Sie sind auf Seite 1von 6

Institute of Physics Publishing doi:10.

1088/1742-6596/34/1/043

Journal of Physics: Conference Series 34 (2006) 258263 International MEMS Conference 2006

Electro-magnetically Actuated Minute Polymer Pump Fabricated using Packaging Technology


G Balaji, A Singh and G K Ananthasuresh Mechanical Engineering, Indian Institute of Science, Bangalore 560012, India {gbalaji, aksingh, suresh}@mecheng.iisc.ernet.in
Abstract. Design, fabrication and preliminary testing of a flat pump with millimetre thickness are described in this paper. The pump is entirely made of polymer materials barring the magnet and copper coils used for electromagnetic actuation. The fabrication is carried out using widely available microelectronic packaging machinery and techniques. Therefore, the fabrication of the pump is straightforward and inexpensive. Two types of prototypes are designed and built. One consists of copper coils that are etched on an epoxy plate and the other has wound insulated wire of 90 m diameter to serve as a coil. The overall size of the first pump is 25 mm 25 mm 3.6 mm including the 3.1 mm-thick NdFeB magnet of diameter 12 mm. It consists of a pump chamber of 20 mm 20 mm 0.8 mm with copper coils etched from a copper-clad epoxy plate using dry-film lithography and milled using a CNC milling machine, two passive valves and the pump-diaphragm made of Kapton film of 0.089 mm thickness. The second pump has an overall size of 35 mm 35 mm 4.4 mm including the magnet and the windings. A breadboard circuit and DC power supply are used to test the pump by applying an alternating square-wave voltage pulse. A water slug in a tube attached to the inlet is used to observe and measure the air-flow induced by the pump against atmospheric pressure. The maximum flow rate was found to be 15 ml/min for a voltage of 2.5 V and a current of 19 mA at 68 Hz.

1. Introduction Miniature pumps are important components of micro-fluidic systems such as lab-on-a-chip, micrototal-analysis systems, implantable drug-delivery devices, etc. Several transduction principles, materials and fabrication methods have been used to develop proof-of-concept miniature pumps. Laser and Santiago [1] recently reviewed the developments from the first reported micro pump in 1986 to 2004. They noted that there are no micro pumps that best suit certain applications despite the rapid development in this field. Flow rates, pressure against which pumping is to take place, precision of the delivered volume, packaged size, power consumed, bio-compatibility, etc., are some of the characteristics that help choose a pump for a given application. The cost of the fabrication and packaging are also equally important. In this paper, we report on a millimetre-thick flat pump that is made of polymer materials (except the magnet and coils) and fabricated using electronic packaging machinery and techniques. Consequently, the cost of fabrication is low. There are a few earlier attempts of polymer micro pumps. One of them was reported by Bohm et al. [2] who used conventional techniques such as punching, milling and injection-molding. They used both piezoelectric and electromagnetic actuation. Boden et al. [3] used paraffin-actuation for plastic pump of about 0.15 cm3 size that gave a flow rate 74 nl/min. Xia et al. [4] used electroactive polymer material for their micro pump. In comparison to the large number of reported micro pumps (more than
2006 IOP Publishing Ltd 258

259

260), there are relatively very few polymer pumps. Polymer materials are attractive for many reasons, which include fabrication without the need for the expensive clean room environment, amenability to use inexpensive macro-machining techniques, bio-compatibility for implantable applications, etc. The strength-stiffness ratio is also very high for polymers when compared with silicon, metals and ceramics. This implies that less power is consumed for the diaphragm-based positive-displacement type pumps for desired change of volume in the pump-chamber. While piezo-actuation is a predominantly used technique for micro pumps, electromagnetic actuation has its own advantages. Unlike piezo-actuation, electromagnetic actuation does not require very high voltages. The power consumed by the pump alone may be more in electromagnetic actuation but when one takes the overall system including the power electronics, it may be even competitive with other actuations [5]. One of the significant advantages is that electromagnetic actuation does not need special materials: a simple permanent magnet and coils will suffice. It is particularly well suited for an all-polymer pump such as the one presented here. The micro pumps reported in [6-8] also use electromagnetic actuation. The pumps in [7] and [8] use Kapton diaphragm integrated with a Low Temperature Co-fired Ceramic (LTCC) tapes. These designs used coil on the diaphragm and had the magnet fixed to the substrate. This was found to be slightly disadvantageous due to the over-heating of the diaphragm and the consequent thermally induced deformation effects. The remainder of the paper is organized as follows. In Section 2, the constructional features of the pump are given along with two different designs that differ in the types of coils used. Section 3 describes the fabrication procedures in detail. Testing and results are in Section 4. Concluding remarks are in Section 5. 2. Design Two different embodiments of the pump are considered in this work. These are shown in figures 1 and 2. Both designs have a stationary coil and a moving magnet mounted on a Kapton diaphragm. The difference between the two lies in the way coils are practically realized. In the first design (see figure 1a), a copper-clad epoxy plate is chosen as the main substrate. This plate is 1.2 mm thick and had copper layer of 17.5 m on one side. A square spiral coil is etched on the copper side of this plate. There exists a milled cavity on the other side. A spacer plates with a square window cutout in the middle along with the milled cavity forms the pumps chamber. Figure 1a shows this along with inlet and outlet passive valves and Kapton diaphragm, which is attached between the first spacer plate and a second spacer plate glued on the top side. A permanent magnet of diameter of 12 mm and thickness of 3.1 mm is attached at the center of the square diaphragm. Passive inlet and outlet check valves were made using strips of Kapton film placed over 1 mm-diameter holes drilled into the pump-chamber plate. Kapton diaphragm Magnet Spacer plate 2 Spacer plate 1 Etched copper coils Bottom epoxy plate with milled cavity

Inlet valve

Outlet valve

Figure 1. Features of the first embodiment of the pump In the second design, coils are wound in an annular recess made in the epoxy plate. For the coil, we used 43-gauge enamel-coated copper wire of 90 m diameter is used. Approximately 400 turns of this wire were manually wound into a coil of average diameter of 9 mm and a height of 2 mm. This coil is buried under the pump chamber. The Kapton diaphragm, the magnet, passive valves and the top-

260

spacer plate are the same as in the previous design. Compared to the etched spiral coil of the first design, the wound coil has the advantage that its constant radius can be chosen to be the optimum radius that has the largest magnetic field near to the magnet. This helps in reducing the electrical power needed to actuate the pump and also helps in preventing undue heating of the pump Kapton diaphragm Magnet Spacer plate 2 Spacer plate 1 Wound copper coils Bottom epoxy plate with milled cavity

Inlet valve

Outlet valve

Figure 2. Features of the second embodiment of the pump

3. Fabrication The fabrication procedures for the first and second types of pumps are slightly different although many of the steps are common to both. The procedure for the first design consists of the following steps. (i) A lithography mask (figure 3a) is made using photo-plotting on a transparency sheet from a Gerber file. The minimum spacing of the 17.5 m thick and 200 m wide coils was 100 m. Using this mask, the image is transferred onto the copper side of the epoxy plate of size 30 mm 30 mm 1.2 mm. The image-transfer can be done using wet or dry lithography methods. In the wet-lithography method, the epoxy sheet is spin-coated with a positive photoresist for one minute. It is then pre-baked and exposed to UV light through the mask followed by photoresist-development and post-baking. In the dry-lithography, the epoxy sheet is laminated with a dry photoresist film. It is then exposed to the UV light for 10 s through the mask. After that it is passed through a developer spraying machine for 10 s. (ii) The patterned copper layer can be etched using ferric chloride solution (250 g of FeCl3 in one litre of DI water and 10% of HCl) in a beaker. Or, the patterned epoxy plate can be passed through a copper etchant (CuCl 2 + HCl + H 2 O 2 )-spraying, conveyerized machine. Then, the photoresist is stripped using acetone or peeled off in NaOH. The etched copper coil is shown in figure 4. (iii) After patterning the copper coating, a pocket was milled into the epoxy plate (figure 3b) to a depth of 0.8 mm. Two holes of 1 mm diameter were drilled to serve as inlet and outlet at the bottom of the milled cavity. (iv) A small strip of 1 mm wide Kapton film was carefully glued on the inside of the milled cavity over the inlet hole. The strip should be movable when the pressure inside the chamber is less than the outside pressure. (v) Another Kapton film was cut to size and glued on top of the first spacer plate which in turn is over the epoxy plate with the coil. It is important to ensure that the Kapton film is in a stretched condition to be effective as the diaphragm. Although it is not imperative, adding a second spacer plate on top of the Kapton diaphragm helps in securing the diaphragm in the taut condition. (vi) The next step is to attach the magnet at the centre of the diaphragm. (vii) A small strip of Kapton film was affixed over the outlet hole on the outer side of the epoxy plate with the coil. This serves as the outlet check valve. (viii) Plastic tubing is epoxy-glued over the inlet and outlet holes.

261

(ix)

The last step is to solder the electrical wires to the ends of the coil. The wires are connected to the breadboard circuit that supplies the required electrical power to the pump.

The fabrication of the second design consists of fewer steps. It begins with Step (iii) wherein a cavity and an annular recess are milled into the epoxy plate and inlet and outlet holes are drilled. The wound coil of enamel-coated 43-gauge wire is placed into the annular recess (figure 5). The inlet check valve is glued as in Step (iv). The remaining steps, i.e., (v)-(ix), are the same as they were for the first design. The prototypes of the two pumps are shown in figures 6a-d.

(b) (a) Figure 3. (a) Photo-mask showing the square spiral (b) the CNC-milled cavity in the epoxy plate.

Outlet check valve

Figure 4. Copper coil etched on the epoxy plate. Thin strip of Kapton acting as the outlet valve is also shown in the figure.

Figure 5. The wound coil of 43-gauge enamel-coated copper wire 4. Testing and results In order to test the pumps for air-flow rate, a breadboard circuit was developed to supply the electrical power to the pump. The circuit amplifies the small-amplitude square or sinusoidal signal of the HP 33220A function generator by using two transistors connected in a cascade (Darlington pair) arrangement. The biasing voltage is supplied by HP 34410A DC power supply. The frequency of the signal can be varied over a range of 0 100 Hz. The range of the voltage is from 1.5 5 V. The current drawn can be measured using the same power supply unit. In order to measure the flow rate, we attached a long plastic tube in which a colored water slug is introduced. The tube is set against a ruler so that the movement of the slug can be timed. This is shown in figure 7. The frequency of the signal is varied at different voltage levels. The measured flow rate against frequency is plotted in figure 8 for pump 2. As it can be seen, it has a maximum flow rate of 4.8 ml/min at a frequency of 68 Hz. The voltage and current for this measurement were 2.5 V and 19 mA respectively. Pump 1 requires much higher voltage, draws more current and consequently suffers from the heat-induced effects and consumes more power for the same performance as pump 2.

262

(a) (b) (c) (d) Figure 6. (a) Pump 1 with etched coils (b) Side-view of pump 1 (c) Pump 2 with wound coils (d) Side-view of pump 2

Figure 7. The test setup for the pump along with the water slug arrangement to measure the flow rate against atmospheric pressure.
16 2.5 V 2.0 V 1.5 V

14

12

Flow Rate (ml/min)

10

0 10

20

30

40

50

60

70

Frequency (Hz)

Figure 8. Air-flow rate of pump 2 at different frequencies and voltages

263

5. Closure The design, fabrication and testing of an electromagnetically actuated minute pump are described in this paper. The key feature of this pump is that it is flat with a thickness of 4 mm that could be reduced even further with design improvements. It is made entirely of polymer materials excepting the magnet and the coil. The measure flow rate was 15 ml/min for a pump of size 5.39 cm3. The fabrication procedure does not need clean room environment and uses mostly electronic packaging machinery. As shown in pump 2, the lithography step may also be avoided. Currently, the winding of the coil in pump 2 is done manually but it too can be automated for a batch-fabrication process. The performance of the pump can be improved further through modeling and design, which will be taken up in our future work. This may help reduce the lateral size as well as thickness of the overall pump. The pump has not yet been tried with the liquids which is our next step. Acknowledgments This work was supported by a grant from the National Program on Smart Materials (NPSM) by the Government of India (Grant # NPSM PARC 5.8). The authors would like to thank Center for Electronics Devices and Technology (CEDT) and Microfabrication Laboratory of the Indian Institute of Science for allowing us to use their facilities and equipment. Technical help by Mr. Anthony Swamy (CEDT) and Mr. Dwarakanath (Microfab) is gratefully acknowledged. References [1] Laser, D. J. and Santiago, J. G., A Review of Micropumps, Journal of Micromechanics and Microengineering, Vol. 14 (2004), pp. R35-R64. [2] Bohm, S., Olthuis, W., and Bergveld, P., A Plastic micropump constructed with conventional techniques and materials, Sensors and Actuators A, 77 (1999), pp. 223-228. [3] Boden, R., Lehto, M., Simu, U., Thornell, G., Hjort, K. and Schweitz, J.-A., A Polymeric Paraffin Actuated High-Pressure Micropump, Sensors and Actuators A, 2006 (article in press). [4] Xia, F., Tadigadapa, S., and Zhang, Q. M., Electroactive Polymer Based Microfluidic Pump, Sensors and Actuators A, 125 (2006), pp. 346-352. [5] Busch-Vishniac, I. J., The Case for Magnetically Driven Microactuators, Sensors and Actuators A, 33 (1992), pp. 207-220. [6] Gong, Q., Zhou Z., Yang, Y., Wang, X., Design, optimization, and simulation on electromagnetic pump, Sensors and Actuators A, 83(2000), pp. 200-207. [7] Kim, M., Ananthasuresh, G. K., and Bau, H. H., Electromagnetic Pump Fabricated with Lowtemperature Co-Fired Ceramic Tapes and Kapton polyimide Film, IMECE 2000, MEMS 2000 Symposium Proceedings, pp. 249-254, 2000. [8] Bau, H. H., Ananthasuresh, G. K., Santiago-Aviles, J., Zhong, J., Kim, M., Yi, M,. EspinozaVallejos, P., and Sola-Laguna, L., Ceramic Tape Based Meso systems Technology, Micro-Electro-mechanical Systems, 1998 Int. Mechanical Eng. Conf. and Exp., pp. 491-498, 1998.

Das könnte Ihnen auch gefallen