Effects and applications of surface tension for fluidic MEMS components Ryan T. Marinis and Ryszard J. Pryputniewicz NEST NanoEngineering, Science and Technology CHSLT Center for Holographic Studies and Laser micro-mechatronics Mechanical Engineering Department Worcester Polytechnic Institute Worcester, MA 01609-2280 Phone: (508) 831-5125, Fax: (508) 831-5713, Email: rmarinis@wpi.edu ABSTRACT In areas of chemical and biological sensing, microelectromechanical systems (MEMS) technology has shown the capability of producing chemical labs-on-a-chip. In order to produce a successful lab-on-a-chip, there must be several functioning subsystems, including the sensing apparatus, fluidic system, and connection to the macro-world. As characteristic size of features is reduced to microns, such as in MEMS components, certain scaling effects must be considered. Specifically, when dealing with fluidic applications, surface tension becomes an important consideration. Surface tension forces scale as L 1, which makes them extremely important as the length (L) scale decreases. If properly considered, such forces may become useful for fluidic control through either pumps or valves. By utilizing the electrocapillary effect of fluids it is possible to control liquids by applying a potential over the liquid and an electrode through electrowetting (EW) or electrowetting on dielectric (EWOD). EW can be applied in a channel wall, essentially producing controlled capillary effect. By controlling fluid through surface tension there is a reduced need for mechanical pumps and valves. Such mechanical features may be difficult to fabricate and also may, as a result of wear, shorten the life of the system. An approach based on analytical, computational, and experimental solutions (ACES) methodology has been applied to characterization of the electrocapillary effects. This methodology allows effective and efficient optimization of designs being developed. Application of the ACES methodology to study of electrocapillary fluid flow is presented and illustrated with representative examples. Keywords: MEMS, electrocapillary, microfluidics, ACES, electrowetting. 1. INTRODUCTION In the growing field of Microelectromechanical Systems (MEMS) there is the possibility to develop complex, mechanical components on a microchip. Applications for this multidisciplinary technology have already been realized with inertial sensors in the automotive industry and with the DLP mirror arrays for projection technology [1]. The overwhelming successes these products have had in a short time have foreshadowed additional uses for this technology with vertical cavity surface emitting lasers (VCSEL) for communications [2], high frequency RF switches [3], and micro total analysis systems (µtas) for biological and chemical analysis [4]. In working on a size scale that is orders of magnitude smaller than traditionally used for mechanical devices there are several considerations that must be made. Forces that scale linearly with size, L 1, become substantial on the small scale while body forces, such as weight, that scale with L 3 become small, and in cases negligible on the micro-scale. Surface tension scales linearly with size, making this an important force when working with MEMS. Surface tension can have detrimental effects during fabrication or beneficial effects for fluid actuation in components. Special considerations are made during the fabrication to reduce stiction, which is the negative effect of surface tension [5]. There are however many beneficial uses for surface tension forces on the small scale. The surface tension force is responsible for capillary flow in small channels. This is generally a passive flow that could be characterized as constantly in the on position. Passive capillary flow can however be controlled by varying the geometry of the channel in which flow occurs [6]. Because surface tension is a reflection of surface energy it may be controlled by thermal or electrical means [6]. At higher temperatures there are less attractive forces between molecules, lowering the surface tension. In general, however, the temperature dependency of surface tension is small, requiring large temperature gradients to create a useful force. The surface energy of the fluid may also be increase with the application of an electric potential. The concept of a capillary flow controlled through the use of a potential difference is known as the electrocapillary effect. Electrocapillary flow can be obtained through several different methods, including electrowetting (EW) and electrowetting-on-dielectric (EWOD).
1.1. Electrowetting Electrowetting is the process of enhancing the wetting of a fluid by applying a potential between the fluid and an electrode. This phenomenon was first explored for fluid control by Matsumoto and Colgate nearly fifteen years prior to this publication [7]. In recent years electrowetting has been explored as a means to drive fluid through channels. Electrowetting is based on the interaction between an electrolyte fluid and an electrode when a potential is applied between them. An electrical double layer (EDL) exists between the fluid and electrode as an electrical insulator. This layer, generally on the order of 10 Å to 100 Å thick, will lower the surface tension of the fluid when a potential is applied over it [6]. If the surface tension of the fluid can be decreased enough it becomes possible to change from a hydrophobic state (contact angle greater than 90º) to a hydrophilic state (contact angle less than 90º). A hydrophilic channel will allow the fluid to fill it, as shown in Fig. 1. Electrolyte Electrode EDL Fig. 1. Electrowetting in microchannel to induce fluid flow. 1.2. Electrowetting-on-dielectric Electrowetting-on-dielectric is based on the same principle as electrowetting, differing in the fact that the EDL is replaced by a thin film capacitor. Typical dielectrics used for this application may be Parylene, Silicon Dioxide, or Teflon [8-11]. In this investigation an alternative dielectric material will be explored for use in EWOD systems. Lot temperature cofired ceramic (LTCC) is common within the microelectronics packaging industry. This material, however, may be useful as a dielectric layer for use in EWOD systems. Because there is no need for an EDL the working fluid does not need to be an electrolyte, increasing the versatility of the technology. However, because the thickness of the film must be substantially greater than the thickness of the EDL to resist electric breakdown there is typically a higher power input for EWOD systems. The process for driving fluids in a channel is similar to that of EW, and are shown in Fig. 2. Aqueous liquid Electrode Dielectric Fig. 2. Electrowetting-on-dielectric in microchannel to induce fluid flow. By considering a fluid segment where the EW or EWOD is applied to only one side of the channel it becomes possible to develop a scenario where a pressure difference in the segment may be produced, resulting in a flow in one direction or another [9]. A schematic of this concept is shown in Fig. 3. Glass Dielectric θ 1 θ 1 d θ 1 θ 2 Substrate Electrode Fig. 3. Fluid segment driven through microchannel with EWOD. In Fig. 3, d represents the channel thickness, θ 1 is the contact angle with no EWOD applied, and θ 2 is the contact angle after EWOD has been applied.
2. METHODOLOGY A hybrid solution methodology has been adopted to fully characterize the system at hand [12]. This methodology is based on analytical, computational, and experimental solutions, and is referred to by the acronym ACES [13]. The analytical solutions are based on exact closed-form equations. Computational solutions are approximations, found by use of finite element methods (FEM), finite difference methods (FDM), and boundary element methods (BEM). The experimental solutions are obtained in the laboratory, often incorporating state-of-the-art interferometric methods yielding displacement and deformation fields with high spatial resolutions and very high measurement accuracy in the full-field-of-view, noninvasive, and in near realtime [14]. This paper will focus on analytical calculations and preliminary computational simulations using Surface Evolver software [15]. 2.1. Analytical considerations To begin analytical considerations of electrocapillary flow the basic governing equations should be studied. In general, fluids create a contact angle when placed on a solid substrate, Fig. 4. γ lg Liquid γ sl θ Gas γ sg Solid Fig. 4. Surface tension forces on a drop of fluid. Through summation of the surface tension forces that exist due to the liquid-gas (γ lg ), solid-liquid (γ sl ), and solid-gas (γ sg ) interfaces we derive Young s equation [6], as follows: γ = γ γ cosθ, (1) sl sg lg where θ is the contact angle of the fluid. Lippmann s equation governs the effect of a potential on the surface tension between the solid and liquid [8], and can be expressed as 1 2 γ sl, 2 = γ sl,1 cv, (2) 2 where γ sl,1 is the surface tension of the solid-liquid interface with no potential applied, γ sl,1 is the surface tension of the solidliquid interface with potential applied, c is the capacitance over the interface, and V is the applied potential. By combining Eqs 1 and 2 the Lippmann-Young s equation is produced, i.e., 1 1 ε r εo 2 cosθ 2 = cosθ1 + V, (3) γ 2 h lg where the capacitance has been replaced by the product of the relative dielectric constant of the insulating material (ε r ) and the dielectric constant of free space (ε o ) divided by the thickness (h). The pressure that is present on each surface of the fluid segment, shown in Fig. 3, can be determined by performing an energy balance between the fluid and gas on each side of the segment [16]. The resultant Laplace equation is p 1 1 = γ lg, Rv R (4) h where p is the pressure on one side of the segment, R v is the radius of curvature in the vertical direction, and R h is the radius of curvature in the horizontal direction. Based on the model introduced in Fig. 3 the radius in the vertical direction on the right side of the segment (R v,r ) can be approximated as
d v, =. (5) θ1 + θ2 θ1 θ2 2cos cos 2 2 R R The vertical radius on the left side of Fig. 3 has a simpler expression than that of Eq. 5, approximated as d v, =. (6) cosθ R L 2 1 By subtracting the pressure of the left side from that of the right side it becomes possible to determine the pressure difference over the fluid segment as γ p = d lg ( θ 2 cosθ ). cos 1 (7) The pressure driving the fluid segment is a function of constants and contact angles. For preliminary analysis it is simpler to conduct open drop analysis to measure the contact angle of a fluid. This consists of a single drop of fluid on an electrode. With this setup it is possible, using zoom lenses, to measure the contact angle the drop makes with the substrate. 2.2. Computational considerations A computational model has been developes of a simple open-drop experiment. Surface evolver has been used to generate a model of a drop on a surface [15]. All parameters that affect the contact angle are adjustable within the software, resulting in a parametric model of a drop on a solid surface. In this preliminary model the contact angle is used as a boundary condition. In future models the surface energy will govern the contact angle. In more complex analysis of flow through a channel this software will be well suited to determine the pressure driving a fluid segment. 2.3. Experimental considerations Experimental analysis of an open drop experiment will involve the measurement of the contact angle a fluid makes with a substrate during the EWOD process. Special consideration will be used to measure hysteresis of multiple load cycles of the voltage, time constant for contact angle change, benefits of a monolayer acting as a lubricant on the dielectric surface, and roughness of dielectric effects. Additionally, the saturation voltage at which contact angle no longer decreases will be investigated. Such results will aid in deriving a more complete analytical model of the EWOD actuation. 3. RESULTS By utilizing Eq. 3 the contact angle as a function of applied potential can be determined for several different materials. Results of Eq. 3 have been plotted for some traditional materials along with LTCC, Fig. 5. For this analysis a thickness of 1 µm has been used. The working fluid that was considered is water, and the initial contact angle for each case was assumed 120º. This contact angle could be obtained for each dielectric with a monolayer acting to condition the surface. The difference in the materials exists in their dielectric constants. Table 1 shows the dielectric constants for the four materials considered for this analysis. Table 1. Dielectric constants for various materials considered for EWOD. Material Dielectric constant Parylene N 2.65 SiO 2 3.90 Teflon 2.00 LTCC 951 7.80 The contact angle as a function of voltage for the four materials is shown in Fig. 5. The lower limit on the contact angle is limited to 60º.
130 CONTACT ANGLE, deg 120 110 100 90 80 Parylene C SiO2 Teflon LTCC 951 70 60 0 10 20 30 40 50 60 70 80 90 VOLTAGE, V Fig. 5. Contact angle as a function of voltage. Surface Evolver representations of the contact angle in an initial state (120º) and a final state (80º) are shown in Fig. 6 depicting the results from an open drop experiment. Fig. 6. Surface evolver simulation of fluid drop with contact angle of predicted initial 120 and final 80. The Surface Evolver simulation results are identical for each of the dielectrics because the contact angle is a boundary condition. As previously stated, future models will be based on surface energy and may be better suited for comparing different materials. 4. DISCUSSION Multiple possibilities for the use of surface tension for fluid actuation have been presented. At this time, the focus of work in this area has been limited to open-drop experiments. ACES methodology has been utilized to study the relation between contact angle and voltage for open drop experiments. By optimizing analytical representation for open drop experiments through ACES methodology more accurate calculations will be possible for complex systems, such as fluid flow in microchannels. Various materials have been considered for use as the dielectric in an EWOD system. Analytical solutions show that LTCC has high potential for an effective electrical insulator for EWOD applications, because its dielectric constant is greater than for most materials traditionally used in this process. LTCC is commonly used in the microelectronics packaging realm, it may be possible to integrate a fluidic network within a package to minimize size and provide a clear link from the macro world to a MEMS device within the package. In order to complete ACES analysis on open-drop EWOD additional computational and experimental analysis will be performed. 5. WORK IN PROGRESS Although we have initiated experimental analysis on EWOD using a Parylene N dielectric, not enough information is available, at this time, for a publication. Continued work with open-drop experiments is planned with Parylene N and LTCC. The goal of
this work is to fully characterize all parameters that have an effect on the contact angle during EWOD actuation. The parameters to be investigated include roughness of the dielectric layer, determination of a saturation contact angle, and effects of monolayer pretreatment on the dielectric. The goal of this work is to adapt EWOD for use in biological sensing apparatus, which will be realized by extending experimentation to biological fluids such as blood. By completing ACES analysis on preliminary open-drop scenarios EWOD may be better characterized. 6. REFERENCES [1] Douglass, M. R., MEMS enables a new reliability paradigm for DLP TM display technology, Proc. 4 th Internat. Symp. on MEMS and Nanotechnology (4 th ISMAN), Charlotte, NC, pp. 21-27, 2003. [2] Matsumoto, C., VCSELs can bring optics to the masses, EE Times, Sept. 12, 2002. [3] Machate, M. S., Furlong, C., and Pryputniewicz, R. J., RF MEMS switch: performance optimization, Proc. 4 th Internat. Symp. on MEMS and Nanotechnology (4 th ISMAN), Charlotte, NC, pp. 12-19, 2003. [4] Thaysen, J., Marie, R., and Boisen, A., Cantilever-based bio-chemical sensor integrated in a microliquid handling system, Proc. of MEMS, Lyngby, Denmark, pp. 401-404, 2001. [5] Tadigadapa, S. A., and Najafi, N., Developments in microelectromechanical systems (MEMS): a manufacturing perspective, J. of Mfg. Sci. and Eng., 125:816-823, 2003. [6] Nguyen, N. T., and Wereley, S. T., Fundamentals and applications of microfluidics, 1 st ed., Artech House, Norwood, MA, pp. 285-292, 2002. [7] Matsumoto, H., and Colgate, J. E., Preliminary investigation of micropumping based on electrical control of interfacial tension, Proc. MEMS 90, 2 nd IEEE Internat. Workshop on Microelectromechanical Systems, Napa Valley, CA, pp. 105-110, 1990. [8] Marinis, R. T., and Pryputniewicz, R. J., EWOD induced fluid flow for biological applications, Proc. 36 th Internat. Symp. on Microelectronics, Boston, MA, pp. 719-724, 2003. [9] Lee, J., Moon, H., Fowler, J., Schoellhammer, T., and Kim, C. J., Electrowetting and electrowetting-on-dielectric for microscale liquid handling, Sensors and Actuators, A95:259-268, 2002. [10] Moon, H., Cho, S. K., Garrell, R. L., and Kim, C. J., Low voltage electrowetting-on-dielectric, J. of Appl. Phys., 92(7):4080-4087, 2002. [11] Saeki, F., Baum, J., Moon, H., Yoon, J. Y., Kim, C. J., and Garrell, R. L., Electrowetting on dielectrics (EWOD): reducing voltage requirements for microfluidics, Polymeric Materials: Science and Engineering, 83:12-13, 2001. [12] Pryputniewicz, R. J., Galambos, P., Brown, G. C., Furlong, C., and Pryputniewicz, E. J., ACES characterization of surface micromachined microfluidic devices, Internat. J. Microcircuits and Electronic Packaging (IJEMP), 24(1):30-36, 2001. [13] Pryputniewicz, D. R., ACES approach to the development of micro-components, MS Thesis, Worc. Poly Inst., Worcester, MA, 1997. [14] Klempner, A. R., Hefti, P., Furlong, C., and Pryputniewicz, R. J., Development of a compact interferometric system for characterization of MEMS, Proc. 30 th Annual Symp. & Exhibition of IMAPS-NE, Boxboro, MA, pp. 190-195, 2003. [15] Blake, K. A., Surface evolver manual, Mathematics Department, Susquehanna University, Selinsgrove, PA, 1999. [16] Adamson, A. W., and Gast, A. P., Physical chemistry of surfaces, 6 th ed., Wiley, New York, NY, pp. 195-199, 1997.