812
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 47, NO. 6, JUNE 2000
Micromachined Pipette Arrays Ian Papautsky, Member, IEEE, John Brazzle, Member, IEEE, Harold Swerdlow, Robert Weiss, and A. Bruno Frazier, Member, IEEE
Abstract—In this paper, the design and characterization of batch fabricated metallic micromachined pipette arrays is described. The process used to fabricate the micromachined pipette arrays (MPA) includes + etch-stop membrane technology, anisotropic etching of silicon in potassium hydroxide, sacrificial thick photoresist micromolding technology, and electrodeposition. Arrays of one to ten pipettes have been fabricated using nickel as the structural material and palladium as the biocompatible coating of inside walls. The inner dimensions of the individual pipettes fabricated to date range from 30 m to 1.5 mm in width, 0.5 mm to several cm in length, and 5–50 m in thickness. The center-to-center spacing of these pipettes varies from 100 m to several centimeters. The MPA have a number of advantages when compared to the current micropipette technology, including the ability to transfer precise volumes of samples in the submicroliter range; the ability to manipulate samples, reagents, or buffers in a highly-parallel fashion by operating hundreds of individual pipettes simultaneously; and the compatibility with the submilimeter center-to-center dimensions of the microscale biochemical analysis systems. The application of the MPA to high lane density slab gel electrophoresis is explored. Sample wells are formed in agarose gels by using micromachined combs (solid MPA) at center-to-center spacing ranging from 250 m to 1.9 mm. The samples are loaded using the MPA. The results of the micro-gel separations compare favorably with the standard mini-gel separations and show a twofold increase in the number of theoretical plates as well as a sixfold increase in lane density. Index Terms—Gel electrophoresis, metallic microchannels, micromachined pipettes, microscale sample loading, pipette arrays.
I. INTRODUCTION
T
HE NUMBER of biomedical applications for micromachining technologies is rapidly growing. Since micromachining technologies are a relatively new and increasing set of manufacturing technologies, there are many critical applications still to be addressed. To date, many advances have been made in biomedical instrumentation development as a result of using micromachining technologies to fabricate (or partially fabricate) the total instrumentation system. The main types of biomedical instrumentation that use micromachining Manuscript received January 14, 1998; revised February 1, 2000. This material is based upon work supported by the University of Utah Technology Innovation Grant and the University of Utah Graduate Research Fellowship. Asterisk indicates corresponding author. *I. Papautsky is with the Department of Electrical and Computer Engineering and Computer Science, University of Cincinnati, 814 Rhodes Hall, P.O. Box 210030, Cincinnati, OH 45221 USA (e-mail:
[email protected]). J. Brazzle is with IntelliSense Corporation, Wilmington, MA 01887 USA. H. Swerdlow is with the Center for Genomics Research, Karolinska Institutent, Stockholm S-171 77, Sweden. R. Weiss is with the Department of Human Genetics, University of Utah, Salt Lake City, UT 84112 USA. A. B. Frazier is with the School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA 30332 USA. Publisher Item Identifier S 0018-9294(00)04417-7.
technologies are neural stimulation/recording systems [1]–[4], biological/chemical analysis systems [5]–[17], glucose and lactate sensors [18]–[20], whole cell systems [21]–[24], surgical instrumentation [25], [26], medical monitoring systems [27], prosthetic devices (e.g., artificial limbs) [28], and tactile sensors [29], [30]. During the past ten years, a large number of biological/chemical analysis techniques have been demonstrated using microscale systems and have been implemented using micromachining technology. Currently, these systems include electrophoresis [11]–[14], free-flow electrophoresis [9], [10], electrical field-flow fractionation (EFFF) [15], polymerase chain reaction (PCR) [16], [17] gas chromatography [5], [6], and liquid chromatography [7], [8]. The rationale for using microfabrication technologies in analytical instrumentation include reduction in instrument size and cost, reduction in sample and reactant volumes, reduction in analysis time, increase in analysis throughput, and possibility of integration of sample preparation and analysis functions. Injection, motion, and placement of liquids is crucial in microfabricated chemical analysis systems [31], [32]. The mixing, reaction, and separation steps performed in these systems require precise volumes of liquid samples to be accurately positioned in the microchannels [31], [32]. Coupling of sample handling and reactions with separation techniques can provide complete biological/chemical analysis, i.e., the micro total analysis system ( TAS) concept. Currently, pipetting is the primary method for reagent and sample manipulation on the macroscale. In pipetting, small sample volumes are transferred into the loading ports of the analysis system. The major disadvantage of this method is the lack of fine control over the actual sample volumes in the submicroliter range. In addition, while parallel pipetting systems are available, large center-to-center spacing makes them incompatible with miniaturized biochemical analysis systems. Another macro method for fluid delivery that is gaining popularity for use in miniaturized analysis systems is pL droplet delivery systems [33]. The droplet dispensing systems that are being used by several of the biochemical analysis companies are based on the droplet dispensing systems developed for the inkjet printing market. The current systems are being used to load reagents into miniaturized biochemical analysis systems. The reagent loading typically occurs prior to final packaging of the biochemical analysis system. While the fluid delivery volumes of the droplet dispensing systems are in the pL to nL range, the dead volume of the total dispensing system is large, thus making the current designs impractical for sample delivery. Electromigration is the predominant method of sample manipulation in miniaturized analysis systems [34], [35].
0018–9294/00$10.00 © 2000 IEEE
PAPAUTSKY et al.: MICROMACHINED PIPETTE ARRAYS
813
Fig. 1. Schematic drawing of a micromachined pipette array (MPA).
Electromigration techniques are typically used in microscale sample delivery approaches. In this case, the microchannel containing the sample is interconnected and integrated with the analysis channel, therefore not addressing the application area of macroscale manipulation of pL to L fluid volumes. In electromigration, samples are loaded into a channel perpendicular to the separation microchannel. Application of an electric field along the length of these channels results in electromigration of the sample and the loading of precise sample volumes. This method requires additional microchannels for the sample loading operation and requires the use of electrically active samples. Given the limitations of the previously mentioned methods, it is important to develop a method for highly parallel microscale sample loading of pL to L volumes that would allow precise handling of the samples and still be compatible with the size dimensions (center-to-center spacing) of the micromachined biochemical analysis systems. In this paper, we present a method for realizing a micromachined interface between macroscale sample preparation formats and microscale analysis systems. A schematic diagram of such a device, a micromachined pipette array (MPA), is illustrated in Fig. 1. There are many advantages of the micromachined pipette arrays, including: 1) the ability to transfer precise volumes of samples in the submicroliter range, and 2) the ability to handle samples/reagents in a highly parallel fashion by manipulating hundreds of samples/reagents simultaneously, 3) a wide range of pipette center-to-center spacing, 4) the ability to individually address the micromachined pipettes in an array for fluid dispensing and extraction operations, and 5) the possibility of adding functionality to the pipettes, such as performing polymerase chain reaction (PCR) within the micropipettes. II. METHODS A. Fabrication The micromachined pipette arrays (MPA) are fabricated using extensions of previously reported micromachining fabrication technologies [36]. The fabrication process is conducted at low temperature and is compatible with integrated circuit (IC) technology as a post process. silicon wafer is RCA cleaned and one side Initially, a is heavily doped with boron using high temperature thermal layer. Next, a layer of silicon diffusion to form a 4–6 m
Fig. 2. Micromachined pipette array fabrication procedure: (a) create silicon membrane using high-temperature p doping and KOH etching; (b) microelectroform bottom shell; (c) apply and pattern thick photoresist; (d) microelectroform side walls and top shell; (e) remove thick photoresist using acetone bath; (f) release pipettes by etching p membrane.
nitride is deposited using plasma enhanced chemical vapor deposition (PECVD). The silicon nitride is used as a mask during the subsequent anisotropic etching of silicon in 20% boron layer serves as an KOH at 60 C [37]. The diffused effective etch stop, forming a sacrificial membrane upon which portions of the pipette shafts are fabricated and subsequently released [Fig. 2(a)]. Next, a metal system of adhesion layers and an electroplating seed layer is electron beam evaporated (Denton Vacuum E-Beam evaporator) over the substrate. The metal system used in this study consists of a 500-Å titanium layer for adhesion to the substrate, an overlying 1500-Å copper electroplating seed layer, and a 1000-Å chromium layer for adhesion between the seed layer and the overlying masking material used to microelectroform the bottom wall of the pipette. Following deposition of the metal system, the micro molding layer (P4620 positive photoresist from Clariant Corp.) for defining the bottom wall of the pipette is spin coated and patterned using standard photolithographic techniques [38]. The top adhesion chromium layer is etched and the bottom wall of the pipette is formed using conventional electrodeposition technology [Fig. 2(b)]. A variety of metals can be used as the primary structural material in pipette fabrication (e.g.,
814
nickel, gold, copper, and palladium). In this work, a nickel bath consisting of NiSO , NiCl , and H BO was primarily used [39]. Typically, a 0.5- m layer of palladium is deposited using a commercially available phosphate-based solution (PallaTech Pd-MI, Lucent Technologies) on top of the electroplated metal used to form the bottom wall. The layer of palladium is used to improve biocompatibility of the pipette inner surface. Once the electroforming is finished, the micromolding photoresist is removed using a 30-s rinse in acetone, a 30-s rinse in propanol, and a 5-min rinse in water at room temperature. The top adhesion layer, the electroplating seed layer, and the bottom adhesion layer are removed from the substrate surface using wet etching [36]. Next, 5–50 m of commercially available thick positive photoresist (P4620) is spin coated and lithographically patterned on top of the electroformed metal. The thick photoresist is used to precisely define the inner dimensions of the pipette and serves as a thick sacrificial layer to be removed later in the process [Fig. 2(c)]. A 1000-Å palladium electroplating seed layer is sputter deposited (sputtering using a Denton Vacuum Model 15) over the thick photoresist. The sputtering process is used due to its superior side wall coverage. After the deposition, P4620 photoresist is spin coated and patterned on top of the sputtered metal to create a micromolding layer. The photoresist is patterned such that the layer remains over the entire substrate, defining the top and side walls of the pipette, except in the regions where the bottom wall and thick photoresist are located. An additional thin layer (e.g., 0.5 m) of palladium is electroformed on top of the sputtered metal to increase thickness of the seed layer. The primary structural material is then electrodeposited to complete the top and side walls of the pipette [Fig. 2(d)]. As in the case of the bottom shell, the palladium layer is used to improve the biocompatibility of the pipette inner surface. Once the walls are formed, an acetone bath is used to remove the micromolding photoresist and the sacrificial thick photoresist from the inside of the structure forming the hollow pipette [Fig. 2(e)]. To electrically isolate the microchannel, reverse sputtering or wet etching is used to remove the initial adhesion and seed layers. The membrane is removed from the membrane regions by SF dry etching (Oxford Plasma Tech Model 80) from the backside of the wafer. Thus, the shafts of the pipettes are released and freely suspended, projecting outward from the substrate [Fig. 2(f)]. B. Packaging Micromachined pipette arrays are packaged by bonding machined acrylic interfaces that connect external macroscale tubing to the pipette input ports. The acrylic interfaces contain important features, including a manifold covering individual or all pipette array input ports for pressure equalization and flow distribution, and a female receptacle for Micro Fingertight fittings (F-125, Upchurch Scientific Inc.) for connection to in (397 m) Teflon tubing. The pipette input ports can be fabricated either in the top wall or by etching through the substrate for backside access. Fig. 3 schematically illustrates interfacing schemes for both cases. The topside input ports are fabricated by modifying the mask for lithographic patterning of the palladium seed layer used for elec-
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 47, NO. 6, JUNE 2000
Fig. 3. Micromachined pipette array packaging schemes: (a) interfacing from the top via top input ports; (b) interfacing from the back via back-side input ports.
troforming the top shell in the Fig. 2(d) step. The fabrication of backside input ports requires modification of two masks-one for silicon nitride patterning prior to KOH etching in the Fig. 2(a) step and another for lithographic patterning of the copper seed layer used for bottom shell formation in the Fig. 2(b) step. The use of the backside input ports allows the macroscale interface between pipettes and macrotubing to be moved from the front of the device to the back. Thus, fluids can be transferred to and from pipettes via macrotubing as shown in Fig. 3(b). III. APPLICATION TO SLAB GEL ELECTROPHORESIS One of the challenges of future miniaturized biological/chemical analysis laboratories is manipulation of small (submicroliter) samples on microscale in a parallel fashion. Since the development and implementation of the 96-well microtiter plate technology, a considerable progress has been made to increase the microtiter plate density. A number of commercially available high-density formats has been developed based on the well-established 96-well format, including 386, 864, and 1536 wells. The commercially available sample handling systems, however, are lagging due to wide center-to-center spacing and inability to handle sample volumes below approximately 0.5 L. In addition, these systems are capable of dispensing only 6–12 pipettes at one time. With the current trends of miniaturizing biochemical analysis techniques (e.g., electrophoresis, chromatography, PCR) and the drive toward the development of the TAS, techniques must be developed to allow precise manipulation of submicroliter sample volumes in a highly parallel manner at small center-to-center spacing. The technique presented in this work addresses these problems by using the MPA for sample loading of miniaturized biochemical analysis systems. The general approach is schematically presented in Fig. 4. Initially, the samples are loaded into the MPA by capillary action. The volume of individual pipettes determines the amount of sample loaded onto the analysis system. Once loaded, the MPA is positioned over the input ports of the miniaturized analysis system and samples are dispensed using a syringe pump.
PAPAUTSKY et al.: MICROMACHINED PIPETTE ARRAYS
815
Fig. 4. Schematic diagram of the micromachined pipette array used for sample delivery to a generic analysis system.
In this work, we demonstrate the application of the MPA to slab gel electrophoresis. An electrophoretic gel (7 cm 10 cm) is prepared by first dissolving 0.56 g of agarose in 70 mL 1 TBE buffer. This is accomplished by heating 70 mL of TBE buffer in a microwave for 2–3 min and allowing to cool for 5–10 min. Next, 5 L (5 mg/ L) of ethidium bromide is added and gel is poured into the plastic form. Combs are then applied to form wells and gel is allowed to set for 30–40 min. Once agarose is polymerized, combs are removed from the gel. Samples are prepared by mixing 31 L of HPLC water, 1 L (1 g/ L) of 1-kB DNA ladder, and 8 L of mapping dye. The mapping dye consists of 25% glycerol, 0.1% BPB, 10-mM Tris adjusted to pH of 7.6, and 5-mM EDTA. Two kinds of combs, and thus sample loading methods, are used with the prepared electrophoretic gels. The first type is the commercially available mini-combs from Fisher Scientific with 4-mm-wide teeth spaced at 6 mm center-to-center. Thus, these gels are called mini-gels. The 5- L samples are loaded into the wells of the mini-gels using a standard P10 pipette. The second type of combs has 600- m-wide teeth spaced at 1 mm center-to-center. These combs are micromachined [40], thus the gels are referred to as micro-gels. The 0.5- L samples are loaded onto the micro-gels using the micromachined pipettes. The samples are initially loaded into the MPA by capillary action by first pipetting DNA samples into a titer dish well and then placing ends of the pipettes into the well. The volume of individual pipettes is 0.5 L, which corresponds to the amount of sample loaded. The samples are dispensed onto the gel by a controllable syringe pump that produces the 1 TBE buffer flow directed to the interface through Teflon tubing. Once gels are loaded, they are placed in 1 TBE buffer and an electrophoretic separation is performed. Following separation, DNA samples are visualized using transmitted ultra-violet (UV) light. IV. RESULTS AND DISCUSSION Arrays of pipettes of different dimensions and center-to-center spacing have been fabricated on top of silicon substrates. The lengths of the pipettes range from 2.5 mm to several cm with 500 m to 10 mm extending from the substrate. The inner widths of the pipettes range from 5 m to 1.5 mm, while the heights, which is also the thick photoresist thickness, ranges from 5 m to 50 m. The center-to-center spacing of these pipette arrays varies from 100 m to 1 cm. The pipette arrays fabricated to date include those electrodeposited from low-stress Watt’s nickel baths, nickel sulphamate, cupric
(a)
(b) Fig. 5. (a) SEM micrograph of an array of pipettes extending from the silicon 1000 30 m (L W H). substrate. The wide sections are 12 750 Pipettes extend 1.5 mm from the substrate and are 400 m wide. The structural material is electroformed nickel with wall thickness of 15 m. (b) Close-up of the end of a pipette; the inner cross-sectional area is 400 30 m :
2
2
2
2
2
sulfate, and palladium phosphate solutions [39]. The thickness of the electroplated walls ranges from 5 m to 25 m. The sacrificial photoresist removal rate is dependant on the bake cycle of the sacrificial photoresist [36]. In order to obtain high rates, the temperature of the sample after the development of the sacrificial photoresist must be kept as low as possible during the remaining processing steps. Typical times required to clear the pipettes are on the order of hours, depending on the pipette dimensions. Consequently, the pipette length is limited to approximately 30 mm as the removal of sacrificial photoresist in longer channels takes an excessively long time. Detailed information regarding clearance of the metallic pipettes and channels can be found in [41]. A typical micromachined pipette array (MPA) fabricated on top of silicon substrate is shown in Fig. 5. Sections of the individual pipettes remaining on the surface of the silicon substrate are expanded in width to increase the pipette volume. The pipettes are 12.75 mm long and 1.5 mm wide, while the pipette shafts extend 2 mm from the silicon substrate and are 500 m wide. The inner thickness of these pipettes is 30 m. The electroplated nickel walls are 15 m thick with 1- m palladium layer lining the inside. The volume of the individual pipettes in this array is 0.5 L. Interfaces machined from acrylic are used to connect the in (397 m) inner diameter Teflon fabricated MPA with tubing. Typical dimensions of the interfaces are approximately
816
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 47, NO. 6, JUNE 2000
(a)
(a)
(b)
Fig. 7. Photographs of a slab gel electrophoretic separations of 1kB DNA ladder using (a) a standard mini-gel and (b) a microscale gel. (b) Fig. 6. (a) Side-view photograph of a micromachined pipette array interfaced from the backside. Teflon tubing with a 1=32 in (397 m) inner diameter connects the acrylic interface to an external pressure source. A penny is provided for size comparison. (b) Top-view photograph of a MPA (seven pipettes) interfaced form the backside.
6 mm in width and 12 mm in height, while length ranges from 6 mm to 12 mm based on pipette dimensions and their number in an array. The acrylic interfaces are bonded to the MPA using a UV-curable adhesive (3321, Loctite Corp.) that provides a leak-tight connection. Teflon tubing connects the acrylic interfaces to a pressure source, typically a syringe pump. The top and side view photographs of an interfaced micromachined pipette array are presented in Fig. 6. Samples are loaded into the MPA using capillary action. Prior to each use, the MPA are cleaned by running 1 TBE buffer through for several minutes. The volume of individual pipettes used in this study is 0.5 L, which corresponds to the amount of sample loaded. Once loaded, the MPA are visually aligned with wells formed in electrophoretic micro-gels and samples are dispensed using a syringe pump. Due to the small size of wells and a nearly transparent gel, the sample dispensing is performed under a 2.25 illuminated magnification lamp (KFM Series, Terra Universal, Inc.). A controllable syringe pump is used to produce 1 TBE buffer flow to expel samples from the MPA during sample dispensing. Initially, the MPA is positioned so that the tips are inside the wells. The MPA is then gradually moved away from the gel, thus allowing to fill the wells. The buffer flow is stopped once the samples are transferred and the MPA tips have cleared the top of the wells.
Fig. 8. Longitudinal scans obtained from both standard mini-gel and a microscale gel.
The electrophoretic separations, both mini-gel and micro-gel, are carried out for 2 hrs in 1 TBE buffer under electric field of 7 V/cm (total voltage of 70 V). Micro-gel separations performed to date include those with a center-to-center spacing ranging from 250 m to 1.9 mm. Photographs of micro-gel and mini-gel taken under a UV light following a separation are shown in Fig. 7. The 5-lane micro-gel separation with 1 mm center-to-center spacing can be seen clearly and is comparable to the mini-gel separation. The longitudinal intensity scans for both systems presented in Fig. 8 indicate that the separations are in fact identical. Based on the sharpness of the peaks, the number of theoretical plates of the micro-gel separation is approximately 25 915. This is nearly a twofold increase compared
PAPAUTSKY et al.: MICROMACHINED PIPETTE ARRAYS
817
The application of the MPA to the high lane density slab gel electrophoresis has been explored. Wells are formed in agarose gel using micromachined combs (solid MPA) at center-to-center spacing from 250 m to 1.9 mm, while samples are loaded using the MPA. The MPA were initially loaded with samples using capillary action, with inner dimensions of individual pipettes determining the volume of the loaded samples. The samples were dispensed into the wells of the electrophoretic gel using a syringe pump. The results of the microscale separations compared favorably with the standard mini-gel separations and showed a twofold increase in the number of theoretical plates and a sixfold reduction in lane spacing. ACKNOWLEDGMENT (a)
The authors would like to acknowledge J. “Griff” Griffin of Clariant Corp. for donating P4620 photoresist and the University of Utah Undergraduate Research Opportunity Program (UROP). REFERENCES
(b) Fig. 9. Transverse scans of separations using (a) a standard mini-gel and (b) a microscale gel.
to the mini-gel system (13 196), meaning that separations can be done in half the time or at twice the resolution. In addition, as seen from Fig. 8, individual lanes of the micro-gel separation are well separated and there is very little cross talk. The transverse scans for both gels are presented in Fig. 9. Overall, the sample size was reduced from 5 L for the mini-gel to 0.5 L for the micro-gel system. V. CONCLUSION Micromachining technologies have been used as a method for batch fabrication of metallic micromachined pipette arrays. Using micromachining technology, pipettes made from a variety of electroplated metals have been fabricated. There are many advantages of the micromachined pipette arrays (MPA), including: 1) the use of low temperature IC compatible fabrication process; 2) the ability to transfer precise volumes of samples in the submicroliter range; 3) the ability to handle samples/reagents in a highly parallel fashion by manipulating hundreds of samples/reagents simultaneously; 4) a wide range of pipette center-to-center spacing; 5) serve as an interface between macroscale and microscale instrumentation; and 6) the possibility of adding functionality to the pipettes, such as polymerase chain reaction (PCR).
[1] K. L. Drake, K. D. Wise, J. Farraye, D. J. Anderson, and S. L. BeMent, “Performance of planar multisite microprobes in recording extracellular single-unit intracortical activity,” IEEE Trans. Biomed. Eng., vol. 35, pp. 719–732, 1988. [2] K. Najafi and K. D. Wise, “A scaled electronically-configurable multichannel recording array,” Sensors Actuators, vol. A21–A23, pp. 589–591, 1990. [3] N. A. Blum, B. G. Carkhuff, H. K. Charles, R. L. Edwards, and R. A. Meyer, “Multisite microprobes for neural recordings,” IEEE Trans. Eng., vol. 38, pp. 68–74, Jan. 1991 . [4] A. B. Frazier, D. P. O’Brien, and M. G. Allen, “A two dimensional metallic micro electrode array,” IEEE J. Microelectromech. Syst., vol. 2, no. 2, p. 87, 1993. [5] S. C. Terry, J. H. Jerman, and J. B. Angell, “A gas chromatographic air analyzer fabricated on a silicone wafer,” IEEE Trans. Electron. Devices, vol. ED-26, pp. 1880–1886, Dec. 1979. [6] R. R. Reston and E. S. Kolesar, “Silicon-micromachined gas chromatography system used to separate and detect ammonia and nitrogen dioxide—Part I: Design, fabrication, and integration of the gas chromatography system,” IEEE J. Microelectromech. Syst., vol. 3, pp. 134–146, Apr. 1994. [7] G. Ocvirk, E. Verpoorte, A. Manz, and H. M. Widmer, “Integration of a micro liquid chromatography onto a silicon chip,” in Proc. Tranducers ’95, Stockholm, Sweden, June 25–29, 1995, pp. 756–759. [8] A. Manz, Y. Miyahara, J. Miura, Y. Watanabe, H. Miyagi, and K. Sato, Sensors Actuators, vol. B1, pp. 249–255, 1990. [9] D. E. Raymond, A. Manz, and H. M. Widmer, “Continuous sample preparation using free-flow electrophoresis on a silicon microstructure,” in Proc. Transducers ’95, Stockholm, Sweden, June 25–29, 1995, pp. 760–765. [10] A. Manz, “Miniaturized chemical analysis systems based on electroosmotic flow,” in Proc. IEEE Micro Electro Mechanical Syst. Conf., Nagoya, Japan, Jan. 26–30, 1997, pp. 1–4. [11] C. S. Effenhauser, A. Manz, and M. H. Widmer, “Glass chips for high-speed capillary electrophoresis separation with submicrometer plate heights,” Anal. Chem., vol. 65, pp. 2637–2643, 1993. [12] Z. H. Fan and D. J. Harrison, “Micromachining of capillary electrophoresis injectors and separators on glass chips and evaluation of flow at capillary interconnections,” Anal. Chem., vol. 66, pp. 177–184, 1994. [13] S. C. Jacobsen, A. W. Moore, and J. M. Ramsey, “Fused quartz substrates for microchip electrophoresis,” Anal. Chem., vol. 67, pp. 2059–2063, 1995. [14] H. Nakanishi, T. Nishimoto, N. Nakamura, S. Nagamachi, A. Arai, Y. Iwata, and Y. Mito, “Fabrication of electrophoresis devices on a quartz and glass substrates using a bonding with HF solution,” in IEEE Micro Electro Mechanical Syst. Conf., Nagoya, Japan, Jan. 26–30, 1997, pp. 299–304.
818
[15] B. K. Gale, A. B. Frazier, and K. Caldwell, “Micromachined electrical field-flow fractionation (-EFFF) system,” in IEEE Micro Electro Mechanical Syst. Conf., Nagoya, Japan, Jan. 25–30, 1997, pp. 119–124. [16] M. A. Northrup, C. Gonzalez, D. Hadley, R. F. Hills, P. Landre, S. Lehew, R. Saiki, J. J. Sninsky, and R. Watson, “A MEMS-based miniature DNA analysis system,” in Proc. Transducers ’95, Stockholm, Sweden, June 25–29, 1995, pp. 764–767. [17] R. C. Anderson, G. J. Bogdan, and R. J. Lipshutz, “Miniaturized genetic-analysis system,” in Proc. IEEE Solid-State Sensor and Actuator Workshop, Hilton Head, SC, June 2–6, 1996, pp. 258–261. [18] G. Volpe, D. Moscone, D. Compagnone, and G. Palleschi, “In vivo continuous monitoring of 1-lactate coupling subcutaneous microdialysis and an electrochemical biocell,” Sensors Actuators, vol. B24–25, pp. 138–141, 1995. [19] T. Kuriyama, “Non-invasive blood glucose monitoring,” presented at the Transducers ’95, Stockholm, Sweden, June 25–29, 1995. [20] N. Ito, T. Matsumoto, S. Nakamoto, H. Fujiwara, Y. Mastumoto, S. Kayashima, T. Arai, M. Kikuchi, and I. Karube, “Transcutaneous lactate monitoring with a micro-planar amperometric biosensor,” presented at the Transducers ’95, Stockholm, Sweden, June 25–29, 1995. [21] R. Austin, R. Carlson, S. Chan, and J. Knight, “Micro and nanopore structures for biological applications,” in Proc. Transducers ’97, Chicago, IL, June 16–19, 1997, pp. 1303–1305. [22] H. Ayliffe, R. Rabbit, P. Tresco, and B. Frazier, “Micromachined celluar characterization system for studying the biomechanics of individual cells,” in Proc. Transducers ’97, Chicago, IL, June 16–19, 1997, pp. 1307–1310. [23] L. Bousse, “Whole cell biosensors,” in Proc. Transducers ’95, Stockholm, Sweden, June 25–29, 1995, pp. 483–486. [24] P. E. Anderson, P. C. H. Li, R. Smith, R. J. Szarka, and D. J. Harrison, “Biological cell assays on an electrokinetic microchip,” in Proc. Transducers ’97, Chicago, IL, June 16–19, 1997, pp. 1311–1314. [25] D. Mathieson, U. Beerschwinger, B. J. Robertson, S. J. Yang, R. L. Reuben, R. A. Lawes, F. N. Goodall, J. Spencer, and R. S. Bartholomew, “Design considerations for surgical microfluid actuators,” Nanobiology, vol. 3, pp. 123–132, 1994. [26] D. Mathieson, B. J. Robertson, U. Beerschwinger, S. J. Yang, R. L. Reuben, and A. J. Addlesee, “Micro-torque measurements for a prototype turbine,” J. Micromech. Microeng., vol. 4, pp. 190–196, 1994. [27] Abbott Critical Care Systems, Our Products Guide Salt Lake City, UT. [28] T. Bell, K. D. Wise, and D. Anderson, “A flexible micromachined electrode array for a cochlear prosthesis,” in Proc. Transducers ’97, Chicago, IL, June 16–19, 1997, pp. 1315–1318. [29] S. Sugiyama, K. Kawahata, M. Yoneda, and I. Igarashi, “Tactile image detection using a 1K-element silicon pressure sensor array,” Sensors Actuators, vol. A21–A23, pp. 397–400, 1990. [30] R. J. DeSouza and K. D. Wise, “A high density tactile imager for reading embossed characters,” in Proc. IEEE Solid-State Sensor and Actuator Workshop, Hilton Head, SC, June 13–16, 1994, pp. 13–16. [31] R. C. Anderson, G. J. Bogdan, Z. Barniv, T. D. Dawes, J. Winkler, and K. Roy, “Microfluidic biochemical analysis system,” in Proc. Transducers ’97, Chicago, IL, June 16–19, 1997, pp. 477–480. [32] M. A. Burns, C. H. Mastrangelo, T. S. Sammarco, F. P. Man, A. R. Kaiser, and D. T. Burke, “Microfabricated structures for integrated DNA analysis,” Proc. Nat. Acad. Sci., vol. 93, pp. 5556–5561, 1996. [33] K. Handique, D. T. Burke, C. H. Mastrangelo, and M. A. Burns, “Nanoliter-volume discrete drop injection and pumping in microfabricated chemical analysis systems,” in Proc. IEEE Solid-State Sensor and Actuator Workshop, Hilton Head, SC, June 8–11, 1998, pp. 346–349. [34] T. D. Boone, H. H. Hooper, and D. S. Soane, “Integrated chemical analysis on plastic microfluidic devices,” in Proc. IEEE Solid-State Sensor and Actuator Workshop, Hilton Head, SC, June 8–11, 1998, pp. 878–92. [35] S. C. Jacobson, C. T. Culbertson, and J. M. Ramsey, “High-speed microchip electrophoresis: Exploring the limits,” in Proc. IEEE Solid-State Sensor and Actuator Workshop, Hilton Head, SC, June 8–11, 1998, pp. 93–96. [36] I. Papautsky, J. Brazzle, H. Swerdlow, and A. B. Frazier, “A low temperature, IC compatible process for fabricating surface micromachined metallic microchannels,” IEEE J. Microelectromech. Syst., vol. 7, no. 2, pp. 267–273, 1998. [37] H. Siedel, L. Csepregi, and A. Heuberger, “Anisotropic etching of crystalline silicon in alkaline solutions,” J. Electrochem. Soc., vol. 37, pp. 3612–3620, 1990. [38] M. W. Putty and K. Najafi, “A micromachined vibrating ring gyroscope,” in Proc. IEEE Solid-State Sensor and Actuator Workshop, Hilton Head, SC, June 13–16, 1994, pp. 213–220.
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 47, NO. 6, JUNE 2000
[39] W. H. Safrenek, The Properties of Electrodeposited Metals and Alloys. New York: Elsevier, 1974. [40] I. Papautsky, J. D. Brazzle, R. B. Weiss, T. A. Ameel, and A. B. Frazier, “Parallel sample manipulation using micromachined pipette arrays,” in Proc. SPIE Microfludic Devices and Systems, Santas Clara, CA, Sept. 21–22, 1998, pp. 104–114. [41] I. Papautsky, “Metallic micro instrumentation for biomedical fluid applications,”, Univ. of Utah, Salt Lake City, Dec. 1999.
Ian Papautsky revived the B.S. degree in biomedical engineering from Boston University, Boston, MA, in 1995 and the Ph.D. degree in bioengineering from University of Utah, Salt Lake City, in 1999. His Ph.D. research involved the microscale instrumentation for biomedical fluid applications at the Micro Instrumentation Research Laboratory. He is currently an Assistant Professor of Electrical and Computer Engineering at the University of Cincinnati, Cincinnati, OH. His research interests include biomedical microelectromechanical systems (BioMEMS), microfluidics, biosensors, microscale genetic analysis systems, and lab-on-a-chip devices.
John Brazzle received the B.S. degree in electrical engineering from the University of Utah, Salt Lake City, in 1997, and the M.S. degree in bioengineering at the Micro Instrumentation Research Laboratory, Department of Bioengineering, University of Utah, in 2000. For his MS thesis, he designed, fabricated, characterized, and packaged hollow metallic micromachined needles for use in drug delivery and fluid extraction. He has authored a number of journal and conference publications in this area. His other research interests include micromachining technologies, semiconductor device fabrication, and microscale self-assembly systems. He is currently a Development Engineer with IntelliSense Corporation, Wilmington, MA.
Harold Swerdlow was born in Brooklyn, NY, in 1957. He received the Ph.D. degree in bioengineering from the University of Utah, Salt Lake City, in 1991. In 1993, he joined the Bioengineering and Human Genetics Departments of the University of Utah as a Faculty Member. In 1998, he relocated to Stockholm, Sweden, to help nucleate the Center for Genomics Research, a newly formed department of the Karolinska Institute. His research focuses on applying instrumental methods to solve molecular biology problems. He has worked for many years in the field of capillary electrophoresis, specifically, the use of multiple-capillary methods to perform high-throughput DNA sequencing. He is also interested in fluidics technologies for robust, efficient sample preparation and loading as well as the application of microfabrication and microarraying methods to separation science and DNA analysis.
Robert Weiss is an Associate Professor in the Department of Human Genetics at the University of Utah, Salt Lake City. He is Co-Director of the Utah Genome Center. His research interests include large-scale DNA sequencing and novel technologies for DNA-based hybridization assays. He trained in molecular genetics at the University of Washington, Seattle, and made contributions to the field of protein synthesis before entering the new arena of genomics. His multidisciplinary team of molecular biologists, engineers, and computer scientists at the Utah Genome Center is one of ten selected to participate in the new National Institutes of Health (NIH)-sponsored Mouse Genome Sequencing Network.
PAPAUTSKY et al.: MICROMACHINED PIPETTE ARRAYS
A. Bruno Frazier received BS and MS degrees in electrical engineering from Auburn University, Auburn, AL, in 1986 and 1987, respectively. In December 1993, he received his Ph.D. in Electrical and Computer Engineering from Georgia Institute of Technology, Atlanta. From 1987 to 1990, he worked for Intergraph Corporation as a Video Graphics Custom Circuit Designer for the UNIX-based computer-aided graphics systems. From 1990 through 1993, he attended Georgia Institute of Technology, Atlanta, and conducted research into micromachining processes for the fabrication of metallic microstructures, development and characterization of micromachining materials, as well as micromachined devices utilizing the previously developed processes and materials. From March 1994 to July 1995, he worked as a Visiting Scholar at the University of Michigan, Ann Arbor. While at the University of Michigan, he focused on the development of integrated hygrometers and the associated mixed signal interface circuitry. From August 1995 to August 1999, he held a joint faculty position as a Professor of Bioengineering and Electrical Engineering at the University of Utah, Salt Lake City. In September 1999, he joined the faculty at Georgia Institute of Technology. He is currently a Professor in the Electrical and Computer Engineering Department. His current research interests are in the area of micro instrumentation systems including biomedical, optical, and magnetic applications of micro systems technology. In general, the micro systems research includes MEMS devices as well as interface circuitry necessary for system operation. His other research interests are related to systems-level issues such as integration of circuitry and packaging of the micro systems.
819