Control of Surface Properties Using Fluorinated Polymer Brushes

New York 14853, Departments of Materials and Chemical Engineering, University of California, Santa Barbara, California ..... Marion K. Brinks , Ar...
0 downloads 0 Views 299KB Size
Download PDF
10498

Langmuir 2004, 20, 10498-10506

Control of Surface Properties Using Fluorinated Polymer Brushes Produced by Surface-Initiated Controlled Radical Polymerization Luisa Andruzzi,† Alexander Hexemer,‡ Xuefa Li,† Christopher K. Ober,*,† Edward J. Kramer,‡,§ Giancarlo Galli,| Emo Chiellini,| and Daniel A. Fischer⊥ Department of Materials Science and Engineering, Cornell University, Ithaca, New York 14853, Departments of Materials and Chemical Engineering, University of California, Santa Barbara, California 93106, Dipartimento di Chimica e Chimica Industriale, Universita` di Pisa, 56126 Pisa, Italy, and Materials Science and Engineering Laboratory, National Institute for Standards and Technology, Gaithersburg, Maryland 20899 Received March 21, 2004. In Final Form: September 5, 2004 Surface-grafted styrene-based homopolymer and diblock copolymer brushes bearing semifluorinated alkyl side groups were synthesized by nitroxide-mediated controlled radical polymerization on planar silicon oxide surfaces. The polymer brushes were characterized by X-ray photoelectron spectroscopy (XPS), near-edge X-ray absorption fine structure (NEXAFS), and time-dependent water contact angle measurements. Angle-resolved XPS studies and water contact angle measurements showed that, in the case of the diblock copolymer brushes, the second block to be added was always exposed at the polymer-air interface regardless of its surface energy. Values of z*/Rg were estimated based on the radius of gyration, Rg, of the grafted homopolymer or block copolymer chains for the grafted brushes and thickness of the brush, z*. The fact that z*/Rg > 1 suggests that all these brushes are stretched. These results support the idea that after grafting the first block onto the surface the nitroxide-end capped polymer chains were able to polymerize the second block in a “living” fashion and the stretched brush so formed was dense enough that the outermost block in all cases completely covers the surface. NEXAFS analysis showed a relationship between the surface orientation of the fluorinated side chains and brush thickness with thicker brushes having more oriented side chains. Time-dependent water contact angle measurements revealed that the orientation of the side chains of the brush improved the surface stability toward reconstruction upon prolonged exposure to water.

Introduction The control of surface properties represents a fundamental issue in numerous applications ranging from coating technology and biotechnology to microelectronics.1-4 Control can, for example, be achieved using “polymer brushes” applied as polymer films onto a substrate. Polymer brushes are classically prepared from block copolymer solutions where one block serves to anchor the polymer to a substrate while the functional block (“brush”) forms the surface.5 However, a drawback of this strategy is that desorption of the polymer brush may occur. Therefore, an alternative approach involves the use of surface-grown polymers, covalently attached to a solid substrate. Several methods have been used to grow polymer brushes from a surface, including radical,6-8 †

Cornell University. Department of Materials, University of California, Santa Barbara. § Department of Chemical Engineering, University of California, Santa Barbara. | Universita ` di Pisa. ⊥ National Institute for Standards and Technology. ‡

(1) Swalen, J. D.; Allara, D. L.; Andrade, J. D.; Chandross, E. A.; Garoff, S.; Israelachvili, J.; McCarthy, T. J.; Murray, R.; Pease, R. F.; Rabolt, J. F.; Wynne, K. J.; Yu, H. Langmuir 1987, 3, 932. (2) Halperin, S. T.; Tirrell, M.; Lodge, T. P. Adv. Polym. Sci. 1992, 100, 31. (3) Milner, S. T. Science 1991, 251, 905. (4) Wang, J.; Mao, G.; Ober, C. K.; Kramer, E. J. Macromolecules 1997, 30, 1906. (5) Dan, N.; Tirrell, M. Macromolecules 1993, 26, 4310. (6) Prucker, O.; Ru¨he, J. Macromolecules 1998, 31, 592. (7) Prucker, O.; Ru¨he, J. Macromolecules 1998, 31, 602. (8) De Boer, B.; Simon, H. K.; Werts, M. P. L.; Van der Vegte, E. W.; Hadziioannou, G. Macromolecules 2000, 33, 349.

anionic,9 cationic10,11 and controlled/living radical polymerization.12-15 Among the many methods, controlled radical polymerization methods can be easily applied in order to produce surface-grown polymer brushes with controlled chemical architecture. Recent reports that describe several approaches to surface-grown polymer brush formation produced by living radical polymerization techniques have appeared in the literature. For example, surface polymerization and consequent brush formation can be accomplished using a surface modified with a chlorosilyl or alkoxysilyl-functionalized nitroxide “living” radical initiator. Hawker and co-workers16 modified the properties of a silicon oxide surface by growing poly(styrene) brushes from surface-bound 2,2,4,4-tetramethylpiperidinyl-1-oxy (TEMPO) controlled radical initiator. Brittain and co-workers have grown AB and ABA-type block copolymer brushes of poly(styrene) and poly(methyl methacrylate) on planar silicate substrates by atomtransfer radical polymerization (ATRP) using a self(9) Jordan, R.; Ulman, A.; Kang, J. F.; Rafailovich, M. H.; Sokolov, J. J. Am. Chem. Soc. 1999, 121, 1016. (10) Zhao, B.; Brittain, W. J. J Am. Chem. Soc. 1999, 121, 3557. (11) Jordan, R., Ulman, A. J. Am. Chem. Soc. 1998, 120, 243. (12) Zhao, B. Brittain, W. J. Macromolecules 2000, 33, 8813. (13) Matyjaszewski, K.; Miller, P. J.; Shukla, N.; Immaraporn, B.; Gelman, A.; Luokala, B. B.; Siclovan, T. M.; Kicklelbick, G.; Vallant, T.; Hoffmann, H.; Pakula, T. Macromolecules 1999, 32, 8716. (14) (a) Von Werne, T.; Patten, T. E. J. Am. Chem. Soc. 2001, 121, 7409. (b) Von Werne, T.; Patten, T. E. J. Am. Chem. Soc. 2001, 123, 7497. (15) Ejaz, M.; Yamamoto, S.; Ohno, K.; Tsujii, Y.; Fukuda, T. Macromolecules 2001, 123, 5934. (16) Husseman, M.; Malmstro¨m, E. E.; McNamara, M.; Mate, M.; Mecerreyes, D.; Benoit, D. G.; Hedrick, J. L.; Mansky, P.; Huang, E.; Russell, T. P.; Hawker, C. J. Macromolecules 1999, 32, 1424.

10.1021/la049264f CCC: $27.50 © 2004 American Chemical Society Published on Web 10/28/2004

Surface Property Control Using Polymer Brushes

assembled monolayer of a bromoisobutyrate initiator.17,18 The authors studied the morphology of the brushes in different swelling solvents for applications in nanopattern fabrication. Brittain and co-workers also used the reversible addition-fragmentation chain-transfer (RAFT) polymerization to synthesize polymer brushes on silicate substrates. The brushes consisted of diblock copolymers of poly(styrene-b-dimethacrylamide) and of poly(dimethylacrylamide-b-methyl methacrylate) grown from silicate surfaces that were modified with azoinitiators and 2-phenyl-sec-propyl dithiobenzoate as the RAFT chain-transfer agent.19 Patten and co-workers carried out surfaceinitiated ATRP of various hydrophilic methacrylate monomers on sub-micrometer-sized silica particles in aqueous media leading to polymer-grafted silica particles whose colloidal stabilization was dependent on the nature of the grafted polymer. These organic-inorganic hybrids represented model colloids for evaluating theories of steric stabilization.20 The chemistry of grafting polymers from latex particles is a relatively new but developing area, and, as an example, poly(N,N-dimethylacrylamide) brushes were grown from poly(styrene) latex particles by aqueous atom-transfer radical polymerization.21 Genzer and coworkers prepared high-density poly(acrylamide) brushes by polymerization from a mechanically assembled monolayer of an ATRP initiator created on a PDMS surface.22 This surface growth allowed for the production of an elastomeric material with a hydrophilic surface. In the control of surface properties, a challenging goal is the creation of hydrophobic non-reconstructing surfaces for applications in coatings technology.23 Fluorinated surfaces represent good candidates for this application because of their low surface energy,24 and the creation of hydrophobic, minimally adhesive surfaces usually involves the use of fluoropolymers.25 However, a common problem with such materials consists of low substrate adhesion, as well as difficult processing due to low solubility in common organic solvents. The production of surfaces bearing tethered fluorinated polymer brushes would overcome such problems, while leading to hydrophobic non-reconstructing surfaces with the possibility of nanopattern fabrication. For example, fluorinated silicon oxide surfaces could be used for the fabrication of analytical devices for marine microbiology studies.26 Such surfaces would allow one to study the interactions between marine cells and low-energy surfaces free from roughness and other defects. Such behavior represents a fundamental issue in the design of marine antifouling coatings for minimum foulant adhesion and thus maximum fouling release. Also, due to their ability to adsorb proteins, fluorinated silicon oxide surfaces could also be useful for (17) Zhao, B.; Brittain, W. J.; Zhou, W.; Cheng, S. Z. D. Macromolecules 2000, 33, 8821. (18) Boyes, S. G.; Brittain, W. J.; Weng, X.; Cheng, S. Z. D. Macromolecules 2002, 35, 4960. (19) Baum, M.; Brittain, W. J. Macromolecules 2002, 35, 610. (20) Perruchot, C.; Khan, M. A.; Kamitsi, A.; Armes, S. P.; von Werne, T.; Patten, T. E. Langmuir 2001, 17, 4479. (21) Jayachandran, K. N.; Takacs-Cox, Aniko; Brooks, D. E. Macromolecules 2002, 35, 4247. (22) Tao, W.; Efimenkco, K.; Genzer, J. Macromolecules 2001, 34, 684. (23) Genzer, J.; Sivaniah, E.; Kramer, E. J.; Wang, J.; Ko¨rner, H.; Xiang, M.; Char, K.; Ober, C. K.; DeKoven, B. M.; Bubeck, R. A.; Chaudhury, M. K.; Sambasivan, S.; Fischer, D. A. Macromolecules 2000, 33, 1882. (24) Wang, J. G.; Ober, C. K. Macromolecules 1997, 30, 7560. (25) Ajroldi, G. Chim. Ind. (Milan) 1997, 79, 484. (26) Callow, M. E.; Rjennings, A.; Brennan, A. B.; Seegert, C. E.; Gibson, A.; Wilson, L.; Feinberg, A.; Baney, R.; Callow, J. A. Biofouling 2002, 18, 237.

Langmuir, Vol. 20, No. 24, 2004 10499

the creation of devices for the separation or microfiltration of proteins.27 In this work, the creation of fluorinated surfaces was addressed by surface-initiated controlled radical polymerization of fluorinated monomers from a TEMPO-based initiator anchored onto a silicon oxide substrate. Surfacegrown homopolymer and diblock copolymer brushes were synthesized and characterized by measurements of water contact angle and X-ray photoelectron spectroscopy (XPS). In parallel, the surface orientation of the fluorinated side chains was studied by near-edge X-ray absorption fine structure (NEXAFS) and correlated to the contact angle measurements. Experimental Section Materials. All reagents and stable free radical 2,2,6,6tetramethylpiperidinyloxy (TEMPO) were purchased from Aldrich and used without further purification unless noted. Styrene was washed several times with a 30% NaOH aqueous solution and then distilled over CaH2 (bp 45 °C/30 mmHg). Diglyme was distilled over sodium (bp 62 °C/17 mmHg). The synthetic procedures for the preparation of silicon oxide surfaces modified with fluorinated homopolymer and block copolymer brushes are illustrated in Schemes 1 and 2. Experimental details are given below. Preparation of Surface-Bound TEMPO-Based Initiators. The surface-bound TEMPO-based initiator 1 was synthesized starting from p-chloromethylstyrene and TEMPO reacted in the presence of [N,N′-bis(3,5-di-tert-butylsalicylidene)-1,2-cyclohexanediaminato]manganese(III) chloride and sodium borohydride. The addition product was then reacted with 5-hexen-1-ol and finally with trichlorosilane to yield a trichlorosilylhexyl derivative that was further reacted with freshly cleaned silicon wafers in dry toluene in the presence of triethylamine as a catalyst. Prior to reaction, silicon wafers were cut into 1 × 1 cm2 pieces and cleaned with piranha solution, first in a mixture of NH4OH/ H2O2/H2O (1/1/4 (v/v)) at 80 °C for 10 min and then with HCl/ H2O2/H2O (1/1/ 4 (v/v)) at 80 °C for 10 min. Finally the silicon pieces were cleaned with NH4OH/ H2O2/H2O (1/1/4 (v/v)) at 80 °C for 10 min, washed with deionized water, flushed with nitrogen, and dried in a vacuum oven at 120 °C for 5 min. The wafers were used within 30 min of cleaning. After surface modification, the silicon wafers with surface-bound initiator were rinsed in dry toluene, dry dichloromethane, and finally sonicated with chloroform for 3 min. Further details of this synthesis can be found in the literature.16 Preparation of Monomers. The fluorinated styrene-based monomers, 3a,b, were synthesized by a phase-transfer reaction of p-chloromethylstyrene with commercially available fluoro alcohols according to a procedure described elsewhere.28 (see Supporting Information). Preparation of |-[PFn]x Brushes. The preparation of |-[PF6]27 is described as an example. A round-bottom flask was charged under nitrogen with a small magnetic bar, a small piece of initiator-modified silicon wafer 1, 8 mg (0.03 mmol) of free TEMPO-based initiator 2, 0.5 g (1.04 mmol) of fluorinated monomer 3a, and 0.5 mL of freshly distilled diglyme. The free initiator was used to allow parallel formation of polymer in solution and thereby estimation of the molecular weight of the grafted chain by assuming that the propagation rates of free and grafted chains are the same. The mixture was then subjected to three freeze-thaw cycles and subsequently reacted at 125 °C for 48 h under nitrogen. After quenching the polymerization with ice water, the silicon wafer was separated from the liquid feed, washed repeatedly with trifluorotoluene, and finally ultrasonicated in 1,1,2-trichlorotrifluoroethane for 4 min. The wafer was then characterized by water contact angle and XPS measurements. In parallel, the soluble polymer was precipitated twice into MeOH from CHCl3, vacuum-dried, and analyzed by gel permeation chromatography (GPC) in tetrahydrofuran (THF). (27) Saito, K. Sep. Sci. Technol. 2002, 37, 535. (28) Andruzzi, L.; Chiellini, E.; Galli, G.; Li, X., Kang, S. H.; Ober, C. K. J. Mater. Chem. 2002, 12, 1684.

10500

Langmuir, Vol. 20, No. 24, 2004

Andruzzi et al.

Scheme 1. Synthesis of Surface-Grown Fluorinated Homopolymer Brushes

Table 1. Characterization Data of Polymer Brushes sample |-[A]n-[B]pa |-[PF6]27 |-[PF8]17 |-[PS]99 |-[PF6]27-[PS]128 |-[PF8]17-[PS]138 |-[PS]99-[PF6]15 |-[PS]99-[PF8]12 |-[PS]99-[PF8]17 |-[PS]99-[PF8]30

Mn(calc)d z* e θA f θRf Mnb (kg/mol) (kg/mol) (nm) (deg) (deg) z*/Rg(calc)g 13 ndc 10 13 14 7 ndc ndc ndc

11.6 9.5 10.4 12.3 13.3 7.6 6.9 9.5 17.2

10 11 8 12 13 16 15 40 48

110 122 94 95 94 108 115 119 120

104 116 80 80 79 103 105 111 111

6.7 10.6 6.9 3.4 3.7 5.2 5.0 13.0 6.03

a Degrees of polymerization (n, p) are estimated from molecular weights measured by GPC of homopolymers separated from the polymerization mixture or, in the case of insoluble homopolymers, from the fraction of unreacted monomer. b Molecular weight of the grafted terminal block estimated by GPC in THF of homopolymers separated from the polymerization mixture. c Not determined due to insolubility of the fluoropolymer in the analysis solvent, THF. d Molecular weight of the last block to be grafted estimated from initial moles of monomer, unreacted monomer, and “free” TEMPObased initiator under the assumption of polydispersity equal to 1 for the homopolymer separated from the polymerization mixture. e Thickness of the grafted polymer brush determined by ellipsometry. f Contact angles determined by goniometry measurements with water as the wetting liquid. g z* is the thickness of the brush and Rg is the radius of gyration of the polymer as formed in solution and is estimated as 0.7(N/6)1/2 with 0.7 being the statistical segment length a of polystyrene and N the degree of polymerization. In the case of block copolymers, Rg is calculated as 〈Rg2〉 ) a2(NA + NB)/6 ) 〈Rg2〉A + 〈Rg2〉B.

In the case of |-[PF8]17 the polymer separated from the polymerization feed was not soluble in the GPC analysis solvent (THF). For this case the molecular weight of the brush chains was estimated from the initial moles of monomer, unreacted monomer, and “free” TEMPO-based initiator (Table 1). The number of moles of free initiator is chosen to greatly exceed any possible upper limit for the number of moles of grafted initiator. This calculation also assumes that the polymerization is living. For cases where the free polymer is soluble in THF, this calculation was also carried out so that we could compare this estimate with the value of molecular weight from GPC.

Preparation of |-[PS]99 Brushes. A round-bottom flask was charged under nitrogen with a small magnetic bar, a small piece of initiator-modified silicon wafer 1, 12 mg (0.046 mmol) of free TEMPO-based initiator 2, and 0.5 g (4.8 mmol) of freshly distilled styrene. The free initiator was used to allow parallel formation of polymer in solution and thereby estimation of the molecular weight of the grafted chains. The mixture was then subjected to three freeze-thaw cycles and subsequently reacted at 125 °C for 48 h under nitrogen. After quenching the polymerization with ice water, the silicon wafer was separated from the feed, washed repeatedly with CHCl3, and finally ultrasonicated in CHCl3 for 4 min. The wafer was then characterized by water contact angle and XPS measurements. In parallel, the soluble polymer was precipitated twice into MeOH from CH2Cl2, vacuum-dried, and analyzed by GPC. Preparation of Block Copolymer Brushes. The preparation of |-[PS]99-[PF6]25 is described as an example. A roundbottom flask was charged under nitrogen with a small magnetic bar, a silicon wafer modified with a TEMPO-end-capped polystyrene brush, 12 mg (0.046 mmol) of free TEMPO-based initiator 2, 0.5 g (1.04 mmol) of fluorinated monomer 3a, and 0.5 mL of freshly distilled diglyme. The free initiator was used to allow parallel formation of polymer in solution and thereby estimation of the molecular weight of the grafted terminal block. The mixture was then subjected to three freeze-thaw cycles and subsequently reacted at 125 °C for 48 h under nitrogen. After quenching the polymerization with ice water, the silicon wafer was separated from the feed, washed repeatedly with trifluorotoluene, and finally ultrasonicated in 1,1,2-trichlorotrifluoroethane for 4 min. The wafer was then characterized by water contact angle and XPS measurements. In parallel, the soluble polymer was precipitated twice into MeOH from CHCl3, vacuum-dried, and analyzed by GPC. In the case of |-[PS]99-[PF8]p (p ) 12, 17, and 20) the polymer separated from the polymerization feed was not soluble in the GPC analysis solvent (THF), so the molecular weight was estimated from the initial moles of monomer, unreacted monomer, and “free” TEMPO-based initiator (Table 1) as described above. These calculations were carried out under the assumption of polydispersity equal to 1 for the polymer separated from polymerization feed. Characterization. Gel permeation chromatography was carried out using four Waters Styragel HT columns operating at 32 °C. The effective molecular weight range of the columns was

Surface Property Control Using Polymer Brushes

Langmuir, Vol. 20, No. 24, 2004 10501

Scheme 2. Synthesis of Surface-Grown Fluorinated Block Copolymer Brushes

from 500 to 107. GPC data were collected by a Waters 490 programmable multiwavelength detector and a Waters 410 RI detector. Molecular weights were quoted with respect to monodisperse polystyrene standards. THF was used as a solvent and the GPC operated at 1 mL/min. Polymers were dissolved in THF at a concentration of 1.0 mg/mL. Polymer solution volumes of 20 µL were used for GPC measurements. Contact angles were measured using a NRL contact angle goniometer Model 100-00 (Rame-Hart Inc.) at 20 °C. The contact angles were averaged over four measurements made on different areas of sample surfaces. The advancing contact angle was read by injecting a 4 µL liquid drop. The receding contact angle was measured after removing 3 µL of liquid from the droplet. The scanning force microscopy (SFM) images were taken on a commercially available multimode SFM (Digital Instruments, Santa Barbara, CA). The measurements were performed in soft tapping mode in air at room temperature. All images (512 × 512 pixels) were height images, collected while maintaining a constant average force. The X-ray photoelectron spectroscopy measurements were performed using an Axis Ultra XPS system (Kratos) with a monochromatic Al KR X-ray source (1486.6 eV) operating at 180 W under 7.0 × 10-9 Torr vacuum with charge compensation carried out by injecting very low energy electrons into the magnetic lens of the electron spectrometer. The pass energy of the analyzer was set at 80 eV. The energy resolution was set at

0.2 eV with a dwell time of 200 ms. The spectra were analyzed using CasaXPS v. 2.1.9 software. The near-edge X-ray absorption fine structure (NEXAFS) experiments were carried out on the NIST/Dow materials characterization end station on the U7A beamline at the National Synchrotron Light Source at Brookhaven National Laboratory. The beam line is equipped with a toroidal mirror spherical grating monochromator that gives this beamline an incident photon energy resolution and intensity of 0.2 eV and 5 × 1010 photons/s, respectively, for an incident photon energy of 300 eV and a typical storage ring current of 500 mA. The X-rays are elliptically polarized, with the electric field vector dominantly in the plane of the storage ring (polarization factor ) 0.85). The end station is equipped with a heating/cooling stage positioned on a goniometer that controls the orientation of the sample with respect to the polarization vector of X-rays. A differentially pumped ultrahigh vacuum compatible proportional counter is used for collecting the fluorescence yield (FY) signal. In addition, the partial electron yield (PEY) is collected using a channeltron electron multiplier with an adjustable entrance grid bias. All the data reported here are for a grid bias of -150 V.

Results and Discussion The goal of this work was to synthesize polymer brush modified surfaces having a hydrophobic and non-reconstructing character. The nitroxide-mediated controlled

10502

Langmuir, Vol. 20, No. 24, 2004

Figure 1. SFM height image of a TEMPO-based initiator modified silicon oxide surface.

radical polymerization (CRP) was chosen to produce such surfaces due to the chemical versatility of this approach. The surface CRP method has been used for the production of tethered polystyrene brushes with controlled architecture, as well as defined surface morphology.10,16 In this work the method was applied to 4-(1H,1H,2H,2H-perfluoroalkyl)oxymethylstyrene monomers synthesized by a phase-transfer-catalyzed reaction as described elsewhere.28 A TEMPO-based silicon oxide-anchored initiator 1 and a “free” TEMPO-based initiator 2 were synthesized according to general procedures reported in the literature16 and used for the creation of both homopolymer and diblock copolymer fluorinated brushes. The initiator-modified surfaces turned out to be quite smooth and homogeneous, as shown by the SFM image in Figure 1. The surfaces showed strong XPS peaks at binding energies of 533 and 285 eV corresponding to the O 1s and C 1s electron binding energies, respectively, indicative of the presence of an organic layer bound onto the surface. Surface-grafted homopolymer brushes were then produced by reaction of the surface-bound initiator 1 in the presence of a fluorinated styrene monomer 3 and a given amount of “free” initiator 2 at 125 °C for 48 h, as shown in Scheme 1. The “free” initiator was used to control the polymerization process, both at the surface and in the polymerization feed,16 and also to allow parallel formation of homopolymer in solution for characterization purposes and comparison with the parent surface-grafted homopolymer.12 A small amount of diglyme was used as a solvent in order to prevent precipitation of the free polymer during reaction. After quenching the reaction the modified surface |-[PFn]m (where the number n of (CF2) units on the monomer was either 6 or 8) was separated from the polymerization feed, repeatedly washed with trifluorotoluene, and finally sonicated for 2 min in 1,1,2-trichlorotrifluoroethane. The final sonication treatment proved to be a crucial step in the purification process leading to clean and smooth surfaces, free from adsorbed polymer particles, as shown by the SFM image in Figure 2. In parallel, the soluble polymer was separated from the polymerization feed by repeated precipitations into methanol. Surface-grafted block copolymer brushes |-[PFn]m-[PS]p and |-[PS]m[PFn]p were similarly produced by reaction of a surfacebound macroinitiator, |-[PFn]m or |-[PS]m, respectively, in the presence of either styrene or fluorinated styrene and a given amount of “free” initiator at 125 °C for 48 h, as

Andruzzi et al.

Figure 2. SFM height image of the |-[PF8]17 brush.

shown in Scheme 2. Also in this case the “free” initiator was used to control the polymerization process both at the surface and in the polymerization feed and to allow formation of homopolymer in solution for characterization purposes and comparison with the surface-grafted second block.12 GPC measurements were carried out on homopolymers formed in solution and showed polydispersities of about 1.1. Due to lack of solubility of [PF8]p polymers in the GPC analysis solvent, molecular weights were also estimated for all polymers from the initial number of moles of monomer, unreacted monomer, and “free” TEMPObased initiator assuming that the polymerization of both grafted and free chains is living. The number of moles of unreacted monomer was determined by weighing the monomer recovered by reduced pressure evaporation of the solvent used to precipitate the polymer formed in the polymerization feed. All relevant data are shown in Table 1. Agreement between GPC-determined molecular weights and calculated weights was good. The molecular weights determined by GPC were considered indicative of the degree of polymerization of the surface-grafted blocks in all cases except for [PF8]p polymers, in which case the molecular weights estimated from unreacted monomer were used. All modified silicon wafers were characterized by water contact angle measurements in order to probe the wettability of the surface. As can be noted in Table 1, the tethered fluorinated homopolymer brushes |-[PFn]m (where n ) 6 and 8) and the polystyrene brush |-[PS]99 showed values of advancing contact angle of about 110-120° and 94°, typical of fluorinated and polystyrene surfaces, respectively. The contact angle of |-[PFn]m brushes showed a dependence on the length of the semifluorinated alkyl side group, with longer side groups having higher contact angles. The tethered block copolymer brushes |-[PFn]m[PS]p (where n ) 6 and 8) showed a contact angle of about 94°, indicating the presence of an upper polystyrene block covering the starting fluorinated block. Furthermore, the tethered block copolymer brushes |-[PS]m-[PFn]p (where n ) 6 and 8) showed a contact angle of about 110-120°, indicating the presence of a fluorinated polystyrene block covering the starting polystyrene block. These data showed that the first grafted block was able to initiate the polymerization of another monomer leading to formation of a second block, which is a typical feature of “living” polymerization processes.

Surface Property Control Using Polymer Brushes

Ellipsometry measurements yielded polymer brush thickness values z* in the range 10-50 nm. Values of z*/Rg were estimated on the basis of Rg of the grafted homopolymer or block copolymer chains. Since the molecular weights were determined for the polymers formed from free initiator in the same solution and were polystyrene equivalent molecular weights, a rough estimate of Rg was carried out by computing a(N/6)1/2, where N is the degree of polymerization of the PS equivalent chain and a is the statistical segment length of PS (about 0.7 nm). For [PF8]p molecular weights estimated from unreacted monomer were used instead. If z*/Rg were less than 1, the grafted chains could be considered to be in the mushroom regime.29 The fact that z*/Rg > 1, as shown in Table 1, suggests that all these brushes are stretched. The large ratio of brush thickness to radius of gyration, as well as the dramatic change in the contact angle in passing from a homopolymer to the corresponding block copolymer pointed to high brush density and high surface coverage. Angle-Resolved XPS Studies. The composition depth profiles of the block copolymer brushes were examined using angle-resolved XPS (AR-XPS). Carbon 1s photoelectron intensities were measured at six different emission angles (0°, 15°, 30°, 45°, 60°, and 75°). The detector acceptance angle is taken to be 0° for our measurements. The spectrum was analyzed by subtracting a background that was assumed to vary as a three-parameter polymer Tougaard function30 over the region of the C 1s edge. The C 1s spectrum is made up of several peaks: a small peak from -C*F3 centered at a binding energy of 294.6 eV, a larger peak from -C*F2- at 291.7 eV, a peak from -C*H2O-CH2- at about 286.5 eV, and peaks from C atoms in the phenyl ring and on the backbone at 284.4 and 285.0 eV, respectively, as shown on the low-resolution spectrum shown in Figure 3a for |-[PS]99-[PF8]17. The fraction of photoelectrons from the carbon atoms bonded to fluorine at each emission angle was calculated from the ratio of the sum of the areas under the 294.6 and 291.7 eV peaks to the total area under all the C 1s peaks and is shown in Figure 3b. The error in this number is estimated to be (2%. The area of a shake-up peak due to the phenyl ring at 284.4 eV in the polymer and equal to 10% of the area of the phenyl ring peak of the polymer was subtracted from the fluorocarbon peaks prior to determining the ratio. The magnitude of this correction was verified from the spectrum measured from a pure polystyrene sample. From the growth mechanism we assume the fraction of the fluorinated block to follow a hyperbolic tangent function with distance away from the surface. This model allows three degrees of freedom. The three fitting parameters were chosen to be the fraction of fluorinated block on the surface, the width of the hyperbolic tangent, and the position of its inflection point from the surface. A photoelectron inelastic mean free path λ (Å) for the carbon 1s electrons was estimated to be 37 Å using a Tanuma, Powell, and Penn algorithm31 (Quases-IMFP-TPP2M). The solid line in Figure 3b corresponds to the best-fit values of the apparent atomic fraction of fluorocarbon versus emission angle for the polymer brush surface. The inset in Figure 3b shows the depth profile of the fluorinated block corresponding to this fit. The depth profile shows a surface almost completely covered by the fluorinated block, a result (29) de Gennes, P. G. Macromolecules 1980, 13, 1069. (30) Tougaard, S. Surf. Interface Anal. 1997, 25, 137. (31) Tanuma, S.; Powell, C. J.; Penn, R. R. Surf. Interface Anal. 1993, 21, 165. (32) Butoi, C. I.; Mackie, N. M.; Barnd, J. L.; Fisher, E. R.; Gamble, L. J.; Castner, D. G. Chem. Mater. 1999, 11, 862-864.

Langmuir, Vol. 20, No. 24, 2004 10503

Figure 3. (a) X-ray photoelectron spectroscopy C 1s spectrum (filled symbols) for the block copolymer brush |-[PS]99-[PF8]17 at 0° emission angle. The deconvoluted spectrum is shown (solid lines) with tags identifying individual peaks. (b) Fraction of C 1s photoelectrons from the carbon atoms bonded to fluorine as a function of emission angle for the block copolymer brush |-[PS]99-[PF8]17. Solid lines represent the calculated fluorocarbon fraction of the C 1s photoelectrons resulting from the depth profile of the fluorocarbon block shown in the inset.

that is in agreement with both the NEXAFS and water contact angle data. Some discrepancies between the predictions of the model and the experiment at the higher emission angles are to be expected since the model takes no account of the fact that the fluorocarbon chains in the fluorocarbon block may be oriented and thus closer to the surface than the carbons in the phenyl ring and polymer backbone, a fact that is clearly demonstrated for this sample by NEXAFS. A similar analysis was also carried out by comparing the F 1s and C 1s portions of the angleresolved XPS spectrum using the relative sensitivity factors for F 1s and C 1s photoelectrons provided by Kratos for our instrument, and this analysis leads to depth profiles similar to those in Figure 3b. The analysis above has less uncertainty however since the photoelectron escape depths are the same over the energy range of the C 1s spectrum and no relative sensitivity factors are required. Surface Structure Studies and Correlation with Wettability Behavior. A study of the surface structure of the tethered brushes was undertaken by near-edge X-ray absorption fine structure analysis. All samples were annealed under vacuum at 100 °C to facilitate selfassembly of side chains at the polymer film-air interface. This technique represents a useful tool for studying molecular orientation at the upper 3 nm surface region of a variety of materials.29,30 NEXAFS can measure surface composition and bond orientation and, in particular for

10504

Langmuir, Vol. 20, No. 24, 2004

Andruzzi et al.

Figure 4. NEXAFS partial electron yield (PEY) intensity for polymer surfaces versus X-ray photon energy for angles θ ) 20° (dashed line) and 90° (solid line) between the electric field vector of the polarized X-rays and the sample normal at EGB ) -150 V: (a) |-[PF8]17; (b) |-[PS]99-[PF8]17; (c) spun-cast PF8 film; (d) |-[PS]99; (e) |-[PF8]17-[PF8]138; (f) spun-cast PS film.

our study, determine the order parameter SC-F of the C-F bonds of fluorocarbon groups at the surface.31 The uniaxial orientational order parameter SF-helix of the CF2 helix as well as the average tilt angle between the helix and the surface normal, τF-helix, can be derived from SC-F by assuming that the fluorocarbon helix is rigid and that all the C-F bonds are oriented normal to it. Figure 4 shows representative partial electron yield (PEY) versus photon energy data at angles θ between the polarization vector of the X-ray beam and the surface normal, 20° (dashed line) and 90° (solid line), of homopolymers and block copolymers as compared to parent spun-cast homopolymer films as references. The brushes |-[PS]99-[PF8]17 and |-[PF8]17 (Figure 4a,b) showed NEXAFS spectra typical of fluorinated surfaces with a major peak at E ) 292.0 eV due to the C 1s f σ* transition of the C-F bond. Other peaks were observed at E ) 284.5 eV due to the C 1s f π* transition of the

phenyl ring and at E ) 287.9 eV and E ) 294.8 eV due to the C 1s f σ* transitions of the C-H and C-C bonds, respectively. The same type of spectrum was observed for the corresponding spun-cast PF8 homopolymer film of 50 nm thickness (Figure 4c). In comparison, the spectra of |-[PF8]17-[PS]138 and |-[PS]99 (Figure 4d,e) showed an intense peak at E ) 284.5 eV due to the C 1s f π* transition of the phenyl ring, but no peak was observed at the energy characteristic of the C 1s f σ* transition of the C-F bond. The same spectral features were observed for the spuncast pure polystyrene film of 50 nm thickness (Figure 4f). A similar NEXAFS spectrum was observed for |-[PF6]27[PS]128, while the NEXAFS spectrum for |-[PS]99-[PF6]15 was similar to that in Figure 4a. These results support the idea that after grafting the first block onto the surface the nitroxide moieties were able to polymerize the second block and the brush so formed was dense enough that the outermost block in all cases completely covers the surface

Surface Property Control Using Polymer Brushes

Langmuir, Vol. 20, No. 24, 2004 10505

Figure 6. Advancing (filled symbols) and receding (open symbols) contact angles ((3°) of |-[PS]99-[PF8]12 (b) and |-[PS]99-[PF8]30 (9) as a function of immersion time in water.

Figure 5 shows NEXAFS spectra of |-[PS]m-[PF8]p brushes of different thickness at angles θ ) 20° (dashed line) and 90° (solid line). It should be noted that the 15 nm thick brush (Figure 5a) did not show any intensity variation of the peak at 292.0 eV, indicative of a lack of -(CF2)- side chain surface orientation, whereas the 40 and 48 nm thick brushes (Figure 5b,c) showed an intensity variation of this peak that increased with brush thickness. The fact that the signal intensity decreased with the angle passing from 20° to 90° clearly showed that the net orientation of the axis of the fluorocarbon helix of the side chains was toward the surface normal, thus forming a -CF3-rich surface. Order parameters SF-helix ) 0.16 and 0.18, as well as average helix tilt angles 〈τ F-helix〉 of 48.4° and 47.7° were calculated for |-[PS]99-[PF8]17 and |-[PS]99[PF8]30, respectively. Compared to the PF8 spun-cast homopolymer film of similar thickness (Figure 4c), the brushes |-[PS]m-[PF8]17 and |-[PS]m-[PF8]30 showed a somewhat lower intensity ratio of the signals at 90° and 20° for the transitions C 1s f σ* of both C-F and C-C bonds. This indicated a slightly higher propensity of the side chains of the homopolymer in the spun-cast film to assemble in an orderly manner at the polymer film-air interface. This greater order could be due to higher mobility of the untethered polymer chains under the annealing conditions. On the contrary, |-[PS]m-[PF6]p brushes did not show any orientation, likely due to the shorter length of the fluorocarbon side chain substituent. To examine the influence of surface orientation of the side chains of the brush on contact angle, |-[PS]99-[PF8]12 and |-[PS]99-[PF8]30 were exposed to water for 1 week and the advancing and receding contact angles were monitored as a function of the water immersion time, as shown in Figure 6. The oriented brush, |-[PS]99-[PF8]30, was found to be more stable toward surface reconstruction, due to an enhanced stability of the ordered CF3 surface toward side group rearrangement. Figure 5. NEXAFS partial electron yield (PEY) intensity for polymer surfaces versus X-ray photon energy for angles θ ) 20° (dashed line) and 90° (solid line) between the electric field vector of the polarized X-rays and the sample normal at EGB ) -150 V of polymer brushes: (a) |-[PS]99-[PF8]12; (b) |-[PS]99[PF8]17; (c) |-[PS]99-[PF8]30.

to a depth of at last 2 nm since few of the Auger electrons constituting the NEXAFS PEY signal emerge from deeper below the surface.33 (33) Sto¨hr, J. NEXAFS Spectroscopy; Springer-Verlag: Berlin, 1992.

Conclusions Surface-tethered styrene-based homopolymer and diblock copolymer brushes bearing semifluorinated alkyl side groups of different length, i.e., -(CH2)2(-CF2)n, n ) 6 or 8, were synthesized by nitroxide-mediated controlled radical polymerization on planar silicon oxide surfaces. Angle-resolved XPS and contact angle measurements (34) Xiang, M.; Li, X.; Ober, C. K.; Char, K.; Genzer, J.; Sivaniah, E.; Kramer, E. J.; Fisher, D. A. Macromolecules 2000, 33, 6106. (35) Li, X.; Andruzzi, L.; Chiellini, E.; Galli, G.; Ober, C. K.; Hexemer, A.; Kramer, E. J.; Fischer, D. A. Macromolecules 2002, 35, 8078.

10506

Langmuir, Vol. 20, No. 24, 2004

showed that in the case of diblock copolymer brushes the second block to be added was always exposed at the polymer-air interface regardless of its surface energy. Values of z*/Rg were estimated on the basis of the radius of gyration of the grafted homopolymer or block copolymer chains for the grafted brushes and on the thickness of the brush. The fact that z*/Rg > 1 suggested that all these brushes are stretched. These combined results support the idea that after grafting the first block onto the surface the nitroxide end-capped polymer chains were able to polymerize the second block in a “living” fashion, leading to formation of a stretched brush. Furthermore, the brush so formed was dense enough that the outermost block in all cases completely covers the surface regardless of its surface energy. For instance, in the case of a nontethered block copolymer with different surface energy blocks, it is not possible to produce a polymer film exposing a layer of the higher surface energy block at the polymer-air interface. The surface behavior observed for these brushes holds very important implications for the stability of such surfaces toward reconstruction when exposed to different environments, which is a crucial requirement in application fields, ranging from the coating technology to biotechnology, where stable and non-reconstructing surfaces are needed. NEXAFS analysis revealed a dependence between the surface orientation of the fluorinated side groups when the number of -CF2 units (n ) 8) and brush thickness, with thicker brushes having the more oriented side chains. Time-dependent water contact angle measurements of these films showed that the orientation of the side chains of the brush improved the surface stability toward reconstruction upon prolonged exposure to water.

Andruzzi et al.

We thus conclude that these block copolymer brushes can be used to precisely tailor the physicochemical properties of a polymer film, allowing precise control over surface stability, molecular structure, and behavior. These concepts can be exploited to accurately design polymer blocks to form surface-tethered brushes for a variety of applications that in addition require properties such as film robustness and dimensional stability. Acknowledgment. This material is based upon work supported by both the National Science Foundation, Division of Materials Research, Polymers Program (Grant Nos. DMR-0307233 and DMR-0208825) and the Office of Naval Research (Grant No. N00014-02-1-0170). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation or any other funding agency. The Cornell Center for Materials Research and the Materials Research Laboratory of UCSB (both funded by the NSF-DMRMRSEC Program under Grant Nos. DMR-0079992 and DMR-0080034, respectively), the National Synchrotron Light Source, Brookhaven National Laboratory, which is supported by the U.S. Department of Energy, Division of Materials Sciences and Division of Chemical Sciences, are also acknowledged for use of their facilities. Supporting Information Available: Experimental details for the preparation of 4-(1H,1H,2H,2H-perfluorohexyl)oxymethylstyrene. This material is available free of charge via the Internet at http://pubs.acs.org. LA049264F