Journal of the Korean Physical Society, Vol. 51, No. 3, September 2007, pp. 1160 1165 Protein Patterning on a Glass Substrate with a Capillary Force Lithography Process Enhanced by Surface Treatment Processes Choonghan Ryu, Chijung Kim and Heeyeop Chae Department of Chemical Engineering, Sungkyunkwan University, Suwon 440-746 Jae Do Nam Department of Polymer Science and Engineering, Sungkyunkwan University, Suwon 440-746 (Received 13 November 2006) Protein patterning on a glass surface was achieved with a capillary force lithography (CFL) process with a self-assembled monolayer surface treatment in this work. A polystyrene line and space pattern was fabricated on the glass substrate by using a polydimethylsiloxane (PDMS) stamp with a capillary force lithography process. With this approach, it is possible to avoid high pressure, UV exposure, and high temperature. The thin polystyrene residual layer on the glass surface was removed by CF 4/O 2 plasma etching to complete the polystyrene pattern, and an O 2 plasma process was applied to make the exposed glass surface hydrophilic. The oxidized polystyrene area was modified into a hydrophobic one by using vaporized toluene to reduce non-specific protein binding. A self-assembled monolayer (SAM), 3-aminopropyltriethoxysiloxane (3-APTES), was used to amine-functionalize glass surface. The protein patterning was achieved and verified by immobilizing fluorescent labeled protein, fluorescein isothiocyanate conjugate - bovine serum albumin (FITC-BSA), on the amine-functionalized microscale line and space pattern. Surface adsorption of protein and 3-APTES was confirmed by analyzing the nitrogen and the oxygen compositions with an X-ray photoelectron spectrometer. The nitrogen peak showed a stronger signal when a 3-APTES layer was formed, and the signal gets stronger with protein adsorption on the 3-APTES layer. As a result, chemical patterning was achieved in this work by using a capillary force lithography process and was verified by protein patterning. PACS numbers: 81.16.Dn, 81.16.Fg, 81.65.-b Keywords: Surface treatment, Capillary force, Protein patterning I. INTRODUCTION The applications for protein patterning are expanding due to increased demand for high throughput screening, micro-arrays of bio-molecules, and diagnosis systems in various fields of the biotechnology [1 3]. Protein patterning is a crucial technique in bio- micro electronic mechanical system (bio-mems) technology and makes it possible to use significantly reduced amount of sample and to get diagnostic results within significantly reduced times. One of the popular approaches for the protein patterning is photolithography, which is widely accepted in the fabrication of highly integrated microelectronic devices. Recently developed new approaches include micro-contact printing (µcp) [4], nano-imprint lithography (NIL) [5], and capillary force lithography (CFL) [6, 7]. Photolithography and a photo-sensitive method E-mail: hchae@skku.edu; Fax: +82-31-290-7272 -1160- [8,9] are limited by high-cost equipment and many processing steps. In this work capillary force lithography is adopted as a patterning process with surface modification. A self assembled monolayer (SAM) is typically defined as a well organized monolayer spontaneously covering a substrate. A SAM is applied to various applications, such as protein immobilization [10, 11], mineralization [12], or surface modification [13] by utilizing two different functional groups, at each end of the SAM molecule. The self-assembled molecules are divided into three parts - head group, end group, and chain part. The head group chemically bonds to the surface, and the end group is typically exposed toward the outside, opposite the surface. 3-aminopropyltriethoxysiloxane (3-APTES) was used as a SAM material to immobilize the bovine serum-albumin-labeled fluorescent material (FITC-BSA) on the glass substrate in this work. The head group bonds with the glass surface, and the end group bonds with proteins in peptide bonding.
Protein Patterning on a Glass Substrate with a Capillary Choonghan Ryu et al. -1161- Fig. 1. PDMS stamp fabrication with photolithography: a) preparation of slide glass (4.8 2.8 cm), b) photo-resistant spin coating, c) placement of the mask on the glass and exposure under UV light, d) removal of the mask, e) pouring the PDMS solution onto the glass patterned, and f) release of the PDMS stamp from the glass after hardening. II. EXPERIMENTS The experimental procedure can be divided into the following steps: i) mold and stamp fabrication, ii) capillary force lithography, iii) plasma etching of the residual layer and surface treatment, and iv) SAM application and protein immobilization. Molds were made by using a photolithography process with an AZ4903 photoresist (AZ electronic materialr) on glass, and polydimethylsiloxane (PDMS) from Dow Corning c was used to make the stamps. The PDMS stamp fabrication process is shown in Figure 1. The AZ4903 photoresist was coated on the slide glass by spin coating with three steps: 2000 rpm for 5 sec, 1000 rpm for 40 sec, and 1000 rpm for 5 sec (Figure 1(b)). The mask was placed on the photoresistcoated glass and was exposed to the UV light for 2 min (Figure 1(c)). The UV-exposed photoresist is developed in the developer solution for 5 to 10 min, followed by high-pressure nitrogen brushing (Figure 1(d)). The PDMS solution was applied on top of the photoresistpatterned surface and hardened at 60 C for 12 hours. The PDMS and the curing agent were mixed in a 10 : 1 ratio (37-g PDMS : 3.7-g curing agent) to get the PDMS solution (Figure 1(e)). Finally, the PDMS stamp was completed by releasing from the slide glass after 12 hours (Figure 1(f)). After the PDMS stamp fabrication, protein pattern was achieved by using the capillary force lithography (CFL) process, followed by the residual removal using a plasma etch and various surface modification pro- Fig. 2. Process flow of protein patterning in this work. The process includes capillary force lithography, etching, surface modification, and protein immobilization with SAM. cesses. CFL was adopted for the polystyrene patterning as shown in Figures 2(a) to (c). A 10 wt% polystyrene solution was made with toluene solvent, and the solution was spin-coated on the slide glass (4.8 2.8 cm 2 ) at 1000 rpm for 10 sec and 2000 rpm for 20 sec. Then, the toluene was volatilized, and the polystyrene remains in the solid phase. (Figure 2(a)) After the PDMS stamp was placed on the glass-coated polystyrene, it was heated at 120 C for 30 min in an oven. The polystyrene solid state changes into rubbery state at this temperature, and polystyrene fills the gap formed by the stamp (Figure 2(c)). The PDMS stamp was released, and the polystyrene pattern was replicated on the glass substrate [7,14]. This gap filling process step leaves an undesired thin residual layer of polystyrene on the glass surface. The plasma etching step in an inductively-coupled tubular plasma reactor (ICP) was adopted to remove the residual from the glass surface as followed condition - 2-sccm CF 4 and 2-sccm O 2, a 260-mTorr chamber pressure, and a 250-W source power for 3 min (Figure 2(d)). The thin residual layer formed after the CFL process was removed at about a 300 nm/min etch rate. Additional multiple surface modification processes were employed to complete the protein patterning at a desired location of the glass surface (Figure 2(e)-(g)).
-1162- Journal of the Korean Physical Society, Vol. 51, No. 3, September 2007 Fig. 3. Surface modification with vaporized toluene. The PS surface become smooth and hydrophobic by using vaporized toluene. Fig. 4. 3-APTES monolayer on the glass surface. Modifying the exposed glass surface into -OH terminated groups is necessary to form the monolayer of 3APTES. A variety of surface modification processes were compared to select the most efficient method for surface modification as shown in Figure 1. As a result, the H2 O2 /H2 SO4 wet process and O2 plasma treatment are shown to be the effective methods. The O2 plasma was selected in this work because it is more consistent for our experimental conditions. Other processes require high temperatures and toxic solutions. The O2 plasma was generated in an inductively coupled tubular reactor, and the O2 treatment was employed to modify the glass surface with the following condition - 7-sccm O2, 0.32-Torr chamber pressure and 10-W source power for 30 min. The glass surface was modified mainly by O radicals generated from the O2 plasma. The polystyrene area, which is modified into a hydrophilic surface after the O2 plasma treatment, tends to cause nonspecific binding of protein on the polystyrene surface, so the oxidized polystyrene area has to be modified into a hydrophobic one to reduce undesired adhesion of protein while the glass surface is kept hydrophilic (Figure 2(f)). The patterned slide glass was placed upside down in a chamber at 120 C to modify the polystyrene area into a hydrophobic one by using vaporized toluene (Figure 3). The self-assembled monolayer material used in this work is 3-aminopropyltriethoxysilane (3-APTES (H2 C=C(CH3 )CO(OCH2 CH2 )no2 CC (CH3 )=CH2 )) from Sigma-Aldrichr, and it forms the monolayer on the glass surface. A two-µmol% 3-APTES solution is prepared by stirring with toluene for 2 hours at room temperature. The monolayer was selectively formed on the glass surface after soaking the patterned slide glass in the 3-APTES solution for 40 min at room temperature [15] (Figure 4). For the final step of the protein patterning, fluorescein Fig. 5. Images of the PDMS stamp under a field-emission scanning electron microscope (FE-SEM). (a) The master mold is a silicon wafer patterned on a scale of 50 µm, and (b) The PDMS stamp is transferred from the master mold. isothiocyanate conjugate - bovine serum albumin (FITCBSA) from Sigma-Aldrichr was used in this work, and FITC-BSA has luminance in the green region of the visible spectrum. Phosphate-buffered saline (PBS) was prepared as a buffer solution, and it had a ph of 7.4. 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC, PierceTM) and 1 mg of N-hydroxy succinimide (NHS, PierceTM) were used as cross-linkers to help the protein bind with the NH2 functional group at the end of the 3-APTES molecule (Figure 2(h)). Five-mg FITC-BSA, 10-mg EDC, 1-mg NHS and 1200-µl PBS were mixed at RT for 30 min without exposure to light. The 150µl FITC-BSA solution was dropped onto the slide glass patterned with polystyrene, and the dropped glass was incubated for 3 hour at RT without exposure to light. After a DI water rinse the adhesion of the protein and the 3-APTES was confirmed by using a fluorescent microscope (Nikon Eclipse 80i). The surface composition was also investigated by using X-ray photoelectron spectroscopy (XPS) for an elemental composition analysis. III. RESULT AND DISCUSSION FE-SEM images of the silicon master mold and the replicated PDMS stamp are shown in Figure 5. The
Protein Patterning on a Glass Substrate with a Capillary Choonghan Ryu et al. -1163- Table 1. Surface treatment effect for protein immobilization (1000-msec exposure time). High value of fluorescent intensity is represented by asterisk (*). Sample Condition Process of Treatment Fluorescent Intensity Note 1 Bare glass Reference 125.7 2 Glass (H2 O2, H2 SO4 10 min) Wet process 254.8 * Toxic 3 Glass (400 C, 4 hour) Thermal process 150.0 High Temp. 4 Glass (Tesla coil) Tesla coil 201.9 Short maintain span 5 Glass (O2 plasma) Plasma process 240.7 * Non-toxicLow temp.long maintain span Fig. 6. Images of the polystyrene pattern on the glass substrate under a FE-SEM. (a) The scale is 50 µm and line : space = 1 : 2. (b) The whole PS thickness is about 3 µm, and the pattern thickness is about 2 µm. structure of the master mold and the PDMS stamp is a 50-µm line and 100-µm space. The depth of stamp is about 2 µm. The depth is perfectly transferred from master mold. Due to the low surface energy of PDMS, the SEM image of the PDMS shows a little dust on the surface. Stiction is a challenge in the use of PDMS, especially for a small ratio of width to depth. The ratio of width to depth, however, is large enough in this work. Figure 6 shows the structure of polystyrene patterned on the glass by using CFL process, as analyzed by using field emission scanning electron microscope (FE-SEM). The 2-µm-deep wall represents the pattern formed by the capillary force in rubbery polystyrene at 120 C. The thickness of the residual layer is about 1 µm. Therefore, an etching process is necessary for the 1-µm residual (Figure 6(b)). In self-assembled monolayer formation, typical major process parameters include the water content, the reaction temperature, and the surface chemistry. The water contents in the SAM solution and the reaction temperature were adopted from other research results [16]. For an effective silanization reaction between 3-APTES and the surface, the surface should be modified so as to become OH-terminated. A variety of surface treatments for OH groups are compared, wet process, thermal process, and plasma process, as shown in Table 1. In wet treatment, the glass surface is treated in a piranha solution (H2 SO4 : H2 O2 = 7 : 3) for 10 min and shows the strongest fluorescent intensity of 254. This wet treatment is popular in semiconductor surface cleaning processes, but it requires toxic solutions, and the solutions can cause undesirable reactions in polystyrene, such as floating of the polystyrene layer. Thermal processes are often used for surface modification, and one of the typical conditions is tested in this work. Thermal processes have advantages of being non-toxic and requiring minimal equipment. The glass surface was treated at 400 C in a chamber for 4 h. The thermal process showed the minimum signal intensity among the tested processes. In this work, the thermal process is not a good candidate because it requires high temperature and can deform the polymer pattern on surface. The Tesla coil is also a popular way for surface modification and was tested here. The glass surface was treated by using a Tesla coil for 5 min and showed a fluorescent intensity of 201. To verify the durability of the surface after Tesla coil treatment, we analyzed the contact angle of the surface as a function of time. The Tesla-coil-treated surface could not maintain a low contact angle for more than a day. The contact angle was significantly increased from 4 to 24 in 24 hours. The O2 plasma surface treatment was verified as being the most efficient technique for forming a 3-APTES monolayer and showed a strong fluorescent intensity of 240. Moreover, the process does not require a high temperature or a toxic solution. The contact angle did not change over a day, unlike with the Tesla coil treatment. As a result, the O2 plasma surface treatment is the most efficient and suitable due to its being a non
-1164- Journal of the Korean Physical Society, Vol. 51, No. 3, September 2007 Fig. 7. Images of the polystyrene pattern and the BSA pattern.(a) FE-SEM image of the polystyrene pattern with a 50-µm line and a 100-µm space (1 : 2) on glass, (b) BSA pattern s fluorescent digital microscope image (exposure time: 500 msec, gain: 10.02) with a bright area mean average of 141 and a dark area mean intensity a 125 (fluorescent intensity of background is 118). toxic, low-temperature process. 3-APTES was applied on the polystyrene-patterned glass surface to immobilize the protein in a peptide bond. The protein was immobilized on the glass surface, and a topographical image obtained by using fluorescent digital microscopy is shown in Figure 7. Figure 7(a) shows the FE-SEM image of a 50-µm polystyrene pattern. Figure 7(b) shows a fluorescent digital microscopy image of the same pattern. Bovine serum albumin (BSA) is luminescent with a green color. As Figure 7(b) shown, the bright green area verifies that the protein (BSA) was immobilized on the glass along the green line. The dark area represents the polystyrene line pattern. The signal intensities of the protein and the polystyrene were 23 and 7, respectively, after the background signal of 118 was subtracted. The fluorescent intensity of the protein is two times stronger than one of the polystyrene. Protein physical immobilization with CFL is reported by Suh and coworkers [17, 18]. Since the protein is bound to the surface by physical adsorption, the bonding strength is believed to be not strong, so the protein can be detached by a solution such as blood or buffer solution in a fluidic system. In this work, the protein Fig. 8. X-ray photoelectron spectrometer analysis to verify monolayer formation and protein adsorption for different samples: (a) graph of the N component and (b) graph of the C component. is chemically immobilized on the glass surface, and the bonding strength is believed to be stronger than physical immobilization. XPS analysis was adopted to verify the 3-APTES layer on the glass and the protein binding on the 3-APTES monolayer. Three surfaces were analyzed in sequence, bare glass, glass surface coated by 3-APTES, and glass surface with protein on a 3-APTES monolayer, as shown in Figure 8. The analyzed components are nitrogen (N) and carbon (C) for the protein and the 3-APTES layer. The N and the C components should increase if the 3APTES and the protein adhere well because the N and the C components should be included in the 3-APTES and the protein. The N peak increases in the following order, as expected, due to adhesion of 3-APTES and protein: bare glass, glass surface coated by 3-APTES, and glass surface with protein on a 3-APTES monolayer. In the carbon case, the C component peak also increases, as predicted, for the same reason (Figure 8(b)). The ad-
Protein Patterning on a Glass Substrate with a Capillary Choonghan Ryu et al. -1165- sorption of 3-APTES and protein was verified by comparing the N and the C peaks. IV. CONCLUSION We have demonstrated a novel process for the patterning of bio-molecules with controlled surface properties on a microscale. A combination of CFL and surface treatment was presented as an economical and useful technique for protein patterning. Various surface treatments were compared, and oxygen plasma treatment was verified as a strong technique for protein patterning. We presented data for successful and reproducible immobilization of FITC-BSA molecules on 50-µm-wide lines with an adhered 3-APTES film. The protein immobilization technique demonstrated in this work can provide a platform for the immobilization of various proteins other than BSA by controlling the surface properties. ACKNOWLEDGMENTS The author gratefully acknowledges Taeil Kim for support with capillary force lithography. This paper was supported by Faculty Research Fund, Sungkyunkwan University, 2004. REFERENCES [1] S. H. Yeo, C. R. Choi, D. Jung and H. Y. Park, J. Korean Phys. Soc. 48, 1325 (2006). [2] A. Kramer, E. Vakalopoulou, W. D. Schleuning and J. S. Merneger, Mol. Immunol. 32, 459 (1995). [3] S. Fodor, L. Stryer, J. Winkler, C. Holmes and D. Solas, US patent 5489678. [4] Y. Xia and G. M. Whitesides, J. Am. Chem. Soc. 117, 3274 (1995). [5] S. Y. Chou, P. R. Krauss and P. J. Renstrom, Science 272, 85 (1996). [6] D. Y. Khang and H. H. Lee, Adv. Mater. 16, 176 (2004). [7] K. Y. Suh, Y. S. Kim and H. H. Lee, Adv. Mater. 13, 18 (2001). [8] A. S. Blawas and W. M. Reichert, Biomaterials 19, 595 (1998). [9] J. F. Mooney, A. J. Hunt, J. R. McIntosh, C. A. Liberko, D. M. Walba and C. T. Rogers, Proc. Natl. Acad. Sci. 93, 12287 (1996). [10] Y. J. Oh, W. Jo, J. A. Kim and S. Park, J. Korean Phys. Soc. 48, 1642 (2006). [11] S. Kim, T. W. Kwon, J. Y. Kim, H. Shin, J. G. Lee and M. M. Sung, J. Korean Phys. Soc. 49, 736 (2006). [12] H. J. Shin, D. K. Jeong, J. G. Lee, M. M. Sung and J. Y. Kim, Adv. Mater. 16, 1197 (2004). [13] R. Maboudian, W. R. Ashurst and C. Carraro, Sens. Actuators A Phys. 82, 219 (2000). [14] K. Y. Suh and H. H. Lee, Adv. Funct. Mater. 12, 6 (2002). [15] A. N. Parikht, D. L. AUara, I. B. Azouz and F. Rondelez, J. Phys. Chem. 98, 7577 (1994). [16] J. B. Brzoska, I. Ben Azouz and F. Rondelez, Langmuir 10, 4367 (1994). [17] K. Y. Suh, J. Seong, A. Khademhosseini, P. E. Laibinis and R. Langer, Biomaterials 25, 557 (2004). [18] A. Khademhosseini, S. Jon, K. Y. Suh, T. T. Tran, George Eng, J. Yeh, J. Seong and R. Langer, Adv. Mater. 15, 1995 (2003).