Pneumatic Flow Switching on Centrifugal Microfluidic Platforms In

Jan 10, 2011 - ABSTRACT: This paper describes a flow switching technique ... for flow control at a T-shaped junction between one inlet channel and two...
0 downloads 0 Views 5MB Size
Download PDF
TECHNICAL NOTE pubs.acs.org/ac

Pneumatic Flow Switching on Centrifugal Microfluidic Platforms In Motion Matthew C. R. Kong and Eric D. Salin* Department of Chemistry, McGill University, 801 Sherbrooke Street West, Montreal, Quebec, Canada, H3A 2K6 ABSTRACT: This paper describes a flow switching technique applicable to centrifugal microfluidic platforms, using a regulated stream of compressed gas. This pneumatic flow switching technique allows for flow control at a T-shaped junction between one inlet channel and two outlet channels. This technique provides a noncontact, robust, and efficient method for switching the direction of fluid flow while a disk is rotating at relatively low frequencies. The switching operation can be implemented reproducibly with applied gas flow rates between 17 and 58 L min-1 and rotational frequencies between 400 rpm (6.6 Hz) and 1200 rpm (20 Hz).

M

icrofluidic devices offer considerable potential for providing analytical systems which are low in cost both to manufacture and to operate. The reduced dimensions can provide short diffusion times resulting in short analysis times. The ultimate goal for many is the development of micro total analytical systems (μTAS)1 which provide a complete or near complete analytical system on a single microfluidic platform. Devices that use centrifugal force for fluid-flow control (pumping) are an important subset of μTAS type devices. Centrifugal microfluidic (CM) devices usually take the form of disks, and unlike most other μTAS devices, do not require direct connections to them. This makes CM devices particularly attractive platforms for portable applications as well as encouraging a high degree of parallelism, because any pumping action is uniform over any given radius of the disk. The pumping action on disks is also relatively insensitive to the type of liquid,2 suggesting that, in addition to needing no connections, liquids with varying properties can be pumped. There are, however, some limitations to CM devices. CM fluid flow operations generally move radially outward on these devices, with the flow often controlled by valves. Passive capillary valves are most commonly used, however, wax valves3 and other hydrophobic valving techniques have previously been described.2 Above a threshold rotational frequency, passive capillary valves allow liquid to flow toward the edge of the CM device. Since fluid flow (other than siphon action) is generally unidirectional, there is very little operational flexibility in fluid flow on CM devices. Therefore, developing a flow switching technique would allow for more sophisticated and flexible sequences of operations (e.g., reagent mixing, separations), as well as possibly altering the analytical sequence based on intermediate results. Clinical and biological analyses performed on CM devices have demonstrated a demand for switching the direction of fluid flow in a controllable and reproducible manner while maintaining continuous rotation of the disks.4,5 Also, being able to control fluid flow for r 2011 American Chemical Society

the detection of biological agents on immunoassays is critical in certain applications.6 Furthermore, Haberle and Bouchard have shown that continuous liquid flow addition to CM devices in motion is possible,7,8 offering enormous potential for building CM devices with a minimum number of reservoirs. To fully take advantage of techniques like this, a switching system which allows for directing of fluid flow from an external source to one of two target chambers may be used. To date, the use of Coriolis force has been demonstrated as a potential method for flow switching.9,10 However, both papers demonstrate that relatively high rotational frequencies (∼15 to 55 Hz) and a change in the disk's rotational direction are required for the switching operation to take place. Generally, it is undesirable that high rotational frequencies are required to effect CM operations as these may trigger capillary valves, thereby limiting design options for the CM device. The dimensions and geometries of disk features were optimized by Kim,10 significantly lowering the frequencies (to ∼15 Hz) at which the Coriolis switch operates; however, very specific dimensions for disk designs were required, reducing the fabrication flexibility as a whole. Also, if a CM device is stopped, as required by the Coriolis switching technique, siphons in the system may be activated, further limiting the options for disk designs. A desirable switching technique would be one that is independent of rotational frequency and the direction of disk rotation. In a recent paper, we discussed a new method of pumping fluids on a CM device while it is in motion.11 This technique was used to pneumatically pump liquids from the outer edge of a disk back toward the center during disk rotation, extending the number of sequential operations that can be performed on a CM system. Using the same philosophy, we report here a Received: September 28, 2010 Accepted: December 13, 2010 Published: January 10, 2011 1148

dx.doi.org/10.1021/ac102563g | Anal. Chem. 2011, 83, 1148–1151

Analytical Chemistry

TECHNICAL NOTE

Figure 1. Schematic of experimental setup. (a) Cylinder containing compressed air. (b) Rotameter used to measure flow rate. (c) Centrifugal microfluidic platform. (d) Servo motor and spindle shaft. Reprinted from ref 11. Copyright 2010 American Chemical Society.

pneumatic CM flow switching technique which operates over a wide range of relatively low rotational frequencies and gas flow rates.

’ EXPERIMENTAL SECTION Reagents. Blue commercial food coloring was mixed with distilled deionized water (DDW) to simulate an analytical sample that was clearly visible. A.C.S. grade methyl red (sodium salt) (Fisher Scientific, NJ, USA) and A.C.S. grade hydrochloric acid (ACP Chemicals, QC, Canada) were used as indicated. Fabrication. A polycarbonate disk platform was fabricated using materials and techniques described in detail by LaCroixFralish.12 Data Acquisition. High-speed digital images were obtained using a motorized stage and strobe system developed by Duford13 supplemented with a color digital camera (GRAS-14S5C-C, Point Gray, BC, Canada). The motorized stage, strobe, and camera were controlled by a LabVIEW program (LabVIEW 8.6, Developer Version, National Instruments, QC, Canada). Absorbance measurements were made using the technique and apparatus described in detail by LaCroix-Fralish,12 which includes a deuterium light source (DT-MINI-2-GS, Ocean Optics, Dunedin, FL, USA) with a USB spectrometer (USB4000-UV-vis, Ocean Optics, Dunedin, FL, USA). Qualitative Experiments. During our qualitative investigation of this pneumatic switching technique, extra dry compressed air (MEGS, QC, Canada) flowing at a steady rate of 0.8 SCFM (standard cubic feet per minute) (∼23 L min-1) flowed through an outlet tube (160 cm in length, 1/8 in. (3.175 mm) inner diameter, 1/4 in. (6.35 mm) outer diameter). The flow rate was controlled by a standard regulator and measured by a rotameter (FL7311, Omega, QC, Canada). The outlet tube was positioned directly above and perpendicular to the disk, at a distance of 7.5 mm from the surface, and at the radial position of one of the two air vent holes depending on the desired switching operation. A schematic of the experimental configuration is shown in Figure 1.11

Figure 2. Schematic and dimensions of the centrifugal microfluidic platform used for flow switching. (a) Configuration with dimensions (mm) of the demonstration pneumatic flow switch design. (b) Schematic of the centrifugal microfluidic platform.

A schematic of the disk design used for flow switching is shown in Figure 2. The design consisted of an Initial Chamber (4.0 mm radius, 1.4 mm deep), into which 68 μL of blue colored DDW was injected, and symmetrical left and right Target Chambers (5.0 mm radii, 1.4 mm deep), into which the sample would flow depending on a chosen switching operation. Connecting the Initial Chamber and the two Target Chambers was a T-shaped junction consisting of an inlet channel (6.0 mm length, 100 μm deep) and two outlet channels (11.6 mm length each, 700 μm deep). The rotational frequency used was 400 rpm, with the inlet channel acting as a passive valve (burst frequency ∼300 rpm) for the Initial Chamber. The air vent holes (0.5 mm radii, 700 μm deep) for the left and right Target Chambers were positioned at 36.2 and 41.2 mm from the center of the disk, respectively. Quantitative Experiments. Four sets of quantitative experiments were carried out to determine the functional limits of the pneumatic flow switch. First, the switching operation of 68 μL of blue colored DDW was tested at an air flow rate of 0.8 SCFM, at rotational frequencies between 400 and 1800 rpm, 1149

dx.doi.org/10.1021/ac102563g |Anal. Chem. 2011, 83, 1148–1151

Analytical Chemistry

Figure 3. Strobed images of the disk spinning (400 rpm) without air pressure applied; t represents time elapsed. (a) Initial state of disk, t = 0 s. (b) Intermediate state of disk, 0 s < t < 60 s. (c) Final state of disk, t = 60 s.

at 100 rpm intervals. Second, the upper limits of rotational frequencies were determined at air flow rates of 0.6, 1.2, 1.4, and 1.6 SCFM. Third, the amount of residual liquid left in the Initial Chamber was investigated by sequentially flowing 68 μL of acidified methyl red solution and 50 μL of water into the right and left Target Chambers, respectively, with an air flow rate of 0.8 SCFM and rotational frequency of 800 rpm. Absorbance measurements were made at 519 nm on both the original methyl red solution and water containing residual amounts of methyl red to determine the amount of residual methyl red solution left in the Initial Chamber. Finally, the switching operation was tested using 10 and 15 μL volumes, with an air flow rate of 0.8 SCFM and rotational frequency of 800 rpm.

’ RESULTS AND DISCUSSION It has been previously demonstrated that, without the application of air pressure, fluid flow would primarily be dependent on centrifugal and Coriolis forces.9,10 However, using our fabrication techniques and disk designs, fluid flow without an applied air pressure was unpredictable and irreproducible. An example is shown in the strobed images in Figure 3, whereby a slightly greater liquid level was observed in the left Target Chamber than in the right Target Chamber. Qualitative Experiments. The qualitative experiments demonstrated that, when air pressure was applied to the air vent hole of the left Target Chamber while the disk was spinning, the switching operation was effective and there was near exclusive transfer of liquid from the initial chamber to the right Target Chamber (Figure 4). Similarly, nearly all the liquid was transferred to the left Target Chamber when air pressure was applied to the air vent hole of the right Target Chamber (Figure 5). Quantitative Experiments. A set of experiments was conducted to determine the range of frequencies and flow rates which were effective for switching. The first set of experiments showed that, with an air flow rate of 0.8 SCFM, flow switching was achieved at rotational frequencies between 400 and 1000 rpm, regardless of the direction of disk rotation (clockwise or counterclockwise). The lower limit of 400 rpm was imposed by the burst valve frequency of the Initial Chamber and was independent of the air flow rate. The second set of experiments showed that, with an air flow rate of 0.6 SCFM, the switching operation was successful up to a maximum rotational frequency of 800 rpm. Even with increased air flow rates of 1.2, 1.4, and 1.6 SCFM, the switching operation was only successful up to 1200 rpm.

TECHNICAL NOTE

Figure 4. Strobed images of disk spinning (400 rpm) with air pressure (0.8 SCFM) applied to the air vent hole of the left target chamber; t represents time elapsed. (a) Initial state of disk, t = 0 s. (b) Intermediate state of disk, 0 s < t < 60 s. (c) Final state of disk, t = 60 s.

Figure 5. Strobed images of disk spinning (400 rpm) with air pressure (0.8 SCFM) applied to air vent hole of the right target chamber; t represents time elapsed. (a) Initial state of disk, t = 0 s. (b) Intermediate state of disk, 0 s < t < 60 s. (c) Final state of disk, t = 60 s.

Between 1300 and 1800 rpm, partial flow switching was achieved at best. The air pressure was unable to overcome the larger centrifugal force experienced by the liquid. The amount of air entering the system was potentially limited by the size of the air vent hole and the widths of the channels, leading to the observed 1200 rpm limit. The quantitative experiments showed that the pneumatic switching technique is effective over a wide range of rotational frequencies (400-1200 rpm) and over a wide range of applied air flow rates (0.6-1.6 SCFM). Third, to determine how much liquid was left behind at 800 rpm, experiments were conducted by following an initial transfer of 68 μL to one side by another transfer of 50 μL of water to the opposite chamber. From the absorbance measurements made, the amount of residual liquid left in the Initial Chamber was determined to be 1.9 μL (2.8%). The fourth set of experiments showed that, under identical conditions as described above, this flow switching technique is also applicable to significantly lower volumes of 10 and 15 μL. The operational limits for the pneumatic flow switch could potentially be improved by optimizing the dimensions of the disk design and the pneumatic apparatus. However, a major advantage to implementing the pneumatic flow switching technique is simplicity of design combined with a wide range for operational parameters. A potential concern involving the use of a pressurized gas is contamination. However, this can be resolved using a filtered gas source or carrying out the experiment in a clean environment. 1150

dx.doi.org/10.1021/ac102563g |Anal. Chem. 2011, 83, 1148–1151

Analytical Chemistry

TECHNICAL NOTE

An important advantage of this pneumatic switch is that it can be effectively implemented while the disk is continuously spinning in either direction, thus eliminating the need to stop the disk during operation. This offers an advantage because stopping the disk could result in undesirable fluid flow caused by the abrupt loss of centrifugal force.

’ CONCLUSIONS A pneumatic flow switch has been demonstrated on a centrifugal microfluidic platform in motion. This flow switch can be implemented at a wide range of gas flows and relatively low rotational frequencies. The design is simple and operates independently of the direction of disk rotation while not requiring that the disk be stopped or suffer a direction change. This flow switch allows for greater control of fluid flow on centrifugal microfluidic platforms and allows operations of greater complexity to be carried out on these platforms.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Tel: þ1-514-398-6236. Fax: þ1-514-398-3797.

’ ACKNOWLEDGMENT The authors wish to thank the Natural Sciences and Engineering Research Council of Canada for support under the Discovery Grants Program. ’ REFERENCES (1) Arora, A.; Simone, G.; Salieb-Beugelaar, G. B.; Kim, J. T.; Manz, A. Anal. Chem. 2010, 82, 4830–4847. (2) Madou, M.; Zoval, J.; Jia, G.; Kido, H.; Kim, J.; Kim, N. Ann. Rev. Biomed. Eng. 2006, 8, 601–628. (3) Park, J.-M.; Cho, Y.-K.; Lee, B.-S.; Lee, J.-G.; Ko, C. Lab Chip 2007, 7, 557–564. (4) Steigert, J.; Grumann, M.; Brenner, T.; Mittenbuehler, K.; Nann, T.; Ruehe, J.; Moser, I.; Haeberle, S.; Riegger, L.; Riegler, J.; Bessler, W.; Zengerle, R.; Ducree, J. JALA 2005, 10, 331–341. (5) Zhang, J.; Guo, Q.; Liu, M.; Yang, J. J. Micromech. Microeng. 2008, 18. (6) En Lin, S. I. Microfluid. Nanofluid. 2010, 9, 523–532. (7) Haeberle, S.; Brenner, T.; Schlosser, H. P.; Zengerle, R.; Ducree, J. Chem. Eng. Technol. 2005, 28, 613–616. (8) Bouchard, A. P.; Duford, D. A.; Salin, E. D. Anal. Chem. 2010, 82, 8386–8389. (9) Brenner, T.; Glatzel, T.; Zengerle, R.; Ducree, J. Lab Chip 2005, 5, 146–150. (10) Kim, J.; Kido, H.; Rangel, R. H.; Madou, M. J. Sens. Actuators, B: Chem. 2008, 128, 613–621. (11) Kong, M. C. R.; Salin, E. D. Anal. Chem. 2010, 82, 8039–8041. (12) LaCroix-Fralish, A.; Clare, J.; Skinner, C. D.; Salin, E. D. Talanta 2009, 80, 670–675. (13) Duford, D. A.; Peng, D. D.; Salin, E. D. Anal. Chem. 2009, 81, 4581–4584.

1151

dx.doi.org/10.1021/ac102563g |Anal. Chem. 2011, 83, 1148–1151