS-COOH can be controlled by a number of variables such as polymer molecular weight, adsorption solvent, and surface chemistry. The topography of the adsorbed layers indicated the presence of aggregate structures on the surface, and the size scale of these aggregates diminished as tris(TMS) surface coverage increased. The wetting properties of an overlying polymer thin film were investigated on these surfaces. As the tris(TMS) coverage increases, and consequently the amount of adsorbed polymer chains decreases, the wettability of these surfaces diminishes as probed by contact angles measured by AFM. Acknowledgment. We thank the University of Massachusetts Materials Research Science and Engineering Center and the Office of Naval Research for financial support. LA010498Y
