Langmuir 2006, 22, 2719-2725
2719
Simple Photografting Method to Chemically Modify and Micropattern the Surface of SU-8 Photoresist Yuli Wang,†,‡,§ Mark Bachman,†,‡ Christopher E. Sims,§ G. P. Li,*,†,‡ and Nancy L. Allbritton*,†,§ Integrated Nanosystems Research Facility, Department of Electrical Engineering and Computer Science, and Department of Physiology and Biophysics, UniVersity of California, IrVine, California 92697 ReceiVed NoVember 24, 2005. In Final Form: January 18, 2006 SU-8 has gained widespread acceptance as a negative photoresist. It is also finding increasing use as a structural material in microanalytical devices. Consequently, methods to tailor the surface properties of SU-8 as well as to micropattern coatings on the surface of SU-8 are needed. The SU-8 photoresist consists of EPON SU-8 resin mixed with the photoacid generator triarylsulfonium hexafluoroantimonate. This photoacid generator can also serve as a photoinitiator generating free radicals when illuminated with UV light. Under the appropriate conditions, sufficient triarylsulfonium hexafluoroantimonate remains within cured SU-8 to act as a source of free radicals and initiate UV-mediated grafting of polymers onto the surface of the SU-8. UV-mediated grafting was used to coat SU-8 surfaces with poly(acrylic acid) and other water-soluble monomers. The SU-8 surface was chemically micropatterned by placing a mask between the UV light and SU-8. The X-Y spatial resolution of micropatterned poly(acrylic acid) on the SU-8 surface was 2 µm. Three applications of these chemically modified SU-8 surfaces were demonstrated. In the first, poly(ethylene glycol) was used to protect the SU-8 from interactions with proteins, yielding a surface resistant to biofouling. In the second demonstration, the SU-8 surface was micropatterned with a cell-resistant layer to guide cellular attachment and growth. In the final application, SU-8 micropallets were encoded with polymer lines. The bar codes were read by either absorbance or fluorescence measurements. Thus, UV-mediated graft polymerization is an efficient and effective method to micropattern coatings onto the surface of SU-8.
Microfabricated lab-on-a-chip devices are evolving rapidly as new manufacturing techniques and materials are expanding their functional repertoire. Initial devices utilized silicon or glass for their fabrication, but because of the ease and cost of manufacturing these materials, polymers such as cyclic olefin copolymer (COC), poly(dimethylsiloxane) (PDMS), and poly(methyl methacrylate) (PMMA) have come into widespread use.1-8 More recently, epoxy-based SU-8 initially developed by IBM as a high-contrast, negative photoresist has found increasing use as a structural component in device manufacturing.9 SU-8 quickly achieved popularity in the electronics industry as a photoresist for micromachining and microelectronic applications because of its compatibility with conventional microfabrication techniques and its ability to produce high-aspect-ratio microstructures.10 Although SU-8 has found widespread use as a photoresist in the microfabrication of bioanalytical devices, it also possesses a number of properties that make it attractive as a structural material. These * Corresponding authors. N.L.A. E-mail:
[email protected]. Tel: 949824-6493. Fax: 949-824-9137. G.P.L. E-mail:
[email protected]. Tel: 949824-4194. Fax: 949-824-3732. † Integrated Nanosystems Research Facility. ‡ Department of Electrical Engineering and Computer Science. § Department of Physiology and Biophysics. (1) McDonald, J. C.; Whitesides, G. M. Acc. Chem. Res. 2002, 35, 491-499. (2) Sia, S. K.; Whitesides, G. M. Electrophoresis 2003, 24, 3563-3576. (3) Grayson, A. C. R.; Shawgo, R. S.; Johnson, A. M.; Flynn, N. T.; Li, Y.; Cima, M. J.; Langer, R. Proc. IEEE 2004, 92, 6-21. (4) Mekaru, H.; Yamada, T.; Yan, S.; Hattori, T. Microsys. Technol. 2004, 10, 682-688. (5) Wei, S. Y.; Vaidya, B.; Patel, A. B.; Soper, S. A.; McCarley, R. L. J. Phys. Chem. B 2005, 109, 16988-16996. (6) Li, C.; Yang, Y.; Craighead, H. G.; Lee, K. H. Electrophoresis 2005, 26, 1800-1806. (7) Lee, D. S.; Yang, H.; Chung, K. H.; Pyo, H. B. Anal. Chem. 2005, 77, 5414-5420. (8) Sohn, Y. S.; Kai, J.; Ahn, C. H. Sens. Lett. 2004, 2, 171-174. (9) Shaw, J. M.; Gelorme, J. D.; LaBianca, N. C.; Conley, W. E.; Holmes, S. J. IBM J. Res. DeV. 1997, 41, 81-94. (10) Lorenz, H.; Despont, M.; Fahrni, N.; LaBianca, N.; Renaud, P.; Vettiger, M. J. Micromech. Microeng. 1997, 7, 121-124.
properties include biocompatibility, rigidity, thermal and chemical stability, and light transparency above 360 nm.9-13 SU-8 has been employed as a structural component in bioanalytical microdevices, for example, microelectrodes, atomic force microscopy tips, microelectrophoresis chips, integrated PCR thermocyclers, cell microculture systems, DNA and protein arrays, microfluidic networks, and chromatographic sample preparation systems.14-22 Despite its many attributes, the hydrophobicity of the SU-8 surface presents a limitation to many biological applications because this property enhances nonspecific adsorption of biomolecules, limits cell attachment, and presents challenges in surface wetting and channel filling.3,23-25 Therefore, the attractiveness of SU-8 as a substrate material for the fabrication of bioanalytical (11) Kotzar, G.; Freas, M.; Abel, P.; Fleischman, A.; Roy, S.; Zorman, C.; Moran, J. M.; Melzak, J. Biomaterials 2002, 23, 2737-2750. (12) Voskerician, G.; Shive, M. S.; Shawgo, R. S.; von Recum, H.; Anderson, J. M.; Cima, M. J.; Langer, R. Biomaterials 2003, 24, 1959-1967. (13) Belanger, M. C.; Marois, Y. J. Biomed. Mater. Res. 2001, 58, 467-477. (14) Liu, J.; Bian, C.; Han, J.; Chen, S.; Xia, S. Sens. Actuators, B 2005, 106, 591-601. (15) Kim, G. M.; Kim, B.; Liebau, M.; Huskens, J.; Reinhoudt, D. N.; Brugger, J. J. Microelectromech. Syst. 2002, 11, 175-181. (16) Zhang, J.; Tan, K. L.; Gong, H. Q. Polym. Test. 2001, 20, 693-701. (17) El-Ali, J.; Perch-Nielsen, I. R.; Poulsen, C. R.; Bang, D. D.; Telleman, P.; Wolff, A. Sens. Actuators, A 2004, 110, 3-10. (18) Wakamoto, Y.; Inoue, I.; Moriguchi, H.; Yasuda, K. Fresenius J. Anal. Chem. 2001, 371, 276-281. (19) Marie, R.; Schmid, S.; Johansson, A.; Ejsing, L.; Nordstrom, M.; Hafliger, D.; Christensen, C.; Boisen, A.; Dufva, M. Biosens. Bioelectron. 2006, 21, 13271332. (20) Kastantin, M. J.; Li, S.; Gadre, A. P.; Wu, L. Q.; Bentley, W. E.; Payne, G. F.; Rubloff, G. W.; Ghodssi, R. Sens. Mater. 2003, 15, 295-301. (21) Tseng, F. G.; Lin, K. H.; Hsu, H. T.; Chieng, C. C. Sens. Actuators, A 2004, 111, 107-117. (22) Carlier, J.; Arscott, S.; Thomy, V.; Camart, J. C.; Cren-Olive, C.; Le Gac, S. J. Chromatogr., A 2005, 1071, 213-222. (23) Nordstrom, M.; Marie, R.; Calleja, M.; Boisen, A. J. Micromech. Microeng. 2004, 14, 1614-1617. (24) Hu, S.; Ren, X.; Bachman, M.; Sims, C. E.; Li, G. P.; Allbritton, N. L. Electrophoresis 2003, 24, 3679-3688. (25) Whitesides, G. M.; Ostuni, E.; Shuichi, T.; Jiang, X.; Ingber, D. E. Annu. ReV. Biomed. Eng. 2001, 3, 335-373.
10.1021/la053188e CCC: $33.50 © 2006 American Chemical Society Published on Web 02/18/2006
2720 Langmuir, Vol. 22, No. 6, 2006
microdevices would be further enhanced by efficient techniques for the chemical modification of the polymer’s surface. The surface of many polymers can be physically masked by adsorption or chemically modified to meet the specific requirements of different applications. Although surface modification techniques of other polymers have been extensively described, SU-8 has received little attention.26 Cured SU-8 can be made hydrophilic by exposure to an oxygen plasma or reaction with ethanolamine.21,23 These bulk approaches improve the wettability of microchannels fabricated from SU-8. In applications such as DNA arrays and immunoassays, it is often desirable to covalently modify the surface on the microscale in order to tailor the surface properties in specific regions. Such selective patterning typically requires a surface modification process in concert with a micropatterning technique. Marie and colleagues have demonstrated contact printing of 100 µm spots for the immobilization of DNA in an array format on SU-8.19 The development of a greater number of strategies to micropattern surface coatings on SU-8 would be of high utility and would further enable the development of SU-8 as a structural material in bioanalytical applications. UV-initiated graft polymerization has seen widespread use throughout polymer chemistry.27,28 It is an effective and efficient method used to modify the surface of a variety of polymers because it has few steps and can be used to modify the internal surfaces of an assembled device.27-33 Typically, UV-mediated graft polymerization requires the creation of reactive sites (radicals) on a surface either by exposure to UV light alone or by exposure to UV light in combination with a photoinitiator.34,35 Radical formation is followed by the covalent linkage of a monomer to the surface, which then acts as the initiation site for free radical polymerization. A wide range of monomers, including mixtures of monomers, are suitable for use in the grafting process. Thus UV-initiated graft polymerization can be applied to impart almost any property to a surface.27 By imposing a mask between the surface and the light source, UV graft polymerization can be used to pattern a surface modification.30,36 The successful application of UV-mediated grafting to photopattern coatings onto SU-8 would increase the potential of SU-8 to serve as an alternative material for bioanalytical microdevices. In the current work, the ability of UV light to initiate the polymerization of a variety of monomers onto SU-8 surfaces is demonstrated. This method relies on the residual photoacid generator within the SU-8 matrix after curing the SU-8. In the presence of a monomer solution and UV light, the remaining photoacid generator initiates the polymerization of the monomer onto the SU-8 surface. In addition, UV-mediated grafting was combined with a mask to pattern polymeric coatings on SU-8. Features with a variety of sizes and shapes were patterned onto the top surface of a planar, microstructured SU-8. The ability of the polymeric coatings to resist biofouling, pattern cell growth, (26) Makamba, H.; Kim, J. H.; Kwanseop, L.; Park, N.; Hahn, J. H. Electrophoresis 2003, 24, 3607-3619. (27) Chan, C. M. Polymer Surface Modification and Characterization; Hanser/ Gardner Publications: Cincinnati, OH, 1994. (28) Garbassi, F.; Morra, M.; Occhiello, E. Polymer Surfaces, 2nd ed.; John Wiley and Sons: New York, 1998. (29) El Kholdi, O.; Lecamp, L.; Lebaudy, P.; Bunel, C.; Alexandre, S. J. Appl. Polym. Sci. 2004, 92, 2803-2811. (30) Hu, S.; Ren, X.; Bachman, M.; Sims, C. E.; Li, G. P.; Allbritton, N. L. Anal. Chem. 2004, 76, 1865-1870. (31) Jagur-Grodzinski, J. Heterogeneous Modification of Polymers; John Wiley and Sons: New York, 1997. (32) Yang, W. T.; Ranby, B. J. J. Appl. Polym. Sci. 1996, 62, 533-543. (33) Rohr, T.; Ogletree, D. F.; Svec, F.; Frechet, J. M. J. AdV. Funct. Mater. 2003, 13, 264-270. (34) Dorman, G.; Prestwich, G. D. Biochemistry 1994, 33, 5661-5673. (35) Inoue, H.; Kohama, S. J. Appl. Polym. Sci. 1984, 29, 877-889. (36) Wang, Y.; Lai, H. H.; Bachman, M.; Sims, C. E.; Li, G. P.; Allbritton, N. L. Anal. Chem. 2005, 77, 7539-7546.
Wang et al.
and optically encode micropallets was demonstrated. The techniques and data presented in this study will facilitate the use of SU-8 as a structural component in biomedical and lab-ona-chip microdevices. Experimental Protocol Materials. The SU-8 10 photoresist and SU-8 developer were purchased from MicroChem Corp. (Newton, MA). EPON resin SU-8 was donated by Resolution Performance Products (Houston, TX). Pre-cleaned glass slides (75 × 25 × 1 mm3) were purchased from Corning Glass Works (Corning, NY). Iron oxide plates (Ferroxoplate, 7.6 × 7.6 × 0.15 cm3) were from Towne Technologies Inc. (Somerville, NJ). Acrylic acid, poly(ethylene glycol) methyl ether acrylate (average Mn ≈ 454), 2-(dimethylamino)ethyl methacrylate, toluidine blue, rhodamine B, and γ-butyrolactone were obtained from Sigma-Aldrich (St. Louis, MO). Chicken anti-mouse IgG was obtained from Invitrogen (Carlsbad, CA) and labeled with Alexa Fluor 647 by using an Alexa Fluor 647 protein labeling kit (Invitrogen, Carlsbad, CA). All of the reagents were used without further purification, except that acrylic acid was distilled under reduced pressure to remove the inhibitor, monomethyl ether of hydroquinone, and stored in a -20 °C freezer until use. Photomask Fabrication. Iron oxide photomasks with various micropatterns were fabricated according to the traditional microfabrication process.37 Fabrication of SU-8 Films and Structures. Glass slides were cleaned by immersing them in freshly prepared piranha solution (3:1 concentrated H2SO4/30% H2O2 by volume) for 30 min. Caution: piranha solution is highly corrosiVe. Extreme care should be taken when handling it. The slides were then rinsed with deionized water and dried in a nitrogen stream. The slides were dehydrated on a 200 °C hotplate for at least 5 min before use. SU-8 films of 30 µm thickness were obtained by spin coating the resist on the glass slides at 100 rpm for 10 s, followed by 1000 rpm for 30 s using a WS-200-4NPP spin coater (Laurell Technologies Corp.). The coated slides were baked on a hotplate at 65 °C for 3 min, followed by a second bake at 95 °C for 7 min to remove organic solvent. After baking, the slides were slowly cooled to room temperature. To prepare structured SU-8 (e.g., micropallets), the SU-8 film was exposed to UV light through a photomask with the designed features for 30 s using an Oriel collimated UV source (7.4 mW/cm2). The postexposure baking was done on a hotplate at 65 °C for 1 min and 95 °C for 3 min. After slowly cooling to room temperature, the SU-8 samples were developed in SU-8 developer for 5 min, rinsed with 2-propanol, and dried in a stream of nitrogen. To prepare SU-8 planar films on glass slides, the film was exposed to UV light as above but without a mask. The postexposure bake was identical to that above, but the subsequent step with the SU-8 developer was omitted. For comparison purposes, pure EPON SU-8 resin films (without photoacid generator) were prepared by spin coating 59 wt % SU-8 resin in γ-butyrolactone on glass slides as described above. The sample was then baked on a hotplate to remove the solvent as described above. These films composed of only SU-8 resin were 30 µm thick. UV-Mediated Polymerization on SU-8 Films. The SU-8 film together with the attached glass surface was cut into 2.5 × 2.5 cm2 pieces. The SU-8-coated glass was fixed on a glass plate (7.6 × 7.6 × 0.15 cm3) using Kapton tape at the four corners. In a fume hood, 40 µL of a 10 wt % acrylic acid solution in distilled water was added by a pipet to the edge of SU-8. The liquid was quickly sucked into the space between the glass plate and SU-8, forming a thin, uniform liquid layer. The thickness of this layer was estimated as described in the next paragraph. The excess liquid around the edge was cleaned with a cotton swab. This SU-8/acrylic acid/glass assembly was moved out of the fume hood and placed inside a Loctite Zeta 7411 UV flood system equipped with a 400 W metal halide lamp, with a distance (37) Rai-Choudhury, P. Handbook of Microlithography, Micromachining, and Microfabrication; SPIE Optical Engineering Press: Bellingham, WA, 1997; Vol. 1.
Method to Modify, Micropattern SU-8 Photoresist between the lamp and glass plate of 16 cm, and a UV intensity of 11.5 mW/cm2 measured by a Traceable UV light meter. The assembly was exposed to UV radiation for 1-15 min. After the reaction, the tape was removed, and SU-8 was carefully detached from the assembly, rinsed with deionized water, and incubated in sodium phosphate buffer (0.1 M, pH 8) overnight to remove poly(acrylic acid) (PAA) that was not covalently attached to the SU-8. The thickness of the liquid monomer layer between the SU-8coated glass and the larger glass plate was estimated as follows. A 2.5 cm × 2.5 cm square SU-8-coated glass was attached to a glass plate using a small piece of Kapton tape placed at each of the four corners of the assembly. The SU-8 layer faced the glass plate. Although no spacer was used between the SU-8 surface and the glass plate, a small gap remained between the two surfaces. To determine the volume of the gap between the SU-8 and the glass plate, various volumes of liquid monomer were pipetted onto the edge of the SU-8. The liquid rapidly entered the gap by capillary action, forming a thin, uniform liquid layer. Twenty microliters of monomer solution did not fill the entire space between the SU-8 and the photomask. Forty microliters of monomer solution resulted in excess liquid around the edges of the assembly. Thirty microliters, however, filled the entire gap between the SU-8 and the photomask with no excess liquid forming around the edges. The thickness of the liquid monolayer was estimated from the area (6.3 cm2) of the gap and the volume (30 µL) needed to fill the gap. The estimated thickness of the liquid monolayer was 50 µm. The monomers, poly(ethylene glycol) methyl ether acrylate and 2-(dimethylamino)ethyl methacrylate, were grafted under the same conditions as those used for acrylic acid. Measurement of the Contact Angle of a Water Droplet. Prior to measuring the contact angle of a water droplet on a surface, the surface was rinsed with sodium phosphate buffer (0.1 M, pH 8). The sample was then washed in water and purged until dry. The contact angle was measured using standard methods.24 Micropatterning of SU-8 by Photografting through a Mask. To micropattern the SU-8 surface with polymer, the grafting steps were identical to those described above for UV-mediated grafting, except that a photomask was used in place of the glass plate. Two UV sources were used: one was uncollimated (Loctite Zeta 7411 UV Flood System), and the other was collimated (Karl Suss MJB3 aligner). The uncollimated light source utilized a 400 W metal halide lamp with an intensity of 11.5 mW/cm2 at the SU-8 surface. The collimated light source was equipped with a 200 W mercury lamp with a UV intensity of 12.5 mW/cm2 at the SU-8 surface. A black cloth was placed below the SU-8/acrylic acid/photomask assembly to eliminate light reflection from adjacent surfaces. The illumination time depended on the size of the mask features. Masks with larger transparent areas utilized exposure times as short as 3 min, whereas masks with smaller features required exposure times as long as 12 min on the mask. Staining SU-8 with a Dye. PAA-coated SU-8 surfaces were observed by microscopy after staining with a visible dye (toluidine blue) or a fluorescent dye (rhodamine B). Both toluidine blue and rhodamine B have positively charged amine groups and adsorb to the negatively charged carboxylate groups of PAA, but not to native SU-8. The samples were immersed in toluidine blue (0.1 wt %) or rhodamine B (0.1 wt %) in sodium phosphate buffer (0.1 M, pH 8) for 5 min, rinsed with water, and then dried in a stream of nitrogen. The samples stained with toluidine blue were imaged under transillumination on an inverted microscope (Nikon TE 300), and the samples stained with rhodamine B were imaged using fluorescence microscopy (excitation 540 ( 20 nm, emission 625 ( 20 nm, Nikon TE 300 microscope). Quantitation of Surface COOH Density. The density of carboxylate groups on the PAA-grafted SU-8 film was measured using the toluidine blue method.38 The sample was first stained with toluidine blue as described above. The toluidine blue was then desorbed from the surface by incubating the stained film in 3 mL of 10 wt % acetic acid for 10 min. The optical absorption (633 nm) (38) Uchida, E.; Uyama, Y.; Ikada, Y. Langmuir 1993, 9, 1121-1124.
Langmuir, Vol. 22, No. 6, 2006 2721 of the toluidine blue released from the SU-8 surface was measured using a spectrophotometer (JASCO model V-530). The amount of toluidine blue originally bound to the SU-8 was then calculated from the optical density. The density of COO- groups on the SU-8 surface (expressed as mol/cm2) was calculated on the basis of the assumption of a 1:1 ratio between COO- and bound toluidine blue. Cell Culture. RBL cells were grown on the micropatterned SU-8 surface at 37 °C in a humidified 5% CO2 atmosphere in Dulbecco’s modified eagle medium supplemented with fetal bovine serum (10%), and L-glutamine (584 mg/L). Penicillin (100 units/mL) and streptomycin (100 µg/mL) were added to the media to inhibit bacterial growth. Cells were plated at concentrations (105 cells/mL) determined empirically to produce approximately one cell per 1000× field of view on the day of the experiment and were allowed to recover for 24 h. Immediately prior to use, the growth medium was removed from the cell chamber and replaced with phosphate-buffered saline (138 mM NaCl, 27 mM KCl, 10 mM PO4, pH 7.4). Cells cultured on the micropatterned SU-8 surfaces were imaged under transillumination on an inverted microscope (Nikon TE 300). Quantitation of Microscopic Bar Codes. The top surface of the SU-8 micropallets was patterned with 10-µm-wide lines composed of PAA and then stained with toluidine blue or rhodamine B. Transilluminated and fluorescence images of the bar-coded micropallets were analyzed using NIH Image J software (http://rsb.info.nih.gov/ ij/), which provided a gray-scale profile across the micropallet surface. Protein Attachment to SU-8. Alexa Fluor 647-labeled, chickenderived, anti-mouse IgG antibody was used as a model to test the adsorption of proteins onto native SU-8 and coated SU-8 micropallets (30 µm thick, 50 µm square, and 25 µm gap). A layer of poly(ethylene glycol) (PEG) was placed on the micropallet surface by photografting in the presence of the poly(ethylene glycol) methyl ether acrylate monomer for 15 min. Then, a chamber was constructed by using Sylgard 184 to attach a silicon O-ring (24 mm outer diameter) to the native SU-8 and modified SU-8 samples. IgG labeled with Alexa Fluor 647 (180 µg/mL in PBS, 200 µL) was incubated in the chamber for 20 h. The samples were then rinsed with PBS and observed by fluorescence microscopy (excitation 640 ( 20 nm, emission 660 ( 20 nm, Nikon TE300 microscope). Fourier Transform Infrared Spectroscopy (FTIR). FTIR spectra of SU-8 and modified SU-8 samples (75 mm × 25 mm) were recorded using a Perkin-Elmer 2000-FTIR spectrophotometer, equipped with an attenuated internal reflection (ATR) accessory, with a resolution of 4.00 cm-1. Atomic Force Microscopy (AFM). The surface topography of the micropatterned SU-8 sample was acquired in air with a Nano-R AFM (Pacific Nanotechnology Inc., Santa Clara, CA) in the materialsensing mode at a scan rate of 1 Hz. The AFM cantilever was composed of silicon nitride with a spring constant of 42 N/m, a length of 125 µm, and a thickness of 4 µm.
Results and Discussion UV-Mediated Grafting of Polymers onto the Surface of SU-8. The UV-mediated grafting of polymers has been used successfully to covalently attach coatings to the surface of many types of substrates.33 A photoinitiator, typically benzophenone, is mixed with a monomer solution to initiate polymerization. However, benzophenone has limited solubility, restricting its use to select applications. For example, benzophenone is insoluble in aqueous solutions of poly(ethylene glycol) methyl ether acrylate, a widely used monomer in biomedical applications. An alternative strategy is to utilize an organic solvent to implant benzophenone into the topmost surface of a substrate. When a monomer is placed over the substrate and the substrate is illuminated with UV light, the implanted photoinitiator initiates the polymerization of the monomer onto the surface. This strategy has been used successfully for the PDMS substrate.36 PDMS is an elastic polymer that swells in many organic solvents, permitting the entry of a photoinitiator into the PDMS substrate.
2722 Langmuir, Vol. 22, No. 6, 2006
Figure 1. Schematic of the SU-8 curing process and UV-mediated grafting of the SU-8 surface. (A) The SU-8 resin is placed onto a glass surface by spin coating or another method. (B) Exposure of the SU-8 surface initiates decomposition of the photoacid generator and cross linking of the SU-8 resin. To form SU-8 microstructures, a mask (not shown) is interposed between the UV light source and the SU-8 surface, and uncured SU-8 is then removed with a solvent rinse. (C) A solution of monomer is placed on the fully cured SU-8 surface. (D) The monomer solution and SU-8 surface are illuminated with UV light to initiate the formation of polymer on the surface of the SU-8. Residual photoacid generator within the SU-8 acts as a photoinitiator. To pattern the polymer graft, a mask (not shown) is placed between the UV light source and the monomer solution. (E) A layer of polymer is formed on the SU-8 surface.
In contrast, the SU-8 photoresist is a rigid material that does not swell in organic solvents, making it difficult to implant photoinitiator into the SU-8 surface. SU-8 photoresist is a composite material consisting of the photoacid generator triarylsulfonium hexafluoroantimonate mixed with EPON SU-8 resin. Exposure of triarylsulfonium hexafluoroantimonate to UV light generates hydrogen ions, which initiate cross linking of the SU-8 resin (Figure 1). Triarylsulfonium hexafluoroantimonate is also a cationic photoinitiator generating free radicals upon photodecomposition.39,40 If the triarylsulfonium hexafluoroantimonate is not fully decomposed during the SU-8 curing process, then this residual photoinitiator may be available to initiate surface grafting of a nearby monomer upon further exposure to UV light (Figure 1). To determine whether sufficient triarylsulfonium hexafluoroantimonate remained in the SU-8 substrate to initiate surface polymerization, a slab of SU-8 was overlaid with a solution of acrylic acid (0 or 10% by wt) in water. The reaction mixture was illuminated with UV light for 10 min. To remove any adsorbed polymer, the SU-8 was sequentially washed with water and sodium phosphate (0.1 M, pH 8). The SU-8 pieces were immersed in toluidine blue (0.1 wt %) for 5 min and rinsed with water. The toluidine blue stained the SU-8 grafted with the acrylic acid monomer (forming PAA) but did not bind to SU-8 illuminated in the absence of the acrylic acid monomer (Figure 2A and B). The positively charged dye bound to the negatively charged carboxylic acid groups of the PAA but not to SU-8. To further (39) Knapczyk, J. W.; McEwen, W. E. J. Org. Chem. 1970, 35, 2539-2543. (40) Dektar, J. L.; Hacker, N. P. J. Am. Chem. Soc. 1990, 112, 6004-6015.
Wang et al.
Figure 2. UV-mediated photografting utilizing residual triarylsulfonium hexafluoroantimonate in the SU-8. (A) The SU-8 film grafted with PAA and then stained with toluidine blue. (B) The SU-8 film treated in an identical fashion to the film in A but with the acrylic acid monomer omitted from the grafting reaction. (C) ATR-FTIR spectra of an unmodified SU-8 film (solid line) and an SU-8 film grafted with PAA (dashed line).
assess the presence of COOH groups on the SU-8, the infrared absorbance of the surface was measured using ATR-FITR. SU-8 grafted with PAA possessed an absorbance peak at 1710 cm-1, which is characteristic of the COOH group acrylic acid (Figure 2C). SU-8, which was not grafted with PAA, possessed little absorbance at 1710 cm-1. To determine whether PAA grafted onto the surface of SU-8 formed a stable covalent bond, the grafted SU-8 was incubated with sodium hydroxide (0.1 M) for 36 h. The SU-8 was then incubated with toluidine blue to determine whether carboxyl groups remained on the surface. The PAA-grafted SU-8 adsorbed toluidine, demonstrating that the graft remained on the SU-8 surface and was covalently bound to the SU-8. To determine whether residual triarylsulfonium hexafluoroantimonate in the SU-8 film played a role in initiating the polymerization reaction, an SU-8 film composed of only EPON SU-8 resin without the photocatalyst was photografted with acrylic acid. The sample was then incubated with toluidine blue and washed. Binding of toluidine blue to the surface was minimal, suggesting that very little PAA formed on the SU-8 surface. These data demonstrate that triarylsulfonium hexafluoroantimonate remaining within the SU-8 was responsible for the covalent attachment and polymerization of PAA on the surface of the SU-8. A potential reaction mechanism for the triarylsulfonium hexafluoroantimonate, SU-8, and acrylic acid is depicted in Figure 3. The photodecomposition of triarylsulfonium hexafluoroantimonate generates free radicals that are thought to be active in hydrogen abstraction (Figure 3A).40 The free radicals on the topmost surface diffuse to the monomer solution to initiate the polymerization of acrylic acid or abstract a hydrogen atom on the SU-8 polymer itself (Figure 3B). The later reaction generates free radicals on the SU-8 polymer chains, which then initiate the polymerization of acrylic acid, producing surface-grafted PAA (Figure 3C).
Method to Modify, Micropattern SU-8 Photoresist
Figure 3. Reaction mechanism of the polymerization reaction at the surface of the SU-8. (A) Photodecomposition of the photoacid generator, triarylsulfonium hexafluoroantimonate, produces free radicals. (B) Free radicals abstract hydrogen from the SU-8 substrate, generating free radicals on the surface of the SU-8 substrate. (C) Free radicals on the SU-8 initiate covalent attachment of acrylic acid to the surface, followed by polymerization of the monomer at the surface.
Figure 4. Dependence of surface graft density on UV photografting time and UV preexposure time. (A) The surface density of PAA was measured after varying the UV illumination time in the presence of the acrylic acid monomer. Both SU-8 photoresist (diamonds) and resin (squares) films were used. The film thickness is 30 µm, and the UV intensity is 11.5 mW/cm2. (B) SU-8 photoresist films were preexposed to UV light for varying times at two different intensities, 7.4 (squares) and 12.5 (diamonds) mW/cm2. The films were then grafted with acrylic acid (10 wt %) by exposure to UV light (11.5 mW/cm2) for 8 min. The density of PAA was then measured and plotted against the total energy per unit area applied during the UV preexposure step.
Influence of Photografting Time on the Extent of Polymer Formation. To determine the time over which PAA was formed, a film of SU-8 was grafted with acrylic acid using varying UV illumination times. The density of COO- on the SU-8 surface was then measured. By 4 min, the maximal density of surface COO- was formed (Figure 4A). As long as the UV irradiation time was less than 10 min, the solution above the SU-8 film did not become viscous, suggesting that surface polymerization dominated at these short times. For irradiation times greater than 10 min, the solution demonstrated a noticeable increase in viscosity, suggesting that PAA was forming in the solution above the SU-8. For comparison, an SU-8 film composed of only EPON SU-8 resin without the photocatalyst was also photografted with acrylic acid for varying times. The density of surface COO- was then measured. The surface density of COO- was very low at all illumination times (Figure 4A). At the longest times used, 10 min, the COO- density was approximately 10% of that for SU-8 films prepared with the photocatalyst. UV irradiation may interact directly with the SU-8 surface to form free radicals, but the
Langmuir, Vol. 22, No. 6, 2006 2723
density is not sufficient enough to form substantial amounts of surface PAA. Influence of the Quantity of the Residual Photoacid Generator on the Extent of Polymer Formation. Because the photoacid generator plays a very important role in the photografting of the SU-8 photoresist, the extent of graft formation should be dependent on the concentration of photoacid generator remaining in the SU-8 after curing. If the SU-8 photoresist is exposed to UV radiation for a prolonged time during curing, then large amounts of the photoacid generator will be consumed, leaving little to initiate subsequent polymerization onto the surface of the SU-8. To test this assumption, SU-8 photoresist films (30-µm thickness) were preexposed to UV irradiation of different intensities (7.4 and 12.5 mW/cm2) for varying times (0-8 min). The films were then illuminated with UV light (11.5 mW/cm2) in the presence of acrylic acid for 10 min. The surface grafting density was measured for each sample. The formation of PAA on the SU-8 surface was highly dependent on the duration of the UV preexposure time and the intensity of the UV light (Figure 4B). The PAA graft density was highest when the SU-8 photoresist was not cured (i.e., when it was not preexposed to UV light). In this instance, all of the triarylsulfonium hexafluoroantimonate mixed with the SU-8 resin was available to initiate the surface grafting reaction. When the SU-8 film was exposed to UV light for longer times or at higher intensities, most of the photoacid generator on the SU-8 surface was consumed, leaving little to catalyze the formation of PAA during the subsequent exposure to UV light (Figure 4B). Fortunately, SU-8 photoresist of up to 250 µm in thickness is easily processed for microfabrication applications using UV light intensities and exposure durations that do not consume all of the triarylsulfonium hexafluoroantimonate.41 Photografting of Other Water-Soluble Monomers. Other water-soluble monomers such as poly(ethylene glycol) methyl ether acrylate and 2-(dimethylamino)ethyl methacrylate were successfully photografted onto the surface of the SU-8 photoresist following 10 min of UV illumination. The surface of the cured SU-8 photoresist is relatively hydrophobic, having a water contact angle of 72°. The contact angle for a droplet of water placed on poly(ethylene glycol)-grafted SU-8 was 45°. Water placed on 2-(dimethylamino)ethyl methacrylate-modified SU-8 possessed a contact angle of 13°. These results demonstrate that the surface properties of SU-8 can be customized with UV-mediated photografting. Surface Micropatterning of SU-8. PDMS has been patterned to a resolution of 5 µm by photografting a benzophenoneimplanted surface through a photomask.36 This resolution, achievable with a simple benchtop method, is sufficient for many bioanalytical applications. The infiltration of benzophenone into the PDMS is key to achieving the 5 µm resolution because it prevents the diffusion of activated photoinitiator that would degrade the spatial resolution.36 An analogous process may also be able to micropattern the SU-8 photoresist because the photoinitiator (i.e., photoacid generator) is already embedded in the SU-8 matrix. To determine whether SU-8 might be photopatterned, a thin layer (50 µm) of acrylic acid (10 wt % in water) was sandwiched between an SU-8 film and a photomask with an array of 30 µm circles. The assembly was illuminated with an uncollimated UV light source (11.5 mW/cm2) for 9 min (Figure 5A). The SU-8 was removed from the assembly and washed with sodium phosphate buffer (0.1 M, pH 8) overnight. The SU-8 piece was then immersed in toluidine blue for 5 min and (41) Datasheet, SU-8 Photoresist Formulations. http://www.microchem.com/ products/su_eight.htm.
2724 Langmuir, Vol. 22, No. 6, 2006
Wang et al.
Figure 5. Surface micropatterning of SU-8. (A) Schematic of the photopatterning process for SU-8. The UV light impinging on the mask was collimated or uncollimated as described in the text. (BE) SU-8 films were grafted with PAA through an iron oxide mask with an array of circular holes 30 µm in diameter. For panels B and C, uncollimated UV light was used for photopatterning, whereas collimated light was employed in D and E. The grafted SU-8 was incubated with toluidine blue and photographed. Close-ups of the circular PAA grafts are shown in C and E.
rinsed with water. The photopattern was successfully transferred to the SU-8 film; however, circles patterned with this method possessed a lightly stained halo around the central circle (Figure 5B and C). This is most likely due to stray light from the uncollimated light source entering the regions shadowed by the mask. Although the resolution of this method is not high, it should still be adequate for many applications, for example, cell patterning. Moreover, this strategy is straightforward and can be performed on a benchtop in a standard laboratory with a simple UV light source. Spatial Resolution of the Surface Photopatterning. To determine whether an optimized light source could improve the pattern resolution, the grafting reaction with acrylic acid was repeated using a collimated UV source (12.5 mW/cm2). A mask with an array of circles 30 µm in diameter was used. The PAAgrafted SU-8 samples were then stained with toluidine blue. An array of circles was present with no visible staining by toluidine blue in the regions between the circles (i.e., the regions shadowed by the mask (Figure 5D and E)). The use of the collimated light source eliminated the lightly staining halo that was present previously. These data suggest that patterns on a mask can be faithfully reproduced when light scattering at the edges of the mask is minimized. Several other polymer patterns were also grafted to demonstrate the general utility of the method. An array of lines (20 µm width) was readily visualized with no observable polymer deposition between the lines (Figure 6A). Similarly, the arm (20 µm wide) of a spiral is crisp and clear (Figure 6B). For all subsequent photopatterning, a collimated UV source was used. To determine the spatial resolution of this photografting method, the topography of the patterned SU-8 surface was measured by AFM. The SU-8 sample was grafted with PAA using a photomask possessing squares with a size of 5 µm. The sample was washed and dried and then imaged by AFM (Figure 5C). The average height of the PAA graft was 79 nm above the SU-8 surface. The size of the squares at their base was 9 µm, whereas the size of the top portion of the square was 5 µm. Thus, photografting occurred a distance of 2 µm into the SU-8 region shadowed by the mask. This was most likely due to the diffusion of free radicals of photoacid generator or acrylic acid into the shadowed regions of the SU-8. In addition, scattering of light at the edges of the mask pattern may have also contributed to the formation of a polymer graft under the mask edges. Nevertheless, photografting is largely localized to the UV-exposed regions and is absent from the intervening regions. The spatial
Figure 6. Micropatterned features and patterning resolution on SU8. (A and B) Features were patterned in PAA and stained with toluidine blue. The scale bars are 100 µm. Shown in A is an array of 20-µm-wide lines spaced 20 µm apart, and shown in B is a spiral arm (20 µm wide). (C) AFM image of an SU-8 surface photopatterned with PAA through a mask with 5 µm squares. The illumination time was 6 min.
resolution of 2 µm for the patterning method is sufficient to meet the needs of a large range of bioanalytical applications. Polymeric Coatings to Block Interactions of SU-8 with Biomolecules. Free amino groups on biomolecules undergo reaction with the surface epoxide groups of SU-8. This condensation reaction has been used to attach molecules such as DNA modified with an amine group to SU-8.19 However, the reaction can also result in the undesirable linkage of biomolecules to SU-8 surfaces during assays. To demonstrate the nonspecific attachment of biomolecules to SU-8 surfaces, square SU-8 micropallets (50 µm side, 30 µm height, 25 µm gap) were fabricated on a glass surface. The micropallets were incubated with Alexa Fluor 647-labeled chicken anti-mouse IgG antibody (180 µg/mL), washed, and visualized by fluorescence microscopy. The fluorescence of the micropallets was over 5 times greater than that of a similarly handled glass surface, suggesting that the Alexa Fluor 647-IgG was now covalently attached to the pallet (Figure 7A). Extensive washing with PBS did not remove the IgG. To determine whether the unwanted reaction of the IgG with SU-8 surface could be prevented, SU-8 micropallets were photografted in the presence of poly(ethylene glycol) methyl ether acrylate (10% in water) by illumination with UV light for 15 min. A thin layer of poly(ethylene glycol) (PEG) was formed on the surface of the SU-8. The PEG-coated micropallets were incubated with Alexa Fluor 647-labeled antibody and washed as described above. When examined under the same fluorescence microscopy conditions as for the nongrafted micropallets, the PEG-grafted pallets possessed 1.2 times the brightness of a similarly stained and washed glass surface (Figure 7B). The small increase in signal of the PEG-grafted pallets relative to a glass sample is similar to that seen for pallets never incubated with the fluorescent antibody. Thus, the PEG layer on the SU-8 prevented the interaction of SU-8 with IgG, suggesting that grafted coatings can substantially reduce the biofouling of SU-8.
Method to Modify, Micropattern SU-8 Photoresist
Figure 7. Controlling protein attachment to and cell growth on SU-8. (A and B) SU-8 micropallets without (A) or with (B) a grafted PEG coating were incubated with Alexa Fluor 647-labeled antibody for 20 h at room temperature. (C) An SU-8 surface was grafted with PAA in all regions except that of a spiral. The photopatterned SU-8 was stained with toluidine blue, revealing the spiral that did not bind toluidine blue. The width of the spiral arm was 35 µm. (D) RBL cell growth on the spiral-patterned surface (without toluidine blue staining), showing cell growth along the spiral arm. The scale bar in each panel is 100 µm.
Patterning Cell Growth on SU-8. In many applications of microdevices, cells must be grown at specific locations but not at others.42 Because media used for the cell culture is rich in proteins, placement of the cell growth media onto SU-8 typically results in the rapid attachment of proteins onto the SU-8 surface. This biofouling of SU-8 provides a surface suitable for the attachment of many cell types, and the SU-8 surface can be entirely covered by cells. To place cells at only selected sites, SU-8 must be coated so that most of the surface resists cell attachment. Uncoated regions can then serve as sites of cell attachment. SU-8 was grafted with PAA using a mask that blocked light access to only a small spiral-shaped region of the SU-8. Thus, the majority of the SU-8 surface was coated with PAA. When stained with toluidine blue, a clear, unstained spiral region was visible (Figure 7C). Rat basophilic leukemia (RBL) cells were cultured on the coated SU-8 surface with the spiral pattern. The cells attached to the uncoated spiral region but not to the PAA-coated area (Figure 7D). Because PAA possesses a high negative charge, the cells are unable to adhere to most of the SU-8 surface. Using the photopatterned coatings, it should be possible to place cells at any location on an SU-8 surface. Bar Coding SU-8 Micropallets. Bar coding is frequently used to identify individual microstructures in multiplexed assays. Encoded microstructures have been used to identify cells on carriers during high-throughput image cytometry and beads used in bioassays.43,44 To demonstrate that UV-mediated grafting could be used to bar code microstructures, oval-shaped SU-8 micropallets (200 µm × 600 µm) were photopatterned with lines (10 µm width) of PAA. The bar-coded micropallets were then incubated with either toluidine blue or rhodamine B. Rhodamine B possesses a positive charge and binds to PAA but not to native SU-8. The bar codes on the micropallets stained with toluidine blue were visualized by transmitted light microscopy, and the bar code was read by scanning the pallet along its longest axis (Figure 8A and C). The fluorescent pallets were visualized and (42) Revzin, A.; Sekine, K.; Sin, A.; Tompkins, R. G.; Toner, M. Lab Chip 2005, 5, 30-37 (43) Beske, O.; Guo, J. J.; Li, J. R.; Bassoni, D.; Bland, K.; Marciniak, H.; Zarowitz, M.; Temov, V.; Ravkin, I.; Goldbard, S. J. Biomol. Screen. 2004, 9, 173-185. (44) Gao, X. H.; Chan, W. C. W.; Nie, S. M. J. Biomed. Opt. 2002, 7, 532-537
Langmuir, Vol. 22, No. 6, 2006 2725
Figure 8. SU-8 micropallets with distinctive optical and fluorescent bar codes on their top surface. Shown in A and B are microscopic images of an SU-8 pallet bearing a bar code created by micropatterning the surface with PAA and then staining with toluidine blue A or rhodamine B. The intensity profiles (absorption or fluorescence) of a line scan through the long axis of the pallets are shown in C (for the pallet in A) and D (for the pallet in B). Note that the bar codes on the pallets in A and B are distinct.
the code read in a similar fashion by fluorescence microscopy (Figure 8B,D). The pallets were readily identifiable by either their absorbance or fluorescence signature. When mixed with pallets possessing other bar codes, these bar codes can be used to reveal the identity or history of that micropallet.
Conclusions We have demonstrated a simple method to covalently modify the surface of SU-8 with polymeric coatings. This simple benchtop method relies on the presence of residual photoacid generator (triarylsulfonium hexafluoroantimonate), which also acts as a photoinitiator during UV-mediated graft polymerization. Consequently, the curing conditions for SU-8 must be carefully controlled so that enough photoinitiator remains within the SU-8 to initiate free radical polymerization. A reduction in the quantity of residual photoinitiator decreased the density of the coating on the SU-8 surface. When the photoinitiator in the SU-8 was fully depleted, little to no coating formed. A variety of water-soluble monomers could be polymerized on the surface of the SU-8. Two demonstrated applications for these coatings were resistance to biofouling and cell adhesion. Polymers could also be micropatterned on the surface of SU-8 by interposing a mask between the UV light source and the SU-8 surface during the grafting reaction. Low-resolution patterns were obtainable with an uncollimated, inexpensive light source. These patterned coatings may have applications in many areas, for example, directing the adhesion of cells and patterning arrays of largesized spots for assays. When a collimated light source was used for photopatterning, an X-Y resolution of 2 µm was obtained for a 79-nm-thick coating. This resolution was better than that obtained for poly(dimethylsiloxane) embedded with photoinitiator (benzophenone) and then photografted with polymer. The improved resolution with SU-8 is likely due to the lower porosity of SU-8 compared to that of PDMS and the decreased solubility of triarylsulfonium hexafluoroantimonate relative to that of benzophenone in aqueous media. The described polymeric coatings will enhance the attractiveness and broaden the utility of SU-8 as a substrate material for the fabrication of bioanalytical microdevices by providing an efficient benchtop method for the chemical modification of the SU-8 surface. Acknowledgment. This research was supported by the NIH (EB004436 and EB004597). LA053188E
