Anal. Chem. 2005, 77, 6494-6499
Passive Microfluidic Control of Two Merging Streams by Capillarity and Relative Flow Resistance Sung-Jin Kim,* Yong Taik Lim,† Haesik Yang,‡ Yong Beom Shin,† Kyuwon Kim,§ Dae-Sik Lee, Se Ho Park, and Youn Tae Kim
Biosensor Group, Electronics and Telecommunications Research Institute (ETRI), Daejeon 305-350, Korea
In the progress of microfluidic devices, a simple and precise control of multiple streams has been essential for complex microfluidic networks. Consequently, microfluidic devices, which have a simple structure, typically use external energy sources to control the multiple streams. Here, we propose a pure passive scheme that uses capillarity without using external force or external regulation to control the merging of two streams and even to regulate their volumetric flow rate (VFR). We accomplish this process by controlling the geometry of two inlets and a junction, and by regulating the hydrophilicity of a substrate. Additionally, we use the relative flow resistance to control the VFR ratio of the merged two streams. Our results will significantly simplify the control of multiple streams without sacrificing precision. Controlling multiple streams is crucial in microfluidic devices, especially for multiple and parallel processing of biochemical reactions.1-4 The devices require the control of merging multiple streams and even the regulation of each stream’s volumetric flow rate (VFR), which is the transported volume of the stream per unit of time, and the VFR ratio between the merged streams. Such a process generally relies on active control by mechanical pressure,5-7 electroosmotic force,8 electrowetting,9,10 and electrochemical reaction.11 These active manipulations enable the streams * To whom correspondence should be addressed. E-mail: yahokim@ etri.re.kr. † BioNanotechnology Research Center, Korea Research Institute of Bioscience and Biotechnology, Daejeon 305-333, Korea. ‡ Department of Chemistry, Pusan National University, Busan 609-735, Korea. § Division of Chemical Metrology and Material Evaluation, Korea Research Institute of Standards and Science, Daejeon 305-600, Korea. (1) Weigl, B. H.; Yager, P. Science 1999, 283, 346-347. (2) Burns, M. A.; Johnson, B. N.; Brahmasandra, S. N.; Handique, K.; Webster, J. R.; Krishnan, M.; Sammarco, T. S.; Man, P. M.; Jones, D.; Heldsinger, D.; et al. Science 1998, 282, 484-487. (3) Jiang, X.; Ng, J. M. K.; Stroock, A. D.; Dertinger, S. K. W.; Whitesides, G. M. J. Am. Chem. Soc. 2003, 125, 5294-5295. (4) Takayama, S.; Ostuni, E.; LeDuc, P.; Naruse, K.; Ingber, D. E.; Whitesides, G. M. Nature 2001, 411, 1016. (5) Thorsen, T.; Maerkl, S. J.; Quake, S. R. Science 2002, 298, 580-584. (6) Linder, V.; Sia, S. K.; Whitesides, G. M. Anal. Chem. 2005, 77, 64-71. (7) Marmottant, P.; Hilgenfeldt, S. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 9523-9527. (8) Harrison, D. J.; Fluri, K.; Seiler, K.; Fan, Z.; Effenhauser, C. S.; Manz, A. Science 1993, 261, 895-897. (9) Cho, S. K.; Moon, H.; Kim, C.-J. J. Microelectromech. Syst. 2003, 12, 7080.
6494 Analytical Chemistry, Vol. 77, No. 19, October 1, 2005
to be controlled in a rapid and precise manner. These manipulations, however, need accurate external energy sources, such as a mechanical pump or power supply, a function generator, and a voltage amplifier. As a result, the manipulations increase the size and cost of the external control unit of microfluidic devices. The number of required energy sources also increases with the increasing number of merging processes for multiple or parallel flow control,5 thereby making the external control unit more complex. Accordingly, active control does not seem to be suitable for portable or miniaturized microfluidic devices such as pointof-care testing devices.12 Capillarity, which is the spontaneous motion of liquids due to their surface tension, needs no external energy sources. Thus, when we use capillarity, we only need a very small and simple external control unit or none at all. However, sophisticated regulation of the streams by only capillarity is nontrivial. To date, several studies have been reported on the use of capillarity to manipulate the streams. These studies, however, dealt only with basic cases, such as the initial driving or stopping of a single stream2,13-17 and the transition from a single stream to divided multiple streams.17,18 In these studies, external energy sources were also used to merge two2,14 or more15,17 streams, to regain the flow of the stopped streams,2,14-17 and to regulate the VFR.2,15 Here, we report on a pure passive scheme for controlling the merging process and the VFR. Without any external energy sources, these processes are controlled by capillarity. In addition, we use the relative flow resistance to regulate the VFR ratio. To control the flow, we use the following four key elements (Figure (10) 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. (11) Gallardo, B. S.; Gupta, V. K.; Eagerton, F. D.; Jong, L. I.; Craig, V. S.; Shah, R. R.; Abbott, N. L. Science 1999, 283, 57-60. (12) Mitchell, P. Nat. Biotechnol. 2001, 19, 717-721. (13) Delamarche, E.; Bernard, A.; Schmid, H.; Michel, B.; Biebuyck, H. Science 1997, 276, 779-781. (14) Melin, J.; Roxhed, N.; Gimenez, G.; Griss, P.; van der Wijngaart, W.; Stemme, G., Sens. Actuators, B 2004, 100, 463-468. (15) Zoval, J. V.; Madou, M. J. Proc. IEEE 2004, 92, 140-153. (16) Man, P. F.; Mastrangelo, C. H.; Burns, M. A.; Burke, D. T. Microfabricated Capillary-Driven Stop Valve And Sample Injector. In Proceedings, IEEE Conference MEMS; Sandmaier, H., Stemme, G., Eds.; IEEE: Heidelberg, Germany, 1998; pp 45-50. (17) McNeely, M. R.; Spute, M. K.; Tusneem, N. A.; Oliphant, A. R. Proc. SPIE 1999, 3877, 210-220. (18) Juncker, D.; Schmid, H.; Drechsler, U.; Wolf, H.; Wolf, M.; Michel, B.; de Rooij, N.; Delamarche, E. Anal. Chem. 2002, 74, 6139-6144. 10.1021/ac0504417 CCC: $30.25
© 2005 American Chemical Society Published on Web 08/24/2005
Figure 1. Schematic of the microfluidic device for the control of two merging streams. This device consists of four key elements: pressure balancer (width 140 µm), aspect ratio (w/h), substrate that presents the varying hydrophilicity, and the relative flow resistance (L2/(L1 + L2)); L1 and L2 are the length of the 200- and 60-µm-width channels, respectively. L1 and L2 are variables, and L1 + L2 is a constant. The hydrophilicity is regulated on the PET substrate. This substrate is selectively coated with the different mixtures of PSA and surfactants. Four walls of the channel are composed of COC, SU-8 3050, and PET.
1): (1) a pressure balancer between two inlets, which makes merged two streams form a parallel laminar flow; (2) the aspect ratio (w/h), which is the width-to-height ratio of the stream at the junction. This ratio determines the stopping or passing of the streams at the junction; (3) a substrate that presents the varying hydrophilicity, which regulates the VFR before and after the merging. The substrate is selectively coated in three zones with different mixtures of pressure-sensitive adhesive and surfactants, and the hydrophilicity of each zone depends on the concentration of surfactants; (4) the relative flow resistance, which controls the VFR ratio between the merged streams. The first three elements pertain to the control of the capillarity. The last element helps regulate the flow resistance. MATERIALS AND METHODS Device Fabrication. To fabricate the microfluidic device, we used standard photolithography and selective spin-coating. Figure 2a shows the microfabrication process of the upper plate. The starting substrate was cycloolefin copolymer (COC) wafer with 1.1-mm thickness, which was made by the injection molding process. Then the 80-µm-thick SU-8 (SU-8 3050, MicroChem, www.microchem.com) was photolithographically patterned as a channel. Inlet and outlet holes were drilled through the upper plate prior to sealing. Figure 2b shows the microfabrication process of the lower plate, consisting of three coating processes. The coating materials were the different mixtures of pressure-
sensitive adhesive (PSA) and surfactants. The PSA (AT4827, Samwon, www.samvinol.co.kr) was used to bond an upper and lower plate, and the surfactants were used to regulate the degree of hydrophilicity. The starting substrate was 50-µm-thick poly(ethylene terephthalate) (PET) film. The first step was the spincoating of the 10-µm-thick mixture of PSA and surfactant B (L31, BASF, www.basf.com). We fixed the concentration of surfactant B as 7% (w/w). The second and last steps were the selective spincoating of the 5-µm-thick mixture of PSA and surfactant A (L77, GE Silicones, www.gesilicones.com) using shadow masks. The shadow masks were manually attached to define zone A and zone C in each coating process; after each coating process, they were manually removed. The shadow mask was polymer film (Crystalex, GMP, www.gmp.com), which is used as a protection cover in commercial PSA-coated film. During each coating step, the coated mixtures were dried in an oven (100 °C, 10 min). Finally, the upper and lower plates were bonded by pressure at room temperature, as shown in Figure 2c. Capillary Pressure. We measured the stop pressure of the working liquid at the junction by the hydrostatic pressure (height gage No. 192-130, Mitutoyo, www.mitutoyo.com). The hydrostatic pressure (∆Ph) was applied by gradually raising the syringe height according to ∆Ph ) Fgz, where F is the density of the working liquid, g is gravity, and z is the syringe height. As the applied pressure was increased past a critical point, the liquid abruptly passed through the junction. This value of the hydrostatic pressure is equal to the stop pressure. This measurement approach had a precision of ∼0.1-mm syringe height, corresponding to 1 Pa in pressure. The motion of the liquids was monitored using conventional microscopy with a 5× objective lens and a digital CCD camera. Contact Angle and Surface Tension. Our working liquid was water containing 3.6% (by weight) food dyes (red and blue colors, Innobiosystem, www.innobiosystem.com). In the experiment, we used a commercial measurement system (Pheonix300, SEO, www.s-eo.com). The setup used to measure the contact angle and surface tension consisted of micromanipulator, syringe, CCD camera, and video recorder. All experiments were done at room temperature, 23 °C. In the measurement of contact angle, we used sessile drop method. The micromanipulator was used to adjust the droplet volume of the working fluid. We selected the droplet volume as 5 µL to minimize the gravity effect. We regulated the gap between syringe tip and the substrate to minimize impact effect when the droplet was released. The used substrates were COC, SU-8 3050, and surfactant-coated PET film. We changed the concentration of surfactants (1-10% (w/w)). The images were recorded with a time interval of 20 s. In the measurement of surface tension, we used the Bashforth-Adams technique from the image of the droplet suspended from the syringe tip. The surface tension of the working liquid was measured as 70.5 ( 1.5 mN/m. Volumetric Flow Rate and VFR Ratio. To measure VFR, we recorded the flow of the merged stream in zone C (Figure 1), through an optical microscope and a CCD camera. These files were saved as digital movie files, which have a time resolution of 1/30 s (Pinnacle Studio, Pinnacle Systems, www.pinnaclesys.com). The channel dimension in zone C is 80 µm high, 480 µm wide, and ∼5 mm long. From the information regarding the position of Analytical Chemistry, Vol. 77, No. 19, October 1, 2005
6495
Figure 2. Fabrication process of the microfluidic device. (a) Fabrication of upper plate. The starting substrate is COC wafer, which is made by the injection molding process. (b) Fabrication of lower plate. The first step is the spin-coating of the 10-µm-thick mixture of PSA and surfactant B; the concentration of surfactant B is fixed as 7% (w/w). The second and last steps are the selective spin-coating of the 5-µm-thick mixture of PSA and surfactant A using shadow masks. In zone A, we fix the concentration of surfactant A as 2% (w/w). The PSA is used to bond an upper and lower plate, and the surfactants are used to regulate the degree of hydrophilicity. (c) Bonding of the two plates at room temperature.
Figure 3. Dependence of flow patterns on the existence of pressure balancer (width 140 µm). Here, w/h ) 2.5 and h ) 80 µm, as shown in Figure 1. Scale bars, 200 µm. (a-c) Without a pressure balancer, the reentrant flows occur, resulting in disturbances of volumetric flow rates ratio in each stream. (d, e) With a pressure balancer, the merged two streams pass through the junction equally, resulting in a parallel laminar flow.
the moving meniscus of the merged stream, we calculated the VFR or the average velocity. To measure VFR ratio, we used recorded images. From recorded images, we measured the occupied width of the two streams. Due to the parabolic velocity profile, the measured widths of each stream were converted into the VFR ratio using analytical equations.19,26 (19) Constantinescu, V. N. Laminar Viscous Flow. Springer-Verlag: New York, 1995; pp 121-124.
6496 Analytical Chemistry, Vol. 77, No. 19, October 1, 2005
RESULTS AND DISCUSSION Pressure Balancer between Two Inlets. We used pipets to inject liquids into the two inlets. Our working liquids were 2.5-µL (20) Handique, K.; Burke, D. T.; Mastrangelo, C. H.; Burns, M. A. Anal. Chem. 2000, 72, 4100-4109. (21) de Lazzer, A.; Langbein, D.; Dreyer, M.; Rath, H. J. Microgravity Sci. Technol. 1996, 9, 208-219. (22) Concus, P.; Finn, R. Acta Math. 1974, 132, 177-198. (23) Kim, E.; Whitesides, G. M. J. Phys. Chem. B. 1997, 101, 855-863.
volume of water containing 3.6% (w/w) food dyes. After we injected the liquids, the streams filled each channel and reached the junction by the capillarity. Without a pressure balancer between the two inlets (compare Figure 3a and d); however, the reentrant flows to the other channel are observed, as shown in Figure 3b and c. Figure 3b shows the stopping stream reentering the incoming stream, and Figure 3c shows the incoming stream reentering the stopping stream. These random motions make it impossible to control the VFR ratio at the junction. The reentrant flows show an imbalance of pressure between the two inlets. To examine this variation, we considered the Young-Laplace equation ∆PYL ) σ(r1-1 + r2-1), where ∆PYL is the pressure inside the droplet, σ is the surface tension, and r1 and r2 are the principal radii of curvature. From this equation, we can say that the geometry of the inlet and the volume of the injected liquids determine the pressure in each of the two inlets. Since we injected two equal volumes of 2.5 µL (imprecision e0.7%) into the two inlets by pipets (Research, Eppendorf, www.eppendorf.com), we deduced that the pressure depends solely on the geometry of the inlet. The inlet hole diameter varied from 1.4 to 1.5 mm due to the manual drilling process. Using the dimensions (Figure 2a), we calculated the inlet pressures20 to be in the range from 11.4 to 26.4 Pa. Thus, if we assume an extreme case, a variation of 15 Pa occurs between the two inlets. (Note that we calculated the variation of pressure loss due to flow resistance to be
