Controlled Grafting of Well-Defined Epoxide Polymers on Hydrogen

(PGMA) film thickness with reaction time, indicating that chain growth from the ... The chemical composition of the GMA graft-polymerized silicon (Si-...
0 downloads 0 Views 122KB Size
Download PDF
8294

Langmuir 2004, 20, 8294-8300

Controlled Grafting of Well-Defined Epoxide Polymers on Hydrogen-Terminated Silicon Substrates by Surface-Initiated ATRP at Ambient Temperature W. H. Yu, E. T. Kang,* and K. G. Neoh Department of Chemical Engineering, National University of Singapore, Kent Ridge, Singapore 119260 Received November 6, 2003. In Final Form: June 5, 2004 Controlled grafting of well-defined epoxide polymer brushes on the hydrogen-terminated Si(100) substrates (Si-H substrates) was carried out via the surface-initiated atom-transfer radical polymerization (ATRP) at room temperature. Thus, glycidyl methacrylate (GMA) polymer brushes were prepared by ATRP from the R-bromoester functionalized Si-H surface. Kinetic studies revealed a linear increase in GMA polymer (PGMA) film thickness with reaction time, indicating that chain growth from the surface was a controlled “living” process. The graft polymerization proceeded more rapidly in the dimethylformamide/water (DMF/ H2O) mixed solvent medium than in DMF, leading to much thicker PGMA growth on the silicon surface in the former medium. The chemical composition of the GMA graft-polymerized silicon (Si-g-PGMA) surfaces were characterized by X-ray photoelectron spectroscopy (XPS). The fact that the epoxide functional groups of the grafted PGMA were preserved quantitatively was revealed in the reaction with ethylenediamine. The “living” character of the PGMA chain end was further ascertained by the subsequent growth of a poly(pentafluorostyrene) (PFS) block from the Si-g-PGMA surface, using the PGMA brushes as the macroinitiators.

1. Introduction Recently, there has been growing interest in the functionalization of oriented single-crystal silicon and other semiconductor surfaces with organic molecules.1,2 The motivation for immobilizing organic molecules on an inorganic semiconductor surface stems from the desire to impart organic and chemical functionalities to a semiconductor device. Molecular properties, such as chirality, molecular recognition, nonlinear optical properties, and biocompatibility, can be readily introduced onto silicon surfaces via covalent attachment of the relevant organic compounds.3-6 Surface-initiated polymerization or surface graft polymerization has been shown to be an effective and versatile means for achieving this purpose.7,8 However, grafting or graft polymerization from initiator-containing self-assembled monolayers (SAMs) is more attractive, since a high density of initiators can be immobilized on the surface and the initiation mechanism is more welldefined. Apart from the monolayers prepared by the Langmuir-Blodgett method, work on SAMs has involved thiols on gold surface and silanes on oxidized silicon surface.9 In addition, covalent attachment of organic monolayers directly to oriented single-crystal silicon surfaces via Si-C bonds has also attracted a great deal of attention, because of the stability of the Si-C bonds.10-15 * To whom correspondence should be addressed. Telephone: +656874-2189. Fax: +65-6779-1936. E-mail: [email protected]. (1) Buriak, J. M. Chem. Rev. 2002, 102, 1272. (2) Wolkow, R. A. Annu. Rev. Phys. Chem. 1999, 50, 413. (3) Yates, J. T. Science 1998, 279, 335. (4) Cui, Y.; Wei, Q.; Park, H.; Lieber, C. M. Science 2001, 293, 1279. (5) Zhu, X. Y. Acta Phys.-Chem. Sin. 2002, 18, 855. (6) Ji, H. F.; Thundat, T.; Dabestani, R.; Brown, G. M.; Britt, P. F.; Bonnesen, P. V. Anal. Chem. 2001, 73, 1572. (7) Zhao, B.; Brittain, W. J. Prog. Polym. Sci. 2000, 25, 667. (8) Kato, K.; Uchida, E.; Kang, E. T.; Uyama, Y.; Ikada, Y. Prog. Polym. Sci. 2003, 28, 209. (9) Ullman, A. Chem. Rev. 1996, 96, 1533. (10) Linford, M. R.; Fenter, P.; Eisenberger, P. M.; Chidsey, C. E. D. J. Am. Chem. Soc. 1995, 117, 3145.

Recently, attention has centered on the synthesis of well-defined polymer by living radical polymerization.16-19 This technique combines the virtues of living polymerization with the versatility and convenience of the free radical polymerization process. Successful examples of the living radical polymerization have included nitroxidemediated radical polymerization,16 atom-transfer radical polymerization (ATRP),17,18 and reversible additionfragmentation chain-transfer (RAFT) polymerization.19 ATRP is particularly versatile for the polymerization and block copolymerization of various vinyl monomers, such as styrene, methacrylates, acrylates, and acrylonitrile.17,18 In addition, ATRP generally does not require stringent experimental conditions, as in the cases of anionic and cationic polymerization. Works on surface graft polymerization of well-defined polymers from various substrates, such as polymer,20 silicon,21-26 and gold,27-29 via ATRP have been well-documented. For the silicon substrates, (11) Sieval, A. B.; Demirel, A. L.; Nissink, J. W. M.; Linford, M. R.; van der Maas, J. H.; de Jeu, W. H.; Zuilhof, H.; Sudholter, E. J. R. Langmuir 1998, 14, 1759. (12) Boukherroub, R.; Bensebaa, S. M. F.; Wayner, D. D. M. Langmuir 1999, 15, 3831. (13) Boukherroub, R.; Wayner, D. D. M. J. Am. Chem. Soc. 1999, 121, 11513. (14) Cicero, R. L.; Linford, M. R.; Chidsey, C. E. D. Langmuir 2000, 16, 5688. (15) Buriak, J. M. Chem. Commun. 1999, 1051. (16) Hawker, C. J.; Bosman, A. W.; Harth, E. Chem. Rev. 2001, 101, 3661. (17) Matyjaszewski, K.; Xia, J. Chem. Rev. 2001, 101, 2921. (18) Kamigaito, M.; Ando, T.; Sawamoto, M. Chem. Rev. 2001, 101, 3689. (19) Chiefari, J.; Chong, Y. K.; Ercole, F.; Krstina, J.; Jeffery, J.; Le, T. P. T.; Mayadunne, R. T. A.; Meijs, G. F.; Moad, C. L.; Moad, G.; Rizzardo E.; Thang, S. H. Macromolecules 1998, 31, 5559. (20) Zheng G. D.; Sto¨ver, H. D. H. Macromolecules 2003, 36, 1808. (21) Edmondson, S.; Huck, W. T. S. J. Mater. Chem. 2004, 14, 730. (22) Ejaz, M.; Yamamoto, S.; Tsujii, Y.; Fukuda, T. Macromolecules 2002, 35, 1412. (23) 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/la036089e CCC: $27.50 © 2004 American Chemical Society Published on Web 07/27/2004

Grafting of Well-Defined Epoxide Polymers

Langmuir, Vol. 20, No. 19, 2004 8295

the initiators for ATRP can be coupled to the silicon or silicon oxide surface via organosilanes.21-25 Covalent bonding of initiators via the formation of Si-C bonds has also been reported.26 As a potential surface linker for biomolecules, poly(glycidyl methacrylate) (PGMA) has promising application in advanced biotechnology, such as DNA separation, targeted drug delivery, enzyme immobilization, and immunological assay, because of the ease in conversion of epoxide groups into a variety of functional of groups, such as -OH, -NH2, and -COOH.30,31 In addition, the epoxide-functionalized polymer brushes are promising molecular adhesives.32,33 In the present work, we report on the graft polymerization of glycidyl methacrylate (GMA) on the hydrogen-terminated Si(100) (Si-H) surface via ATRP, using a covalently bonded initiator on the silicon surface. The ATRP was carried out at room temperature, and the reaction rate was enhanced by the use of an aqueous solvent mixture. The chemical composition and topography of the pristine and functionalized Si-H surfaces are studied by X-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM), respectively. The effect of water on polymerization rate was investigated. Block copolymerization of pentafluorostyrene (FS), using the PGMA brushes as the macroinitiators, was carried out to ascertain the “living” character of the PGMAgrafted silicon surface. Functionalization of the grafted PGMA brushes was demonstrated by the reaction of epoxide groups with ethylenediamine. 2. Experimental Section 2.1. Materials. (100)-oriented single-crystal silicon, or Si(100), wafers, having a thickness of about 0.7 mm and a diameter of 150 mm, were obtained from Unisil Co. of Santa Clara, CA. The as-received wafers were polished on one side and doped as n-type. The silicon wafers were sliced into rectangular strips of about 1 cm × 3 cm in size. To remove the organic residues on the surface, the silicon substrate was washed with “piranha” solution, a mixture of 98 wt % concentrated sulfuric acid (70 vol %) and hydrogen peroxide (30 vol %).34 Caution: Piranha solution reacts violently with organic materials and should be handled carefully! After rinsing with copious amounts of doubly distilled water, the Si(100) strips were blow-dried with purified argon. The clean Si(100) strips were immersed in 10 vol % hydrofluoric acid solution in individual Teflon vials for 5 min to remove the oxide film and to leave behind a hydrogen-terminated Si(100) surface (Si-H surface).15 All the chemical reagents were purchased from Aldrich Chemical Co., Milwaukee, WI. Glycidyl methacrylate (GMA) was distilled under reduced pressure and stored in an argon atmosphere at -10 °C. 2,3,4,5,6-Pentafluorostyrene (FS) was passed through a column with the inhibitor-remover replacement packing (Aldrich) and then stored in an argon atmosphere at (24) Matyjaszewski, K.; Miller, P. J.; Shukla, N.; Immaraporn, B.; Gelman, A.; Luokala, B. B.; Siclovan, T. M.; Kickelbick, G.; Vallant, T.; Hoffmann, H.; Pakula, T. Macromolecules 1999, 32, 8716. (25) Mori, H.; Boker, A.; Krausch, G.; Mu¨ller, A. H. E. Macromolecules 2001, 34, 6871. (26) Yu, W. H.; Kang, E. T.; Neoh K. G.; Zhu, S. J. Phys. Chem. B 2003, 107, 10198. (27) Jones D. M.; Huck, W. T. S. Adv. Mater. 2001, 16, 1256. (28) Huang, W. X.; Kim, J. B.; Bruening M. L.; Baker G. L. Macromolecules 2002, 35, 1175. (29) Kim, J. B.; Huang, W. X.; Bruening, M. L.; Baker, G. L. Macromolecules 2002, 35, 5410. (30) For GMA applications: see, for example, www.dow.com/acrylic/ products/gma.html. (31) May, C. A. Epoxy Resins: Chemistry and Technology; Dekker: New York, 1988. (32) Wu, S. Y.; Kang, E. T.; Neoh, K. G.; Han H. S.; Tan, K. L. Macromolecules 1999, 32, 186. (33) Zhang, Y.; Tan, K. L.; Liaw, B. Y.; Liaw, D. J.; Kang, E. T.; Neoh, K. G. J. Vac. Sci. Technol. A 2001, 19, 547. (34) Zhang, J. F.; Cui, C. Q.; Lim, T. B.; Kang, E. T.; Neoh, K. G. Chem. Mater. 1999, 11, 1061.

Figure 1. Schematic diagram illustrating the three-step process of immobilization of the surface initiators and the process of surface graft polymerization via ATRP from the bromoester-functionalized silicon surface. -10 °C. 10-Undecylenic acid methyl ester was prepared according to the method described in the literature.11 Diethyl ether was distilled over LiAlH4 before use. Copper(I) bromide was purified according to procedures described in the literature.24 Ethyl bromoisobutyrate (EBiB), 2,2′-bipyridine (bpy) and other chemical reagents were used as received. 2.2. Immobilization of the Initiator on the HydrogenTerminated Silicon Surface. Immobilization of the initiator on the Si-H surface was achieved in three consecutive steps: (i) UV-induced coupling of 10-undecylenic methyl ester with the Si-H surface to obtain the ester-terminated Si-R1COOCH3 surface, (ii) reduction of the Si-R1COOCH3 surface by LiAlH4 to obtain the hydroxyl-terminated Si-R2OH surface, and (iii) esterification of the hydroxyl terminal groups with 2-bromoisobutyryl bromide to obtain the 2-bromoester-terminated Si-R3Br surface. The three-step immobilization process for the ATRP initiator is shown schematically in Figure 1. Details of the reaction involved in each step have been described earlier.26 From the thickness and mass density of the initiator layer and molecular weight of the initiator, the surface graft density of the initiators was estimated to be about 2 units/nm2. The stability and chemical roburstness of the Si-C bond had been reported earlier.35 2.3. Surface-Initiated Atom-Transfer Radical Polymerization. For the preparation of PGMA brushes on the Si-R3Br (35) Xu, D.; Kang, E. T.; Neoh, K. G.; Zhang, Y.; Tay, A. A. O.; Ang, S. S. J. Phys. Chem. B 2002, 106, 12508.

8296

Langmuir, Vol. 20, No. 19, 2004

surface, GMA (1.9 mL, 14 mmol), CuBr (10 mg, 0.07 mmol), CuBr2 (3.9 mg, 0.0175 mmol), and bpy (27.5 mg, 0.175 mmol) were added to 4 mL of a mixed solvent (DMF:water, 2:1 (v/v) (DMF ) dimethylformamide)). The solution was stirred and degassed with argon for 20 min. The Si-R3Br substrate was then added to the reaction mixture. The reaction flask was sealed and kept at room temperature for a predetermined period of time. After the reaction, the silicon substrate with surface-grafted poly(glycidyl methacrylate) (Si-g-PGMA substrate) was removed from the reaction mixtures and washed thoroughly with excess acetone. For the subsequent grafting of poly(pentafluorostyrene) (PFS) blocks on the Si-g-PGMA surface, FS (1.94 mL, 14 mmol), CuBr (20 mg, 0.14 mmol), CuBr2 (7.8 mg, 0.035 mmol), and bpy (55 mg, 0.35 mmol) were added to 3 mL of xylene.36 The mixture was stirred and purged with argon for 20 min. The Si-g-PGMA substrate was then introduced into the reaction mixture. The reaction flask was sealed and placed in a 110 °C oil bath for a predetermined period of time. After the reaction, the silicon substrate with surface-grafted PGMA brushes and PFS blocks (Si-g-PGMA-b-PFS surface) was removed from the reaction mixture and washed thoroughly with excess xylene. 2.4. Preparation of the Amine-Derivatized Si-g-PGMA Surface: The Si-g-PGMA-NH2 Surface. The Si-g-PGMA-NH2 surface was prepared by treatment of the Si-g-PGMA surface with excess ethylenediamine. The Si-g-PGMA substrate was immersed in 10 mL of 4 M ethylenediamine DMF solution in a conical flask. The reaction mixture was kept at room temperature for 10 h. The strip was then washed thoroughly with copious amounts of absolute acetone, prior to being dried under reduced pressure. 2.5. Materials and Surface Characterization. The XPS measurements were performed on a Kratos AXIS HSi spectrometer using a monochromatized Al KR X-ray source (1486.6 eV photons) at a constant dwell time of 100 ms and a pass energy of 40 eV. The samples were mounted on the standard sample studs by means of double-sided adhesive tapes. The core-level signals were obtained at a photoelectron takeoff angle (R, measured with respect to the sample surface) of 90°. The X-ray source was run at a reduced power of 150 W (15 kV and 10 mA). The pressure in the analysis chamber was maintained at 10-8 Torr or lower during each measurement. All binding energies (BE’s) were referenced to the C 1s hydrocarbon peak at 284.6 eV. Surface elemental stoichiometries were determined from the spectral area ratios, after correcting with the experimentally determined sensitivity factors, and were reliable to within (10%. The elemental sensitivity factors were calibrated using stable binary compounds of well-established stoichiometries. Reflectance FT-IR spectra of the surface-functionalized silicon substrates were obtained from a Bio-Rad FTS 135 FT-IR spectrophotometer using a ZnSe prism with an incident angle of 70°. Each spectrum was collected by accumulating 150 scans at a resolution of 8 cm-1. The topography of the pristine and graft-polymerized silicon surfaces was studied by atomic force microscopy (AFM), using a Nanoscope IIIa AFM from the Digital Instruments Inc., Santa Barbara, CA. In each case, an area of 5 µm × 5 µm square was scanned using the tapping mode. The drive frequency was 330 ( 50 kHz, and the voltage was between 3 and 4.0 V. The drive amplitude was about 300 mV, and the scan rate was 0.5-1.0 Hz. An arithmetic mean of the surface roughness (Ra) was calculated from the roughness profile determined by AFM. The thickness of the polymer films grafted on the silicon substrates was determined by ellipsometry. The measurements were carried out on a variable-angle spectroscopic ellipsometer (model VASE, J. A. Woollam Inc., Lincoln, NE) at incident angles of 70 and 75° in the wavelength range of 500-1000 nm. The refractive index of the dried PGMA films at all wavelengths was assumed to be 1.5. All measurements were conducted in dry air at room temperature. For each sample, thickness measurements were made on at least three different surface locations. Each thickness data reported were reliable to (1 nm. Data were recorded and processed using the WVASE32 software package. Static water contact angles of the pristine and functionalized Si-H surfaces were measured at 25 °C and 60% relative humidity (36) Jankova, K.; Hvilsted, S. Macromolecules 2003, 36, 1753.

Yu et al. by the sessile drop method, using a 3-µL water droplet in a telescopic goniometer (Rame-Hart, Model 100-00-(230), manufactured by the Rame-Hart, Inc., Mountain Lakes, NJ). The telescope with a magnification power of 23× was equipped with a protractor of 1° graduation. For each sample, at least three measurements from different surface locations were averaged. Each angle reported was reliable to (3°.

3. Results and Discussion 3.1. Surface-Initiated ATRP of GMA from the R-Bromoester Functionalized Silicon Surface. Figure 1 outlines the synthesis pathway for the preparation of PGMA brushes from Si-H surface. The initiators were immobilized on the Si-H surface by the three-step process, as described previously.26 In the preparation of covalently bonded polymer brushes on the silicon surface, a uniform and dense layer of initiators covalently tethered to the silicon surface is indispensable. The main difference between ATRP from a surface and ATRP in bulk or solution is the low absolute number of the initiators immobilized on the surface, in comparison with the amount of initiators present in the bulk of the polymerization solution. The low surface initiator concentration prevents the formations of a sufficient concentration of the deactivating species (Cu(II) complex) at the beginning of polymerization. The deactivating species is crucial to the rapid establishment of an equilibrium between the dormant and the active chains. Usually, the Cu(II) species can be obtained by the reaction of Cu(I) complex with the initiator or by addition at the beginning of reaction. In previous studies, two approaches have been proposed to solve the problem of low initiator concentration. One of the approaches is the addition of the free initiator at the beginning of the reaction,22,23 and the other is the addition of the deactivating Cu(II) complex.24 In the present study, the latter approach is adopted. Previous studies have shown that the use of polar solvents for ATRP can result in an increase in polymerization rate and dissolution of the copper complex.17,37,38 Thus, a mixed solvent (DMF/water ) 2/1 (v/v)) was chosen for the surface-initiated ATRP of GMA on the Si-R3Br substrate. The reaction medium remains homogeneous throughout the polymerization process. Neither copper nor polymer precipitate was observed during polymerization. To ascertain that the catalyst system (CuBr/CuBr2/ bpy)39,40 can provide living character for the ATRP of GMA in the mixed solvent at room temperature, synthesis of GMA homopolymer in DMF/water (2/1 (v/v)) mixture was carried out under identical conditions, except the Si-R3Br substrate was replaced by ethyl 2-bromoisobutyrate, at a monomer to initiator molar ratio of 200. After 2 h of polymerization, a GMA polymer (PGMA) sample with a M h n of about 12 400 and polydispersity index (PDI) of 1.25 was obtained. At 3 h of polymerization, the M h n and PDI were 20 264 and 1.32, respectively. With the increase in molecular weight, PGMA began to precipitate in the mixed solvent. A PGMA layer of about 9 nm in thickness, as measured by the ellipsometry, was obtained after the surfaceinitiated ATRP of GMA from the Si-R3Br surface at room temperature for 2 h (Table 1). Control experiments, under conditions similar to those used for the surface graft polymerization via ATRP, were carried out on the following (37) Pascual, S.; Coutin, B.; Tardi, M.; Polton, K.; Vairon, J. P. Macromolecues 1999, 32, 1432. (38) Perrier, S.; Haddleton, D. M. Macromol. Symp. 2002, 182, 261. (39) Lowe, A. B.; McCormick, C. L. Aust. J. Chem. 2002, 55, 367. (40) Osborne, V. L.; Jone D. M.; Huck, W. T. S. Chem. Commun. 2002, 1838.

Grafting of Well-Defined Epoxide Polymers

Langmuir, Vol. 20, No. 19, 2004 8297

Table 1. Chemical Composition, Contact Angle, Film Thickness, and Surface Coverage of the Graft-Polymerized Silicon Surfaces Si-g-PGMA surface

[O]/[C]a

1d 2e

0.40 (0.43) 0.42 (0.43)

chem composn [C-H]/[C-O]/[OdC-O]b 3.1/3.0/1.0 (3/3/1) 3.0/3.0/1.0 (3/3/1)

static contact angle ((3°)

film thickness ((1 nm)

surface coveragec (mg/m2)

68 67

9 28

9 28

a Determined from XPS sensitivity factor corrected core-level spectral area ratio. Value in parentheses is the theoretical ratio. b Determined from the XPS curve-fitted C 1s core-level spectra. Value in parentheses is the theoretical ratio. c Surface coverage ) film thickness × density. (1.0 g/cm3 for PGMA). d Reaction conditions: [PGMA]/[CuBr]/[CuBr2]/[Bpy] ) 200/1/0.25/2.5; [GMA] ) 2.4 M; solvent, DMF/water (2/1 (v/v)); temp, 25 °C; reaction time, 2 h. e Reaction conditions are the same as sample 1 except the polymerization time was 6 h.

Figure 2. XPS (a) wide scan and (b) C 1s core-level spectra of the Si-R3Br surface subjected to ATRP of GMA at room temperature in a mixed DMF/H2O medium for 2 h.

surfaces: the Si-H surface, the Si-R1COOCH3 surface, and the Si-R2OH surface. No increase in thickness of the organic layer on any of these control surfaces was discernible. The results confirm that the increase in thickness observed is the result of graft polymerization from the 2-bromoisobutyrate-functionalized silicon surface, or the Si-R3Br surface. The surface coverage is defined as the amount of the grafted polymer per square meter of the surface and is calculated from the product of thickness and density of the grafted polymer layer. For simplicity, the density of the corresponding bulk polymer is used as the density of the grafted polymer film. Thus, a surface coverage of about 28 mg/m2 was obtained for the PGMA brushes of 28 nm in thickness on the Si-H surface after 6 h of reaction at room temperature (Table 1). The relatively low surface coverage of the PGMA layer (9 nm in thickness) did not allow the cleavage of sufficient materials from the silicon substrate surface for molecular characterization. Nevertheless, the molecular weight of GMA polymer obtained by bulk polymerization under similar conditions has been obtained (see above) 3.2. Characterization of the Si-g-PGMA Surface. The presence of grafted GMA polymer on the silicon surface is ascertained by XPS analysis after the silicon substrate has been washed exhaustively in acetone. Figure 2 shows the wide scan and C 1s core-level spectra of the Si-g-PGMA surface with a thickness of about 9 nm. The wide scan spectrum is dominated by the C 1s and O 1s signals. The fact that the Si signals are barely discernible in the widescan spectrum of the Si-g-PGMA surface indicates that the thickness of the PGMA layer is comparable to the probing depth of the XPS technique.41 The C 1s core-level spectrum of the Si-g-PGMA surface can be curve-fitted (41) Tan, K. L.; Woon, L. L.; Wong, H. K.; Kang, E. T.; Neoh, K. G. Macromolecules 1993, 26, 2832.

Figure 3. FT-IR spectra of (a) the Si-g-PGMA surface (surface coverage ) 28 mg/m2) and (b) the surface after being subjected to reaction with 4 M ethylenediamine in DMF at room temperature.

with three peak components having BE’s at about 284.6, 286.2, and 288.5 eV, attributable to the C-H, C-O, and OdC-O species, respectively.42,43 Table 1 summarizes the surface analysis results of the GMA-graft-polymerized silicon substrates via surface-initiated ATRP at two different reaction times. As shown in Table 1, the [O]/[C] ratio and the [C-H]/[C-O]/[OdC-O] ratio of the two Si-g-PGMA surfaces, obtained from XPS analysis, are in good agreement with the corresponding theoretical ratios of 3/7 and 3/3/1. Thus, the chemical structure, in particular the epoxide functional group, of the GMA polymer is preserved during the surface-initiated ATRP process (see also below). Figure 3a shows the surface reflectance FT-IR spectrum of the Si-g-PGMA surface with a surface coverage of 28 mg/m2. The appearance of the absorption band at 1735 cm-1, attributable to the carbonyl stretching vibration of the ester group of PGMA,33 confirms the presence of a PGMA layer on the silicon surface. In addition, absorption peaks around 2968 cm-1 are attributable to the C-H stretching of the aliphatic compounds. The changes in topography of the silicon surfaces after modification by the surface-initiated ATRP were studies by AFM. A representative AFM image of the Si-g-PGMA surface (PGMA thickness of about 9 nm) is shown in Figure 4a. As reported previously, the root-mean-square surface (42) Beamson, G.; Briggs, D. High-Resolution XPS of Organic Polymers: The Scienta ESCA300 Database; Wiley: Chichester, U.K., 1992; p 277. (43) Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D. In X-ray Photoelectron Spectroscopy; Chastian, J., Ed.; Perkin-Elmer: Eden Prairie, MN, 1992.

8298

Langmuir, Vol. 20, No. 19, 2004

Figure 4. AFM images of (a) the Si-g-PGMA surface and (b) the Si-g-PGMA-NH2 (PGMA thickness ) 9 nm).

Figure 5. Dependence of the thickness of the PGMA layer, grown from the Si-R3Br surface via ATRP on polymerization time in (a) DMF/water medium and (b) DMF.

roughness (Ra) of the pristine Si-H and Si-R3Br surfaces is only about 0.44 and 0.52 nm, respectively.26 After surface graft polymerization of GMA via ATRP, the Ra value increases slightly to about 1.8 nm. Although, the ellipsometry data still indicate that the grafted PGMA film exhibits nanoscopic uniformity in thickness, the slow growth rate of about 9 nm in 2 h is consistent with the presence of a relatively low initiator graft concentration (about 2 initiator graft units/nm2) on the silicon surface, arising from the multistep surface reaction. Nevertheless, the ATRP has proceeded uniformly on the Si-R3Br surface, and the grafted GMA polymer chains on the silicon surface exist as a distinctive overlayer. 3.3. Effect of Solvent on the Rate of SurfaceInitiated ATRP of GMA. The effect of water on the rate of graft polymerization of GMA by ATRP is also investigated. The thickness of polymer brushes grown on the silicon surface as a function of polymerization time in two different reaction media is shown in Figure 5. An approximately linear increase in thickness of the PGMA graft layer on the Si-R3Br surface with polymerization

Yu et al.

time, albeit not at the same rate, is observed both in the mixed DMF/water medium and in pure DMF. These results indicate that the process of surface-initiated ATRP of GMA is controlled in both solvent systems. Surface-initiated polymerization of GMA in pure DMF gave rise to only a 5 nm thick PGMA film in 6 h, compared to the 28 nm thick film obtained when the graft polymerization was carried out in the DMF/water mixture. This marked difference in surface graft polymerization rate indicates that the ATRP process is facilitated by the presence of an aqueous medium. Recently, several reports on the unexpectedly rapid, yet controlled, ATRP in aqueous media have appeared.27-29 For example, synthesis of hydrophobic polymer brushes from gold surface was facilitated in methanol/water medium.27 Poly(2-hydroxyethyl methacrylate) (PHEMA) film of about 700 nm in thickness has been synthesized on gold surface in an aqueous medium in just 12 h.28 We have also obtained polymer brushes over 100 nm in thickness from surfaceinitiated ATRP of poly(ethylene glycol) monomethacrylate (PEGMA) on Si-H surface in 12 h under ambient temperature in an aqueous medium. Several factors have been proposed to account for the accelerated reaction rate in aqueous media in the previous studies. An increase in polarity of the reaction medium might have a marked effect on the stability and structure of the reactants and intermediates.38 The most probable catalytic species in aqueous media is the simple monocationic [Cu(I)(bpy)2]+ complex with a halide counterion.17,44 The presence of water is believed to promote the formation of the more active Cu(I) catalyst. However, in conventional bulk or nonaqueous ATRP, the catalyst systems are the neutral, binuclear Cu(I) complexes with bridging halide ligands.17,45 The difference in catalyst structure probably accounts for the difference in the overall rate of polymerization. 3.4. Block Copolymerization of Pentafluorostyrene with the Si-g-PGMA Surface. Additional evidence on the controlled nature of the present graft polymerization process is obtained by using the GMA polymer brushes as the macroinitiators for the second round of ATRP of the same or another monomer. Recently, homopolymer and block copolymer of FS have been synthesized by ATRP using the copper/bipyridine system.36 Block copolymerization of FS with the Si-g-PGMA surface was selected in the present study to investigate the “living” character of the PGMA brushes. The F 1s signal of the FS polymer (PFS) is expected to provide a sensitive marker in the XPS analysis. The Si-g-PGMA substrate, with a PGMA layer thickness of about 9 nm, was placed in the FS monomer solution at 110 °C for 6 h to prepare the Si-gPGMA-b-PFS surface. An increase in thickness of the grafted polymer layer by about 4 nm was observed after the ATRP of FS, confirming the presence of living chain ends on the Si-g-PGMA surface. The substantial decrease in the rate of reinitiation from the Si-g-PGMA surface, nevertheless, suggests that chain termination on the surface, followed by disproportion reactions that consume the active chains, may have become important on the initial Si-g-PGMA surface with increasing polymerization time. Parts a and b of Figure 6 show the respective C 1s and F 1s core-level spectra of the Si-g-PGMA-b-PFS surface. A new F 1s peak component at the BE of about 688 eV has appeared. The C 1s core-level spectrum of the Si-g-PGMA-b-PFS surface can be curve-fitted with three peak components having BE’s at about 284.6, 286.4, and (44) Wang, X. S.; Armes S. P. Macromolecues 2000, 33, 6640. (45) Kickelbick, G.; Reinohl, U.; Ertel, T. S.; Weber, A.; Bertagnolli, H.; Matyjaszewski, K. Inorg. Chem. 2001, 40, 6.

Grafting of Well-Defined Epoxide Polymers

Langmuir, Vol. 20, No. 19, 2004 8299

Figure 6. XPS (a) C 1s and (b) F 1s core-level spectra of Si-g-PGMA-b-PFS surface.

Figure 7. XPS (a) C 1s and (b) N 1s core-level spectra of the Si-g-PGMA-NH2 surface.

288.3 eV, attributable to the C-H, C-O/C-C5F5, and OdC-O/C-F species, respectively.42,43 The [F]/[C] ratio and [C-H]/[C-O and C-C5F5]/[OdC-O and C-F] ratio are about 0.3 and 1/1.1/0.95, respectively, as determined from the XPS F 1s and curve-fitted C 1s core-level spectral area ratios. Compared to the C 1s core-level spectrum of the Si-g-PGMA surface shown in Figure 2b, the high BE component (about 288.3 eV) of the curve-fitted C 1s corelevel spectrum of the Si-g-PGMA-b-PFS surface is enhanced, as shown in Figure 6a, due to contribution of the C-F species from PFS. The ratio of the PFS blocks to the PGMA blocks within the XPS sampling depth is defined as the number of the FS repeat units (NFS) per GMA repeat unit (NGMA) and calculated from the expression

chain reaction (PCR), and cDNA molecules.46 Furthermore, peptides, proteins (for instance, antibodies), cells, and tissues can be immobilized on epoxy microarrays or microslides.47 The surface-functionalized silicon substrates with a high density of epoxide groups are, thus, promising for applications in hybrid molecular-semiconductor devices or chemical biosensors. The room-temperature reaction of the Si-g-PGMA surface with an amine-containing compound (ethylenediamine) was investigated to ascertain, in a more quantitative manner, the preservation of the epoxide groups and their reactivity toward further functionalization. The room-temperature reaction of PGMA brushes with excess ethylenediamine results in the opening of the epoxide rings. Parts a and b of Figure 7 show the respective C 1s and N 1s core-level spectra of the resulting Si-g-PGMA-NH2 surface from immersion of the Si-g-PGMA substrate (PGMA surface coverage: 28 mg/m2) in 4 M ethylenediamine solution. A new N 1s peak component at the BE of about 399.2 eV, attributable to the amine nitrogen,42 is observed for the Si-g-PGMA-NH2 surface. In addition, a new peak component at the BE of 285.5 eV, attributable to the C-N species,42 is discernible in the curve-fitted C 1s core-level spectrum of the Si-PGMA-NH2 surface, when compared to the C 1s corelevel spectrum of the Si-g-PGMA surface (Figure 2b). The [N]/[C] ratio and [C-H]/[C-N]/[C-O]/[OdC-O] ratio of the Si-g-PGMA-NH2 surface, obtained from XPS analysis, are about 0.2 and 3.0/2.9/2.0/1.0, respectively. The plausible reactions of the epoxy groups with ethylenediamine are shown in Scheme 1. Reaction of 2 mol of epoxide with 1 mol of ethylenediamine will result in a [C-N]/[C-H] ratio of 2/3, while reaction of 1 mol of epoxide with 1 mol of ethylenediamine will result in a [C-N]/[C-H] ratio of 3/3. The C-N peak component in the curve-fitted C 1s core-level spectrum of the Si-g-PGMA-NH2 surface is contributed solely by ethylenediamine. Thus, the extent of conversion of epoxide groups within the XPS sampling depth can be determined simply from the [C-N]/[C-H] ratio. The observed [C-N]/[C-H] ratio of about 0.96 is indicative of opening of almost all the epoxide rings, and each ethylenediamine molecule has reacted with only one epoxide group. The conclusion that each ethylenediamine molecule has reacted with only one epoxide group, and not two or more epoxide groups, is further supported by

NFS ) NGMA (area of F 1s peak)/5 ((area of C 1s peak) - (area of F 1s peak) × 8/5)/7 where the stoichiometric factors of 5 and 8 are introduced to account, respectively, for the five F atoms and eight C atoms in each PFS unit. The factor 7 accounts for the seven C atoms per GMA unit. The spectral areas were corrected using their respective sensitivity factors. A ratio of 4/5 was obtained for the PFS/PGMA blocks. These results confirm that some of the dormant sites at the grafted PGMA chain ends allow the reactivation of the graft polymerization process, resulting in the formation of the block copolymer brushes on the silicon surface. After FS has been block-copolymerized with the PGMA brushes, the contact angle of the Si-g-PGMA-b-PFS surface increases from about 70 to 110°. The latter contact angle is comparable to those of the fluoropolymer films41 and further confirms that the hydrophobic PFS blocks have been grafted on the Si-g-PGMA surface. 3.5. Functionalization of the Si-g-PGMA Surface. An important feature of the Si-g-PGMA surface is the preservation of the reactive epoxide groups during the ATRP process. In principle, all nucleophilic groups, such as, -NH2, -SH, -OH, and -COOH groups, will react readily and irreversibly with the epoxy groups in subsequent surface functionalization. Epoxide-based microarray is especially suited for the covalent immobilization of oligonucleotides (10-80 bases), the products of polymerase

(46) Stears, R. L.; Martinsky T.; Schena, M. Nat. Med. (N.Y.) 2003, 9, 140. (47) MacBeath, G.; Schreiber, S. L. Science 2000, 289, 1760.

8300

Langmuir, Vol. 20, No. 19, 2004

Yu et al. Scheme 1

the fact that the [N]/[C] ratio of the surface is about 0.2, which is very close to the theoretical ratio of 0.22 for the 1/1 curing reaction. The 1/1 curing reaction will also be favored in the presence of a large excess of the amine curing agent, as in the present case. Furthermore, although ethylenediamine is tetrafunctional, the secondary amines are expected to be less reactive with epoxide groups than the primary amines at room temperature. The complete opening of the epoxide rings is further ascertained by the fact that the [C-O]/[OdC-O] ratio of the Si-g-PGMA-NH2 surface has decreased to 2/1, from 3/1 for the original Si-g-PGMA surface. The reflectance FT-IR spectrum of the Si-g-PGMA-NH2 surface is shown in Figure 3b. The appearance of a broad band in the 3360 cm-1 region is attributable to the O-H and N-H stretching and further confirms the ring-opening reaction of the epoxide groups on the silicon surface by ethylenediamine. The topography of the Si-g-PGMA-NH2 surface was also studied by AFM. A representative AFM image of the Si-g-PGMA-NH2 surface is shown in Figure 4b. After surface functionalization with ethylenediamine, the Ra value increases to about 2.6 nm, which is slightly higher than that of the corresponding Si-g-PGMA surface shown in Figure 4a.

4. Conclusions Controlled grafting of well-defined GMA polymer (PGMA) brushes could be carried out by surface-initiated ATRP of GMA on the hydrogen-terminated Si(100) substrates (Si-H substrates) at ambient temperature after surface functionalization with R-bromoester. The rate of graft polymerization was enhanced in the aqueous mixture (DMF/water medium). The epoxies functional groups on the resulting Si-g-PGMA surface were preserved quantitatively. Kinetic studies revealed a linear increase in thickness of the surface-graft-polymerized film with reaction time, indicating that the chain growth from the surface was a controlled process with a “living” character. The living character of the grafted PGMA chain ends was further confirmed by the subsequent growth of a poly(pentafluorostyrene) block from the Si-g-PGMA surface. The grafted PGMA brushes can be further functionalized in high yield via reaction of their epoxide functional groups. The PGMA brushes thus provide opportunities for tailoring the oriented single-crystal silicon surface for a variety of potential applications, such as covalent immobilization of oligonucleotides, PCR products, and DNA. LA036089E