ACES Characterization of Surface Micromachined ... - IMAPS

Intl. Journal of Microcircuits and Electronic Packaging

ACES Characterization of Surface Micromachined Microfluidic Devices Ryszard J. Pryputniewicz, Paul Galambos*, Gordon C. Brown, Cosme Furlong, and Emily J. Pryputniewicz NEST – NanoEngineering, Science, and Technology CHSLT - Center for Holographic Studies and Laser micro-mechaTronics Mechanical Engineering Department Worcester Polytechnic Institute Worcester, Massachusetts 01609-2280 Phone: 508-831-5536 Fax: 508-831-5713 e-mail: [email protected] * Intelligent Micromachines Department Sandia National Laboratories Albuquerque, New Mexico

Abstract Recent breakthroughs in surface micromachining led to development of the first generation surface micromachined microfluidic devices. The simplest microfluidic device is a microfluidic channel (microchannel). A typical surface micromachined microchannel is about 2 µm deep, its width and depth depend on the specific application, while its top is less than 1 µm thick silicon nitride membrane. Almost all microfluidic devices are concerned with flow through very small passages and unintentional restrictions that lead to high-pressure gradients within the microchannels, which, in turn, cause deformations of the membranes. Knowledge of these deformations facilitates design and optimization of the microfluidic devices. The researchers have developed a hybrid approach for accurate and precise characterization of microfluidic devices. This approach is based on Analytical, Computational, and Experimental Solutions (ACES) methodology. The experimental aspects of this methodology are based on laser interferometric measurements yielding displacement and deformation fields, while analytical and computational aspects are based on exact (closed form) and approximate (FEM) solutions, respectively. Comparison of analytical, computational, and experimental results indicates correlation to within 1%, for maximum deformations on the order of 1 µm. Continued advances in the methodology for characterization of the microfluidic devices will lead to development of robust devices, which will allow their integration with electronic actuation to enable a broad range of revolutionary new applications achievable with surface micromachining.

Key words:

1. Introduction

MEMS, Microfluidics, Surface Micromachining, Microchannel, ACES Methodology, Laser Interferometry, and Optoelectronic Laser Interferometry Microscope.

The field of microfluidics is currently undergoing rapid growth in terms of new device and system developments. As a result, a variety of microfluidic devices are being developed for applications ranging from Micro-Total (chemical) Analysis System (µTAS)1 to ink jet printing2. These devices are fabricated using a wide range of technologies including bulk micromachining, high

The International Journal of Microcircuits and Electronic Packaging, Volume 24, Number 1, First Quarter, 2001 (ISSN 1063-1674) 30

© International Microelectronics And Packaging Society

ACES Characterization of Surface Micromachines Microfluidic Devices aspect ratio (HAR) micromachining, laser micromachining, and, more recently, surface micromachining. Surface micromachining has significant potential advantages over other fabrication techniques in some microfluidic applications3. One advantage is volume minimization. A typical bulk micromachined microfluidic device has a channel depth on the order of 100 µm, width of 500 µm, and 1000 µm long channel, or a volume on the order of 50 nl. A typical surface micromachined channel is only 2 µm deep, 200 µm wide, and 1000 µm long, and has a volume of 0.4 nl, which is more than 2 orders of magnitude smaller than that of a typical bulk micromachined channel. In applications where volume minimization is important (such as µTAS), this difference may be significant in reducing mixing times and reagent requirements. Furthermore, surface micromachined microfluidic devices can be potentially integrated with electronics to produce electrofluidic devices. In addition, integration of microfluidics onto a single silicon substrate containing MEMS can lead to a new class of devices, electro-microfluidic MEMS on a chip3. This class of devices has a potential to accomplish a wide variety of functions, such as, power generation, µTAS, hydraulic actuation and control, in a very compact package. However, in order to use these devices as part of a larger system on a chip or to interface them with devices external to the chip, performance characteristics of microfluidic device must be quantified. The simplest microfluidic device is a microfluidic channel (microchannel). This paper addresses design, fabrication, and characterization of a microchannel.

2. Design and Fabrication Although surface micromachined microfluidic devices offer significant advantages over devices produced by other methodologies, the very small sizes of channels pose difficulties in device and system packaging and characterization. In order to explore the potential advantages, Sandia has developed first generation surface micromachined microfluidic devices, Figure 1, containing a number of microchannels.

A typical surface micromachined microchannel, Figure 2, is about 2 µm deep, its bottom and sides are lined up with 0.3 µm thick silicon nitride while its top is less than 1 µm thick silicon nitride membrane. 200µm Figure 2. Single microchannel with inlet and outlet vias. At the beginning of fabrication of microfluidic channels, thermal oxide (TEOS) hard mask is patterned onto a silicon wafer and 2 µm trenches are etched in silicon4. After etching the trenches, the TEOS mask is stripped and 0.3 µm thick layer of low stress silicon nitride is deposited using Low Pressure Chemical Vapor Deposition (LPCVD). Then, a sacrificial oxide (SACOX) refill is deposited and trenches are chemical-mechanically polished (CMP) flat. Following CMP, 0.8 µm thick low stress silicon nitride membrane (channel cover) is deposited over the SACOX. The membrane is patterned with etch release holes using another TEOS hard mask, and the release holes are etched using a dry etch. After stripping the hard mask, the structures are released in an acid bath; the solution is highly selective for etching the SACOX while not affecting the silicon nitride. Finally, the etch release holes are sealed using another LPCVD silicon nitride deposition. The resulting channels are approximately 2 µm deep. These channels are 200 µm wide at the inlet and outlet and neck down to various widths in between, depending on the specific design. For example, the narrow section of the microchannel shown in Figure 2 is 2 µm deep, 10 µm wide, and 200 µm long. In order to introduce fluid into the channel, inlet and outlet vias, Figure 2, are etched through the wafers from the back using a Deep Reactive Ion Etch (DRIE) process.

3. Characterization Methodology

Almost all microfluidic devices are concerned with flow through very small passages and unintentional restrictions, that lead to high-pressure gradients within the microchannels, which, in turn, cause deformations of the membranes. Knowledge of these deformations facilitates design and optimization of microfluidic devices. The researchers have developed a hybrid approach for accurate and precise characterization of microfluidic devices. This approach is based on Analytical, Computational, and Experimental Solutions (ACES) methodology5. In this paper, the experimental aspects of the ACES methodology are based on optoelectronic laser interferometric microscope (OELIM) method, which measures deformations of the microfluidic devices with nanometer accuracy6, while analytical and computational aspects are based on exact (closed form) and approximate (FEM) soluFigure 1. Sandia microfluidic test-package containing tions7, respectively. multiple devices. The International Journal of Microcircuits and Electronic Packaging, Volume 24, Number 1, First Quarter, 2001 (ISSN 1063-1674) © International Microelectronics And Packaging Society

31

Intl. Journal of Microcircuits and Electronic Packaging I i (x, y ) = 2 I 0 (x, y ){1 + cos[φ (x, y ) + θ i ]},

3.1. OELIM Method In the OELIM method, a beam of collimated coherent light is brought into the system and is directed into a spatial filter (SF) assembly consisting of a microscope objective and a pinhole filter, Figure 3. The resulting, expanded, light field is then collimated by lens L1, and redirected by the directional beam splitter (DBS) through the long distance microscope objective lens (MO) to illuminate the microfluidic device. The proximal beam splitter (PBS) is placed close to the device. The reflected light is transmitted back through MO, DBS, and the relay lens to the CCD camera.

(1)

where the subscript i (= 1, 2, ....., n) indicates the i-th interferogram – with n being the total number of interferograms, x,y are the spatial coordinates, I 0 (x, y ) is the average light intensity, φ (x, y ) is the unknown phase, and θ i is the known phase step defined as follows, θi =

i −1 π 2

.

(2)

Based on the set of n equations of the type of Equation (1), spatial phase distribution can be determined employing the fundamental equation, !!!!!!!"(x, y) = "!(x, y)[I1(x, y), I2(x, y), ...., In (x, y)],

(3)

and used to compute displacement vector fields L(x, y ) from the following relation,

Mirror

L(x, y ) = L(x, y )[K (x, y ), φ (x, y )],

(4)

where K (x, y ) is the sensitivity vector defining illumination and observation (i.e., laser light propagation) geometry of the OELIM system used to record interferograms of the microfluidic devices.

3.2. Analytical and Computational Methods

Figure 3. OLIM system for quantitative measurements of microfluidic devices-optical configuration: SF/LI is the spatial filter/collimating lens assembly “shaping” the illumination beam, DBS is the directional beam splitter, MO is the microscope objective, PBS is the proximal beam splitter, CCD is the host computer controlled image acquisition camera. Illumination of the microfluidic device, by the laser used in the OELIM system, allows recording of the interferograms of the device under its operating condition. These recordings are made using phase stepping technique to acquire several interferograms that describe a specific state of the object; a known phase step is introduced for each interferogram8,9. Then, the sequence of the phase-stepped interferograms is processed to obtain detailed quantitative information on the displacements and deformations of the microfluidic devices. Spatial intensity distribution, I x, y , within the interferograms recorded by the CCD camera can be described as follows,

(

)

Analytical methods are characterized by exact, closed form, solutions and make use of infinitesimal elements in modeling the microfluidic device. Exactness of the analytical results depends on assumptions made during the solution. Computational methodologies make use of finite size elements in discretization of the physical domain and provide approximate solutions. The degree of approximation depends on the type and size of the elements used. In this paper, computational modeling was performed using Finite Element method (FEM).

4. Results Typical interferogram of a deformed membrane of a microfluidic device, obtained using the OELIM method, is shown in Figure 4. The deformations were caused by loadings produced by compressed air supplied to the microfluidic device. Quantitative interpretation of interferometric fringe pattern shown in Figure 4 yields maximum deformation of 1.048 µm. The analytically determined deformation of 1.044 µm, for the same loading conditions of the microfluidic device, as those used to record the interferogram of Figure 4, is in good agreement, well within the uncertainties based on the analytical model - with



ACES Characterization of Surface Micromachines Microfluidic Devices the experimentally measured deformation. FEM results, based on the model developed during this study, show that the corresponding maximum deformation is 1.059 µm, Figure 5.

Figure 6. FEM determined deformations of the membrane over half of the same microfluidic device as that shown in Figures 4 and 5. Magnitude of the maximum FEM determined deformation of the membrane, shown in this Figure, is 1.059µm.

Figure 4. OELIM interferogram of a deformed membrane of the microfluidic device of Figure 2, showing part of the membrane to the right of the 10µm narrow section. Maximum deformation of the membrane, determined from this interferogram, is 1.048 µm.

Figure 5. FEM determined deformations of the same part of the membrane of the microfluidic device as that shown in Figure 4. Maximum FEM determined deformation of the membrane, shown in this Figure, is 1.059µm. Comparison of the analytical, computational, and experimental results, for the case shown in Figures 4 and 5, indicates that the correlation between them is within 1%. Similar ACES correlation was also obtained by comparing the results obtained for several other cases of different loading conditions applied to the microfluidic devices. Figure 6 shows FEM determined deformations of the entire part of the membrane to the right of the 10 µm narrow section (i.e., half of the microfluidic device) of the same microfluidic device as that displayed in Figures 4 and 5. Clearly, maximum deformations of 1.059 µm are seen in the center of the membrane with the narrow section showing negligible deformations.

Representative results shown in Figures 4 to 6 indicate that good correlation is obtained between analytical, computational, and experimental results if geometry, dimensions, material properties, initial conditions, boundary conditions, and loading conditions are known. However, if any of these parameters change and accuracy of their values is not known, correlation between the results may not be easy. Nevertheless, differences between various results can be vividly noticed. For example, Figure 7 shows a microchannel and its inlet via with a drop of water trapped at the entrance to the narrow section. This drop forms an unintentional restriction in the microchannel and obstructs flow of air from left to right. As a result of this restriction, high-pressure gradients develop within the microchannel and these, in turn, cause deformations of the membrane, subject to the boundary conditions associated with the presence of the drop of water at the inlet to the narrow section.

Figure 7. Detail of a microchannel and its inlet via, note drop of water trapped at the entrance to the 100µm narrow section. The microchannel of Figure 7 was loaded by air at 4 psig, OELIM fringe patterns of the deformed membrane were recorded, and their quantitative interpretation yielded results shown in Figure 8. The displacements shown range from 0 nm to 722 nm.

The International Journal of Microcircuits and Electronic Packaging, Volume 24, Number 1, First Quarter, 2001 (ISSN 1063-1674) © International Microelectronics And Packaging Society

33


Figures 10 and 11 show longitudinal and transverse profiles of the deformed membrane of Figures 7 and 8. These profiles clearly indicate influence of the water drop, trapped at the entrance to the narrow section of the microchannel.

Figure 10. Longitudinal profile of the deformed membrane, measured along line L shown in Figure 8a.

Figure 8. Representative deformations of the microfluidic membrane, based on the OELIM fringe pattern corresponding to a load caused by air supplied at 4 psig: 3D planar representation (a) and 3D isometric representation (b). Displacements of the membrane range from 0 to 722 nm and their distribution is affected by the drop of water (see Figure 7) trapped at the entrance to the narrow section. The FEM determined deformations of the membrane of the microfluidic device shown in Fig. 7, excluding the water drop, are shown in Figure 9. For this condition, the maximum deformation is 1.045 µm, while the analytical model indicates 1.018 µm.

Figure 11. Transverse profiles of the deformed membrane, measured along lines T1 through T4 shown in Figure 8a: symbols indicate actual data points measured from the OELIM interferogram, while continuous lines display analytical representations of the corresponding transverse profiles.

5. Conclusions

Figure 9. Deformations of the membrane of the microfluidic device shown in Figure 7, subjected to air pressure of 4 psig, based on FEM half-model computations: 3D planar representation (a) and 3D isometric representation (b), corresponding to Figures 8a and 8b, respectively. Since the drop of water (see Figure 7) was not included in the model, magnitude and distribution of the FEM computed displacements agree with the analytical results, but differ from the experimental ones shown in Figure 8.

A new hybrid approach for accurate and precise quantitative characterization of microfluidic devices was presented in this work. This approach is based on ACES methodology. In this study, deformations of surface micromachined microfluidic devices, subjected to different loading conditions, were characterized. ACES correlation between the analytical, computational, and experimental results was within 1%, for an unobstructed channel. The ACES methodology is also applicable to characterization of the microfluidic devices fabricated by other methods than surface micromachining. In addition, it is applicable to characterization of microdevices other than microchannels. In summary, continued advances of methods for characterization of the microfluidic devices will lead to introduction of nonsimple boundary conditions into the analytical and computational models. Following experimental validation of these models, new tools for development of robust microfluidic devices will be available. These tools will allow integration of the microfluidic devices with electronic actuation to enable a broad range of revolutionary new applications achievable with surface micromachining.



ACES Characterization of Surface Micromachines Microfluidic Devices

Acknowledgments

About the authors

The microfluidic devices were fabricated at and provided by Sandia National Laboratories. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy under Contract DE-AC04-94AL85000. This study was partially supported by the ME-CHSLT NEST Program at WPI.

Ryszard J. Pryputniewicz, educated both in Poland and the United States, is Professor of Mechanical Engineering and founding Director of the Center for Holographic Studies and Laser micromechaTronics (CHSLT) at Worcester Polytechnic Institute (WPI) in Worcester, Massachusetts, since 1978; previously, a faculty member and Director of the Laser Research Laboratory at the School of Engineering and the Health Center of the University of Connecticut (6 years); member of the Aerospace technical staff (4 years). His research interests concentrate on theoretical and applied aspects of MicroElectroMechanical Systems (MEMS), smart sensors and structures, and, in particular, holographic interferometry. In this work, he emphasizes unification of analytical, computational, and experimental solution (ACES) methodologies, especially when they can be merged to provide solutions where none would be obtainable otherwise, to ease the solution procedure, or to attain improvements in the results.

References 1. D. J. Harrison, “Micro-Total Analysis Systems”, Kulver Academic Publishers, Boston, Massachusetts, 1998. 2. S. Kamisuki, “A Low Power Electrostatically Driven Commercial Inkjet Head,” MEMS’98, ASME, New York, 1998. 3. P. Galambos, W. P. Eaton, R. Shul, C. G. Willison, J. J. Sniegowski, S. L. Miller, and D. Gutierrez, “Surface Micromachined Microfluidics: Design, Fabrication, Packaging, and Characterization,” MEMS’99, ASME, New York, pp. 441-448, 1999. 4. E. J. Garcia and J. J. Sniegowski, “Surface Micromachined Microengine,” Sensors and Actuators A, Vol. 48, pp. 203214, 1995. 5. D. R. Pryputniewicz, “ACES Approach to the Development of Microcomponents”, M.S. Thesis, Worcester Polytechnic Institute, Worcester, Massachusetts, 1997. 6. G. C. Brown, “Laser Interferometric Methodologies for Characterizing Static and Dynamic Behavior of MicroElectroMechanical Systems (MEMS)”, Ph.D. Dissertation, Worcester Polytechnic Institute, Worcester, Massachusetts, 1999. 7. C. Furlong, “Hybrid, Computational and Experimental, Approach for the Efficient Study and Optimization of Mechanical and Electromechanical Components”, Ph.D. Dissertation, Worcester Polytechnic Institute, Worcester, Massachusetts, 1999. 8. G. C. Brown and R. J. Pryputniewicz, “New Test Methodology for Static and Dynamic Shape Measurements of Microelectromechanical Systems”, Optical Engineering, Vol. 39, pp. 127-136, 2000. 9. C. Furlong and R. J. Pryputniewicz, “Absolute Shape Measurements Using High-Resolution Optoelectronic Holography Methods,” Optical Engineering, Vol. 39, pp. 216-223, 2000.

Cosme Furlong received his Mechanical Engineering Degree from the University of the Américas – México, in 1989, and his Master of Science and Ph.D. Degrees in Mechanical Engineering from Worcester Polytechnic Institute (WPI), Worcester, Massachusetts, in 1992, and 1999, respectively. He is currently Assistant Professor of Mechanical Engineering at WPI working in the field of NanoEngineering, Science, and Technology (NEST). His professional interests include: combination of modeling and simulation (CAD/CAE) with quantitative optical techniques, fiber optics, opto-electronic holography, nondestructive testing, image processing, numerical analysis, materials characterization, and optimization of mechanical and electro-mechanical components.

Emily J. Pryputniewicz received her Bachelor of Science and Master of Science Degrees in Mechanical Engineering from Worcester Polytechnic Institute (WPI) in Worcester, Massachusetts, in 1999, and 2000, respectively. While an undergraduate, she was named NASA Scholar three times and conducted her undergraduate research on development of optical methodology for shape measurements at NASA-Langley facilities. Her graduate research was conducted at the Sandia National LaboThe International Journal of Microcircuits and Electronic Packaging, Volume 24, Number 1, First Quarter, 2001 (ISSN 1063-1674) © International Microelectronics And Packaging Society

35


ratories, Albuquerque, New Mexico, where she concentrated on the studies of MicroElectroMechanical Systems (MEMS). Currently, she is continuing her work on MEMS as a Research Associate at the Institute for Defense Analysis, Alexandria, Virginia. Gordon C. Brown holds a Bachelor of Science Degree in Applied Geophysics from Michigan Technological University, and Master of Science and Ph.D. Degrees in Mechanical Engineering from Worcester Polytechnic Institute (WPI), Worcester, Massachusetts, received in 1994 and 1999, respectively. From 1999 to 2000 he was Assistant Professor at WPI and conducted research in holographic interferometry techniques for nondestructive testing of MicroElectroMechanicalSystems (MEMS) at the Center for Holographic Studies and Laser micro-mechaTronics (CHSLT). Currently, he is continuing to work on MEMS at the CorningLasertron in Bedford, Massachusetts. Paul Galambos received his BSME Degree from Bradley University in 1982, MSME Degree from The University of Texas at Austin in 1987, and his Ph.D. Degree in Mechanical Engineering from the University of Washington in Seattle, Washington in 1998. In between periods of education he has worked at Caterpillar Tractor Co. as a design engineer, Texas Instruments as an electronics packaging engineer, and Lockheed/Martin as an aerospace engineer on the NASP program. He is currently a staff member at Sandia National Laboratories in Albuquerque, New Mexico, working in the area of microsystems development, where he is conducting research in the area of microfluidics with special emphasis on developing microfluidic systems for biological applications.

