Anal. Chem. 2005, 77, 943-947
Surface Modification Method of Microchannels for Gas-Liquid Two-Phase Flow in Microchips Akihide Hibara,†,‡,§ Shinobu Iwayama,† Shinya Matsuoka,† Masaharu Ueno,†,‡,§ Yoshikuni Kikutani,‡,§ Manabu Tokeshi,‡,§ and Takehiko Kitamori*,†,‡,§
Department of Applied Chemistry, School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-8656, Japan, Integrated Chemistry Project, Kanagawa Academy of Science and Technology, 3-2-1 Sakado, Takatsu, Kawasaki, Kanagawa 213-0012, Japan, and PRESTO and CREST, Japan Science and Technology Agency, 4-1-8 Honcho Saitama, Japan
A capillarity restricted modification method for microchannel surfaces was developed for gas-liquid microchemical operations in microchips. In this method, a microstructure combining shallow and deep microchannels and the principle of capillarity were utilized for chemical modification of a restricted area of a microchannel. A hydrophobic-hydrophilic patterning in microchannels was prepared as an example for guiding gas and liquid flows along the respective microchannels. Validity of the patterning was confirmed by measuring aqueous flow leak pressure from the hydrophilic microchannel to the hydrophobic one. The leak pressure of 7.7-1.1 kPa agreed well with that predicted theoretically from the Young-Laplace equation for the microchannel depth of 8.6-39 µm. In an experiment to demonstrate usefulness and effectiveness of the method, an air bubble was first introduced into the hydrophilic microchannel and purged from the hydrophobic-hydrophilic patterned microchannels. Next, the patterning structure was applied to remove dissolved oxygen by contacting the aqueous flow with a nitrogen flow. The concentration of dissolved oxygen decreased with contact time, and its time course agreed well with numerical simulation. These demonstrations showed that the proposed patterning method can be used in general microfluidic gas-liquid operations. Microchemical systems have been investigated for a wide variety of applications in analysis, organic synthesis, and so on.1-8 * To whom correspondence should be addressed. Fax: +81-3-5841-6039. E-mail:
[email protected]. † The University of Tokyo. ‡ Kanagawa Academy of Science and Technology. § Japan Science and Technology Agency. (1) Proceedings of Micro Total Analysis Systems 2002; Baba, Y., Shoji, S., van den Berg, A., Eds.; Kluwer Academic Publishers: Dordrecht, 2002. (2) Proceedings of Micro Total Analysis Systems 2003; Northrup, M. A., Jensen, K. F., Harrison, D. J., Eds.; Transducer Research Foundation: San Diego, 2003. (3) Reyes, D. R.; Iossifidis, D.; Auroux, P. A.; Manz, A. Anal. Chem. 2002, 74, 2623-2636. (4) Auroux, P. A.; Iossifidis, D.; Reyes, D. R.; Manz, A. Anal. Chem. 2002, 74, 2637-2652. (5) Hansen, C.; Quake, S. R. Curr. Opin. Struct. Biol. 2003, 13, 538-544. (6) Fletcher, P. D. I.; Haswell, S. J.; Pombo-Villar, E.; Warrington, B. H.; Watts, P.; Wong, S. Y. F.; Zhang, X. L. Tetrahedron 2002, 58, 4735-4757. (7) Weigl, B. H.; Yager, P. Science 1999, 283, 346-367. 10.1021/ac0490088 CCC: $30.25 Published on Web 12/31/2004
© 2005 American Chemical Society
From among these applications, we developed the concepts of microunit operations (MUOs) and the continuous-flow chemical process to utilize the advantages and characteristics of microfluidic devices,9,10 and we designed complicated chemical systems in a microchip. To further promote the applicability of microchemical systems, MUOs concerned with gas-liquid flows are required and there have been a few reports on controlling or operating a gas-liquid flow.11-15 In ordinary macroscale experiments, gasliquid operations such as bubbling are commonly used. But gasliquid flow systems have difficulties in operating in a micrometersized restricted space because the phase transition from liquid to gas or from gas to liquid is accompanied by a very large volume change. In addition, gas-phase flow is compressive flow while liquid-phase flow can be treated as incompressible flow, and viscosity of gas flow is 2 or 3 orders lower than that of liquid flow. From another viewpoint, gas-liquid flow operations are indispensable for general microchemical systems. For example, air bubble contamination from connectors or reservoirs often occurs and it disturbs the stability of the systems. Therefore, microstructures for complete gas-liquid separation are necessary to prevent operation errors in microchemical systems. One of the most distinctive features of microchannels in the microchemical systems is the high specific interface between fluid-fluid and fluid-wall. By utilizing this feature, applications of parallel multiphase flows of immiscible solvents have been demonstrated.16-21 For example, we showed that microchannel (8) Kenis, P. J. A.; Ismagilov, R. F.; Takayama, S.; Whitesides, G. M. Acc. Chem. Res. 2000, 33, 841-847. (9) Tokeshi, M.; Minagawa, T.; Uchiyama, K.; Hibara, A.; Sato, K.; Hisamoto, H.; Kitamori, T. Anal. Chem. 2002, 74, 1565-1571. (10) Sato, K.; Hibara, A.; Tokeshi, M.; Hisamoto, H.; Kitamori, T. Adv. Drug Delivery Rev. 2003, 55, 379-391. (11) Tokeshi, M.; Kanda, K.; Hibara, A.; Kitamori, T. In Proceedings of Micro Total Analysis Systems 2002; Baba, Y., Shoji, S., van den Berg, A., Eds.; Kluwer Academic Publishers: Dordrecht, 2002; pp 356-358. (12) de Mas, N.; Gunther, A.; Schmidt, M. A.; Jensen, K. F. Ind. Eng. Chem. Res. 2003, 42, 698-710. (13) Guenther, A.; Khan, S. A.; Thalmann, M.; Trachsel, F.; Jensen, K. F. Lab Chip 2004, 4, 278-286. (14) Huh, D.; Tkaczyk, A. H.; Bahng, J. H.; Chang, Y.; Wei, H. H.; Grotberg, J. B.; Kim, C. J.; Kurabayashi, K.; Takayama, S. J. Am. Chem. Soc. 2003, 125, 14678-14679. (15) Gordillo, J. M.; Perez-Saborid, M.; Ganan-Calvo, A. M. J. Fluid Mech. 2001, 448, 23-51. (16) Hibara, A.; Tokeshi, M.; Uchiyama, K.; Hisamoto, H.; Kitamori, T. Anal. Sci. 2001, 17, 89-93.
Analytical Chemistry, Vol. 77, No. 3, February 1, 2005 943
surface modification patterning stabilizes the multiphase flow very well.19-21 In these example applications, modified and nonmodified plates were sealed only by pressure and the multiphase flow was formed in a top-and-down configuration, not in a side-by-side configuration. Although the pressure-sealed method is very effective for spectroscopic measurement of interfaces in multiphase flow, it is not a conclusive method for general control of multiphase flow. A photolithographic method for patterning of chemical modification was proposed,22 but it does not seem applicable to general gas-liquid operations. Thus, a simple surface patterning method, which should be easily applied to gas-liquid operations, is desirable. In this paper, we developed a capillarity restricted modification (CARM) method for patterning of the microchannel surface modification. In the CARM method, only microchannel depth needs to be controlled and only capillarity is utilized for modification patterning. Furthermore, the depth control provides favorable features for gas-liquid operations because Laplace pressure, which is the principle for operations, is enhanced by contrast of the channel depth. With the CARM method, we realized hydrophobic/hydrophilic patterning in microchemical systems and successfully demonstrated bubble separation in the microchannel. Furthermore, the patterning structure was applied to get stable contact of gas-liquid flow and realize dissolved gas control; this was demonstrated through removal of dissolved oxygen in pure water by nitrogen gas flow. EXPERIMENTAL SECTION Fabrication. Microchannels having an asymmetric cross section were fabricated by a two-step photolithographic wetetching technique. The procedure is shown in Figure 1. Mechanically polished 0.7-mm-thick Pyrex glass plates were used (top and bottom plates). After inlet and outlet holes were drilled by ultrasonic sandblasting on the top plate, both top and bottom plates were annealed at 570 °C for 5 h. For good contact between the substrates and the photoresist and protection of the substrates during glass etching, 20-nm-thick Cr and 100-nm-thick Au layers were sputtered on the bottom plate. A 2-µm-thick positive photoresist was spin-coated on the Au metal layer and baked at 90 °C for 30 min. First, UV light was shone through a photomask to transfer only the deep microchannel pattern onto the photoresist. The photoresist was developed, and a pattern with 10-µm-wide lines was obtained. The Au and Cr layers were etched with I2/ NH4I and Ce(NH4)2(NO3)6 solutions. The bare glass surface with this deep microchannel pattern was etched with a 50% HF solution at an etching rate of 13 µm/min. Next, UV light was shone again through another photomask to transfer the shallow microchannel pattern. The photoresist was developed, and the Au and Cr layers and the glass were etched in the same way. Then, the remaining (17) Hisamoto, H.; Horiuchi, T.; Uchiyama, K.; Tokeshi, M.; Hibara, A.; Kitamori, T. Anal. Chem. 2001, 73, 5551-5556. (18) Surmeian, M.; Slyadnev, M. N.; Hisamoto, H.; Hibara, A.; Uchiyama, K.; Kitamori, T. Anal. Chem. 2002, 74, 2014-2020. (19) Hibara, A.; Nonaka, M.; Hisamoto, H.; Uchiyama, K.; Kikutani, Y.; Tokeshi, M.; Kitamori, T. Anal. Chem. 2002, 74, 1724-1728. (20) Hibara, A.; Nonaka, M.; Tokeshi, M.; Kitamori, T. J. Am. Chem. Soc. 2003, 125, 14954-14955. (21) Aota, A.; Nonaka, M.; Hibara, A.; Kitamori, T. In Proceedings of Micro Total Analysis Systems 2003; Northrup, M. A., Jensen, K. F., Harrison, D. J., Eds.; Transducer Research Foundation: San Diego, 2003; pp 441-444. (22) Zhao, B.; Moore, J. S.; Beebe, D. J. Science 2001, 291, 1023-1026.
944 Analytical Chemistry, Vol. 77, No. 3, February 1, 2005
Figure 1. Fabrication scheme of two-step photolithographic wetetching technique. (a) A Pyrex glass substrate was coated with Cr, Au, and photoresist layers. The deep microchannel pattern was transferred by irradiating UV light through a photomask. (b) The deep microchannel pattern was etched by HF solution after development of the resist, Au, and Cr layers. (c) The shallow microchannel pattern was transferred by irradiating UV light through a photomask. (d) Both the deep and shallow microchannel patterns were etched with HF solution. (e) After stripping the resist, Au, and Cr layers, the fabricated substrate was thermally bonded with a cover glass plate.
photoresist was removed in acetone and metals were removed in I2/NH4I and Ce(NH4)2(NO3)6 solutions. The fabricated microchannels were covered with another substrate by a thermal bonding method. Chemicals. Octadecyltrichlorosilane (ODS) was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Toluene, sodium hydroxide, and methylene blue were purchased from Kanto Kagaku Co., Ltd. (Tokyo, Japan). Sulfuric acid, potassium sodium tartrate tetrahydrate, and ammonium iron(II) sulfate hexahydrate were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). All were used without further purification. The distilled and deionized water used had resistivity values of more than 1.7 × 107 Ω‚cm at 25 °C. Concentrations of dissolved O2 in water were determined based on the protocol of the Miller method23 as modified by the Japanese Industrial Standards Committee.24 A 350-g sample of sodium potassium tartrate tetrahydrate and 100 g of sodium hydroxide were dissolved in 1 L of pure water (tartrate solution). A 0.1-g sample of methylene blue was dissolved in 100 mL of pure water (methylene blue solution). A 5-mL sample of sulfuric acid was added to 100 mL of pure water and then 5.4 g of sodium potassium tartrate tetrahydrate was dissolved and diluted with pure water to 1 L. Furthermore, a 100mL portion of this solution was diluted to 1 L (titration solution). A 4-mL aliquot of sample water was collected in a test tube, which was capped with liquid paraffin to avoid contact between the sample and air. The sample water was diluted with 16 mL of pure (23) Miller, J. J. Soc. Chem. Ind. 1914, 33, 185-186. (24) Testing methods for industrial water JIS-K0101; Japan Standard Association: Tokyo, revised 1998.
Figure 2. Modification procedures by CARM method and corresponding photographs. (a) The shallow and deep microchannels have separate inlet holes and contact points in the microchip. (b) A solution containing modification compounds was introduced from the inlet of the shallow channel by capillarity. (c) The solution was not leaked to the deep microchannel and only the shallow microchannel was modified. (d) The solution was pushed away with air pressure from the deep microchannel. The red lines indicate the modified surface. (e) A sectional illustration along the s-s’ dashed line in (d). A movie file is available as Supporting Information.
water. Next, a drop of the methylene blue solution and 2 mL of the tartrate solution were added to the sample. Then, the dissolved oxygen was titrated until the blue color died away. Apparatus. Flows in microchannels were observed by a phasecontrast microscope (Eclipse TS100, Nikon). Photos and movies were taken by a CCD camera (ExwaveHAD, Sony). RESULTS AND DISCUSSION The main idea of the CARM method is to introduce the modification solution to a restricted part of a microchannel by utilizing physical shape and capillarity. The procedures are shown in Figure 2a-c. A portion of an ODS/toluene solution (1 wt %) was dropped onto the inlet hole of the shallow channel, and the solution was spontaneously drawn into this channel by capillary action. The solution was stopped at the boundary between the shallow and deep channels by the balance between the solidliquid and gas-liquid interfacial energies. Therefore, the solution did not enter the deep channel. It remained at the boundary for several minutes and then was pushed from the deep channel side by air pressure. This procedure ensured only the shallow microchannel was modified with ODS groups and the deep channel
remained as a bare glass surface. Figure 2e shows a cross section across the s-s′ dashed line in Figure 2d. Contact angle θ of water on ODS-modified glass was measured as 115°. To test the hydrophobic-hydrophilic patterning, aqueous flow leakage from the deep main microchannel to the shallow side microchannel filled with air was measured as a function of depth of the side channel, d. Glass microchips having the side channel depth d of 8.6-39 µm were prepared. The results are presented in Figure 3, where the obtained flow rate of the leakage (in µL/ min units) was converted to pressure difference by utilizing the Hagen-Poiseuille relationship. To verify the obtained leak pressure, we calculated the theoretical values by considering balance between hydrodynamic and Laplace pressures. In this calculation, the Young-Laplace equation
∆P )
γ 2γ sin(θ - 90°) ) R d
was used, where γ is interfacial tension and R is radius of liquid curvature. The literature value of 72 mN/m was used for γ.25 The solid line in Figure 3 shows theoretical values. The experimental Analytical Chemistry, Vol. 77, No. 3, February 1, 2005
945
Figure 3. (a) Illustration of leak pressure for aqueous flow. (b) Leak pressure for aqueous flow from the deep hydrophilic microchannel to the shallow one. The solid line indicates the theoretical prediction from the Young-Laplace equation.
and theoretical values agreed well with each other. This meant that partial modification of the microchannel was successfully carried out and that the hydrophobic-hydrophilic patterning for gas-liquid operations functioned as intended. To demonstrate patterning effectiveness by the CARM method, we designed a bubble introduction and removal chip as shown in Figure 4a. The red line (Figure 4b) indicates the shallow channel, which had a width of 170 µm and a depth of 10 µm and was modified with ODS groups. Contact length between the deep main channel having a width of 200 µm and depth of 100 µm, and the side channels was 30 mm. Pure water was introduced from inlet 1, and air was introduced from inlet 2. At the junction within the rectangular box c, air bubbles were introduced, utilizing the shallow and narrow connection structure (Figure 4b). The flow rates of pure water and air were, respectively, 10 and 10 µL/min. Almost all air bubbles were removed from the main channel within the contact length of 2.5 mm (Figure 4c). At 3.5 mm downstream from the beginning of contact, no air bubbles were observed (Figure 4d). A movie of the air bubble introduction and removal is available in Supporting Information. In actual use of microchannel systems, bubbles are often introduced from a connector part or generated in the liquid phase by a chemical reaction or temperature increase. System errors (25) CRC Handbook of Chemistry and Physics, 83th ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 2002.
946 Analytical Chemistry, Vol. 77, No. 3, February 1, 2005
Figure 4. (a) Concept of air bubble introduction and purge system in microchannels. (b) Design of air bubble introduction and removal experiment. (c-d′) Optical microscope images of bubble introduction and removal experiment where the rectangular areas denoted as c and d in (b) correspond to the images of (c) and (d), (d′), respectively. A movie file is available as Supporting Information.
due to air bubble introduction or generation can be avoided with hydrophobic-hydrophilic patterning by the CARM method. As an example of a dissolved gas control process, we demonstrated dissolved oxygen removal. In ordinary macroscale experiments, bubbling of nitrogen or argon is commonly performed. As an effective way to remove dissolved oxygen in microchannels, we designed a continuous countercurrent contact channel (Figure 5a and 5b). The red line indicates the shallow
The oxygen concentration in the collected water is plotted in Figure 5c as a function of gas-liquid contact time in the microchannel. The solid line indicates simulated mean concentration of dissolved O2 based on Fick’s law by a finite element method, where the concentration at the gas-liquid interface was set to zero as a boundary condition. The theoretical and experimental values correlated well with each other, which indicated that transport phenomena in the gas-liquid contact structure were controlled only by diffusion of the gas component in the liquid phase. Although only dissolved O2 removal was demonstrated, the hydrophobic-hydrophilic patterning can be applied to general dissolved gas control including purification of gas flow usually performed in a laboratory.
Figure 5. (a) Design of dissolved gas control experiment. (b) Layout of microchannels. (c) Dissolved O2 concentration as a function of gas-liquid contact time. The solid line indicates numerically simulated concentration based on Fick’s law.
channel, which had a width of 170 µm and a depth of 10 µm and was modified with ODS groups. Contact length between the deep main channel having a width of 200 µm and depth of 100 µm and the side channels was 34 mm. Pure water, which had been saturated with air and contained 8.8 ppm O2,25 was introduced into inlet hole 1 and the flow from outlet hole 3 was collected. The flow rates of pure water were set to 2, 10, and 40 µL/min, which corresponded to gas-liquid contact times of 19, 3.8, and 0.94 s, respectively. Nitrogen gas was introduced from inlet 4 with a fixed flow rate of 4 µL/min to form a countercurrent gas-liquid flow.
CONCLUSIONS We developed the CARM method for controlling gas-liquid operations in microchemical systems. The CARM method is a simple and reproducible method for surface modification of a restricted area; it needs only channel depth control in the fabrication process, and it utilizes only the principle of capillarity. We succeeded in microchannel modification of the shallow microchannel while the deep microchannel remained hydrophilic. The validity of this hydrophobic and hydrophilic patterning was confirmed by analyzing depth dependence of the leak pressure. Next, we demonstrated air bubble introduction and removal from aqueous flow in the deep channel to the shallow channel. We also demonstrated dissolved gas control by utilizing the hydrophobichydrophilic patterning. This method will be applicable to various operations, such as bubble removal in cell cultivation chips, adjustment of pH by bringing a buffer solution into contact with a CO2 gas flow, chemical reactions inducing gas generation in microchannels, a bubbling mixer, etc. Of course, the CARM method will be also effective for controlling micromultiphase flows of immiscible solvents, which are frequently used in microchemical systems. ACKNOWLEDGMENT This research was partially supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan. SUPPORTING INFORMATION AVAILABLE Movies of CARM method and of air bubble introduction and removal. These materials are available free of charge via the Internet at http://pubs.acs.org.
Received for review July 7, 2004. Accepted November 2, 2004. AC0490088
Analytical Chemistry, Vol. 77, No. 3, February 1, 2005
947
