Multilayer Transfer Printing on Microreservoir-Patterned Substrate

Aug 2, 2006 - The mixture of inherently hard NOA 63 and hydrogel precursor, PEGDA, ... Employing the NOA 63-PEGDA composite mold, a polyelectrolyte ...
3 downloads 0 Views 520KB Size
Download PDF
Langmuir 2006, 22, 7689-7694

7689

Multilayer Transfer Printing on Microreservoir-Patterned Substrate Employing Hydrophilic Composite Mold for Selective Immobilization of Biomolecules Nae Yoon Lee, Ju Ri Lim, Min Jung Lee, Sungsu Park, and Youn Sang Kim* Center for Intelligent Nano-Bio Materials, DiVision of Nano Sciences (BK 21) and Department of Chemistry, Ewha Womans UniVersity, 11-1 Daehyun-dong, Seodaemun-gu, Seoul 120-750, Korea ReceiVed February 1, 2006. In Final Form: June 28, 2006 In this study, we introduce a hydrophilic composite mold with elasticity and moderate water permeability, suitable for transferring water-soluble polar molecules such as polyelectrolyte multilayer. This composite mold is constructed from two UV-curable polymerssNorland Optical Adhesives (NOA) 63, a urethane-related polymer, and poly(ethylene glycol) diacrylate (PEGDA). The mixture of inherently hard NOA 63 and hydrogel precursor, PEGDA, resulted in an optically transparent mold with some degree of elasticity and enhanced water permeability upon UV polymerization. Employing the NOA 63-PEGDA composite mold, a polyelectrolyte multilayer comprising alternate thin layers of poly(acrylic acid) (PAA) and poly(acrylamide) (PAAm) was transfer-printed onto arrays of microreservoir-patterned substrate to selectively prevent unwanted adsorption of biomolecules on the protruding surface. Antibody was immobilized selectively inside the microreservoirs where multilayer was not transferred, and a specific antibody binding reaction was detected inside the microreservoirs. Furthermore, the potential of this composite mold as a convenient tool for constructing a biosensor for detecting Escherichia coli (E. coli) O157:H7 was explored.

Introduction Ultrathin film coatings have gained much attention due to their many potential applications, such as the selective micropatterning of the functional nanoparticles and biomolecules used for establishing biochemical sensors in an economical fashion. Among the various techniques which can be used for constructing multiple layers of thin films, the contact-printing method utilizing a polymeric mold has been widely adopted, owing to its process simplicity and wide applicability. Choosing the right material for the mold, therefore, has emerged as a primary concern for the successful transfer of charged functional polymers such as polyelectrolyte multilayers.1-10 So far, poly(dimethylsiloxane) (PDMS)2,11,12 and agarose13-15 stamps have been widely used for the micropatterning and transfer of polar inks onto a substrate. PDMS is highly suitable for microcontact printing, owing to its * To whom correspondence should be addressed. Tel: +82-2-3277-4131. Fax: +82-2-3277-3419. E-mail: [email protected]. (1) Zheng, H.; Lee, I.; Rubner, M. F.; Hammond, P. T. AdV. Mater. 2002, 14, 569. (2) Park, J.; Hammond, P. T. AdV. Mater. 2004, 16, 520. (3) Lee, I.; Hammond, P. T.; Rubner, M. F. Chem. Mater. 2003, 15, 4583. (4) Park, J.; Kim, Y. S.; Hammond, P. T. Nano Lett. 2005, 5, 1347. (5) Feng, J.; Wang, B.; Gao, C.; Shen, J. AdV. Mater. 2004, 16, 1940. (6) Ahn, J. S.; Hammond, P. T.; Rubner, M. F.; Lee, I. Colloids Surf., A 2005, 259, 59. (7) Yang, S. Y.; Mendelsohn, J. D.; Rubner, M. F. Biomacromolecules 2003, 4, 987. (8) Yang, S. Y.; Lee, D.; Cohen, R. E.; Rubner, M. F. Langmuir 2004, 20, 5978. (9) Berg, M. C.; Yang, S. Y.; Hammond, P. T.; Rubner, M. F. Langmuir 2004, 20, 1362. (10) Zheng, H.; Berg, M. C.; Rubner, M. F.; Hammond, P. T. Langmuir 2004, 20, 7215. (11) Bernard, A.; Renault, J. P.; Michel, B.; Bosshard, H. R.; Delamarche, E. AdV. Mater. 2000, 12, 1067. (12) Donzel, C.; Geissler, M.; Bernard, A.; Wolf, H.; Michel, B.; Hilborn, J.; Delamarche, E. AdV. Mater. 2001, 13, 1164. (13) Weibel, D. B.; Lee, A.; Mayer, M.; Brady, S. F.; Bruzewicz, D.; Yang, J.; DiLuzio, W. R.; Clardy, J.; Whitesides, G. M. Langmuir 2005, 21, 6436. (14) Campbell, C. J.; Smoukov, S. K.; Bishop, K. J. M.; Grzybowski, B. A. Langmuir 2005, 21, 2637. (15) Mayer, M.; Yang, J.; Gitlin, I.; Gracias, D. H.; Whitesides, G. M. Proteomics 2004, 4, 2366.

conformal contact with solid substrate and low surface energy which facilitates its release. However, due to its inherently poor wetting and strong hydrophobicity, it sometimes requires either fast processing with ambient humidity11 or additional steps to modify the surface of the PDMS stamp,12 in order to facilitate the physical adsorption of water-soluble molecules such as proteins and polyelectrolyte multilayers, thereby limiting the choice of chemicals or molecules that can be transferred. On the other hand, agarose16,17 is highly suitable as a mold for transferring water-soluble biomolecules because of its high permeability to water so that multiple stamping becomes possible without the need for the intermediate re-inking of the stamp. However, because it is likely to shrink in dry air, some challenges remain to be overcome before the successful assembling and positioning of biomolecules with high resolution can be achieved. In this study, we introduce a new composite mold with outstanding water permeability and enhanced surface wettability which enables the transfer of hydrophilic polar ink such as polyelectrolyte multilayer. This mold is composed of two highly transparent, UV-curable polymerssNorland Optical Adhesives (NOA) 63,4,18-20 a urethane-related commercial polymer which is cured by means of mercapto-related cross-linking reagents and is generally adopted as an optical adhesive, and poly(ethylene glycol) diacrylate (PEGDA) which forms a hydrogel upon its reaction with a photoinitiator.21,22 Of these two materials, NOA 63 provides high mechanical rigidity and the PEGDA hydrogel provides high permeability to water. When these two materials (16) Smoukov, S. K.; Bishop, K. J. M.; Klajn, R.; Campbell, C. J.; Grzybowski, B. A. AdV. Mater. 2005, 17, 1361. (17) Grzybowski, B. A.; Bishop, K. J. M.; Campbell, C. J.; Fialkowski, M.; Smoukov, S. K. Soft Matter 2005, 1, 114. (18) Kim, Y. S.; Lee, N. Y.; Lim, J. R.; Lee, M. J.; Park, S. Chem. Mater. 2005, 17, 5867. (19) Mogensen, K. B.; Petersen, N. J.; Hubner, J.; Kutter, J. P. Electrophoresis 2001, 22, 3930. (20) Zauscher, S.; Klingenberg, D. J. J. Colloid Interface Sci. 2000, 229, 497. (21) Revzin, A.; Russell, R. J.; Yadavalli, V. K.; Koh, W.-G.; Deister, C.; Hile, D. D.; Mellott, M. B.; Pishko, M. V. Langmuir 2001, 17, 5440. (22) Chan-Park, M. B.; Yan, Y.; Neo, W. K.; Zhou, W.; Zhang, J.; Yue, C. Y. Langmuir 2003, 19, 4371.

10.1021/la060305c CCC: $33.50 © 2006 American Chemical Society Published on Web 08/02/2006

7690 Langmuir, Vol. 22, No. 18, 2006

Lee et al.

Scheme 1. Overall Scheme for the Transfer of PAA/PAAm Polyelectrolyte Multilayer onto the Microreservoir-Patterned NOA 63 Substrate Using the Composite Mold and Subsequent Immobilization of Antibody Inside the Microreservoirsa

a (a) Negative pattern transfer from Si wafer onto NOA 63 prepolymer; (b) UV curing (λ ) 365 nm, 30 min); (c) Si wafer release from the patterned NOA 63 substrate; (d) oxygen plasma treatment (50 W, 0.1 Torr, 30 s); (e) APTMS treatment (5%, 5 min); (f) stacking of 10.5 bilayers of PAA/PAAm on the mold; (g) PAA/PAAm transfer onto NOA 63 substrate; (h) rabbit IgG (1 mg/mL) immobilization inside the microreservoirs (4 °C, overnight); and (i) specific binding of rabbit IgG with FITC-tagged anti-rabbit IgG (room temperature, 40 min).

are blended at an appropriate ratio, the resulting composite mold possesses some degree of elasticity and water permeability, which are the key factors for the development of an appropriate mold for microcontact printing and the polymer on polymer stamping (POPS) of polar chemicals. To examine the water permeability of this composite mold, transfer of hydrophilic polyelectrolyte multilayer using a flat composite mold was performed onto a microreservoir-patterned polymeric substrate. Because it is reported7-9 that a polyelectrolyte multilayer, composed of alternate poly(acrylic acid) (PAA)/poly(acrylamide) (PAAm) layers, tends to resist protein adsorption, we contact-printed the multilayer selectively onto the protruding surface of the microreservoir-patterned substrate using the composite mold and immobilized the antibody exclusively inside the microreservoirs. As an ultimate objective, we examined the potential of this mold to be used as a simple tool for the construction of a biosensor platform for detecting Escherichia coli (E. coli) O157:H7. Experimental Section Materials. Negative photoresist, SU-8 2007, was purchased from MicroChem Corporation. Poly(dimethylsiloxane) (PDMS, Sylgard 184) was purchased from Dow Corning. Poly(ethylene glycol) diacrylate (PEGDA, Mw ) 575), 2-hydroxy-2-methylpropiophenone (HOMPP), 3-aminopropyltrimethoxysilane (APTMS, 97%), poly(acrylic acid) (PAA, Mw ) 100 000), and poly(acrylamide) (PAAm, Mw ) 10 000) were purchased from Aldrich. Norland Optical Adhesives (NOA) 63 was purchased from Norland Company. Rabbit immunoglobulin G (IgG), fluorescein isothiocyanate (FITC)-tagged anti-rabbit IgG, FITC-tagged anti-goat IgG, and N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDAC) were purchased from Sigma. N-Hydroxysuccinimide (NHS) was purchased from Fluka. E. coli O157:H7 antibody was purchased from Kirkegaard

& Perry Laboratories (KPL, Gaithersburg, MD). E. coli O157:H7 43894 was purchased from the American Type Culture Collection (ATCC). Composite Mold Fabrication. The composite mold was prepared by blending the NOA 63 prepolymer and PEGDA hydrogel precursor solution at a ratio of 6:4 (w/w). The PEGDA hydrogel precursor solution was prepared by mixing PEGDA and the photoinitiator, HOMPP, at a ratio of 95:5 (v/v). This ratio of PEGDA and the photoinitiator was chosen in order to endow the mold with high transparency after UV polymerization. The 6:4 (w/w) mixture of NOA 63 prepolymer and PEGDA hydrogel precursor solution was poured into a Petri dish to a depth of 1 mm and was homogenized via thorough blending using a plastic stick. After the mixture was left undisturbed for approximately 5-10 min to evacuate air bubbles, it was cured under UV light (λ ) 365 nm) for 30 min using UV lamps (Philips, TL 8W) (see also the Supporting Information). The total power of the lamp was 56 W, and the distance between the lamp and the substrate was 6 cm. The polymerized composite mold was cut into pieces with dimensions of 1 cm × 1 cm. The water permeability of the mold was calculated using the equation,23 M ) (Ww - Wd)/Wd × 100, where M is the absorbed moisture (wt %) in the mold, Wd is the weight of the dry mold, and Ww is the weight of the wet mold after immersion in water for 30 min. Microreservoir Fabrication. Arrays of 8 µm square-shaped microreservoirs with approximate depths of 5 µm were created on a Si wafer by a conventional photolithographic method. As shown in Scheme 1, the negative pattern of the arrays of microreservoirs on the Si wafer was transferred onto NOA 63 prepolymer on a glass substrate (Scheme 1a) and cured under UV light for 30 min (Scheme 1b). After peeling off the Si wafer, microreservoir-patterned NOA 63 polymeric substrate was obtained (Scheme 1c). Multilayer Stacking and Transfer. The surface of the microreservoir-patterned NOA 63 substrate was treated with oxygen plasma (50 W, 0.1 Torr, 30 s) (Scheme 1d) and dipped into a 5%

Hydrophilic Composite Mold for Multilayer Transfer

Langmuir, Vol. 22, No. 18, 2006 7691

Table 1. Comparisons of Young’s Modulus and Water Permeability of Four Moldsa

mold NOA 63 mold PEGDA hydrogel mold NOA 63-PEGDA composite mold PDMS mold

Young’s modulus (MPa) 1655 8.5∼9.1 2.4∼4.5

water permeability (%) 0.11 5.7 1.3 0.05

a These results depend on the curing conditions and the size of the mold. These are the averages of triplicate measurements.

solution of APTMS in methanol for 5 min to generate the amineterminated surface (Scheme 1e). The above-prepared composite mold was soaked in deionized water for 30 min, and 11 layers of PAA and 10 layers of PAAm were stacked alternatively (10.5 bilayers of PAA/PAAm) by both dipping and spinning methods (Scheme 1f). In the dipping method, the composite mold was dipped into an aqueous solution of PAA (0.5 wt %, pH ) 2.82, 15 min), rinsed three times (2, 1, 1 min in separate bins) with deionized water and then dipped into an aqueous solution of PAAm (5 wt %, pH ) 3.57, 15 min), followed by the same rinsing procedures. The pH of the deionized water used for rinsing was adjusted to 3.0 with dilute HCl (0.01 M). In the spinning method, the water-soaked composite mold was inked with 10.5 bilayers of PAA (0.5 wt %)/PAAm (5 wt %) by alternately spin-coating PAA and PAAm at 3000 rpm for 40 s and rinsing with deionized water by spinning at 3000 rpm for 40 s after every spin-coating step. The multilayer-stacked composite mold was pressed lightly onto aminosilane-treated microreservoirpatterned NOA 63 substrate for 3 min by applying homogeneous pressure over the whole area of the composite mold to selectively transfer the multilayer onto the raised outer surrounding regions of the microreservoir-patterned NOA 63 substrate via the POPS technique (Scheme 1g). The transfer of the PAA/PAAm multilayer onto the amine-terminated NOA 63 substrate was confirmed by electrostatic attachment of positively charged amino-functionalized polystyrene (PS) beads onto both the NOA 63 substrate and the composite mold after the transfer. Antibody Immobilization. A total of 1 mg/mL of rabbit IgG was dissolved in a 1:1 (v/v) mixture solution of 150 mM EDAC and 60 mM NHS dissolved in phosphate buffered saline (PBS; 10 mM sodium phosphate buffer, 2.7 mM KCl, and 137 mM NaCl; pH 7.4), and evenly distributed over the multilayer-coated substrate. After overnight immobilization of antibody inside the microreservoirs at 4 °C (Scheme 1h) and washing out unimmobilized antibody, immobilized antibody was reacted with fluorescein isothiocyanate (FITC)-tagged anti-rabbit IgG (110 µg/mL) or FITCtagged anti-goat IgG (110 µg/mL) at room temperature for 40 min (Scheme 1i). E. coli O157:H7 Culture. Green fluorescent protein (GFP) plasmid was transformed into E. coli O157:H7 43894 and incubated in an autoclaved Luria-Bertani (LB) broth at 37 °C until the optical density (OD) reached 1.77, and this culture broth was used directly as an indication for E. coli O157:H7 without further dilution. Measurements. Scanning electron microscopy (SEM) images were obtained using a JSM-6700F (JEOL). The optical micrographs and the fluorescent micrographs were obtained using an inverted fluorescence microscope (Axiovert 200 MAT, Carl Zeiss).

Results and Discussion Characterization of NOA 63-PEGDA Composite Mold. The Young’s modulus and water permeability of four types of moldssNOA 63, PEGDA hydrogel (PEGDA:HOMPP ) 95:5 (v/v)), NOA 63-PEGDA (6:4 (w/w)), and PDMSswere calculated as shown in Table 1. The 6:4 (w/w) blend was chosen as a representative of the composite mold among various blend ratios because it displayed the highest flexibility. The Young’s modulus obtained in this experiment revealed that the NOA 63-PEGDA composite mold used in this study

possessed some degree of elasticity, which seemed to originate from blending NOA 63, which is an inherently hard polymer with a Young’s modulus of 1655 MPa and tensile strength of 34.5 MPa, with the hydrogel precursor, PEGDA. The water permeability of the NOA63-PEGDA composite mold was 26 times higher than that of PDMS, making it an ideal platform for the transfer printing of various polar inks. The amount of water absorbed for 30 min reached approximately 1.3% of its dry weight, but the water absorption did not deform the mold. The enhanced water permeability of the composite mold seemed to originate from the inherently hydrophilic nature of PEGDA hydrogel. The advancing water contact angles on the surfaces of UV-polymerized NOA 63 mold, UV-polymerized PEGDA mold, and UV-polymerized composite mold were measured to be 76°, 40°, and 62°, respectively (see also the Supporting Information). The contact angle of the UV-polymerized composite mold was approximately the intermediate value of those measured on the UV-polymerized NOA 63 mold and the UV-polymerized PEGDA mold, signifying that phase separation has not taken place. The water contact angle of the 6:4 (w/w) composite mold, 62°, indicates that the surface is hydrophilic. Because of this inherent hydrophilicity of the new composite mold, the contact angle of the 6:4 (w/w) composite mold showed no significant change over time and remained constant. Therefore, unlike the hydrophobic PDMS mold, the composite mold enables the transfer or contact printing of polar ink without the pretreatment of the mold surface. The elasticity and water permeability of a mold are two important factors that determine successful transfer of hydrophilic polar inks via soft lithographic techniques. The NOA 63 prepolymer and PEGDA hydrogel precursor are highly transparent and viscous polymers, both of which cure under the same UV light wavelength of 365 nm. Once they are blended completely and cured, they did not exhibit any phase separations either before or during the UV polymerization, and they formed a totally new material which is optically transparent. As for the evidences that phase separation has not taken place, we observed the external appearance of the mixture of NOA 63 prepolymer and PEGDA hydrogel precursor solution by leaving undisturbed the thoroughly blended homogeneous mixture for 20 h under dark room condition. As a result, the external appearance as well as the transparency remained unchanged within the time frame of 20 h. After UV polymerization, we could obtain a highly transparent mold. Also, to examine whether the surface retains homogeneous surface property, water wettability was observed from region to region by dropping an equivalent amount of water (50 µL). The water wettability seemed almost the same on the entire surface of the composite mold, signifying even further that phase separation has not taken place (see also the Supporting Information). The resulting mold displayed some degree of elasticity and exhibited enhanced water permeability, making it an outstanding mold for various soft lithographic techniques. Fabrication of Microreservoirs. Figure 1 shows the SEM images of the arrays of 8 µm square-shaped microreservoirs patterned on the NOA 63 substrate. NOA 63,4,18 originally developed as an optical adhesive, has excellent optical properties over a wide spectral range (transmission above 98% in the wavelength range from 360 to 1260 nm) and strong mechanical properties (Young’s modulus, E ) 1655 MPa). Approximately 15 700 microreservoirs were successfully fabricated on a 1 cm × 1 cm area of the NOA 63 substrate with high fidelity. NOA 63 substrate is highly resistant to swelling in various organic solvents such as methanol, toluene, and trichloroethylene. This

7692 Langmuir, Vol. 22, No. 18, 2006

Figure 1. SEM images of (a) arrays of 8 µm square-shaped microreservoirs and (b) their enlarged image patterned on NOA 63 polymeric substrate.

high resistance to organic solvents is especially advantageous when performing surface modification of the mold to immobilize various biomolecules because the surface modification process involves the use of organic solvents. Stacking and Transfer of the Polyelectrolyte Multilayer. The film thickness of 10.5 bilayers of PAA/PAAm was approximately 15.4 nm, when measured using a Gaertner ellipsometer operating at 632.8 nm. AFM measurements also revealed that the film thickness was approximately 19.8 nm. Both results were in reasonable agreement with the result reported by Yang et al.7

Lee et al.

Figure 2 shows the results of the polyelectrolyte multilayer stacking and subsequent transfer onto the microreservoir-patterned NOA 63 substrate. Multilayer transfer phenomenon was investigated by electrostatically attaching positively charged aminofunctionalized PS beads (∼400 nm) on the surface of the NOA 63 substrate and that of the mold after the transfer. Optical micrographs were obtained for the untreated NOA 63 substrate (Figure 2a) and multilayer-transferred NOA 63 substrate (Figure 2b). As can be seen, 10.5 bilayers of PAA/PAAm were successfully transferred onto the raised outer surrounding regions of the microreservoirs. The cracking and wrinkling observed in Figure 2b is due to the shrinking of the polyelectrolyte multilayer,2 which seems to be a common phenomenon when handling polyelectrolyte multilayer. This is because water molecules escape when PAA and PAAm are assembled into a film. Cracking and shrinking also occur because of the gel-like nature of PAAm. Nevertheless, the shrinkage of the multilayer seemed to have little effect on the function of the multilayer. Parts c and d of Figure 2 show the results of the bead attachment on the multilayertransferred NOA 63 substrate when the top layers of the transferred multilayer were PAAm and PAA, respectively. As expected, no beads were attached when the outermost layer was PAAm (Figure 2c), while the beads were attached selectively on the surfaces of the raised outer surrounding regions of the microreservoirs

Figure 2. Optical micrographs of (a) bare NOA 63 substrate, (b) (PAA/PAAm)10.5 multilayer-transferred NOA 63 substrate, (c) positively charged bead attachment on multilayer-transferred NOA 63 substrate when PAAm was the outermost layer, and (d) bead attachment after multilayer transfer when PAA was the outermost layer. Optical micrographs of the surface of the NOA 63-PEGDA composite mold coated with the beads (e) before and (f) after the multilayer transfer.

Hydrophilic Composite Mold for Multilayer Transfer

when the outermost layer was PAA (Figure 2d). Similar results were obtained for the multilayer transfer using both the dipping and spinning processes. In this way, it was revealed that PAA/ PAAm multilayers were selectively transferred onto the raised outer surrounding regions of the microreservoirs via the POPS technique. To further ensure that the whole multilayer was detached from the mold completely, amine-functionalized PS beads were attached on the surface of the composite mold before and after the transfer of 10.5 bilayers of PAA/PAAm, with PAA being stacked as both the starting and ending layers. Before the multilayer transfer, the whole surface of the composite mold was attached with beads with high homogeneity (Figure 2e). After the transfer, however, it was found that no beads were attached on the region of the composite mold which came into direct conformal contact with the raised regions on the NOA 63 substrate because multilayer was removed from the mold and that the beads were attached exclusively on the region of the composite mold which came into contact with the microreservoirs (Figure 2f) because multilayer was not removed from the mold. These results strongly suggest that the whole 10.5 bilayers of PAA/ PAAm were completely transferred onto the raised regions of the NOA 63 substrate. Although not shown, this phenomenon was observed over the entire area of the mold, signifying high multilayer transfer performance. The PAA layer, which was the first stacking layer, is physically adsorbed on the surface of the water-soaked composite mold by hydrophobic interaction. On the other hand, the electrostatic interaction dominates between the charged PAA and PAAm layers. Since the electrostatic interaction is much stronger than the hydrophobic interaction, the stacked multilayer was easily detached from the composite mold and transferred in its entirety onto the surface of the amine-terminated NOA 63 substrate. In this way, the binding of the protein on the multilayer-coated, raised outer surrounding regions of the microreservoirs could be prevented and the target biomolecules could be immobilized exclusively inside the microreservoirs. Unlike PDMS mold which poses a problem of sagging24,25 when performing POPS process due to its high elasticity, the possibility of mold sagging during the POPS process could be prevented with the composite mold (E ) 8.5∼9.1 MPa) because it is less elastic than PDMS (E ) 2.4∼4.5 MPa) at room temperature. Also, as we have already shown in previous reports4,18 that micro- and nanostructures can be successfully fabricated with NOA 63, we think that selective immobilization of biomolecules could be further extended to nanometer-scale reservoirs. Antibody Immobilization inside the Microreservoirs. Figure 3 shows the results of the antibody immobilization inside the microreservoirs brought about by the amide linkage between the carboxyl-terminated Fc region of the antibody and the amineterminated surface inside the microreservoirs on NOA 63 substrate, after transferring the multilayer onto the raised outer surrounding regions of the substrate. As can be seen in Figure 3, the FITC-tagged anti-rabbit IgG (Figure 3a) was selectively bound to the rabbit IgG, immobilized inside the microreservoirs, as indicated by the relatively high green fluorescence intensity inside the microreservoirs compared to that on the outer surrounding regions where the immobilization of the antibody was prevented by the transferred multilayer, (23) Ishisaka, A.; Kawagoe, M. J. Appl. Polym. Sci. 2004, 93, 560. (24) Xia, Y.; Whitesides, G. M. Angew. Chem., Int. Ed. 1998, 37, 550. (25) Rolland, J. P.; Hagberg, E. C.; Denison, G. M.; Carter, K. R.; De Simone, J. M. Angew. Chem., Int. Ed. 2004, 43, 5796.

Langmuir, Vol. 22, No. 18, 2006 7693

Figure 3. Fluorescent micrographs of the binding of (a) FITCtagged anti-rabbit IgG and (b) FITC-tagged anti-goat IgG when rabbit IgG was immobilized inside the microreservoirs and the multilayer was transferred onto the raised outer surrounding regions (PAA as the outermost layer). (c) Fluorescent micrographs of the binding of FITC-tagged anti-rabbit IgG when the multilayer was transferred onto the raised outer surrounding regions (PAAm as the outermost layer) and rabbit IgG was immobilized inside the microreservoirs.

signifying highly specific antibody reaction with its complementary anti-antibody. However, the FITC-tagged anti-goat IgG (Figure 3b) was not bound to the rabbit IgG, as indicated by the absence of fluorescence inside the microreservoirs. Although not significant, nonspecific binding of the antibody was also observed on the multilayer-coated, raised outer surrounding regions. This might have resulted from various interactions, such as the electrostatic, hydrogen, van der Waals, and hydrophobic interactions between the FITC-tagged antibody and the polyelectrolyte multilayer. Nevertheless, the difference in the fluorescence intensity inside the microreservoirs and on the outer surrounding regions was clearly demonstrated, thereby confirming the easiness and convenience of this technique for patterning physiologically vulnerable biomolecules, such as antibodies, selectively inside the microreservoirs. The sensitivity of the system is determined by the density and the orientation of the antibody immobilized inside the microreservoirs, and versatile strategies26-29 were employed to enhance the reaction between the antibody and a target molecule by controlling the orientation of antibody on the solid substrate. It is general that antibodies are immobilized on a substrate via the amide linkage between the amine-groups on the antibody and the carboxyl-terminated substrate. However, the Fab regions

7694 Langmuir, Vol. 22, No. 18, 2006

Lee et al.

To explore the applicability of the microreservoir-patterned NOA 63 substrate as a potential biosensor platform for detecting E. coli O157:H7, E. coli O157:H7 antibody was immobilized inside the microreservoirs and reacted with E. coli O157:H7 (ATCC 43894), expressing GFP. Fluorescence was measured at excitation wavelength range of 475∼490 nm because the optimum excitation wavelength at which the fluorescence of GFP becomes visible is 488 nm. As can be seen in Figure 4a, the green fluorescence was detected inside the microreservoirs. To confirm that the green fluorescence inside the microreservoirs is a genuine evidence for the presence of bound E. coli O157:H7, a second image (Figure 4b) was taken at the same spot, but at a different excitation wavelength range, 510∼560 nm, at which no fluorescence should be visible. As shown in Figure 4b, no red fluorescence was observed inside the microreservoirs. These results signify that the source of the green fluorescence in Figure 4a was truly GFP, and that the E. coli O157:H7 antibody has been properly immobilized inside the microreservoirs and reacted with the target, E. coli O157:H7, with distinguishable sensitivity.

Conclusions

Figure 4. Fluorescent micrographs showing bound E. coli O157:H7 inside the microreservoirs, obtained using filter blocks for (a) blue excitation (475∼490 nm) and (b) green excitation (510∼560 nm).

(Scheme 1h) of the antibody are also terminated with amine groups which could possibly bond with the carboxyl-terminated substrate, and this could limit the number of available binding sites for target molecules because Fab regions are involved in the antigen-antibody reaction. In this study, therefore, the carboxyl terminus which exists in the Fc region as a basic subunit of every antibody was activated via carbodiimide chemistry and adopted to immobilize the antibody on an amine-terminated solid support26 by forming amide linkage. In this way, many of the Fab regions were made available. To prevent antibody immobilization on the outer surrounding regions, we transferred polyelectrolyte multilayer onto NOA 63 substrate in such a way that PAA becomes the outermost layer. When the outermost layer of the transferred multiplayer was PAAm, the carboxyl-terminus of the rabbit IgG was immobilized on the entire surface of the microreservoirs and the whole surface was reacted with anti-rabbit IgG, displaying high fluorescence intensity (Figure 3c). However, when the outermost layer was PAA, the rabbit IgG was immobilized exclusively inside the microreservoirs (Figure 3a). (26) Choi, J.-W.; Chun, B. S.; Oh, B.-K.; Lee, W.; Lee, W. H. Colloids Surf., B 2005, 40, 173. (27) Starodub, N. F.; Pirogova, L. V.; Demchenko, A.; Nabok, A. V. Bioelectrochemistry 2005, 66, 111. (28) Peluso, P.; Wilson, D. S.; Do, D.; Tran, H.; Venkatasubbaiah, M.; Quincy, D.; Heidecker, B.; Poindexter, K.; Tolani, N.; Phelan, M.; Witte, K.; Jung, L. S.; Wagner, P.; Nock, S. Anal. Biochem. 2003, 312, 113. (29) Yakovleva, J.; Davidsson, R.; Lobanova, A.; Bengtsson, M.; Eremin, S.; Laurell, T.; Emne´us, J. Anal. Chem. 2002, 74, 2994.

We fabricated an elastic and hydrophilic composite mold which is ideal for transfer-printing polar inks, and we established a simple strategy for immobilizing antibody exclusively inside the microreservoirs by transferring hydrophilic polyelectrolyte multilayer using this composite mold. Polar multilayer film was selectively transfer-printed onto a microreservoir-patterned NOA 63 polymeric substrate with great simplicity, and antibodies were immobilized exclusively inside the microreservoirs where multilayer has not been transferred, making this system suitable as the basis of a biosensor. Due to enhanced hardness of the composite mold, submicrometer features could be patterned on its surface and any hydrophilic substances such as biomolecules, cells, and living organisms could also be microcontact-printed in nanoscale. Our novel composite mold is expected to have versatile patterning applications, including the patterning of biofunctionalized polymers with polyelectrolyte blocks and surface-directed assembly with bio-related functionalities, thus paving the way for the establishment of a simple methodology for the construction of a biosensor. Acknowledgment. This work was equally supported by the SRC program of the Korea Science and Engineering Foundation (KOSEF) through the Center for Intelligent Nano-Bio Materials at Ewha Womans University (Grant R11-2005-008-02003-0) and by the Nano/Bio Science & Technology Program (M10536090002-05N3609-00210) of the Ministry of Science and Technology (MOST) of Korea. The authors are supported by the Brain Korea 21 (BK 21) fellowship from the Ministry of Education of Korea. Supporting Information Available: Photos explaining the procedures for the fabrication of the composite mold, contact angle measurement data, and photos showing that phase separation has not taken place in the process of composite mold fabrication are provided. This material is available free of charge via Internet at http://pubs.acs.org. LA060305C