Imaging of Electroosmotic Flow in Plastic Microchannels - Analytical

May 1, 2001 - Eileen T. Dimalanta, Alex Lim, Rod Runnheim, Casey Lamers, Chris Churas, Daniel K. Forrest, Juan J. de Pablo, Michael D. Graham, Susan N...
1 downloads 14 Views 284KB Size
Anal. Chem. 2001, 73, 2509-2515

Imaging of Electroosmotic Flow in Plastic Microchannels David Ross, Timothy J. Johnson, and Laurie E. Locascio*

National Institute of Standards & Technology, 100 Bureau Drive, Gaithersburg, Maryland 20899

We have characterized electroosmotic flow in plastic microchannels using video imaging of caged fluorescent dye after it has been uncaged with a laser pulse. We studied flow in microchannels composed of a single material, poly(methyl methacrylate) (acrylic) or poly(dimethylsiloxane) (PDMS), as well as in hybrid microchannels composed of both materials. Plastic microchannels used in this study were fabricated by imprinting or molding using a micromachined silicon template as the stamping tool. We examined the dispersion of the uncaged dye in the plastic microchannels and compared it with results obtained in a fused-silica capillary. For PDMS microchannels, it was possible to achieve dispersion similar to that found in fused silica. For the acrylic and hybrid microchannels, we found increased dispersion due to the nonuniformity of surface charge density at the walls of the channels. In all cases, however, electroosmotic flow resulted in significantly less sample dispersion than pressure-driven flow at a similar velocity. The majority of microfluidic devices reported in the literature as well as those now commercially available have been fabricated using silica-based materials.1-3 The advantage of these materials is that their surface and electroosmotic properties have been well characterized; therefore, the performance of these devices can be easily predicted and understood. Polymeric materials are increasingly being used in the fabrication of microfluidic devices, because they offer the possibility for the fabrication of inexpensive, disposable devices.4-8 Although some measurements of electroosmotic mobility have been made in polymer capillaries,9-11 the surface and electroosmotic proper* Corresponding author. E-mail: [email protected]. (1) Harrison, D. J.; Manz, A.; Fan, Z. H.; Ludi, H.; Widmer, H. M. Anal. Chem. 1992, 64, 1926-32. (2) Woolley, A. T.; Mathies, R. A. Proc. Natl. Acad. Sci. U.S. A. 1994, 91, 1134852. (3) Jacobson, S. C.; Hergenroder, R.; Koutny, L. B.; Ramsey, J. M. Anal. Chem. 1994, 66, 2369-73. (4) Roberts, M. A.; Rossier, J. S.; Bercier, P.; Girault, H. Anal. Chem. 1997, 69, 2035-42. (5) Martynova, L.; Locascio, L. E.; Gaitan, M.; Kramer, G. W.; Christensen, R. G.; Maccrehan, W. A. Anal. Chem. 1997, 69, 4783-89. (6) Locascio, L. E.; Perso, C. E.; Lee, C. S. J. Chromatogr.; A 1999, 857, 27584. (7) Xu, J. D.; Locascio, L.; Gaitan, M.; Lee, C. S. Anal. Chem. 2000, 72, 193033. (8) McCormick, R. M.; Nelson, R. J.; Alonsoamigo, M. G.; Benvegnu, J.; Hooper, H. H. Anal. Chem. 1997, 69, 2626-30. (9) Schutzner, W.; Kenndler, E. Anal. Chem. 1992, 64, 1991-95. 10.1021/ac001509f Not subject to U.S. Copyright. Publ. 2001 Am. Chem. Soc.

Published on Web 05/01/2001

ties of polymerssparticularly the sheet polymers that are used to make plastic microfluidic devicessare still, in general, not well known. Ultimately, this makes predicting the performance of polymeric devices more difficult. The performance of microfluidic devices is often based on the behavior of the fluid when it is subjected to electrokinetic pumping. This is because electrokinetic pumping is currently the simplest (because it requires no moving parts) and the most common method used to transport fluids within these devices.3,6,8,12-14 Pumping is achieved by applying an electric field along the length of a buffer-filled microchannel, with flow resulting from the motion of ions in the applied electric field. This phenomenon is referred to as electroosmosis, and it results from the formation of an electric double layer at the microchannel walls. The electric double layer arises when certain chemical groups, such as carboxylate groups, at the surface of the polymer are ionized by their contact with the buffer in the microchannel. Because these charge groups are part of the polymer substrate, they are immobile. However, oppositely charged ions in the solution are drawn toward the layer of immobile surface charges, so that a mobile layer of charge is formed near the surface. When an electric field is applied along the length of the microchannel, the ions in this mobile layer migrate toward the appropriate electrode, and the bulk fluid within the microchannel is carried along by viscous drag. Typically, the width of the electric double layer is very small (less than 1 µm), so that the electroosmotic flow can be considered to be a wall-driven phenomenon. For a perfectly uniform distribution of charge groups on the microchannel walls, this leads to a flow velocity that is constant everywhere in the microchannel. This is termed perfect plug flow and is characterized by dispersion that is dependent only on the molecular diffusion constant of the sample. Conversely, if the surface charge on the walls of the microchannel is nonuniform, the flow velocity will also be nonuniform, and the sample dispersion will be greater than that due to molecular diffusion. Obviously, increased dispersion degrades the performance of electrokinetic separations by impacting the resolution of closely eluted peaks. It also affects the ability to control the movement of discrete samples within a (10) Rohlicek, V.; Deyl, Z.; Miksik, I. J. Chromatogr.; A 1994, 662, 369-73. (11) Bayer, H.; Engelhardt, H. J. Microcolumn Sep. 1996, 8, 479-84. (12) Manz, A.; Effenhauser, C. S.; Burggraf, N.; Harrison, D. J.; Seiler, K.; Fluri, K. J. Micromech. Microeng. 1994, 4, 257-65. (13) Woolley, A. T.; Hadley, D.; Landre, P.; Demello, A. J.; Mathies, R. A.; Northrup, M. A. Anal. Chem. 1996, 68, 4081-86. (14) Waters, L. C.; Jacobson, S. C.; Kroutchinina, N.; Khandurina, J.; Foote, R. S.; Ramsey, J. M. Anal. Chem. 1998, 70, 158-62.

Analytical Chemistry, Vol. 73, No. 11, June 1, 2001 2509

microfluidic device. One of the reasons that electrokinetic pumping is so often used in microfluidics is that much lower sample dispersions can be achieved compared with pressure-driven pumping. Consequently, to successfully develop microfluidic devices made of polymeric materials, it is essential to understand the effects of different materials and fabrication methods on the density and distribution of surface charges that ultimately determine the rate of dispersion under electrokinetic flow. Microfluidic devices are typically fabricated in two steps. First, the network of channels for the device is etched, imprinted, or otherwise machined into the face of one sheet of material. Then a second sheet of material is placed over the channels to seal them, and the two pieces of material are bonded together. With plastics, it may be difficult to do this using two pieces of the same material, because the heat or solvent used for bonding will often degrade the structure of the microchannels. For this reason, polymeric microchannels are often made using two or more different materials.4,7 Since the different materials will, in general, have different surface charge densities,8 this could lead to flow profiles that differ from an ideal plug flow. In addition, it has been shown that some microchannel fabrication methods can result in markedly nonuniform surface charge densities even in channels fabricated entirely of the same material.15 A few groups, including ours, have examined electroosmotic flow in polymeric microchannels;4,6,16,17 however, the reported measurements are only of the average bulk electroosmotic velocities and do not investigate sample dispersion. As yet, the only detailed studies involving flow profiling and sample dispersion in micrometer-sized systems have been performed in fused-silica capillaries.18-20 Here, we have adapted the flow visualization technique of Paul et al.18 for use in plastic microfluidic devices. The results presented here examine in detail electroosmotic flow in straight microchannels made of acrylic and PDMS, as well as hybrid acrylic/PDMS channels. EXPERIMENTAL SECTION Chemicals and Materials. Fluorescein bis(5-carboxymethoxy2-nitrobenzyl) ether dipotassium salt (CMNB-caged fluorescein dye), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDAC), and 5-(aminoacetamido)fluorescein were used as supplied by Molecular Probes (Eugene, OR). All buffer solutions were made using deionized water from a Millipore Milli-Q system (Bedford, MA). Carbonate buffer, 200 mmol/L at pH 9.4, was diluted to make the 20 mM buffer used for most of the work described here. Caged dye solutions were prepared by dissolving ∼1 mg of CMNB-caged fluorescein solid in 500 µL of the carbonate buffer and were filtered before use with syringe filters (pore size 0.8 µm). Microchannels were made using UV-transparent poly(methyl methacrylate) (acrylic) sheet (Acrylite OP-4, Cyro Industries, Mt. (15) Branham, M. L.; MacCrehan, W. A.; Locascio, L. E. J. Capillary Electrophor. Microchip Technol. 1999, 6, 43-50. (16) Wang, S. C.; Perso, C. E.; Morris, M. D. Anal. Chem. 2000, 72, 1704-6. (17) Henry, A. C.; Tutt, T. J.; Galloway, M.; Davidson, Y. Y.; McWhorter, C. S.; Soper, S. A.; McCarley, R. L. Anal. Chem. 2000 72, 5331-7. (18) Paul, P. H.; Garguilo, M. G.; Rakestraw, D. J. Anal. Chem. 1998, 70, 245967. (19) Herr, A. E.; Molho, J. I.; Santiago, J. G.; Mungal, M. G.; Kenny, T. W.; Garguilo, M. G. Anal. Chem. 2000, 72, 1053-7. (20) Tallarek, U.; Rapp, E.; Scheenen, T.; Bayer, E.; Van As, H. Anal. Chem. 2000, 72, 2292-301.

2510

Analytical Chemistry, Vol. 73, No. 11, June 1, 2001

Arlington, NJ) and poly(dimethylsiloxane) (PDMS) prepared according to product information from a Sylgard 184 silicone elastomer kit (Dow Corning, Midland, MI) and cured at room temperature for 1-3 weeks. Fused-silica capillary tubing (360µm o.d.; 50-µm i.d.) was obtained from Polymicro Technologies Inc. (Phoenix, AR). Microchannel Fabrication. Three different types of plastic microchannels were made. In general, the microchannels were made by forming the imprint of the channel into the face of one sheet of material and then using a second sheet of material, the “lid”, to cover and seal the channel. Circular holes in the lid, 3-mm diameter, provided access to the channels and served as fluid reservoirs. Acrylic microchannels were hot-imprinted at 110 °C and 5.1 × 106 Pa (740 psi) for 1 h using an anisotropically etched silicon template5 as the imprinting tool. After imprinting, the acrylic microchannel and lid were rinsed with ethanol and dried. Then the lid was placed on top of the channel, and the two pieces were clamped together between glass slides and bonded by placing them in a circulating air oven at 103 °C for 12 min. To form microchannels in PDMS, the Sylgard 184 silicone elastomer was poured onto a similar silicon template and allowed to cure. The PDMS channels were sealed with lids of cured PDMS. Hybrid microchannels were prepared using imprinted acrylic channels and PDMS lids. PDMS lids naturally adhered to the flat surfaces of the imprinted acrylic and PDMS sheets so that no additional bonding step was necessary. Typical channels (2-5 cm in length) had a trapezoidal cross section defined by the silicon template5 and were 30 µm deep, 20 µm wide at the bottom, and 75 µm wide at the top. The fused-silica capillaries used in this study were stripped of their polyimide coating and embedded in the middle of a sheet of PDMS to reduce beam steering effects due to the curved walls of the capillary and to provide an object geometrically similar to the plastic microchannels. Fluorescent Labeling of Microchannels. The EDAC and fluorescein were dissolved in 100 mM phosphate buffer, pH 7, to make a solution with a final concentration of 0.5 mM EDAC and 0.5 mM fluorescein. The EDAC was used to facilitate the binding between the carboxylic acid groups on the surface of the polymer and the amino group of the fluorescein. The use of EDAC has been reported for the probing of carboxylic acid groups of a cyclosporin derivative21 and sugar carboxylates.22 The protocol used for labeling the polymer substrate microchannels with EDAC and aminofluorescein was a modification of a previously reported procedure.23 The sample was immersed in the EDAC/aminofluorescein solution for 15 h followed by an agitated rinse with 100 mM phosphate buffer, pH 7, then 138 mM carbonate buffer, pH 9.5, and then a final agitated rinse with the phosphate buffer. The channels were then filled with the phosphate buffer under a glass cover slip. The fluorescence associated with the carboxylate surface groups was then viewed by fluorescence microscopy using a mercury lamp and a fluorescein filter set and a CCD camera for detection. (21) Paprica, P. A.; Margaritis, A.; Petersen, N. O. Bioconjugate Chem. 1992, 3, 32-36. (22) Kobayashi, M.; Chiba, Y. Anal. Chem. 1994, 33, 189-94. (23) Hermanson, G. Bioconjugate Techniquies; Academic Press: San Diego, CA, 1996.

Figure 1. Schematic view of the imaging and uncaging apparatus.

Image Acquisition. A schematic of the experimental apparatus is shown in Figure 1. Uncaging of the caged fluorescein dye in the microchannels was performed using the output of a pulsed nitrogen laser (337 nm, pulse duration 1 month) show clear signs of aging effects. They are generally less pliable and adhere less to surfaces than PDMS samples cured for just 1 or 2 weeks.

4.6 ( 0.3

a

Listed uncertainties are the standard deviation resulting from a linear least-squares fit to the data of Figure 7 (or similar data) for field strengths less than 200 V/cm. Additional systematic uncertainty, primarily from the determination of the applied field, is estimated to be less than 3% of the stated values.

In all cases, the mobility is not constant, even at low fields. This result is because the electroosmotic mobility is also proportional to 1/η and, therefore, increases with increasing buffer temperature. It is important to note that the variation of µEO with field strength is small at low electric fields, so that plots of velocity versus field appear linear. The values of µEO extrapolated to zero field for the various microchannels studied are listed in Table 1. Additional data was taken with 200 mM carbonate buffer, pH 9.4, and 20 mM phosphate buffer, pH 7. These data are also listed in Table 1.

ACKNOWLEDGMENT The authors acknowledge the financial support of the NRC/ NIST postdoctoral Research Program. The authors also acknowledge Michael Gaitan from the Semiconductor Electronics Division at NIST for the production of the silicon template. Disclaimer: Certain commercial equipment, instruments, or materials are identified in this report to specify adequately the experimental procedure. Such identification does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.

Received for review December 20, 2000. Accepted March 5, 2001. AC001509F

Analytical Chemistry, Vol. 73, No. 11, June 1, 2001

2515