Anal. Chem. 2008, 80, 6206–6213
Microprocessing of Liquid Plugs for Bio/chemical Analyses Fumihiro Sassa, Junji Fukuda, and Hiroaki Suzuki* Graduate School of Pure and Applied Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8573, Japan A microfluidic device and operation to handle liquid plugs for biochemical analyses were developed for efficient handling of plugs of many solutions. A major part of the device was a T-junction consisting of a main flow channel and a handling flow channel. Unit operations including attachment of plugs, division of a plug, sorting of plugs, and formation of plugs of various lengths enabled controlled sequential reactions in a microflow channel. Rapid mixing could easily be achieved by moving a plug formed by merging two plugs back and forth. The device could be used for efficient characterization of performance in bio/chemical sensing. In experiments using L-glutamate oxidase, plugs containing an enzyme or a substrate were formed, mixed sequentially, and the intensity of fluorescence from plugs of different concentrations of L-glutamate or pHs could be measured simultaneously. Cross-contamination of plugs by neighboring plugs poses a problem in using the same flow channel repeatedly. However, the influence could be minimized by using a cleansing plug placed between them in a sufficiently hydrophobic flow channel and by processing the plugs at a low velocity. The device can be a critical component for microprocessing in various bio/chemical analyses. The advance of micro total analysis system (µTAS) technology has already produced a number of highly advanced systems with high functionalities.1–6 With the recent on-chip microfluidic technology, it is getting more realistic to conduct all necessary procedures on a single chip. Although previous microsystems have been designed for a particular purpose such as the analysis of enzyme activity, the detection of molecules such as DNA and proteins, and chemical synthesis,7,8 it will be more timesaving if the same system can be used for general purposes * To whom correspondence should be addressed. Phone: +81-29-853-5598. Fax: +81-29-853-4490. E-mail:
[email protected]. (1) Erickson, D.; Li, D. Anal. Chim. Acta 2004, 507, 11–26. (2) Delamarche, E.; Juncker, D.; Schmid, H. Adv. Mater. 2005, 17, 2911– 2933. (3) Lee, C.-C.; Sui, G.; Elizarov, A.; Shu, C. J.; Shin, Y.-S.; Dooley, A. N.; Huang, J.; Daridon, A.; Wyatt, P.; Stout, D.; Kolb, H. C.; Witte, O. N.; Satyamurthy, N.; Heath, J. R.; Phelps, M. E.; Quake, S. R.; Tseng, H.-R. Science 2005, 1793–1796. (4) El-Ali, J.; Sorger, P. K.; Jensen, K. F. Nature 2006, 42, 403–411. (5) Song, H.; Chen, D. L.; Ismagilov, R. F. Angew. Chem., Int. Ed. 2006, 45, 7336–7356. (6) Weibel, D. B.; Whitesides, G. M. Curr. Opin. Chem. Biol. 2006, 10, 584– 591.
6206
Analytical Chemistry, Vol. 80, No. 16, August 15, 2008
many times and the system is operated by programming as seen in electronic microprocessors. In previous microfluidic systems, continuous liquids have mostly been used as seen in systems based on electroosmosis and pressure-driving. However, handling of solutions using this approach is accompanied by non-negligible dead volume. This is particularly serious when the device is connected to external pumps such as microsyringe pumps using connecting tubes. The existence of dead volume nullifies the advantage expected from µTAS. Continuous liquids also make handling of many solutions difficult. In contrast, liquid plugs enable very efficient handling of solutions and promote large-scale integration.5,9,10 Unlike continuous fluids, only necessary amounts of solutions can be used, which is particularly advantageous in the application to chemical analysis. Moreover, the method is suitable to handle many solutions in a limited space.9 A row of many plugs of different solutions can be mobilized in a flow channel simultaneously by applying a pressure from one end of the flow channel. Furthermore, mixing of solutions in a plug is rapid. The excellent properties have been demonstrated in screening of protein crystallization conditions,11 synthesis of nanoparticles,12–14 screening of solubility diagrams,15 analysis of DNA,16,17 application to immunoassays,18 and analysis of chemical reactions.19 In addition to practical applications, the expectations for high-throughput processing have stimulated research in constructing plug-based devices that simulate electronic digital microprocessing.20–23 (7) Satoh, W.; Hosono, H.; Suzuki, H. Anal. Chem. 2005, 77, 6857–6863. (8) Nashida, N.; Satoh, W.; Fukuda, J.; Suzuki, H. Biosens. Bioelectron. 2007, 22, 3167–3173. (9) Thorsen, T.; Maerkl, S. J.; Quake, S. R. Science 2002, 298, 580–584. (10) Shiu, J.-Y.; Chen, P. Adv. Mater. 2005, 17, 1866–1869. (11) Zheng, B.; Roach, L. S.; Ismagilov, R. F. J. Am. Chem. Soc. 2003, 125, 11170–11171. (12) Gu ¨ nther, A.; Kahn, S. A.; Thalmann, M.; Trachsel, F.; Jensen, K. F. Lab Chip 2004, 4, 278–286. (13) Shestopalov, I.; Tice, J. D.; Ismagilov, R. F. Lab Chip 2004, 4, 316–321. (14) Hung, L.-H.; Choi, K. M.; Tseng, W.-Y.; Tan, Y.-C.; Shea, K. J.; Lee, A. P. Lab Chip 2006, 6, 174–178. (15) Laval, P.; Lisai, N.; Salmon, J.-B.; Joanicot, M. Lab Chip 2007, 7, 829–834. (16) Obeid, P. J.; Christopoulos, T. K.; Crabtree, H. J.; Backhouse, C. J. Anal. Chem. 2003, 75, 288–295. (17) Srisa-Art, M.; deMello, A. J.; Edel, J. B. Anal. Chem. 2007, 79, 6682–6689. (18) Linder, V.; Sia, S. K.; Whitesides, G. M. Anal. Chem. 2005, 77, 64–71. (19) Song, H.; Tice, J. D.; Ismagilov, R. F. Angew. Chem., Int. Ed. 2003, 42, 768–772. (20) Fuerstman, M. J.; Garstecki, P.; Whitesides, G. M. Science 2007, 315, 828– 832. (21) Prakash, M.; Gershenfeld, N. Science 2007, 315, 832–835. (22) Cheow, L. F.; Yobas, L.; Kwong, D.-L. Appl. Phys. Lett. 2007, 90, 054107. 10.1021/ac800492v CCC: $40.75 2008 American Chemical Society Published on Web 07/16/2008
In this study, we have fabricated a microfluidic device by taking advantage of the excellent features of liquid plugs to realize a liquid-handling device for general purposes. The situation encountered with our case is unique. Previous plug-based devices have been developed for the handling of a lot of plugs of relatively few components. In contrast, our device handles a relatively small number of plugs of many components. To this end, we used a simple T-junction. Although a straight flow channel alone cannot carry out complicated handling of solutions, adding an auxiliary flow channel makes such handling possible. There are several useful procedures for the handling of plugs. These include mixing, merging, separation, and exchange. We will show that the threeterminal device can be used as a microfluidic processor for biochemical analyses. EXPERIMENTAL SECTION Reagents and Materials. Reagents and materials used for the fabrication of the device were obtained from the following commercial sources: glass wafers (No. 7740, 3 in., 500-µm thick) from Corning Japan (Tokyo, Japan); a thick film photoresist, SU8, from Microchem (Newton, MA); precursor solution of poly(dimethylsiloxane) (PDMS), KE-1300T, from Shin-Etsu Chemical (Tokyo, Japan); water repellant, Fluorosurf (FS-1060TH-2.0), from Fluorotechnology (Aichi, Japan) as a gift. Food dyes (Red No. 102, Blue No. 1, Yellow No. 4) were obtained from Kyoritsu Foods (Tokyo, Japan). A green solution was prepared by mixing the blue and yellow dyes. Fluorescein and resorufin were obtained from Wako Pure Chemical Industries (Osaka, Japan) and Aldrich Chemical Co. (Milwaukee, WI), respectively. Bovine serum albumin labeled with fluorescein isothiocyanate (BSA-FITC), AF977, was obtained from Elastin Products Co. (Owensville, MO). Enzymes and their substrates were obtained from the following commercial sources: L-glutamate and horseradish peroxidase (HRP, EC 1.11.1.7, 100 U mg-1) from Wako Pure Chemical Industries; Amplex Red (10-acetyl-3, 7-dihydroxyphenoxazine) from AnaSpec (San Jose, CA); L-glutamate oxidase (L-GlOx, EC 1.4.3.11 from Streptomyces sp., 330 U mL-1) from Yamasa (Chiba, Japan) as a gift. All solutions were prepared with distilled-deionized water. Fabrication of the Device. A simple T-junction structure was a major critical component. To form the structure, grooves were formed on a PDMS substrate (typically 2 mm thick) and were closed with a glass substrate adhered to the PDMS substrate. The PDMS structure was formed by replica molding using a template formed with the thick-film photoresist, SU-8. The flow channels consisted of a main flow channel and a handling flow channel branched from the main flow channel. The width and depth of the former were 600 and 200 µm, whereas those of the latter were 200 and 200 µm, respectively. The handling flow channel was narrowed and elongated to make the measurement of the volume more accurate. The first requirement in using the microfluidic system is that the inner walls of the flow channels must be sufficiently hydrophobic so that plugs can exist stably. To this end, the walls were coated with the surface modifier (Fluorosurf), unless otherwise noted. The modifier solution was (23) Kou, S.; Lee, H. N.; van Noort, D.; Swamy, K. M. K.; Kim, S. H.; Soh, J. H.; Lee, K.-M.; Nam, S.-W.; Yoon, J.; Park, S. Angew. Chem., Int. Ed. 2008, 47, 872–876.
Figure 1. Device used for the handling of liquid plugs. (A) Device with a single T-junction. The picture shows a plug placed in the main flow channel coated with Fluorosurf. (B) Device with three T-junctions for parallel processing.
permeated into the completed flow channels and spread well on the glass and PDMS surfaces. After baking for 60 min at 100 °C, a coating layer was formed with a final thickness of 1-3 µm. The static contact angle of a sessile drop of water placed on a wide flat surface of Fluorosurf was 131°. Silicone tubes of 500-µm inner diameter were connected to the inlet of the main flow channel and the handling flow channel (Figure 1A). The flow channel structure can be changed depending on the purpose. Figure 1B shows three T-junctions connected to a main flow channel. The basic structure and dimensions of the structure were the same as those of a single junction. Unit Operations To Handle Plugs. In this device, liquid plugs were separated by air plugs. This is because the use of the air plugs minimized wastes and had no detrimental influence on other optional procedures such as detection and pH regulation, which may be incorporated in advanced systems and can be carried out advantageously based on electrochemical principles.24–26 Figure 2 shows unit operations for the handling of liquid plugs. Some of the procedures have been used in previous devices.13,27 In the figures, the arrows indicate that the corresponding flow channel is connected to a microsyringe and a positive or negative pressure is applied to the indicated direction. The “×” mark indicates that the end of the flow channel is closed. The other flow channels without the marks indicate that their end is open. The first step is to prepare liquid plugs of predetermined lengths (unit operation 1). To this end, the handling flow channel is used. A scale is incorporated in the form of short segments of the same length connected with turns. The measurement of the length or volume of a plug is carried out by withdrawing a solution to an appropriate position. Prior to this step, a plug can also be divided into two plugs of appropriate lengths by using unit operation 2 or 3. In unit operation 2, an unprocessed plug is placed at the inlet (24) Suzuki, H. Electroanalysis 2000, 12, 703–715. (25) Suzuki, H.; Upadhyay, S.; Loughran, M. In Encyclopedia of Sensors; Grimes, C. A., Dickey, E. C., Pishko, M. V., Eds.; American Scientific Publishers: Stevenson Ranch, CA, 2006; Vol. 6, pp 161-187. (26) Morimoto, K.; Toya, M.; Fukuda, J.; Suzuki, H. Anal. Chem. 2008, 80, 905–914. (27) Hosokawa, K.; Fujii, T.; Endo, I. Anal. Chem. 1999, 71, 4781–4785.
Analytical Chemistry, Vol. 80, No. 16, August 15, 2008
6207
be stored temporarily in the handling flow channel (unit operation 8). Pressure Driving and Detection. Two microinjectors (IM9B, Narishige, Tokyo, Japan) were connected to the end of the main flow channel and the handling flow channel using the silicone tubes attached to the end of the flow channels. Another silicone tube was also connected to the other end. For experiments that demonstrate the operations, the velocity of plugs in the main flow channel did not exceed 13 mm/s. When precise control of the velocity of plugs in the main flow channel was required, a microsyringe pump (model 210, KD Scientific, Boston, MA) controlled by a computer was also used. Prior to processing, solutions of ∼0.5 µL were introduced into the main flow channel using the microinjector. At this stage, the volume of the injected solutions was not measured precisely. Fluorescence images were obtained by using a fluorescence microscope (VB-G25, Keyence, Osaka, Japan) equipped with a CCD detection system (Keyence VB-7000/7100). For the excitation and detection of fluorescence, band-pass filters (XF1067 (530-590 nm) and XF3081 (610-700 nm), Omega Optical, Brattleboro, VT) were used. The fluorescence image and intensity were analyzed using software (Scion Image, Scion Corp., Frederick, MD). The experiments were conducted at room temperature.
Figure 2. Unit operations to handle liquid plugs. The arrows indicate the direction of applied pressure. The crossed circles indicate that the end is closed. The end of the flow channel without the marks is open.
of the handling flow channel and a part of it is withdrawn there. The other part is separated in the main flow channel, and the measured plug in the handling flow channel is returned to the main flow channel. In unit operation 3, an unprocessed plug is placed at the inlet of the handling flow channel. Then, air is injected into the middle of the plug to divide it into two separated plugs. To mix two solutions, there are two options that use unit operation 4 or 5. In unit operation 4, a plug is first withdrawn into the handling flow channel in such a way that the end of the plug comes to the edge of the main flow channel. When the second plug is placed under the plug and a slight pressure is applied, the two plugs merge. Otherwise, the two plugs can be connected by removing air that exists between the two plugs (unit operation 5). In sorting plugs, the plug in the middle is withdrawn into the handling flow channel along with the neighboring air plugs to avoid mixing of solutions (unit operation 6). Then, the other plugs are mobilized to the upper or lower stream. Then, the plug in the handling flow channel is returned to the main channel. More plugs can be sorted in the same manner by repeating the operation. The distance between two plugs can also be adjusted by placing the air plug at the handling flow channel and injecting or withdrawing air through the flow channel (unit operation 7). Moreover, plugs prepared or handled in the preceding steps can 6208
Analytical Chemistry, Vol. 80, No. 16, August 15, 2008
RESULTS AND DISCUSSION Formation and Processing of a Row of Plugs. Plugs of sufficient lengths did not collapse spontaneously in the hydrophobic flow channels, maintaining their initial shape (Figure 1A). The volume of the plug in the main flow channel was 120 nL/ mm along the axial direction. Since the walls of the flow channels are surrounded by PDMS, volume changes caused by evaporation of water may pose a problem. This will actually occur in the time scale of hours. With plugs of typical sizes used in the following experiments, however, volume change caused by evaporation was less than 5% within 1 h. In practice, each plug is processed in the order of seconds or at most within several minutes, so there will be no serious influence from evaporation. In processing solutions, the first step is to prepare plugs of precise lengths or volumes. Figure 3A shows how short plugs of different lengths were formed from a long plug. First, the plug was placed at the inlet of the handling flow channel (Figure 3A (1)). Then, a part of the solution was withdrawn there and the volume was measured (Figure 3A (2)). After the excess solution that remained in the main channel was separated (Figure 3A (3)), the solution in the handling flow channel was returned into the main flow channel (Figure 3A (4)). Then, half of the solution was again withdrawn into the handling flow channel, the volume was measured, the other half of the solution was separated, and the solution in the handling flow channel was returned into the main flow channel (Figure 3A (5)). By repeating the steps and changing the volume of the solution withdrawn into the handling flow channel, a series of plugs whose length was half of that of the neighboring left plug could be formed (Figure 3A (6)-(8)). The minimum volume that could be formed from the handling flow channel was ∼40 nL. However, with this volume, the plug collapsed when it was mobilized in the main flow channel at velocities higher than 10 mm/s. On the other hand, a plug of 80 nL did not collapse even at 100 mm/s. Needless to say, the minimum volume is influenced by the dimensions of the main
Figure 3. (A) Formation of liquid plugs of various lengths. (1)-(3) Formation of a plug using the handling flow channel. (4)-(8) Formation of the other plugs by the repetition of the operation. (B) Sorting of plugs of different components. (1)-(3) Plug 1 is withdrawn into the handling flow channel. (4)-(6) The solution is injected between plugs 4 and 5. (7) The same operation is repeated for the other green plugs.
flow channel. By reducing the size of the flow channels, the volume of the plugs that can be handled in the device is reduced overall. In many bio/chemical analyses, solutions are injected into a reaction chamber and removed from there one by one to allow a reaction to proceed or to remove unnecessary reagents. The procedure can be carried out efficiently by forming a row of plugs placed in a predetermined order and transporting them one by one into a reaction chamber. As an example that needs this procedure, plugs of antibodies, cleansing buffer solutions, and a substrate for a modified enzyme may be used for an enzyme immunoassay. Other examples are found in an enzyme analysis
described in the next section. For these purposes, sorting of plugs becomes necessary. The picture shown in Figure 3B is a demonstration of the sorting operation. Three plugs of the red dye and three plugs of the green dye had been formed in the main flow channel as mentioned earlier (Figure 3B (1)). First, the plug denoted as 1 was withdrawn into the handling flow channel (Figure 3B (2), (3)) and the other plugs were moved to the right (Figure 3B (4)). Next, plug 1 in the handling flow channel was inserted between plugs 4 and 5 in the main flow channel (Figure 3B (5), (6)). The same steps were repeated to mobilize plug 2 between plugs 5 and 6 (Figure 3B (7)). Mixing of solutions is a unit operation carried out in many chemical analyses. In the mixing of continuous liquids flowing in parallel in a microflow channel, a lot of effort has been made in designing microstructures to disturb the sheath flows.28–33 In contrast, when a plug is mobilized in a flow channel, recirculating flows are generated inside the plug by shearing interactions between the plug and the walls.12,19,27,34 Therefore, mixing of components is promoted. Figure 4A shows fluorescence observed after a plug containing fluorescein and another plug containing resorufin were merged. Without mobilizing the merged plug, mixing by diffusion was negligibly slow and green and red fluorescence were observed from each side of the plug (Figure 4A (2)). We then simply mobilized the plug by 6 mm (Figure 4A (3)). With only this step, the fluorescence distributed more uniformly in the plug. In the case of long plugs, Taylor dispersion to the direction of movement becomes effective in addition to the internal flows by the shearing interactions. As seen in Figure 4B,35 the fluorescence became uniform in the same manner after moving the merged plug back and forth several times with an amplitude of 7 mm. The cycle could be finished within a few seconds. (See also Supporting Information.). As a variation to the mixing procedure, a solution can also be diluted rapidly and efficiently. Solutions of any ratios of dilution can be prepared by attaching a plug of a solution of a length that contains a component to be diluted and a blank solution of a different length and moving the merged plug in the flow channel. The initial length of the plug can also be maintained. In the procedure shown in Figure 4C, a plug containing fluorescein and another plug of a buffer solution (50 mM Tris-HCl, pH 7.4) were mixed in a 1:1 ratio. Then, the length of the new plug was halved by unit operation 2. The plug was then attached to another plug of the buffer solution of the same volume. The operation was repeated. Figure 4D shows the change in the concentration of fluorescein. The concentration was calculated from a calibration plot obtained by fluorometry. The expected decrease in concentration was observed by repeating the operation. (28) Bessoth, F. G.; deMello, A. J.; Manz, A. Anal. Commun. 1999, 36, 213– 215. (29) Liu, R. H.; Stremler, M. A.; Sharp, K. V.; Olsen, M. G.; Santiago, J. G.; Adrian, R. J.; Aref, H.; Beebe, D. J. J. Microelectromech. Syst. 2000, 9, 190–197. (30) Johnson, T. J.; Ross, D.; Locascio, L. E. Anal. Chem. 2002, 74, 45–51. (31) Stroock, A. D.; Dertinger, S. K. W.; Ajdari, A.; Mezic´, I.; Stone, H. A.; Whitesides, G. M. Science 2002, 295, 647–651. (32) Kim, D. J.; Oh, H. J.; Park, T. H.; Choo, J. B.; Lee, S. H. Analyst 2005, 130, 293–298. (33) Sudarsan, A. P.; Ugaz, V. M. Proc. Natl. Acad. Sci.U. S. A. 2006, 103, 7228–7233. (34) Tice, J. D.; Song, H.; Lyon, A. D.; Ismagilov, R. F. Langmuir 2003, 19, 9127–9133. (35) Squires, T. M.; Quake, S. R. Rev. Mod. Phys. 2005, 77, 977–1026.
Analytical Chemistry, Vol. 80, No. 16, August 15, 2008
6209
Processing for an Enzymatic Reaction. In general, the performance of bio/chemical sensors and systems are characterized by changing the concentration of reagents, temperature, pH, etc. Although the characterizations are usually carried out one by one by repeating steps including filling of a background solution, injection of an analyte, and discarding of the solution, it is a very time-consuming and elaborate job. However, the device realizes efficient characterization. This is particularly true in cases in which the parameters to be changed are many. Here, a series of reactions involving L-GlOx and HRP were used.36–38 L - GlOx
L - glutamate + O2+H2O 98 R-ketoglutarate + NH3+ H2O2
(1)
HRP
H2O2 + Amplex Red 98 resorufin + O2
Figure 4. Mixing after merging two plugs (A, B) and dilution (C, D) of components. (A) Change of fluorescence after a plug containing fluorescein and a plug containing resorufin (1.3 mm each) were attached and mobilized to the right by 6 mm. The arrows indicate the direction of the movement of the plugs. (B) The same experiment with plugs of a different length (8 mm each). The merged plug was oscillated with an amplitude of 7 mm. (C) Operation for sequential dilution. (D) Dependence of fluorescein concentration on the repetition of dilutions. 6210
Analytical Chemistry, Vol. 80, No. 16, August 15, 2008
(2)
First, solutions of L-glutamate, L-GlOx (330 U/mL), HRP (100 U/mL), and Amplex Red (1 mM) were introduced into the main flow channel (Figure 5A, (1)). As for L-glutamate, solutions of four different concentrations were injected. As explained later, a slight exchange of solution components occurs. Therefore, to suppress the progress of the enzymatic reactions until the final stage, a water plug was placed between the groups of plugs of L-glutamate and the other plugs. The length of the plugs was adjusted by using the operation mentioned earlier (Figure 5A (2)). The three plugs containing L-GlOx, HRP, Amplex Red were merged and mixed well in a 1:1:1 ratio (Figure 5A (3)). The mixed solution was separated into plugs of an appropriate length (Figure 5A (4)). The plugs were then mixed with the plugs of L-glutamate of different concentrations in a 1:1 ratio one by one (Figure 5A (5)). Fluorescence images of all the mixed plugs were recorded in a computer, and the fluorescence intensities were measured from the images using software (Figure 5A (6)). Figure 5B shows the relation between the fluorescence intensity and the concentration of L-glutamate. Reproducibility of the measured values of intensity was satisfactory with a relative standard deviation of 9.0% even in the worst case (5 µM). Although we had prepared standard solutions of different concentrations and injected them into the flow channel, the plugs can also be formed in the flow channel from a long plug of a concentrated solution by the separation and dilution operations mentioned earlier. Other than the dependence on the concentration, the influence of changes in pH, temperature, etc., can also be examined. To obtain the data for the pH dependence of the output collectively, solutions of two components of a phosphate buffer solution, Na2HPO4 (0.2 M) and KH2PO4 (0.2 M), were used. Plugs of appropriate lengths were prepared, merged, and the length was adjusted. The other plugs that contained L-GlOx (330 U/mL), HRP (100 U/mL), Amplex Red (1 mM), and L-glutamate (5 µM) were also prepared as in the previous experiment and the fluorescence (36) Kusakabe, H.; Midorikawa, Y.; Fujishima, T.; Kuninaka, A.; Yoshino, H. Agric. Biol. Chem. 1983, 47, 1323–1328. (37) Kusakabe, H.; Midorikawa, Y.; Kuninaka, A.; Yoshino, H. Agric. Biol. Chem. 1983, 47, 179–182. (38) Zhou, M.; Diwu, Z.; Panchuk-Voloshina, N.; Haugland, R. P. Anal. Biochem. 1997, 253, 162–168.
Figure 5. Plug-based chemical analysis. (A) Operation for the determination of L-glutamate. (1)-(6) show steps to prepare, separate, and merge the plugs. (B) Dependence of the fluorescence intensity on the concentration of L-glutamate. Five runs were made, and averages and standard deviations are shown. (C) Dependence of the fluorescence intensity on pH. Plugs of different pHs were prepared by changing the mixing ratio of 0.2 M Na2HPO4 and 0.2 M KH2PO4. O, data obtained using this device.b, reference data obtained by a conventional method.
intensity was measured. Figure 5C shows the dependence of the fluorescence intensity on the pH of the solution, which is known from the mixing ratio of Na2HPO4 and KH2PO4. The fluorescence intensity of the plug with the mixed components increased with the increase in pH but decreased at pH 8.0. The same experiment was carried out in a 96-well plate using a microplate reader (infinite F200, Tecan, Switzerland). The tendencies were very close
between the two cases. The activity of L-GlOx is highest in a pH range between 7 and 8 and decreases gradually out of this region with the increase or decrease of pH.36,39 On the other hand, fluorescence of resorufin is observed at pH >6, increases with the increase in pH, and levels off at pH 7.5.40 We also confirmed that the fluorescence intensity of resorufin produced from Amplex Red by the enzymatic reaction of HRP increased in the examined pH range with the increase in pH. The tendencies observed in Figure 5C reflect the respective tendencies well. Cross-Contamination by Neighboring Plugs. Our final goal is to realize a device that carries out necessary procedures many times repetitively like an electronic microprocessor. In moving many plugs with different components in the same flow channel, however, cross-contamination of each plug by neighboring plugs may occur. Although no serious influence has been recognized in the experiments described up to now, in which the number of procedures was limited, cross-contamination may be an impediment when many plugs are used in the same flow channel repetitively. To examine the influence, two plugs of 2 mm in length were formed and were separated by 2 mm in a flow channel coated with Fluorosurf. One of them contained fluorescein, and the other was a blank buffer solution (50 mM Tris-HCl, pH 7.4). The plugs were mobilized in the main flow channel back and forth at a constant velocity of 6.7 mm/s with amplitude of 6 mm. The change in the fluorescence intensity from the plug of the buffer solution was measured, and the concentration of fluorescein in the plug was calculated from a calibration plot. Figure 6A shows the dependence of the relative amount of fluorescein that had moved to the plug of the buffer solution on the number of cycles of displacement. With 1 mM fluorescein, cross-contamination was observed clearly as an increase in the fluorescence intensity. Visual inspection under a microscope revealed that the plugs were repelled well from the surface and separated by air between them. Nevertheless, the obtained result indicates that there actually existed residues on the walls and a part had moved to the neighboring plug. We then decreased the concentration of fluorescein to 16 µM. However, the tendency was very close to that of 1 mM. With reference to the result, a very hydrophobic surface will be preferable for repetitive measurements. However, the hydrophobic walls promote nonspecific adhesion of proteins. Therefore, we then examined cross-contamination by proteins using a plug containing 25 mg/mL BSA-FITC and a plug of the buffer solution. To examine the influence in more detail, a bare PDMS flow channel without coating was also used. The result was significantly different between the surface coated with Fluorosurf and that of PDMS. In the former, the cycle could be repeated more than 100 times (Figure 6A). On the other hand, in the latter, the plugs collapsed and merged even after a few cycles during the procedure. The increase in the fluorescence intensity was more significant than the case with fluorescein. To examine this point in more detail, we immersed the substrate coated with Fluorosurf and the PDMS substrate in 50 mg/mL BSA for 10 min, rinsed with distilled-deionized water, dried, and the changes in the static contact angle were examined. The contact angle on Fluorosurf (39) Arima, J.; Tamura, T.; Kusakabe, H.; Ashiuchi, M.; Yagi, T.; Tanaka, H.; Inagaki, K. J. Biochem. 2003, 134, 805–812. (40) Towne, V.; Will, M.; Oswald, B.; Zhao, Q. Anal. Biochem. 2004, 334, 290– 296.
Analytical Chemistry, Vol. 80, No. 16, August 15, 2008
6211
Figure 7. Production of plugs of different colors by parallel processing using the three T-junctions. (1, 2) Plugs of red, blue, and yellow were formed in the three handling flow channels and injected into the main flow channel. (3, 4) The plugs were merged one by one to form plugs of different colors. Figure 6. Cross-contamination by neighboring plugs. (A) Dependence of the relative amount of the contaminant in the buffer solution on the number of cycles of displacement. O, ∆, ], in a flow channel coated with Fluorosurf. 2, b, in a PDMS flow channel. The source of the contaminant contained fluorescein (∆ and 2, 16 µM; O and b, 1 mM) or BSA-FITC (], 25 mg/mL). (B) Dependence of the relative amount of fluorescein after 100 cycles of displacement on the velocity or Ca. (C) Comparison of the relative amount of fluorescein in each plug of the buffer solution after 100 cycles of displacement. The velocity in (A) and (C) was 6.7 mm/s. The amplitude of displacement was 6 mm for (A) and (B) and 13 mm for (C). In the experiments, the plugs were 2 mm in length and were separated by 2 mm.
decreased from 131° to 104°, whereas that on PDMS decreased from 110° to 91°. The fatal collapse of plugs in the flow channel within a few cycles is explained well with the change. 6212
Analytical Chemistry, Vol. 80, No. 16, August 15, 2008
The behavior of an array of plugs moving in a flow channel has been characterized in terms of capillary number, Ca. With a high Ca condition, plugs tend to break into small pieces and can be a cause of cross-contamination.41 Figure 6B shows the dependence of the relative amount of fluorescein moved to the plug of the buffer solution on the velocity or Ca. The increase in the velocity clearly had an adverse influence and the change in the ratio increased with the increase in Ca. When the velocity was decreased by 1 order of magnitude, the ratio decreased to approximately a half. The result suggests that reducing the velocity is effective at reducing the cross-contamination, although it makes the processing time longer. (41) Adamson, D. N.; Mustafi, D.; Zhang, J. X. J.; Zheng, B.; Ismagilov, R. F. Lab Chip 2006, 6, 1178–1186.
To circumvent the cross-contamination, a direct solution is to remove the contaminants on the walls using a cleansing plug.16,18 To check the effect, a plug of a solution containing fluorescein was placed in the main flow channel along with four cleansing plugs of the buffer solution placed on one side (numbered from 1 to 4 from the plug nearest to the fluorescein plug). The group of the plugs was mobilized back and forth at 6.7 mm/s (Ca, 8.6 × 10-5) with amplitude of 13 mm. Figure 6C shows changes in the relative amount of fluorescein in each plug after 100 cycles. The amount of fluorescein was measured from the fluorescence intensity and was normalized to the initial amount in the fluorescein plug. With the present device, the fluorescein moved to the other four cleansing plugs. The amount of fluorescein moved to cleansing plug 1 was less than 30%. In the other plugs, the amount was less than half of the neighboring plug with a larger amount. Comparison between fluorescein plug 0 and cleansing plug 2 indicates that the influence of contaminants can be reduced significantly by placing a cleansing plug (plug 1) between them. Moreover, comparison between fluorescein plug 0 and cleansing plug 4 suggests that it is more effective to use several cleansing plugs. It should be noted that this is an extreme case, in which the number of cycles was increased to enhance the contamination. In practice, when the number of cycles is limited to several times, the contamination by neighboring plugs will be suppressed almost completely by using only one cleansing plug. Parallel Processing Using T-Junctions. In the experiments explained up to now, only a single T-junction was used. This is a burden to the structure when the solutions to be handled are many. In addition, the cross-contamination should be minimized. An effective solution to these problems will be to use more T-junctions and make them play independent specific roles like transistors in an electronic circuit. The device with three Tjunctions shown in Figure 1B carries out such parallel processing. Procedures that had been carried out with a single junction can be divided into several groups and carried out in parallel. Figure 7 shows a procedure to form plugs of seven different colors. First, an array of liquid plugs of the same color was formed and stored in the handling flow channel of each T-junction. The plugs were injected into the main flow channel and mobilized to the left. Finally, they were merged with four different combinations. In (42) Satoh, W.; Shimizu, Y.; Kaneto, T.; Suzuki, H. Sens. Actuators, B 2007, 123, 1153–1160. (43) Suzuki, H.; Yoneyama, R. Sens. Actuators, B 2003, 96, 38–45.
the last picture of Figure 7, the mixed plugs of four different colors are placed in the main flow channel along with the initial plugs. We also tried the determination of L-glutamate carried out in the previous experiments by conducting the respective procedures using different junctions. Indeed, data close to those obtained using the single T-junction have been obtained using the device, which are not presented here again. CONCLUSIONS Microfluidic devices based on the handling of liquid plugs provide a possibility to conduct many analyses efficiently and repetitively on a chip. To handle the plugs for bio/chemical analyses, a T-junction that consists of a main flow channel and a handling flow channel can play a key role. Formation of plugs of various lengths, mixing of solutions, and sorting of plugs can be conducted on the chip. The device and the method can be used for the determination of L-glutamate or the optimization of experimental conditions such as pH. Although complete elimination of the cross-contamination between plugs is a current issue, the use of very hydrophobic walls and a cleansing plug along with processing at a lower velocity could minimize the influence. As with the electronic transistor, it will be possible to carry out more complicated procedures by increasing the number of T-junctions. In this study, plugs were mobilized manually using microsyringes or by switching microsyringe pumps. The procedure can be carried out using a program. The reciprocating movement can be generated using micropumps that operate based on electrochemical bubble formation.42,43 Therefore, on-chip automatic manipulation of plugs based on the procedures will be possible. Furthermore, with a storage device comparable to a semiconductor memory,9,10 more complicated processing of solutions will be possible. ACKNOWLEDGMENT This study was supported by a Grant-in-Aid for Scientific Research B (No. 19360369) under the Japan Society for the Promotion of Science. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review March 8, 2008. Accepted May 13, 2008. AC800492V
Analytical Chemistry, Vol. 80, No. 16, August 15, 2008
6213