Tunable Wetting of Nanoparticle-Decorated Polymer Films - Langmuir

Publication Date (Web): June 12, 2009. Copyright © 2009 ... Gregory M. Su , Katherine Best , Thangamani Ranganathan , Todd Emrick , and Alfred J. Cro...
0 downloads 0 Views 2MB Size
pubs.acs.org/Langmuir © 2009 American Chemical Society

Tunable Wetting of Nanoparticle-Decorated Polymer Films Marla D. McConnell, Alice W. Bassani, Shu Yang,* and Russell J. Composto* Department of Materials Science and Engineering and Laboratory for Research on the Structure of Matter, University of Pennsylvania, Philadelphia, Pennsylvania 19104 Received April 15, 2009. Revised Manuscript Received May 21, 2009 In this paper, amine-modified silica nanoparticles (NPs) with diameters (d) from 15 to 230 nm are covalently linked to poly(styrene-random-acrylic acid) (P(S-ran-AA)) films, and wettability is studied as a function of diameter and NP surface coverage. During attachment, films swell and exhibit long and short scale roughness, consisting of a ridged, honeycomb structure, ∼1 μm wide and 45-50 nm deep, which encircles nanoscale features 10-15 nm high and ∼50 nm apart. A maximum NP coverage of ∼70% was achieved for d less than or nearly equal to the nanoscale roughness induced by surface swelling. For d several times greater than this nanoscale roughness, the maximum coverage was limited by interparticle repulsion and reached only ∼30%. For NPs with diameters of 15-106 nm, the water contact angle increased from 75° to 120° as NP coverage increased from 0 to 70%. At low and high NP coverage, the Wenzel and Cassie models, respectively, accurately describe the data. However, at intermediate NP coverage, neither model is satisfactory. An increase in surface roughness alone cannot account for this discrepancy. Atomic force microscopy (AFM) studies show that the NPs partially embed into the swollen P(S-ran-AA) surface, suggesting that the aminecoated NPs are wet by the copolymer, exposing low surface energy styrene. These studies demonstrate that control over surface properties of coatings, such as wetting, can be achieved by selecting NP sizes that complement film roughness.

Introduction Surface wettability, a topic of interest for over half a century, depends on both surface composition (i.e., surface energy) and physical properties (i.e., roughness).1,2 Many studies have investigated the wettability of surfaces that exhibit superhydrophilic and superhydrophobic behavior.3-8 Modification of surfaces with nanoparticle (NP) coatings allows for precise and independent control over both the surface chemistry and roughness. For example, NP-modified surface properties can be tuned by varying the NP characteristics, that is, surface functionality, size, cover*Corresponding authors. E-mail: [email protected] (R.J.C.); [email protected] (S.Y.). (1) Wenzel, R. N. Ind. Eng. Chem. 1936, 28, 988–994. (2) Cassie, A. B. D.; Baxter, S. Trans. Faraday Soc. 1944, 40, 546–551. (3) Feng, X.; Jiang, L. Adv. Mater. 2006, 18, 3063–3078. (4) Genzer, J.; Efimenko, K. Biofouling 2006, 22, 339–360. (5) Lee, W.; Jin, M.-K.; Yoo, W.-C.; Lee, J.-K. Langmuir 2004, 20, 7665–7669. (6) Bok, H.-M.; Shin, T.-Y.; Park, S. Chem. Mater. 2008, 20, 2247–2251. (7) Onda, T.; Shibuichi, S.; Satoh, N.; Tsujii, K. Langmuir 1996, 12, 2125–2127. (8) Oner, D.; McCarthy, T. J. Langmuir 2000, 16, 7777–7782. (9) Tsai, H.-J.; Lee, Y.-L. Langmuir 2007, 23, 12687–12692. (10) Tsai, P.-S.; Yang, Y.-M.; Lee, Y.-L. Langmuir 2006, 22, 5660–5665. (11) Ming, W.; Wu, D.; van Benthem, R.; de With, G. Nano Lett. 2005, 5, 2298– 2301. (12) Cebeci, F. C.; Wu, Z.; Zhai, L.; Cohen, R. E.; Rubner, M. F. Langmuir 2006, 22, 2856–2862. (13) Musick, M. D.; Keating, C. D.; Keefe, M. H.; Natan, M. J. Chem. Mater. 1997, 9, 1499–1501. (14) Boettcher, S. W.; Strandwitz, N. C.; Schierhorn, M.; Lock, N.; Lonergan, M. C.; Stucky, G. D. Nat. Mater. 2007, 6, 592–596. (15) Musick, M. D.; Keating, C. D.; Lyon, A.; Botsko, S. L.; Pena, D. J.; Hooliway, W. D.; McEvoy, T. M.; Richardson, J. N.; Natan, M. J. Chem. Mater. 2000, 12, 2869–2881. (16) Kim, J. Y.; Osterloh, F. E. J. Am. Chem. Soc. 2006, 128, 3868–3869. (17) Tokareva, I.; Minko, S.; Fendler, J. H.; Hutter, E. J. Am. Chem. Soc. 2004, 126, 15950–15951. (18) Phely-Bobin, T. S.; Muisener, R. J.; Koberstein, J. T.; Papadimitrakopoulostrast, F. Adv. Mater. 2000, 12, 1257–1261. (19) Darling, S. B.; Yufa, N. A.; Cisse, A. L.; Bader, S. D.; Sibener, S. J. Adv. Mater. 2005, 17, 2446–2450. (20) Sun, S.; Anders, S.; Hamann, H. F.; Thiele, J.-U.; Baglin, J. E. E.; Thomson, T.; Fullerton, E. E.; Murray, C. B.; Terris, B. D. J. Am. Chem. Soc. 2002, 124, 2884–2885.

11014 DOI: 10.1021/la901331q

age, spacing, patterning, and attachment chemistry.9-12 Nanoparticle-decorated surfaces have been widely investigated for many applications requiring tunable surface properties.13-25 Nanoparticle-decorated polymer films are of interest because both the polymer and nanoparticle functionalities can be tailored to control the chemical and physical characteristics of the surface. To date, many studies have focused on nanoparticle-polymer composite surfaces that display surface wettabilities ranging from superhydrophilic to superhydrophobic. Layer-by-layer (LBL) assembly of alternating layers of polycations with silica nanoparticles has been successfully utilized to create antifogging, superhydrophilic surfaces.12 LBL techniques have also been adapted to form superhydrophobic surfaces by creating porous nanoparticle/ polymer films.26-28 Nanoparticle-polymer blend films have been shown to exhibit both superhydrophilic and superhydrophobic behavior that depends on nanoparticle loading.29-31 Additionally, an elastomeric polymer with a nanoparticle coating has been utilized to create surfaces with mechanically tunable wetting.32 We have previously demonstrated a new synthetic strategy to attach 15 nm diameter, amine-modified silica nanoparticles (21) Baron, R.; Huang, C.-H.; Bassani, D. M.; Onopriyenko, A.; Zayats, M.; Willner, I. Angew. Chem. 2005, 117, 4078–4083. (22) Lee, S. W.; Drwiega, J.; Wu, C. W.; Mazyck, D.; Sigmund, W. M. Chem. Mater. 2004, 16, 1160–1164. (23) Zhong, Z.; Lin, J.; Teh, S.-P.; Teo, J.; Dautzenberg, F. M. Adv. Funct. Mater. 2007, 17, 1402–1408. (24) Walter, N.; Selhuber, C.; Kessler, H.; Spatz, J. Nano Lett. 2006, 6, 398–402. (25) Blummel, J.; Perschmann, N.; Aydin, D.; Drinjakovic, J.; Surrey, T.; Lopez-Garcia, M.; Kessler, H.; Spatz, J. P. Biomaterials 2007, 28, 4739–4747. (26) Zhai, L.; Cebeci, F. C.; Cohen, R. E.; Rubner, M. F. Nano Lett. 2004, 4, 1349–1353. (27) Han, J. T.; Kim, S.; Karim, A. Langmuir 2007, 23, 2608–2614. (28) Han, J. T.; Zheng, Y.; Cho, J. H.; Xu, X. R.; Cho, K. J. J. Phys. Chem. B 2005, 109, 20773–20778. (29) Manoudis, P. N.; Karapanagiotis, I.; Tsakalof, A.; Zuburtikudis, I.; Panayiotou, C. Langmuir 2008, 24, 11225–11232. (30) Hou, W.; Wang, Q. J. Colloid Interface Sci. 2007, 316, 206–209. (31) Wu, Z.; Han, H.; Han, W.; Kim, B.; Ahn, K. H.; Lee, K. Langmuir 2007, 23, 7799–7803. (32) Lin, P.; Yang, S. Soft Matter 2009, 5, 1011–1018.

Published on Web 06/12/2009

Langmuir 2009, 25(18), 11014–11020

McConnell et al.

Article

covalently onto surface acrylic acid moieties in poly(styrenerandom-acrylic acid), P(S-ran-AA), films.33 By varying the mole fraction of acrylic acid, the nanoparticle concentration in suspension, and the reaction time, the coverage and spacing of particles can be controlled reproducibly up to ∼70% coverage, in contrast to the 30% coverage obtained on a flat silicon surface.33,34 We attributed the high particle coverage on P(Sran-AA) films to increased surface roughness resulting from swelling of AA-rich domains. These observations motivate the present studies, which are designed to probe the wetting behaviors of these composite surfaces, as well as to determine if there is a critical particle size at which the swollen surface is unable to facilitate particle crowding, and whether this behavior is correlated to the nanoscale feature size. Herein, we report the effects of particle diameter and coverage on the wettability of nanoparticle-decorated polymer surfaces. For APTES-coated NPs on P(S-ran-AA) that achieve 70% coverage (15, 50, and 106 nm), the static contact angle increases from ∼75° to ∼117° with increasing coverage, in sharp contrast to the ∼42° value observed for a planar APTES surface. Second, P(S-ran-AA) surfaces decorated with 15, 50, and 106 nm diameter NPs exhibit identical maximum contact angles, ∼117°, at 70% coverage. In order to understand this wetting behavior, we compared the experimental results with the Wenzel and Cassie models. The contact angle (CA) on the polymer surfaces with 70% coverage of NPs (CA=117°) can be described with Cassie wetting behavior; however, low coverages of small particles, which result in CA < 90°, show Wenzel behavior. For intermediate particle coverages (90° < CA < 117°), the contact angles are described by a transitional state between the Wenzel and Cassie wetting regimes. This transitional wetting behavior is also exhibited by 230 nm diameter NPs, which achieve a maximum coverage of 30%. Atomic force microscopy (AFM) studies show that NPs sink into the swollen P(S-ran-AA) and suggest that a layer of polymer wets the NP surface. Because acrylic acid is attracted to the amine-rich NP surface, this wetting exposes the hydrophobic styrene segments to the environment, resulting in the observed increase in contact angle. These studies demonstrate that control over surface properties of coatings, such as wetting, is achieved by selecting NP sizes that complement film roughness. These nanoparticle-decorated polymer surfaces provide a unique platform for applications requiring controlled wettability, including antifouling surfaces.

Experimental Section Materials. Dimethyl sulfoxide (DMSO) (anhydrous, 99.9+%), ethanol (200 proof), (3-aminopropyl)triethoxysilane (APTES) (99%), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC) (g98.0%), N-hydroxysuccinimide (NHS) (98%), and tetraethyl orthosilicate (TEOS) (98%) were purchased from Aldrich and used as received. Toluene (HPLC grade) and water (DIUF) were purchased from Fisher Scientific and used as received. Nanoparticle Synthesis. Silica nanoparticles were synthesized according to the St€ ober method.35 For example, 230 ( 19 nm particles were synthesized by adding 5 mL of ethanol solution containing 2.4 M TEOS to a 35 mL ethanol solution of water and ammonia. The 40 mL mixture containing 0.3 M TEOS, 2.3 M H2O, and 1.0 M NH3 was stirred at 30 °C for 5 h. The 138 ( 17 nm (33) McConnell, M. D.; Yang, S.; Composto, R. J. Macromolecules 2009, 42, 517–523. (34) Grabar, K. C.; Smith, P. C.; Musick, M. D.; Davis, J. A.; Walter, D. G.; Jackson, M. A.; Guthrie, A. P.; Natan, M. J. J. Am. Chem. Soc. 1996, 118, 1148– 1153. (35) Stober, W.; Fink, A.; Bohn, E. J. Colloid Interface Sci. 1968, 26, 62–69.

Langmuir 2009, 25(18), 11014–11020

particles were synthesized similarly. Particle diameters were characterized by scanning electron microscopy (SEM), and reported diameters are the average of 100 particles measured with ImageJ. Nanoparticle Modification. Silica nanoparticles, 15 ( 3.5, 50 ( 10.2, and 106 ( 7.3 nm in diameter (Nissan Chemical, 30 wt % in isopropyl alcohol) and 138 (17 and 230 (19 nm (synthesized as described above), were pelletized by centrifugation at 11 000 rpm for 3 h. The solvent was then exchanged to anhydrous DMSO, and the suspension was sonicated for at least 15 min to redisperse the particles. The pelletization/solvent exchange procedure was repeated two more times. The particles were then amine functionalized via reaction with APTES. An excess of 5% (v/v) solution of APTES was added to the nanoparticle suspension in anhydrous DMSO and allowed to react for 8 h. The nanoparticles were then purified via centrifugation using the same procedure as above and solvent exchanged into ethanol. Substrate Preparation. Silicon substrates were immersed in piranha solution (3:7 v:v H2O2/H2SO4) at 80 °C for 20 min. The substrates were subsequently washed with DIUF water, dried in nitrogen, and exposed to UV-ozone (UVO-Cleaner model 42, Jelight Company Inc.) for 10 min to form a uniform oxide layer. A 50:50 mol/mol P(S-ran-tBA) polymer was previously synthesized using reversible addition - fragmentation chain transfer (RAFT) polymerization.33 To prepare P(S-ran-AA) films on silicon, polymer solutions (toluene, 5 wt %) were spin-coated (2000 rpm, 60 s) onto cleaned silicon substrates to form 200 nm thick films, as determined by ellipsometry. The polymer films were dried overnight in a fume hood and then annealed at 185 °C for 15 h under argon. These annealing conditions resulted in complete deprotection of the t-butyl groups, as characterized using attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy (Nicolet Nexus 470 with MCT-B detector, Harrick GATR). The spectra of the as-cast P(S-ran-tBA) films show the characteristic bending and stretching modes associated with t-butyl acrylate and styrene. The absorption bands originating from poly(t-butyl acrylate) are located at 1730 cm-1 (νCdO, ester), 1394/1368 cm-1 (νCH3), 1277/1258 cm-1 (νC-O-O), and 1160 cm-1 (νC-O). Those bands originating from polystyrene are located between 1600 and 1400 cm-1 (νCdC, aromatic). After annealing the P(S-ran-tBA) films at 185 °C for 15 h, the peaks associated with the t-butyl protecting group disappeared completely, and the carbonyl stretching mode underwent a blue shift of about 30 cm-1 and broadened. These changes in the carbonyl peak are consistent with deprotection of the t-butyl group to form acrylic acid.33

Coupling Amine-Modified Nanoparticles to P(S-ranAA). The nanoparticle attachment chemistry to P(S-ran-AA)

films was previously described.33 There are two steps in the covalent attachment of amine-modified silica nanoparticles to the acid groups in the P(S-ran-AA). In the first step, the P(S-ranAA) films were immersed in solution of EDC (0.1 M) and NHS (0.2 M) in DIUF water for 1 h, in order to activate the acrylic acid groups with the NHS. The FTIR spectrum of NHS-activated acrylic acid in the P(S-ran-AA) has a strong band at 1740 cm-1 which is attributed to the succinimidyl carbonyl (νCdO). In the second step, the activated substrates were immersed into suspensions of amine-modified nanoparticles (15, 50, 106, and 230 nm) in ethanol for varying times (0-540 min). The substrates were immersed upside-down in the nanoparticle suspensions to prevent clusters of particles from falling onto the substrates. This geometry minimized the number of particles that nonspecifically adhered to the substrates, and it is responsible, in part, for the reproducibility of particle coverage over macroscopic (mm2) areas, as well as uniform particle spacing. Once the substrates were removed from the nanoparticle suspensions, they were immediately swirled in ethanol for 5 min and then washed vigorously with DIUF water and dried in stream of nitrogen. Covalent attachment of the nanoparticles by amide bond formation was confirmed by ATR-FTIR of the P(S-ran-AA)-nanoparticle substrate, where the band at 1653 cm-1 is attributed to the DOI: 10.1021/la901331q

11015

Article

McConnell et al.

amide carbonyl (νCdO). Each set of conditions (nanoparticle diameter, immersion time, and nanoparticle concentration) was repeated to generate two sample sets on different days, and the reported data were obtained from both sets.

Characterization of Nanoparticle Decorated Surfaces. Samples prepared by the covalent coupling of amine-modified nanoparticles to the surface of P(S-ran-AA) polymer films as a function of time as well as nanoparticles deposited by spin coating onto polystyrene films were characterized with scanning electron microscopy (SEM, FEI Strata DB235). Percent particle coverage was measured from 10 images of nonoverlapping regions on each individual sample (20 images for each set of conditions) using ImageJ. Error bars in the percent coverage data represent one standard deviation. Characterization of Polymer Substrate Surfaces. The surface morphology evolution of P(S-ran-AA) was monitored using atomic force microscopy (AFM, PicoPlus Molecular Imaging, Agilent Technologies). The tips had nominal spring constants of 48 N/m and diameters < 10 nm. In situ imaging was performed in a liquid cell in ethanol. Prior to imaging, the polymer surface was exposed to pH 8 phosphate buffer for 2 h, followed immediately by DIUF water (to simulate the EDC/NHS coupling step) for 1 h. Images were taken in tapping mode, with scan speeds of 0.65-0.8 Hz. Tip resonant frequencies were 190 and 75 kHz in air and solution, respectively. The surfaces of the polymer/nanoparticle composites were also imaged using AFM in air under ambient conditions, and r values (where r is the ratio of the real surface area to the projected surface area) were determined using Gwyddion 2.10 software. Contact Angle Measurements. The wettabilities of the nanoparticle-decorated polymer film surfaces were characterized by measuring the static, advancing, and receding contact angles over five unique spots using a Rame-Hart goniometer model 200. DIUF water (5 μL drops) was used in the static contact angle measurement. DROPimage software was used to control the volume of the water droplet when measuring advancing and receding contact angles and was also used to determine the resulting contact angles.

Results and Discussion 1. Effect of Nanoscale Roughness on Maximum NP Coverage. In previous studies, we found that the coverage of NPs on a polymer surface can be controlled by varying the reaction time.33 The kinetics of covalent attachment of nanoparticles to a flat, unroughened surface follows a diffusion-limited, t1/2 growth law at early time points, while at longer times interparticle repulsions and spatial limitations result in a plateau in coverage around 30-35%. For 15 nm diameter NPs reacting with a swollen P(S-ran-AA) surface, the kinetics of attachment follow t1/2 behavior, although the maximum achievable coverage was significantly larger (∼70%) than predicted by the model. The significantly higher plateau in particle coverage is attributed to swelling and subsequent roughening of the polymer film during the reaction. The appropriate length scale of these surface features, relative to the NP size, is essential for attaining a high surface coverage of NPs. Thus, for a fixed nanoscale roughness, a transition from high to low maximum coverage is expected as particle size increases. Additionally, a similar transition from high to low maximum coverage is expected as particle size decreases. The equilibrium coverage for nanoparticles (d = 15, 50, 106, 138, and 230 nm) coupled to P(S-ran-AA) was followed by allowing the polymer films to react with 0.005-1 wt % nanoparticle suspensions for between 30 s and 540 min. The percent coverage of particles on the P(S-ran-AA) films was determined by quantitative analysis of SEM images. SEM images of the nanoparticle-decorated polymer films can be seen in Figure 1, where 11016 DOI: 10.1021/la901331q

Figure 1. SEM images of equilibrium nanoparticle coverage for 15, 50, 106, 138, and 230 nm diameter particles. Scale bars are 500 nm for the images of the 15 and 50 nm particles and 1 μm for the 106, 138, and 230 nm particles.

the particle diameters are 15, 50, 106, 138, and 230 nm. The equilibrium particle coverage was 71.1 ( 3%, 70.5 ( 4%, and 68.9 ( 3% for particle diameters of 15, 50, and 106 nm, respectively. In contrast, the equilibrium NP coverage was only 32.5 ( 4% and 31.2 ( 3% for the 138 and 230 nm particles, respectively. In all cases, the particles were uniformly dispersed across the P(S-ran-AA) surface. In situ AFM was used to interrogate the morphological changes on the surface of the P(S-ran-AA) films under the same conditions used in the nanoparticle coupling reaction. These experiments were performed to determine the equilibrium surface morphology and roughness encountered by the particles during attachment. Figure 2A shows the equilibrium morphology of the P(S-ran-AA) film. The surfaces were also imaged at each step in the coupling process, and their morphologies are described below. First, the random copolymer films were imaged in air, under ambient conditions, where the dry film exhibited a featureless surface with a rms roughness around 9 A˚. Next, a pH 8 phosphate buffer solution was introduced into the liquid cell sample holder. Within 30 min, protrusions about 2.5-3 nm in height (52 ( 10 nm center-to-center spacing) began to emerge from the surface of the polymer films. These acrylic-acid-rich protrusions remained the same size for the remainder of the 2 h buffer treatment. The polymer surface was then imaged in DIUF water for 1 h, simulating the conditions for activating the acrylic acid groups with EDC/NHS. No significant change in the number or dimensions of the protrusions was observed. Finally, the surfaces were imaged in ethanol, Figure 2A, which is the solvent used in the nanoparticle coupling reaction. During ethanol exposure, the protrusions swelled and spread, resulting in an increase in the roughness of the P(S-ran-AA) film. As represented by Figure 2A, two length scales of roughness are observed: one due to the acrylic-acid-rich protrusions described earlier (∼10-15 nm in height), and the other due to a continuous honeycomb morphology (∼45-50 nm high and 1 μm wide) resulting from swelling the films in ethanol. Based on SEM (e.g., Figure 1) and surface roughness (e.g., Figure 2A) analysis, the interplay between nanoparticle diameter and the feature size of the swollen polymer surface can be understood. Figure 2B is a schematic showing how the length scales resulting from polymer swelling affect the maximum coverage that can be achieved for a particular particle size. The cartoon displays a 1 μm region across the swollen P(S-ran-AA) film (purple), 230 nm particles at 30% coverage (top center), and 50 nm particles at 70% coverage (bottom center), and the approximate Langmuir 2009, 25(18), 11014–11020

McConnell et al.

Article

Figure 2. AFM image and schematic representation of the interplay between polymer film swelling and maximum particle coverage. (A) In situ AFM image of P(S-ran-AA) after 2 h in pH 8 buffer, 1 h in water, and 5 h in ethanol (15 μm  15 μm, Δz = 100 nm). (B) Schematic representation of the influence of roughness on the maximum particle coverage. For particles smaller than or equal to 106 nm, the nanoscale roughness of the polymer reduces interparticle repulsions and results in a particle coverage of 70%.

roughness these particles encounter from the underlying polymer. If both the lateral and vertical features of the swollen surface (the 10-15 nm acrylic-acid-rich protrusions) are on the order of the particle size, then the maximum surface coverage of those NPs will reach 70% as shown by the lower right cartoon. However, for NPs much larger (>10) than these surface features, interparticle repulsions will not be shielded by the height variations in the polymer film. Thus, the maximum achievable NP coverage of large particles is 30%. 2. Wetting Properties of Nanoparticle Decorated Films. Nanoparticle-decorated polymer surfaces present a unique platform for studying how nanoscale roughness affects wettability. Although wetting behavior on roughened surfaces has been widely studied, the effect of nanoscale roughness on the contact angle of water and the ability of water to penetrate between nanoscale features (i.e., Wenzel and Cassie behavior) is not well understood. Because they offer precise control over the nanoroughness, nanoparticle-decorated polymer surfaces serve as a platform to address these fundamental questions systematically. Water contact angles were measured on surfaces containing “small” particles (15, 50, 106 nm) with coverages up to ∼70% and “large” particles (230 nm) with coverages up to ∼30%. The static contact angle data, as seen in Figure 3A, reveal two trends. First, for all particle sizes, the contact angle increases with increasing particle coverage, ranging from 75° to 117° for the 15 and 50 nm particles, 95° to 117° for the 106 nm particles, and 90° to 100° for the 230 nm particles. This trend was unexpected, because APTES, the surface modifier on the nanoparticles, is hydrophilic. Planar monolayers of APTES exhibit a contact angle of 42°, which is much lower than the contact angles of the nanoparticle-decorated surfaces. The second trend is that all the “small” particle surfaces exhibit the same maximum contact angle of ∼117° at 70% coverage. For these “small” particles, surface coverage and not particle diameter determines wetting behavior. The wettability of a surface depends on the surface topography and the Young’s contact angle on a flat substrate, θ0, as described by the Wenzel1 and Cassie2 models. In the Wenzel model, it is assumed that the liquid droplet completely wets the surface, without any trapped air in the grooves of the surface (c.f., Figure 7B, top). The Wenzel contact angle, θW, is described as cos θW ¼ r cos θ0

ð1Þ

where r, the surface roughness, is defined as the ratio of the actual surface area to the geometrically projected surface area. Alternatively, the Cassie model assumes that air is trapped between Langmuir 2009, 25(18), 11014–11020

surface features (e.g., nanoparticles) and the water droplet sits on top of the features at the solid-air interface (c.f., Figure 7B, bottom). The Cassie contact angle, θC, is defined as cos θC ¼ f ðcos θ0 þ 1Þ -1

ð2Þ

where f is the fraction of solid area on the mixed solid-air surface. To understand why contact angle increases with particle coverage, the Wenzel and Cassie models were compared with the experimental results, Figure 3A. To determine the surface roughness, r, in the Wenzel model, AFM was used to characterize the surface morphology of dried, NP-decorated polymer films, Figure 4. Figure 4A and B shows the surface of the P(S-ran-AA) film decorated with 106 nm particles at 18% and 70% coverage, respectively. In Figure 4C and D, 230 nm particles cover 10% and 32% of the P(S-ran-AA) surface, respectively. In particle-free areas of Figure 4C, wrinkles (25-30 nm high) are observed in the dried film that result from swelling of the P(S-ran-AA) during particle attachment. To test the Wenzel model, the static contact angles have been plotted versus the r values determined from AFM, Figure 3B. The r values range from 1.01, for the lowest coverage of 15 nm particles, to 1.51, for the highest coverage of 230 nm particles. The swollen P(S-ran-AA) without nanoparticles has an r value of 1.01, measured with in situ AFM, Figure 2A. Plotted in Figure 3B (solid line) is the Wenzel model calculation for a surface with a Young’s contact angle of 91°, the measured contact angle for PS. In order to determine the maximum contact angle for our experimental roughnesses calculable from the Wenzel model, the particles and the bare surface are assumed to be completely covered by PS. Because the maximum r value for the composite surfaces was 1.51, the maximum possible contact angle predicted by the Wenzel model is only 91.5°. For surfaces with low coverages of 15 or 50 nm particles, the contact angles are less than the values predicted by the Wenzel model using a Young’s contact angle value of 91°. This is an expected result, because much of the P(S-ran-AA) surface (initial contact angle ∼76°) is exposed at low particle coverages. However, as the 15 and 50 nm particle coverages approach 70%, the contact angle values are greater than those predicted by the Wenzel model. Conversely, all the contact angle data for surfaces containing 106 and 230 nm diameter particles are near or greater than the values predicted by the Wenzel model. The contact angle hysteresis, that is, the difference between the advancing and receding contact angle, decreases for all particle diameters as the particle coverage increases, Figure 3C. At the lowest particle coverage, the hysteresis is ∼30° for all particle sizes. At the highest particle coverage, the hysteresis decreases to ∼20°, suggesting Cassie wetting DOI: 10.1021/la901331q

11017

Article

McConnell et al.

Figure 4. AFM images of 106 nm particles with (A) 18% and (B) 70% coverage and 230 nm diameter particles with (C) 10% and (D) 32% coverage. All images are 5 μm  5 μm; Δz = 200 nm for (A) and (B); Δz = 500 nm for (C) and (D).

be calculated using eq 3, f ¼

Figure 3. (A) Static water contact angle versus percent particle coverage for (black square) 15 nm, (red circle) 50 nm, (green triangle) 106 nm, and (blue inverted triangle) 230 nm diameter particles on P(S-ran-AA) films. Water contact angles increase with increasing NP coverage for all NP sizes. (;) Cassie contact angle of a model surface composed of polystyrene coated NPs and air. (B) Static water contact angle versus r for (black square) 15 nm, (red circle) 50 nm, (green triangle) 106 nm, and (blue inverted triangle) 230 nm diameter particles on P(S-ran-AA) films. (;) Wenzel contact angle of a model surface composed of polystyrene coated NPs. (C) Advancing, static, and receding contact angles for low (L), medium (M), and high (H) NP coverages of 15, 50, and 106 nm diameter particles, as well as low and medium NP coverages of 230 nm diameter particles.

behavior on these surfaces. To test the Cassie model, we calculated the Cassie contact angle of a model surface composed of polystyrene and air (solid line in Figure 3A). This calculation assumes that the particles are covered completely with polystyrene. The experimental contact angles (117°) for the highest coverage of the 15, 50, and 106 nm particles are slightly greater than the predicted values. A possible explanation is presented later in the discussion. Using eq 2, it was possible to calculate the f value, using a Cassie contact angle of 117° and a Young’s contact angle of 91°. The calculated f value is 0.56. Accordingly, it is possible to calculate the wetting level of the water droplet down the sides of the nanoparticles, or the azimuthal angle φ (see Figure 7B). According to Lin and Yang,32 the azimuthal angle can 11018 DOI: 10.1021/la901331q

2πR2 ð1 - cos φÞN 2πR2 ð1 - cos φÞN þ ð1 - πðR sin φÞ2 NÞ

ð3Þ

where R is the radius of the nanoparticle and N is the number density of particles. Using the average radius (R = 7.5, 25, and 53 nm) and number density of nanoparticles (N7.5 = 3.96  1015, N25 = 3.57  1014, N53 = 7.94  1013 particles/m2) calculated from SEM images and an f value of 0.56, the azimuthal angle was calculated to be 57° for all particle diameters. Because all three particle diameters (15, 50, 106 nm) have the same azimuthal angle (57°) at 70% coverage, the resulting measured contact angles on all three surfaces should be the same and are shown to agree in Figure 3A. Because neither the Wenzel nor Cassie model fully captured the contact angle variation with particle coverage, additional experiments were performed to investigate the contributions of the total NP-decorated surface roughness (i.e., the contributions from the underlying film and the particles) and chemical composition (i.e., from the exposed film regions and the nanoparticle surfaces) to the measured contact angles. In the first experiment, solutions of APTES-modified, 15 nm diameter nanoparticles in ethanol (0.005-2 wt %) were spun cast onto flat polystyrene films. The surfaces were characterized with SEM to determine the particle coverage (Figure 5A). The contact angles were measured for each of the surfaces and plotted versus particle coverage (Figure 5B), where 0% coverage represents a polystyrene film with no particles and 100% coverage represents a polystyrene film completely covered by a multilayer of APTESmodified particles. The increased particle coverage on the polystyrene film led to a simultaneous increase in surface roughness and hydrophilicity; correspondingly, the contact angle decreased from 91° to 30° as coverage increased from 0% to 100%, respectively, as shown in Figure 5B. These contact angles are in very good agreement with the calculated values from the Wenzel model (solid line in Figure 5B) using r values determined with AFM. Because these results are in direct contrast to those obtained for the nanoparticle-decorated P(S-ran-AA) surfaces, it can be concluded that the surfaces of the nanoparticles become hydrophobic due to the PS wetting layer during attachment to the surface of the P(S-ran-AA) films. Langmuir 2009, 25(18), 11014–11020

McConnell et al.

Article

Figure 5. (A) SEM image of APTES-modified NPs spin coated onto polystyrene film with 2.9% coverage. (B) Static contact angle versus percent particle coverage for (9) 15 nm diameter APTES particles on polystyrene films. (;) Calculated Wenzel contact angles using r values determined with AFM for amine-modified particles on a polystyrene surface.

Figure 6. Percent of nanoparticles embedded into the P(S-ran-AA) film as a function of time for (black square) 15 nm, (red circle) 50 nm, (green triangle) 106 nm, and (blue inverted triangle) 230 nm diameter particles. A schematic is shown defining the initial h and final h0 heights of the particles.

To investigate the influence of nanoparticle surface composition on wetting, a thin film of polystyrene (30 nm) was deposited onto a P(S-ran-AA) surface with 70% coverage of 106 nm particles. This sample was then annealed to ensure that the PS film achieved a conformal coating over the particles. The contact angle of the annealed sample, 109°, was in excellent agreement with the contact angle calculated using the Cassie model, 108.2°, assuming a Young’s contact angle of 91° (PS) and a liquid-solid contact area of 70%. This contact angle is slightly less than that obtained for the nanoparticle-decorated P(S-ran-AA) sample before deposition of the PS film (116°). AFM was used to measure the r value of the NP-decorated polymer film (r = 1.22), which was compared to the r value obtained after the PS film was added and annealed (r = 1.14). This decrease in r value likely occurs because annealing smooths the honeycomb features of the underlying P(S-ran-AA) film and therefore the contour of the overlaid particle surface. This decrease in roughness results in the observed decrease in contact angle. These two studies indicate that the increase in contact angle with increasing amine-modified particle coverage is likely due to the wetting of polymer chains on the surface of the nanoparticles, exposing polystyrene to the air. This Langmuir 2009, 25(18), 11014–11020

thin layer of polystyrene is likely responsible for the unexpected trend in increasing contact angle with increasing particle coverage for the particles on the P(S-ran-AA). Additional evidence for the mobility of polymer chains at the surface of the film came from AFM measurements of the relative height, (1 - h/h0)  100, of APTES-modified nanoparticles on P (S-ran-AA) films (200 nm thick) as a function of reaction time. After 5 h, the smaller particles (15, 50, and 106 nm) sank by 4045%, indicating that the particles had partially penetrated into P (S-ran-AA). In contrast, the 230 nm particles only sank by 15% after 5 h. The particle height change as a function of time is plotted in Figure 6. The kinetics of sinking is similar for particles with diameter from 15 to 110 nm. Further studies are needed to determine if particles completely embed into P(S-ran-AA). According to Deshmukh and Composto,36 a thin wetting layer of about two to three monomers of polystyrene (independent of molecular weight) rapidly formed on the surface of a metal nanoparticle on top of a polystyrene film upon annealing above the glass transition temperature. Although they were not (36) Deshmukh, R. D.; Composto, R. J. Langmuir 2007, 23, 13169–13173.

DOI: 10.1021/la901331q

11019

Article

McConnell et al.

Figure 7. Schematic representations of (A) the maximum coverage of NPs, partial sinking of a NP into the swollen P(S-ran-AA), and the wetting of P(S-ran-AA) chains on the surface of the NPs during attachment and (B) the interface between water and NPs for Wenzel (top) and Cassie (bottom) wetting, as well as intermediate wetting. A representation of the angle φ, from eq 4, is shown in the Cassie wetting schematic.

annealed, our P(S-ran-AA) films underwent significant swelling, as observed by in situ AFM, and were plasticized by the reaction solvent. Because the acrylic acid groups in P(S-ran-AA) are attracted toward the amine groups on the APTES-modified NPs, the acrylic acid repeat units (red) will tend to wet the particles, whereas the hydrophobic styrene units (blue) will be exposed to the air, as illustrated in Figure 7A. This proposed mechanism of polymer wetting is consistent with the large contact angles observed for the nanoparticle-decorated P(S-ran-AA) films presented in Figure 3A. The contact angle measurements (Figure 3A) along with control experiments (Figures 5 and 6) indicated that the wetting of nanoparticle-decorated polymer surfaces depends on both the coverage of nanoparticles and the nanoparticle surface chemistry. For 15 and 50 nm nanoparticle-decorated polymer surfaces with coverage less than 50%, contact angles smaller than 90° were observed, consistent with Wenzel wetting behavior (Figure 7B, top). Longer reaction times for these particles resulted in both the saturation coverage of ∼70% and copolymer wetting of the particle surfaces. This wetting exposed styrene monomers to the environment (air), resulting in a Wenzel to Cassie wetting transition. For surfaces with 10-60% coverage of 106 and 230 nm particles, an intermediate wetting regime (90-117°) was observed, which was consistent with the partial filling of water into the grooves between nanoparticles, as illustrated in Figure 7B.

Conclusion We have studied the effect of surface roughness and nanoparticle diameter on the maximum achievable surface coverage of

11020 DOI: 10.1021/la901331q

amine-modified nanoparticles covalently linked to P(S-ran-AA) films, and their combined effects on contact angle. Specifically, the maximum coverage of particles on the polymer films, ∼70%, was achieved when the particle size was less than 10 times the height of the nanoscale surface features (i.e., 15-106 nm in diameter). Particles with diameters between 138 and 230 nm exhibited a plateau in coverage at ∼30%, because surface roughness was too small to reduce interparticle repulsion. The nanoparticle-decorated surfaces exhibited contact angles that depended on both particle coverage and particle diameter. Surfaces with contact angles less than 90° were achieved using particles with diameters less than 100 nm at coverages less than 50%, and showed Wenzel wetting. Surfaces with intermediate contact angles (between 90° and