Cytometry and Velocimetry on a Microfluidic Chip Using

Mar 2, 2005 - Chemistry, Sungshin Women's University, 249-1 Dongsun-dong, Seongbuk-gu, Seoul ... This paper reports a polyelectrolytic salt bridge-bas...
0 downloads 0 Views 224KB Size
Download PDF
Anal. Chem. 2005, 77, 2490-2495

Cytometry and Velocimetry on a Microfluidic Chip Using Polyelectrolytic Salt Bridges Honggu Chun,† Taek Dong Chung,*,‡ and Hee Chan Kim*,†

Department of Biomedical Engineering, College of Medicine and Institute of Medical and Biological Engineering, Medical Research Center, Seoul National University, 28 Yongon-dong, Chongno-gu, Seoul 110-744, Korea, and Department of Chemistry, Sungshin Women’s University, 249-1 Dongsun-dong, Seongbuk-gu, Seoul 136-742, Korea

This paper reports a polyelectrolytic salt bridge-based electrode (PSBE), which is a key embedded unit in a microchip device that can size-selectively count microparticles and measure their velocities. The construction of salt bridges at specific locations within a microfluidic chip enables dc-driven electrical detection to be performed successfully. This is expected to be a competitive alternative to the optical methods currently used in conventional cell sorters. The PSBEs were fabricated by irradiating ultraviolet light over a patterned mask on the parts of interest, which were filled with an aqueous monomer solution containing diallyldimethylammonium chloride. A pair of such PSBEs was easily formed at the two lateral branches perpendicular to the main microchannel and was found to be very useful for dc impedometry. The human blood cells as well as the fluorescent microbeads passing between the two PSBEs produced impedance signals in proportional to their size. The information about the velocity of a microparticle was extracted from a doublet of the dc impedance signals, which were generated when cells or microbeads sequentially passed through two PSBE pairs separated from each other by a fixed distance. The plot of peak amplitude versus velocity of the moving microbeads and cells indicated only a slight correlation between the size and the velocity, which means that the peak amplitude of the dc impedance signals alone can provide information about the size of the cells in a mixture. The experimental results showed a screening rate of over 1000 cells s-1 and a velocity of the cells of over 100 mm s-1. Compared with the previously suggested electrical detection system based on metal electrodes, the sensitivity and selectivity in cell detection were remarkably improved. In addition, the detection unit including the operating circuit became innovatively simple and the whole device could be miniaturized. The modern concept of micro total analysis systems (µ-TAS) dates back to the early 1990s when capillary electrophoresis was * To whom correspondence should be addressed. E-mail: [email protected]; [email protected]. † Seoul National University. ‡ Sungshin Women’s University.

2490 Analytical Chemistry, Vol. 77, No. 8, April 15, 2005

developed on a glass chip by Manz et al.1 As has been welldocumented, chemical and biological processes on a microchip have appreciable benefits that are unattainable using macroscale process configurations. The advantages of µ-TAS are summarized to the tiny sample volume for analysis, low cost, easy automation, and high throughput by parallel processing.2 Therefore, the applications of µ-TAS are expanding to many different areas including clinical diagnostics,3,4 single-chip polymerase chain reaction,5,6 DNA separation,7,8 DNA sequencing,9,10 biological and chemical analysis,11-13 cell analysis,14-16 and flow cytometry.16-27 (1) Manz, A.; Graber, N.; Widmer, H. M. Sens. Actuators, B 1990, 1, 244. (2) Erickson, D.; Li, D. Anal. Chim. Acta 2004, 507, 11-26. (3) Tian, H. J.; Jaquins-Gerstl, A.; Munro, N.; Trucco, M.; Brody, L. C.; Landers, J. P. Genomics 2000, 63, 25-34. (4) Pasas, S. A.; Lacher, N. A.; Davies, M. I.; Lunte, S. M. Electrophoresis 2002, 23, 759-766. (5) Waters, L. C.; Jacobson, S. C.; Kroutchinina, N.; Khandurina, J.; Foote, R. S.; Ramsey, J. M. Anal. Chem. 1998, 70, 5172-5176. (6) Woolley, A. T.; Hadley, D.; Landre, P.; deMello, A. J.; Mathies, R. A.; Northrup, M. A. Anal. Chem. 1996, 68, 4081-4086. (7) Shi, Y. N.; Simpson, P. C.; Scherer, J. R.; Wexler, D.; Skibola, C.; Smith, M. T.; Mathies, R. A. Anal. Chem. 1999, 71, 5354-5361. (8) Ueda, M.; Kiba, Y.; Abe, H.; Arai, A.; Nakanishi, H.; Baba, Y. Electrophoresis 2000, 21, 176-180. (9) Liu, S. R.; Shi, Y. N.; Ja, W. W.; Mathies, R. A. Anal. Chem. 1999, 71, 566573. (10) Paegel, B. M.; Emrich, C. A.; Weyemayer, G. J.; Scherer, J. R.; Mathies, R. A. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 574-579. (11) Beebe, D. J.; Mensing, G. A.; Walker, G. M. Annu. Rev. Biomed. Eng. 2002, 4, 261-286. (12) Jakeway, S. C.; de Mello, A. J.; Russell, E. L. Fresenius J. Anal. Chem. 2000, 366, 525-539. (13) Chovan, T.; Guttman, A. Trends Biotechnol. 2002, 20, 116-122. (14) Stuart, J. N.; Sweedler, J. V. Anal. Bioanal. Chem. 2003, 375, 28-29. (15) Lin, Y. C.; Jen, C. M.; Huang, M. Y.; Wu, C. Y.; Lin, X. Z. Sens. Actuators, B 2001, 79, 137-143. (16) Fu, A. Y.; Spence, C.; Scherer, A.; Arnold, F. H.; Quake, S. R. Nat. Biotechnol. 1999, 17, 1109-1111. (17) Ayliffe, H. E.; Frazier, A. B.; Rabbitt, R. D. J. MicroElectroMechanical Syst. 1999, 8, 50-57. (18) Gawad, S.; Schild, L.; Renaud, P. Lab Chip 2001, 1, 76-82. (19) Larsen, U. D.; Blankenstein, G., Microchip Coulter particle counter. In Proceedings of Transducer 97; IEEE, Chicago, IL, 1997; pp 1319-1322. (20) Gawad, S.; Batard, P.; Seger, U.; Metz, S.; Renaud, P. Leukocytes discrimination by impedance spectroscopy flow cytometry. In Proceedings of the µ-TAS 2002, Nara, Japan, 2002; pp 649-651. (21) Wolff, A.; Perch-Nielsen, I. R.; Larsen, U. D.; Friis, P.; Goranovic, G.; Poulsen, C. R.; Kutter, J. P.; Telleman, P. Lab Chip 2003, 3, 22-27. (22) Eyal, S.; Quake, S. R. Electrophoresis 2002, 23, 2653-2657. (23) Sobek, D.; Young, A. M.; Gray, M. L.; Senturia, S. D. In Micro Electro Mechanical Systems, Proceedings-An Investigation of Micro Structures, Sensors, Actuators, Machines, and Systems; IEEE: New York, 1993; pp 219-224. (24) Schrum, D. P.; Culbertson, C. T.; Jacobson, S. C.; Ramsey, J. M. Anal. Chem. 1999, 71, 4173-4177. 10.1021/ac048535o CCC: $30.25

© 2005 American Chemical Society Published on Web 03/02/2005

Flow cytometry is a technology that measures some of the properties of cells as they move or flow in liquid suspension. The miniaturization of a flow cytometry system for a point-of-care test (POCT) has great importance not only for cell biological research but also for clinical uses, which include the separation of stem cells from the peripheral blood and the control of the white blood cell (WBC) level for juvenile leukemia patients. In most flow cytometers, the cells traveling in the interrogation region are detected either by optical or electrical methods. A fluorescenceactivated cell sorter (FACS) adopts the former, and the Coulter counter is based on the latter.28 There have been many reports on microchip-based flow cytometers using both detection methods. For example, Wolff et al. presented a highly integrated chip for a high-throughput FACS.21 Ayllife et al. reported a chip-based electrical analyzer that was operated by microchannel impedance spectroscopy.17 Despite the significant advances reported on the FACS on a microchip16,22-26 and electrical counter,18-20 flow cytometers on microchips have only been partially successful in practical applications. Combining the conventional FACS concept with a microfluidic chip obviously aims the miniaturization and simplification of a device performing similar tasks to those currently performed by commercialized large systems. However, this idea has a few serious problems. First, it still requires cell modification by markers or antibodies, which can lead to changes in the system under study. Second, it is difficult for optical parts to be reduced to sizes as small as the microchip itself. Even if fluidic parts become substantially small by introducing microfluidic technology, the whole system including other parts such as the detection unit is still too large to be a practical device for POCT. Third, the equipment for detection is rather expensive and hard to operate. Optical equipment is rarely as low cost as electronic devices and requires fine alignments. The electrical detection method has been considered to be an alternative to the optical technology in this regard. The electronic device is the most probable choice in terms of miniaturization, simplification, and cost-effectiveness, assuming that it works as well as the optical setup. However, there are fundamental challenges for the electrical method to accomplish before it can function as well as the optical detection in FACS. Theoretical calculations suggest that two electrodes need to face each other at opposite sides of the channel wall to obtain the best sensitivity and precision of the impedance response.18 There have been many attempts to realize such systems. For instance, two glasses with planar metal band electrodes were overlaid to face each other20 and conducting metals on two planar electrodes were electroplated at the lateral branch channels of a microchannel.17 Unfortunately, those trials have had only limited success for the following reasons. First, the fabrication of such electrodes is tricky. Even if they can be made, the reproducible geometry and characteristics as electrodes cannot be guaranteed. The most critical problem is related to the electrode material and the (25) McClain, M. A.; Culbertson, C. T.; Jacobson, S. C.; Ramsey, J. M. Anal. Chem. 2001, 73, 5334-5338. (26) Kruger, J.; Singh, K.; O’Neill, A.; Jackson, C.; Morrison, A.; O’Brien, P. J. Micromech. Microeng. 2002, 12, 486-494. (27) Altendorf, E.; Zebert, D.; Holl, M.; Vannelli, A.; Wu, C. C.; Schulte, T. In Proceedings of the µ-TAS 1998, Dordrecht, The Netherlands, 1998; pp 7376. (28) Durack, G.; Robinson, J. P. Emerging tools for single-cell analysis; WileyLiss; New York, 2000.

frequency applied. Because the electrical properties of a cell membrane are close to that of a capacitor, the impedance signal should be inversely proportional to the frequency applied. This suggests that a lower frequency generates a larger change in impedance. The most desirable frequency is zero, namely, the dc signal. However, metal electrodes are not compatible with dc or a low-frequency electric potential bias. The impedance changes due to the cells become insignificant upon dc or low-frequency inputs because of the electric double layer, Faradaic reactions on the metal electrode surfaces, or both. As long as conventional metal electrodes are used, an ac input at a high frequency, where the sensitivity of cell detection is obviously limited, is the only option. Another problem is the fact that the data for the cell size correlate with the velocity. That is why the calibration process with respect to velocity was suggested for the estimation of cell size.18,22 In terms of cell sorting, there is one more challenge that should be addressed. In FACS, the moving cells are hydrodynamically focused in 2-D so that the velocities are uniform within a limited error. That makes effective cell sorting possible in FACS. However, 1-D (horizontal) focusing on the microfluidic chip cannot produce as good a flow as generated in the conventional FACS system. The variation in the velocity may mislead the micropump to wrongfully push or pull the cells to be sorted. Considering the fact that the fast velocimetry of the cells on the microchips is one of the critical issues toward automatic cell sorting on a microfluidic chip with a high throughput, the quick and accurate velocimetry of moving cells on a spot can greatly assist in sorting the cells in a flow cytometer. A simple method previously reported for measuring the velocity of flowing cells is video image densitometry,29 where the velocity is estimated by observing the displacement of a cell within a known time interval. However, there are limitations in terms of the accuracy and cost because the video frame interval obviously regulates its temporal resolution. Shah convolution Fourier transform is another method for extracting velocity information on a microfluidic chip.30 Using this method, a mask with a periodic array of slits modulates the excitation beam in space and the cells underneath the mask undergo spatially modulated excitation. Fourier transforming the modulated fluorescent signals produces data containing the velocity information. The mask with periodic slits can be replaced by a waveguide beam splitter for the purpose of integration on a microchip.31 Another method for the velocity measurement is to use the time interval of the fluorescent peaks from two adjacent areas excited by an acoustooptic modulator (AOM).22 However, both the Shah convolution Fourier transform and AOM methods increase the complexity of the instrument and calculation and have limitations in terms of miniaturization because of the space needed for the optical system to be integrated. This paper describes the fabrication and performance of a flow cytometer and velocimetry chip using a polyelectrolytic salt bridgebased electrode (PSBE). The PSBEs in this study were fabricated by employing a photopolymerization technique. Beebe et al. previously used the photopolymerization of a prepolymer mixture (29) Barker, S. L. R.; Ross, D.; Tarlov, M. J.; Gaitan, M.; Locascio, L. E. Anal. Chem. 2000, 72, 5925-5929. (30) Crabtree, H. J.; Kopp, M. U.; Manz, A. Anal. Chem. 1999, 71, 2130-2138. (31) Mogensen, K. B.; Kwok, Y. C.; Eijkel, J. C. T.; Petersen, N. J.; Manz, A.; Kutter, J. P. Anal. Chem. 2003, 75, 4931-4936.

Analytical Chemistry, Vol. 77, No. 8, April 15, 2005

2491

of two kinds of monomers and a photoinitiator by irradiating UV light over a mask to manufacture microfluidic systems.32 Using patterned masks, photopolymerization has the advantage of forming a polymer at the specific spot of interest. On the other hand, the concept of a salt bridge was not so common in microfluidic chip research. Khandurina et al. applied a porous silicate film as a salt bridge for electrophoresis.33 Takamura et al.34 and Brask et al.35 developed a low-voltage cascade electroosmotic pump based on salt bridges. However, the polymer-based salt bridge has not been used as an electrode for detecting moving cells. The PSBE can easily be fabricated at the microchannel walls and make it possible to implement dc impedance analysis. Furthermore, two pairs of the PSBEs separated by a fixed length along a microchannel provide the data containing the velocity of the cells on the same chip. It is expected that PSBE can offer a new opportunity to accomplish both the size-selective detection and the simultaneous velocimetry of the cells traveling along a microchannel without a large or complex peripheral setup. EXPERIMENTAL SECTION Chip Fabrication. Corning 2947 precleaned slide glasses (75 mm by 25 mm, 1 mm thick) were used as substrates. A slide glass was cleaned in piranha solution (H2SO4:H2O2 ) 3:1) for 1 h before washing the slide glass with deionized (DI) water (NANOpure Diamond, Barnstead) and cleaning with acetone (CMOS grade, J. T. Baker), methanol (CMOS grade, J. T. Baker), and DI water twice sequentially. The cleaned slide glass was dehydrated on a hot plate at 150 °C for 10 min and was cooled to room temperature. Hexamethyldisilazane (HMDS; Clariant) was spin-coated (Won Corp.) at 4000 rpm for 30 s on the slide glass, on which spincoating of the photoresist (PR) of AZ5214-E (Clariant) was subsequently carried out at 4000 rpm for 30 s. After soft baking the PR on a hot plate at 100 °C for 60 s, the slide glass was cooled to room temperature and aligned under a pattern mask. Exposing the slide to UV light (365 nm) with an intensity of 16 mW cm-2 for 6.5 s (MDE-4000, Midas) was followed by developing the PR with AZ300MIF (Clariant) for 45 s. The slide glass was then washed with DI water, and the PR was hard-baked on a hot plate at 105 °C for 15 min. The HMDS and PR layers were spin-coated on the other side of the slide glass and baked in the same way as described above in order to protect it from the etching solution. The slide glass was etched with 6:1 buffered oxide etch solution (J. T. Baker) for 40 min at 25 °C. The washing processes consisted of a few successive steps; rinsing with DI water and acetone, sonicating in acetone for 5 min in an ultrasonic cleaner (3510EDTH, Bransonic), and soaking in methanol and DI water. Another flat slide glass used to cover the etched glass was drilled at the positions for the reservoirs with a 2-mm-diameter diamond drill at 18 000 rpm. The flat slide glass was then cleaned in a piranha solution for 1 h. The pair of etched and flat slide glasses were permanently attached by thermal bonding. When the two slide (32) Beebe, D. J.; Moore, J. S.; Yu, Q.; Liu, R. H.; Kraft, M. L.; Jo, B. H.; Devadoss, C., Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 13488-13493. (33) Khandurina, J.; Jacobson, S. C.; Waters, L. C.; Foote, R. S.; Ramsey, J. M. Anal. Chem. 1999, 71, 1815-1819. (34) Takamura, Y.; Onoda, H.; Inokuchi, H.; Adachi, S.; Oki, A.; Horiike, Y. Low voltage electroosmosis pump and its application to on-chip linear stepping pneumatic pressure source. In Proceedings of the µ-TAS 2001, Monterey, CA, 2001; pp 230-232. (35) Brask, A.; Goranovic, G.; Bruus, H. Sens. Actuators, B 2003, 92, 127-132.

2492 Analytical Chemistry, Vol. 77, No. 8, April 15, 2005

Figure 1. Microfluidic glass chip used for the cell detection study. One pair of PSBEs was formed at the circular regions in the enlarged picture. The microchannel between the two polymerization spots was protected from UV light by a mask.

glasses contacted each other for bonding, DI water between the glasses kept away the air bubbles. The glasses were heated to 600 °C in a furnace (CRF-M15, CEBER), and the temperature was maintained at 600 °C for 6 h, which was followed by slowly cooling in the furnace to room temperature over 10 h. PSBE Fabrication. Diallyldimethylammonium chloride (DADMAC) was chosen as the monomer of the polyelectrolytic polymer for the salt bridge. A 65% DADMAC aqueous solution was polymerized to yield poly-DADMAC (PDADMAC) by shedding UV light in the presence of 2% photoinitiator (2-hydroxy-4′-(2hydroxyethoxy)-2-methylpropiophenone) and 2% cross-linker (N,N′methylenebisacrylamide). The high charge density makes PDADMAC hold many anions inside its structure so that transport of the mobile anions is facile and the apparent resistance of the polymer plug decreases. The stationary charge in the PDADMAC is independent of the pH in the medium. As a result, PDADMAC possesses good properties as a salt bridge. DADMAC, the photoinitiator, and the cross-linker were purchased from SigmaAldrich (St. Louis, MO). The salt bridge fabrication process using the photopolymerization technique is as follows: The microchannel network of a microfluidic glass chip was filled with the DADMAC solution with a composition described above. The chip was aligned under the mask and subsequently exposed to UV light (365 nm) with an intensity of 16 mW cm-2 for 5.0 s. The DADMAC monomers were polymerized to form PDADMAC at the two circular regions shown in Figure 1. After photopolymerization, the microchannel was cleaned with a 1 M KCl solution to remove the remaining DADMAC monomers. In Figure 1, the moving cells are detected by impedance analysis using the PSBEs, while the sample solution flows from left to right. The microchannel was 50 µm wide and 22 µm deep. Electrical Properties of the PSBEs. The microchannel network of a microfluidic glass chip was filled with an isotonic 0.92% NaCl solution. Two Ag/AgCl wires were immersed in the corresponding reservoirs in Figure 1 and were connected to an LCR meter (Precision Component Analyzer 6440A, Wayne Kerr). The impedance between the two reservoirs of the salt bridge was recorded as frequency continuously changing from dc to 3.0 MHz. The test results showed a flat impedance property of 30 kΩ throughout the whole frequency range. Sample Preparation. The performance of the cytometry microchip with the PSBEs was evaluated by dc impedance analysis with fluorescent microbeads of 9.95 (P(S/V-COOH), (480, 520),

Figure 2. dc impedance analysis of the microchannel with the PSBEs. dc bias of 0.4 V was applied between two Ag/AgCl electrodes that were electrochemically connected to two PSBEs through an isotonic NaCl solution. The impedance was calculated from the electric current that was measured under a fixed bias voltage. The amplified impedance signals were acquired and stored in a PC for further analysis.

Bangs Laboratory, Fishers, IN) and 5.70 µm (P(S/5.5% divinylbenzene/5% MAA), (480, 520), Bangs Laboratory) in diameter. The fluorescent microbeads were diluted in an isotonic NaCl solution to 0.025 and 0.005 wt % for 9.95- and 5.70-µm beads, respectively. The human blood was sampled from healthy subjects and centrifuged to separate the red blood cells (RBC) and WBC for the tests. The RBC and WBC were diluted to 0.0025 cells pL-1 in RPMI 1640 medium (1×, Jeil biotechservices, Seoul, Korea) before being used in the experiment. Signal Detection and Data Acquisition. A diluted fluorescent microbead solution was injected into the microchannel using a syringe pump (KDS100, KD Scientific, Holliston, MA). Figure 2 shows the configuration of the microfluidic glass chip on which the dc impedance analysis was implemented. The PSBEs were connected to an external dc impedance analyzer via their respective isotonic NaCl solutions and Ag/AgCl electrodes. A 0.4-V dc bias generated a ∼13-µA dc current, which was maintained at least 1 h without reversing the polarity. The experiments could be extended for longer than 1 h, if necessary, by just switching the bias polarity. The impedance signals fluctuating when the cells or microbeads passed through the microchannel between the PSBEs were converted to voltage signals and then amplified with the total gain of 2000. For estimating the velocity of moving cells, two pairs of the PSBEs were fabricated with a 1-mm separation. To prevent crosstalk between two pairs of the PSBEs, two isolated power supplies served the impedance analyzing circuits for the respective pairs of PSBEs, the concept of which is shown in a block diagram in Figure 3. Otherwise, the leakage current between the two pairs of PSBEs may cause serious cross-talk. The signal outputs from the two impedance analyzing circuits were digitized with Lab-PC1200 (National Instruments, Austin, TX) at a sampling frequency of 30 kHz. Safety Precautions. The piranha solution is a powerful oxidizer and reacts violently with organic materials or solvents and should be handled with extreme care.

Figure 3. Concept of data acquisition for a velocimetry using two pairs of PSBEs. Two pairs of PSBEs were located with a 1-mm space to measure the velocity of a flowing cell. Impedance analyzing circuits should be electrically isolated to eliminate potential cross-talk.

Figure 4. Impedance variation between one pair of PSBEs caused by 9.95-µm fluorescent microbeads passing through the detection volume of the microchannel. Each downward peak corresponds to a microbead between the two PSBEs facing each other.

RESULTS AND DISCUSSION Detection of Moving Microbeads. Figure 4 shows the amplified signal of the impedance between the two PSBEs responding to the 9.95-µm fluorescent microbeads randomly moving along the microchannel. Each downward peak corresponds to a single microbead. The standard deviation of the background signals is 0.0085 V. The screening rate of the developed cytometry microchip can be estimated from the width of the peak signal for a cell passing through the detection volume between the PSBEs. The half-power widths of the signals revealed that the maximum screening rate is higher than 3000 cells s-1. The experimental screening rate is up to 1000 cells s-1. The fast screening rate was partly attributed to the quick response of the dc impedance analysis, which is possible only with the PSBEs. This rate was still only 10% or less than that attainable with high-speed FACS, the several tens of thousands of cells per second. However, parallel processing using multiple processing units on a single microfluidic chip is expected to increase the screening rate. Analytical Chemistry, Vol. 77, No. 8, April 15, 2005

2493

Figure 5. Histogram of the peak amplitude of the impedance change obtained using 9.95- and 5.70-µm-diameter fluorescent microbeads. The clearly separated groups show that the peak amplitude alone can distinguish one group of microbeads from the other.

Figure 5 displays a histogram of the peak amplitude for the two types of fluorescent microbeads with different diameters. Correlation between the peak height and the impedance was investigated using two types of fluorescent microbeads, 9.95 and 5.70 µm. The test results showed that the distribution of the peak amplitude was (m, σ2) ) (1.0612, 0.12362) for the 9.95-µm microbeads and (m, σ2) ) (0.1635, 0.02632) for the 5.70-µm microbeads. In a high-frequency ac impedance analysis between coplanar metal electrodes, the amplitudes are severely dependent on the altitude of the cell in the microchannel even when the flow in the microchannel network is hydrodynamically focused. However, dc impedance analysis using the PSBEs gives impedance signals that are unaffected by the position of the cells in the analysis region. As a result, the dependence of the impedance signals on the relative position of a cell from the electrodes can be reduced markedly, which means that the problems arising from the hydrodynamic focusing on a microfluidic chip become less critical. Velocimetry with Two Pairs of PSBEs. Two adjacent pairs of the salt bridge electrodes illustrated in Figure 3 will provide impedance peak signals with a certain time interval when a single cell passes through themselves successively. The velocity of a moving cell is then calculated by dividing the fixed distance between the two pairs of PSBEs, 1 mm, with the time elapsed. A velocity up to 100 mm s-1 was determined by using the velocimetry microchip developed in this study. Neither the response time of the analysis circuit nor the salt bridge electrodes regulated this upper limit of the velocity measurements. The pump capacity used in these experiments is the limiting factor. On the other hand, Figure 6 also shows that the correlation between the peak amplitude and the flow velocity is fairly low. The Pearson correlation value for each size is -0.244 and -0.207 for 9.95 and 5.70 µm, respectively. Counting Human Blood Cells. The performance of the developed cytometry microchip was evaluated with the RBC and WBC from the human blood samples. The size of the blood cell 2494 Analytical Chemistry, Vol. 77, No. 8, April 15, 2005

Figure 6. Scatterplot of the velocity and the peak amplitude obtained for the two types of fluorescent microbeads, 9.95 and 5.70 µm in diameter. The velocity of beads was intentionally varied by changing the flow rate in order to see if there is correlation between the peak voltage and the velocity. The two groups are clearly separated. The amplitude of the impedance peak for each group is independent of the flow velocity, which suggests that two different sized microbeads can be classified solely by the peak amplitude.

Figure 7. Scatterplot of the velocity and peak amplitude for the RBC and WBC moving along the microchannel on the velocimetry microchip. The velocity of cells was intentionally varied by changing the flow rate in order to see if there is a correlation between the peak voltage and the velocity.

was distributed between 6-9 and 12-18 µm for the RBC and WBC, respectively. According to the results from the experiments with the microbeads, it should be possible for the RBC and WBC to be classified according to their difference in size. Figure 7 shows a scatterplot of the velocity and the peak amplitude obtained from human blood cells. The test result showed that peak amplitude distribution is (m, σ2) ) (0.3135, 0.03832) and (m, σ2) ) (0.8319, 0.17922) for the RBC and WBC, respectively. It is clear that the peak amplitude does not correlate with the velocity in the range of 1-100 mm s-1. Therefore, a reliable classification of the RBC and WBC is possible with the peak amplitude only. The results using human blood cells show that complete blood cell counting with a hand-held device is possible instantly. This

means that the developed cytometry microchip is applicable to a POCT-type cell counter for many clinical applications including WBC level control for the juvenile leukemia patients. CONCLUSIONS A fabrication technology for PSBE in µ-TAS was developed, and its performance as a flow cytometry glass microchip was also evaluated. It was demonstrated that the PSBE developed provides a better solution for impedance analysis in a microfluidic glass chip than the conventional metal electrode. The developed PSBEs were embedded on the cytometry and velocimetry microchips and evaluated using both the fluorescent microbeads and human blood cells. The test results show that (1) a screening rate of over 1000 samples s-1, (2) measurement of the cell velocity up to 100 mm s-1, and (3) velocity-independent classification according to the particle size are possible. The PSBE in this report suggests many applications. For example, this device can be used in electrochemical cells that need

to be integrated on microchips or on a decoupler in chip-based electrophoresis for electrochemical detection. The PSBEs are practically suitable for the miniaturization of both cell counting and velocimetry systems, which means that they may be applicable to small-sized POCT devices. The developed technology offers a promising chance toward future microsystems for clinical uses, e.g., a miniaturized stem cell collector. ACKNOWLEDGMENT This work was supported by the R&D Program for Fusion Strategy of Advanced Technologies in Korea.

Received for review October 1, 2004. Accepted January 25, 2005. AC048535O

Analytical Chemistry, Vol. 77, No. 8, April 15, 2005

2495