Symposium Organizers
Frank W. DelRio National Institute of Standards and Technology
Maarten P. de Boer Carnegie Mellon University
Christoph Eberl Karlsruhe Institute of Technology (KIT)
Evgeni P. Gusev Qualcomm MEMS Technologies
S1: Fabrication Methods
Session Chairs
Monday PM, November 29, 2010
Room 207 (Hynes)
9:30 AM - **S1.1
Transfer Printing With Advanced Stamps For Applications in MEMS and Other Areas.
John Rogers 1
1 , University of Illinois, Urbana, Illinois, United States
Show AbstractFabrication in MEMS often demands formation of complex, three dimensional layouts and heterogeneous collections of materials. This talk summarizes some work on the use of soft, elastomeric stamps for transfer printing solid objects of silicon and other materials relevant to MEMS. Advanced stamp designs that incorporate strategically located microtips enable large, passive switching of adhesion with a contrast ratios approaching 1000 times, thereby providing unmatched capabilities in transfer. We present theoretical and experimental studies of the mechanics of these stamps, and demonstrate their use in forming two and three dimensional structures of silicon, metal and oxides in representative devices.
10:00 AM - S1.2
Fabrication and Replication of Hierarchically Textured Polymer Microstructures Using Carbon Nanotube Master Molds.
Davor Copic 1 , Sameh Tawfick 1 , Sei Jin Park 1 , Michael De Volder 1 2 , A. John Hart 1
1 Mechanical Engineering, University of Michigan, Ann Arbor, Michigan, United States, 2 , IMEC, Leuven Belgium
Show AbstractFabrication of polymer structures and surfaces having controlled shape and texture at a hierarchy of length scales is essential to understanding and controlling the behavior of soft matter, such as liquid wetting, dry adhesion, and cell-substrate interactions. The advent of soft lithography using PDMS transformed our capability for cost-effective fabrication of structures for these applications; however, master mold features and the resulting replica structures typically have straight cross-sections, and it is particularly difficult to create vertical and multi-directional nanoscale textures on microstructures. To meet this need, we introduce use of carbon nanotube (CNT) composite master molds to fabricate textured polymer microstructures and polymer surfaces. These novel master molds are made by infiltration of vertically aligned CNT microstructures with SU-8, followed by replication using standard soft lithography methods. The master molds are made by (1) CNT growth by thermal CVD from a lithographically patterned catalyst; (2) capillary densification of the CNTs by condensation and evaporation of a solvent onto the substrate; and (3) infiltration of the densified CNTs with SU-8. The densification step critically improves the robustness of the CNT microstructures while maintaining the aligned nanoscale texture of the sidewalls of the microstructures, which is precisely copied into the final polymer replicas as quantified by atomic force microscopy. Uniform centimeter-scale arrays of structures having critical dimensions ranging from 5-1000 micrometers have been made in both SU-8 and PDMS. These include high-aspect-ratio needles, micro-foams with hexagonal cells, and thin sheets with radially anisotropic surface texture. The hierarchical roughness of the master and replica surfaces is controlled by the CNT diameter and packing density, and is demonstrated to enable uniform and tunable SERS enhancement on Au-coated and Ag-coated replicas. This study indicates promise for batch fabrication and replication of polymer features that capture complex nanoscale shapes and textures, while retaining compatibility with existing microfabrication methods.
10:15 AM - S1.3
Carbon Nanotube Based NEMS Actuators.
Michael Forney 1 , Jordan Poler 2
1 Nanoscale Science, UNC Charlotte, Charlotte, North Carolina, United States, 2 Chemistry, UNC Charlotte, Charlotte, North Carolina, United States
Show AbstractCarbon nanotubes (CNTs) have been widely studied because of their superior mechanical properties as well as their ballistic one-dimensional electronic behavior. We have grown vertically aligned CNTs (VA-CNTs) grown onto Octosensis microcantilever arrays, which provide an architecture for novel actuators. SEM and Raman spectroscopy indicate that the CNTs are small multi-walled carbon nanotubes. Electrical characterization shows reasonable electrical resistance (~ 70 kΩ) from the chip body to the microcantilever tips. By applying a voltage, we load charge onto the array of VA-CNTs. The electrostatic repulsion among the charged CNTs provides surface stress that induces microcantilever deflection. COMSOL Multiphysics modeling results for the actuator design show that only a few electrons per CNT are needed to produce measureable deflections, and experimental actuators are currently being characterized using SEM, Raman spectroscopy, an I-V probe station, and an AFM optical lever system. Other microcantilever actuators in the literature have been successful at inducing deflections of tens or hundreds of nanometers, and we expect larger deflections, coupled with faster response times. Actuators based on this architecture could be used for nano-manipulation, release of drugs from a capsule, or nano-valves that are controlled by electrical inputs.
10:30 AM - S1.4
Environmental Effects on the Elasticity of Cross-linked Poly(methyl methacrylate) Nano-wires Produced by Two-photon Lithography.
Satoru Shoji 1 , Tomoki Hamano 1 , Shota Kuwahara 1 , Thomas Rodgers 1 , Satoshi Kawata 1
1 Department of Applied Physics, Osaka University, Suita, Osaka Japan
Show AbstractRecent novel technologies allow us to study the intrinsic properties of polymers in the micro/nano-scale, which can be quite different from those of macro-scale polymers. Previously, we presented experimental evidence by our laser lithography and laser manipulation techniques [1-3] that the elastic modulus of polymers starts to show size-dependence when the size of the polymers is less than a micrometer. The size-dependent mechanical properties of poly(methyl methacrylate) (PMMA) nano-wires fabricated by two-photon lithography were observed as an enhancement of the elasticity and a concurrent decrease of the phase transition temperature. Although the mechanism of these size-dependent features is not fully understood yet, it is clear that the size-dependence is seen only when the polymer is thinner than a certain critical size. One of the possible factors is the effect of molecules surrounding and/or penetrating into the surface of the polymer. In this presentation, we show a comparison of the elasticity of PMMA nano-wires in different circumstances. We prepared crosslinked-PMMA polymer nano-wires in the shape of coil springs with different radii from 100 to 500 nm. We performed stress tests for each nano-wire by means of laser trapping technique and/or atomic force microscopy for two cases where the nano-wires are dried in air, and are wet in ethanol. We found that the elastic modulus of the polymer nano-wires significantly drops by three orders of magnitude when they are immersed in ethanol. This significant atmosphere-dependent change of the elasticity was seen when the size of the polymer wires is smaller than 500 nm. It is known that, in general, ethanol does not have a good affinity to PMMA. We speculate that, although it is very weak, the interaction of ethanol with PMMA macromolecules in a thin layer of the polymer surface could induce such significant modulation of elasticity. In the presentation we also introduce our methods for the preparation and the tensile test of the polymer nano-wires with high precision. [1]S. Shoji, S. Nakanishi, T. Hamano, and S. Kawata, MRS Proc. 1224, 1224-FF06-05-DD06-05 (2009). [2]S. Nakanishi, S. Shoji, H. Yoshikawa, Z. Sekkat, and S. Kawata, J. Phys. Chem. B 112, 3586 (2008). [3]S. Nakanishi, S. Shoji, S. Kawata, and H.-B. Sun, Appl. Phys. Lett. 91, 063112(2007).
10:45 AM - S1.5
Xenon Difluoride Etching of Germanium for Free-standing III-V Heterostructures.
Garrett Cole 1 , Yu Bai 2 , Markus Aspelmeyer 1 , Eugene Fitzgerald 2
1 Physics, University of Vienna, Vienna, Vienna, Austria, 2 Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States
Show AbstractWe have developed a novel fabrication technique for the realization of free-standing monocrystalline AlGaAs heterostructures of arbitrary aluminum content. This process is enabled by recent advances in high quality III-V/Ge epitaxy and utilizes the noble gas halide, xenon difluoride (XeF2), in order to rapidly and selectively remove a sacrificial germanium (Ge) underlayer in a room temperature gas-phase etching procedure. Combining polar and non-polar semiconductor multilayers allows for the first time the extension of dry selective undercutting, as has been pioneered in silicon-based microstructures, to high performance monocrystalline compound semiconductor devices. We demonstrate two possibilities for exploiting this unique procedure: 1) bulk micromachining of an optomechanical resonator consisting of an epitaxial GaAs/AlAs distributed Bragg reflector (DBR) grown on a Ge substrate, and 2) epitaxial lift-off (ELO) of submicron-thickness GaAs films via removal of an embedded Ge sacrificial layer.All samples are prepared using a low-pressure metal organic CVD system. The optomechanical resonators are fabricated from an epitaxial DBR consisting of 40.5 periods of alternating GaAs (high index) and AlAs (low index), grown on a 2” diameter, epi-ready (100) Ge substrate, offcut 6 degrees toward the [011] direction. The use of the offcut substrate is necessary to obtain high quality GaAs epitaxy on Ge by inhibiting anti-phase boundaries. Resonator fabrication entails a single-mask bulk micromachining process utilizing a pulsed XeF2 etching system used for selective underetching of the Ge substrate. In order to explore the limits of this process, we further demonstrate ELO of sub-micron thickness, single-crystal GaAs films. Two distinct materials structures are used in this experiment: both structures are grown on an epi-ready (100) GaAs substrate offcut 6 degrees toward the [011] direction, followed by 190 nm of Ge (1 μm in the second sample), capped with 180 nm (or 320 nm) of GaAs. We record respective lateral etch rates of 30 and 50 μm/min in these samples; a factor of 5 larger than that typically found in silicon.Post etch characterization yields selectivities between 0.26×10^6 and 1.8×10^6 for Ge versus AlGaAs etching. Compared with typical results for hydrofluoric-acid-based ELO, our demonstration XeF2 process shows similar selectivity, is an order of magnitude faster, and additionally enables the release of epitaxial films with arbitrary aluminum content. Along with a rapid etch rate and high selectivity to all AlGaAs alloy compositions, further advantages of this etchant include the elimination of surface tension forces in the release of suspended structures, and the alleviation of ion-induced damage associated with plasma exposure.
11:30 AM - **S1.6
Production and Characterization of Molded Micro-parts Made of Metals and Ceramics.
Oliver Kraft 1
1 izbs, Karlsruhe Institute of Technology, Karlsruhe Germany
Show AbstractIn the past decades, miniaturization of micro electro-mechanical systems (MEMS) based on silicon technology has led to ever smaller, cheaper and better devices demonstrating strong economic success in a broad range of technological applications. Meanwhile, functional systems to be developed for biological or medical applications, for micro-chemical engineering or in many other relevant areas of science and technology call for a much wider selection of materials and for parts with true three-dimensional shapes. For this, the development of production routes other than typical silicon processing is required. In particular, for the fabrication of micro-components, made of metals and ceramics, manufacturing processes known from the macro world are to be scaled down. Therefore, it has been the major goal of a Collaborative Research Center at the Karlsruhe Institute of Technology to investigate the limits of this approach and to develop the full process chain for micro-molding including powder injection-molding as well as micro-casting. This talk will give an overview of the major achievements with a particular emphasis on the processing-microstructure- property relationship of micro-parts made from cast alloys as well as high strength ceramics. For mechanical characterization at the micro-scale, a number of micro-testing methods have been developed to study strength, toughness and fatigue endurance of the produced materials. Overall, it turns out that at a scale between microns and a few hundred microns scaling effects play an important role for the mechanical behavior as samples and components have a large surface to volume ratio. However, an equally important aspect relates to the fact that the microstructure and the resulting properties of the produced materials are quite sensitive to details of the production route. Nevertheless, it is shown by the successful manufacturing of complex parts and demonstrators that micro-molding allows for processing a large variety of materials at small scale.
12:00 PM - S1.7
All-Oxide MEMS Devices Based on Free-standing Structures of Epitaxial Transition Metal Oxides.
Luca Pellegrino 1 , Michele Biasotti 1 2 , Renato Buzio 1 , Emilio Bellingeri 1 , Nicola Manca 2 , Cristina Bernini 1 , Antonio Sergio Siri 2 1 , Daniele Marre 2 1 , Teruo Kanki 3 , Hidekazu Tanaka 3
1 , CNR SPIN , Genova Italy, 2 Physics Department, University of Genova, Genova Italy, 3 Institute of Scientific and Industrial Research, Osaka University, Osaka Japan
Show AbstractMost of the applications envisaged for Transition Metal Oxides (TMO) concern electronic or optoelectronic devices such as memories, transistors, LED. An additional route toward the realization of devices employing the rich properties of TMOs is the fabrication of free-standing elements of epitaxial TMO thin films for applications in smart Microelectromechanical Systems (MEMS) devices. Our fabrication process starts from the deposition of epitaxial oxide films or multilayers that are micromachined by conventional microlithography. A key factor of the process is the use of crystalline oxide sacrificial layers that are selectively removed by acids thus leaving free-standing elements. As an example, the combined use of HF and HCl allowed the fabrication of free-standing structures made of crystalline SrTiO3 (001) films using sacrificial layers of (La,Sr)MnO3 [1,2]. In some cases, combination of dry etching and acids is employed like in the fabrication of epitaxial rutile TiO2 (110) microcantilevers. After a brief description of the fabrication protocol used to obtain microcantilevers and suspended bridges based on different TMOs, we will show possible applications of suspended structures based on epitaxial (La,Sr)MnO3, (La,Sr)CoO3 and VO2 films. Strain generator devices, in which reversible modulation of the electrical resistance of TMO thin films is induced by crystalline SrTiO3(001) microcantilevers, will be illustrated. By bending downward the microcantilever through an AFM tip or gate electric fields, tensile strain is produced at the upper surface of the cantilever and the overgrown film is thus strained through epitaxial lock. MEMS-based strain devices are expected to influence properties of TMOs inducing shifts of Metal-Insulator Transitions - thus affecting properties of applicative interest such as the temperature coefficient of resistance (TCR) - or even modifying magnetic ordering. Further perspectives such as the realization of micro-hotplates and micro bolometers based on functional epitaxial oxides will be introduced as well as the possibility of measuring interacting forces between correlated oxides and external fields using oxide micromechanical oscillators. The combination of strain, mechanical vibration and high thermal insulation together with the enormous possibilities offered by TMO epitaxial heterostructures is a fascinating field that could open interesting perspectives for applications of such materials. [1] L. Pellegrino, M. Biasotti, E. Bellingeri, C. Bernini, A. S. Siri, D. Marré Adv. Mater. 21, 2377 (2009)[2] M. Biasotti, L. Pellegrino, E. Bellingeri, C. Bernini, A. S. Siri, D. Marré Procedia Chemistry 1 839–842 (2009)
12:15 PM - S1.8
A New Route To Fabricating PDMS-Based Microfluidic Components.
Roger Diebold 1 2 , David Clarke 2
1 Materials Department, UC Santa Barbara, Santa Barbara, California, United States, 2 School of Engineering and Applied Sciences, Harvard University, Cambridge, Massachusetts, United States
Show AbstractIn the efforts to build standardized microfluidic components for ‘Lab-on-a-chip’ purposes, Polydimethylsiloxane (PDMS) is a structural material which has gained much attention. However, the low surface energy of cured PDMS limits the use of optical lithography since many of the traditional Novolak photoresists used in optical lithography dewet. Several different materials (metal, parylene, etc) have been applied as adhesive layers on cured PDMS for the purpose of lithographic patterning but in doing so necessitate undesirable alterations to the PDMS surface via oxygen plasma treatment. In addition, soft lithographic methods used in fabricating microfluidic devices also require oxygen plasma treatment of the PDMS surface for bonding purposes, which has been previously shown to have irreproducible bond strength and inherently imprecise feature alignment. In this paper, the authors will present an alternative route for microfluidic device fabrication which is completely solution processable, does not require plasma treatment, and fully compatible with basic optical lithography equipment.Polydimethylglutarimide (PMGI) is a commercially available deep UV and electron beam photoresist that provides sufficient adhesion to cured PDMS surfaces permitting traditional optical lithography techniques to be applied, allowing the fabrication of multilayer elastomeric structures useful in microfluidic applications. As an adhesive underlayer supporting a positive Novolak-based photoresist, PMGI is completely solution processable and easily removed using N-methyl-2-pyrrolidone-based photoresist strippers. To demonstrate the utility of PMGI in fabricating microfluidic devices, the authors have fabricated a PDMS-based, electrostatically actuated peristaltic pump without alteration of the PDMS surface. We anticipate that the technique presented will be applicable in fabricating other microfluidic device components.
12:30 PM - S1.9
Integrating Nanomaterials with Micromachined Structures Using Electron-beam Lithography.
Kaushik Das 1 , Pascal Hubert 1 , Srikar Vengallatore 1
1 Mechanical Engineering, McGill University, Montreal, Quebec, Canada
Show AbstractIntegrating nanomaterials (in the form of quantum dots, nanotubes, nanowires, nanocrystalline thin films, and nanocomposite films) with micromachined structures and devices can enable the design of microelectromechanical systems with multiple functionalities, improved performance, and higher reliability. However, achieving this integration is a daunting challenge because the processing techniques used for synthesizing the nanostructures are usually different from, and often incompatible with, the standard methods used for micromachining. Nanomaterials can be grown, or self-assembled, on micromachined structures, but it is difficult to control their dispersion, alignment, pattern density, and location. These parameters can be controlled if the nanostructures are patterned on micromachined devices using high-resolution lithography or direct-write techniques. Here, we report an approach for integrating polymeric and metallic nanowires directly on microcantilevers platforms that are commonly employed in MEMS-based sensors. Single-crystal silicon microcantilevers were micromachined using photolithography, anisotropic wet etching, and deep reactive-ion etching. These beams were then sputter-coated with thin aluminum films, spray coated with poly methyl methacrylate (PMMA), and then patterned using electron-beam lithography. The nanostructures comprise of an array of oriented trenches that are 200 nm wide and 450 nm deep, and the spacing between adjacent trenches ranges from 1 μm to 5 μm. A metal lift-off process was used to fabricate an array of oriented chromium nanowires (10 nm thick and 130 nm wide) with controlled spacing of 2 μm and 5 μm between adjacent lines. By using deposition after patterning, aluminum/PMMA/aluminum nanocomposites were also fabricated on the microcantilever. The use of these nanostructures for applications in resonant sensing, and for fundamental studies of idealized nanocomposite materials, will be discussed.
12:45 PM - S1.10
C-MEMS Structures as Three Dimensional Current Collectors for Micro-supercapacitors.
Majid Beidaghi 1 , Wei Chen 1 , Chunlei Wang 1
1 Mechanical & Materials Engineering, Florida International University, Miami, Florida, United States
Show AbstractDevelopment of miniaturized electronic systems has stimulated the demand for miniaturized power sources that can be integrated into such systems. Micro-supercapacitors with high power density can be coupled with energy harvesting devices to store the generated energy. Moreover, they can also be paired with micro-batteries to provide the peak power and improve the cycle lifetime. Electrically conducting polymers, such as polyaniline (PANI), polypyrrole (Ppy) and their derivatives, and transition metal oxides, such as RuO2 and MnO2 are promising electro-active materials for supercapacitors. In this work, we are aiming to develop on-chip supercapacitors based on interdigitated C-MEMS electrode microarrays, which are employed as three dimensional (3D) current collectors of pseudo-capacitive materials. Fabrication of C-MEMS structures involves a two-step photolithography on silicon oxide wafer followed by a pyrolysis step. Ppy and MnO2 were electrochemically deposited on the 3D interdigitated C-MEMS electrodes. Effects of different experimental parameters on the performance of micro-supercapacitor cells are investigated by Cyclic Voltammetry (CV), Galvanostatic Charge-discharge and Electrochemical Impedance Spectroscopy (EIS). Detailed results will be presented at the conference.
S2: Materials Development and Optimization
Session Chairs
Monday PM, November 29, 2010
Room 207 (Hynes)
2:30 PM - **S2.1
Towards the Development of Ni-base Superalloys as High Temperature MEMS Materials.
Devin Burns 1 , Michael Teutsch 2 , S. Suresha 1 3 , Klaus Bade 4 , Jarir Aktaa 2 , Sara Johnson 5 , Tresa Pollock 6 , Kevin Hemker 1
1 Mechanical Engineering, Johns Hopkins University, Baltimore, Maryland, United States, 2 Institute for Material Research II, Karlsruhe Institute of Technology, Karlsruhe Germany, 3 National Center For Electron Microscopy, Lawrence Berkeley National Laboratory, Berkeley, California, United States, 4 Institue for Microstructure Technology, Karlsruhe Institute of Technology, Karlsruhe Germany, 5 Materials Science and Engineering, University of Michigan, Ann Arbor, Michigan, United States, 6 Materials, University of California Santa Barbara, Santa Barbara, California, United States
Show AbstractElectrodeposited LIGA Ni micro-structures offer an attractive balance of toughness, stiffness, and room temperature strength, but the elevated temperature strength of traditional LIGA Ni components and molds are far from optimal. The technological motivation for the work to be presented is derived from a desire to expand the temperature capabilities of current MEMS materials. Cast and wrought Ni-base superalloys with highly developed two-phase (γ−γ’) microstructures are ubiquitous in land-based and aero turbines. The processing routes required to shape MEMS and NEMS components with sub-micron precision do not, however, lend themselves to traditional casting and forging processes. Two alternative processing routes will be presented and contrasted. The first involves control and optimization of electrodeposition parameters to uniformly entrain Al nano-particles in ED Ni micro-components, the development of heat treat schedules required to transform the as-deposited green composites into γ−γ’ superalloy microstructures, and characterization of the resultant microstructures and attendant mechanical properties. The second employs vapor-phase aluminization of existing LIGA Ni micro-structures to achieve desired alloy compositions, as well as subsequent heat treating and characterization of the resultant microstructures and mechanical properties. This talk will review the properties of traditional LIGA Ni micro-structures, outline recent attempts to develop solid solution and oxide dispersion strengthened LIGA Ni alloys, and then focus on the development of processing routes for the fabrication of Ni-base superalloys for use in MEMS and NEMS applications.
3:00 PM - S2.2
Mechanical and Piezoelectric Behavior of Thin Film PZT Composites for MEMS.
Ioannis Chasiotis 1 , Sivakumar Yagnamurthy 1
1 Aerospace Engineering, University of Illinois at Urbana-Champaign, Urbana, Illinois, United States
Show AbstractThe mechanical and piezoelectric properties of freestanding PZT composite films, comprised of SiO2, Pt and PZT layers, were measured from microscale tension specimens. Since PZT thin films for MEMS are not fabricated in freestanding form, thin film stacks were fabricated in combination with SiO2 and Pt. The specimens tested were stacks of SiO2-TiPt-PZT-Pt, SiO2-TiPt-PZT, and individual SiO2 and Pt thin films with gauge lengths of 1,000 μm and widths of 50-100 μm and thicknesses of 400-1,200 nm. Full-field strain measurements were conducted with the aid of a fine speckle pattern (1 μm particle size) generated on the samples and analyzed by digital image correlation. The PZT films demonstrated very high tensile strengths and non-linear responses at strains higher than 0.35%, which was due to domain switching. Furthermore, the d31 coefficient was calculated from the out-of-plane deflection of biased PZT specimens in conjunction with analytical solutions for the bending response of multilayer piezoelectric beams. The field induced in-plane stress hysteresis loops were asymmetric at small in-plane stresses becoming of similar magnitude as the stress increased beyond 300 MPa. Similarly, the intersection of hysteresis loops shifted from negative to positive electric field at stresses larger than 150 MPa. Finally, the applied stress resulted in reduction of the hysteresis magnitude due to mechanical constraints imposed on 90° domain switching. The aforementioned measurements on the hysteresis and stress controlled piezoelectric response of PZT films are critical for the design and reliability of active MEMS devices.
3:15 PM - S2.3
Nanomechanical Characterization of Thin, Compliant Coatings.
Michelle Oyen 1
1 Engineering Dept. , Cambridge University, Cambridge, 0, United Kingdom
Show AbstractExpansion of the functionality of advanced MEMS and NEMS devices requires the inclusion of a diverse materials set. In modern MEMS applications, particularly but not exclusively in the domain of biomedical applications, there is a need for coatings of polymeric materials and hydrogels with significant mechanical integrity. This is particularly challenging because the materials themselves are relatively compliant (and thus also typically exhibit time-dependent mechanical behavior) such that traditional mechanical techniques for material properties characterization are not available. In this study, polymer and hydrogel coatings are characterized using nanoindentation techniques. Material constitutive laws are considered as viscoelastic for polymers and poroelastic for hydrogels. Both analytical and finite element analyses are used to ascertain the effects of coating thickness on observed mechanical response and to deconvolute true material properties of the coatings. These techniques show great promise for MEMS development in the context of robust polymer-based functional layers.
3:30 PM - S2.4
Pt/TiO2 Growth Templates for PZT Films and Quality MEMS Devices.
Daniel Potrepka 1 , Glen Fox 2 , Luz Sanchez 1 3 , Ronald Polcawich 1
1 RDRL-SER-L, U.S. Army Research Laboratory, Adelphi, Maryland, United States, 2 , Fox Materials Consulting, LLC, Colorado Springs, Colorado, United States, 3 Materials Science, University of Maryland, College Park, Maryland, United States
Show AbstractThe crystallographic texture of PZT thin films strongly influences the piezoelectric properties used in MEMS applications. When PZT films are poled to saturation, the piezoelectric response increases sequentially on transforming from random orientation to {111} texture to {001} texture. Textured growth can be achieved by relying on crystal growth habit, but it can also be initiated by the use of a seed layer that provides a heteroepitaxial template. The choice of template and the process used to form it determines the structural quality and ultimately influences performance and reliability of MEMS PZT devices such as switches, filters, and actuators. This study focuses on how {111} textured PZT is generated by a combination of crystal habit and templating mechanisms that occur in the PZT/bottom electrode stack . The sequence begins with {002} oriented Ti deposited on PECVD or thermally grown SiO2 on a Si wafer. The Ti is then converted to {200} TiO2 with the rutile structure through thermal oxidation and then {111} Pt can be grown to act as a template for {111} PZT. The Ti and Pt are deposited by DC magnetron sputtering. Optimization of the TiO2 and Pt film textures and structure were studied by variation of sputtering deposition times, temperatures and power levels, and anneal conditions. The relationship between Ti, TiO2, and Pt texture and their impact on PZT growth will be presented.
3:45 PM - S2.5
Contact Resistivity of Laser Annealed SiGe for MEMS Structural Layers Deposited at 210°C.
Joumana El Rifai 1 2 3 , Sherif Sedky 2 4 , Ahmed Abdel Aziz 2 , Robert Puers 1 3 , Chris Van Hoof 1 3 , Ann Witvrouw 1
1 , IMEC, Leuven Belgium, 2 Youssef Jameel Science and Technology Research Center, The American University in Cairo, Cairo Egypt, 3 , Katholieke Universiteit Leuven, Leuven Belgium, 4 Physics Department, The American University in Cairo, Cairo Egypt
Show AbstractThis work examines, for the first time, the contact resistivity between a laser annealed SiGe MEMS structural layer and a bottom TiN electrode. By using laser induced crystallization a TiN-SiGe contact resistivity as low as 2.14×10-3 Ωcm2 was obtained for SiGe deposited at 210°C.Lowering the SiGe deposition temperature will enable a transition from rigid Si substrates to flexible polymer or other temperature sensitive substrates, thus increasing the integration flexibility of SiGe-based MEMS [1]. However, depositing SiGe at 210°C by plasma enhanced chemical vapor deposition will yield amorphous films with high stress, stress gradient and resistivity. A post-deposition laser annealing treatment has already been used to produce SiGe films with improved stress gradient and low resistivity suitable for MEMS applications [2,3]. In this paper we investigate an additional property of the laser annealed SiGe, namely its contact resistivity to 100 nm thick TiN layers. As previously investigated, increasing the laser energy leads to a decrease in electrical resistivity and an increase in crystallization depth. The same is transferable to contact resistivity, as shown in this work. The laser annealing with a 248 nm KrF excimer laser of the SiGe layer was carried out both in air and vacuum. Laser energy densities up to 100 and 240 mJ/cm2 in air and vacuum, respectively, were applied. Using higher laser energy densities leads to damage in the Si-oxide dielectric layer situated in between the SiGe and TiN layers and eventually the SiGe layer itself.It was found that an energy density of 100 mJ/cm2 applied to a 1.0 µm thick SiGe layer is sufficient to produce a contact resistivity of 3.02×10-1 Ωcm2 for a 263 µm2 contact area if the laser annealing is conducted in an air ambient. Changing the annealing environment to a vacuum ambient will produce layers with a reduced surface roughness and allows the usage of higher laser energies [4]. Whereas the previous sample can only withstand a maximum of 100 mJ/cm2 in air before damage, it can be subjected to a 240 mJ/cm2 laser pulse in vacuum and produce a contact resistivity of 1.88×10-1 Ωcm2. Using a soft sputter etch and a Ti-TiN layer between SiGe and the TiN interface leads to a further reduction in the contact resistivity. SiGe layers with a 0.7 µm thickness and a Ti-TiN interlayer had a contact resistivity as low as 2.14×10-3 Ωcm2 for a 3 µm2 contact area after laser annealing in vacuum with a single pulse of 200 mJ/cm2. This beneficial effect of the Ti/TiN interlayer on contact resistivity has also been seen for poly-SiGe layers deposited at 450°C [5].References[1]A. Witvrouw, Mater. Res. Soc. Symp. Proc., vol. 1075, 2008.[2]S. Sedky et al., JMEMS, vol. 16, no. 3, pp. 581-588, 2007.[3]J. El-Rifai et al., Mater. Res. Soc. Symp. Proc., vol. 1153, 2009.[4]A. T. Voustas et al., J. Electrochem. Soc., vol. 146, no. 9, pp. 3500-3505.[5]G. Claes et al., Mater. Res. Soc. Symp. Proc., vol. 1222, 2010.
S3: Strength Characterization
Session Chairs
Monday PM, November 29, 2010
Room 207 (Hynes)
4:30 PM - **S3.1
Making Silicon Stronger.
Brad Boyce 1
1 , Sandia National Laboratories, Albuquerque, New Mexico, United States
Show AbstractSilicon microfabrication has seen many decades of development, yet the structural reliability of microelectromechanical systems (MEMS) is far from optimized. The fracture strength of Si MEMS is limited by a combination of poor toughness and nanoscale etch-induced defects. A MEMS-based microtensile technique has been used to characterize the fracture strength distributions of both standard and custom microfabrication processes. Recent improvements permit 1000’s of test replicates, revealing subtle but important deviations from the commonly assumed 2-parameter Weibull statistical model. Subsequent failure analysis through a combination of microscopy and numerical simulation reveals salient aspects of nanoscale flaw control. Grain boundaries, for example, suffer from preferential attack during etch-release thereby forming failure-critical grain-boundary grooves. We will discuss ongoing efforts to quantify the various factors that affect the strength of polycrystalline silicon, and how weakest-link theory can be used to make worst-case estimates for design. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000.
5:00 PM - S3.2
Mechanical and Electromechanical On-chip Testing of Mono- and Poly-crystalline Silicon Nanobeams.
Jean-Pierre Raskin 2 3 , Vikram Passi 2 , Umesh Bhaskar 2 , Azeem Zulfiqar 1 2 , Thomas Pardoen 1
2 ICTEAM, université catholique de Louvain, Louvain-la-Neuve Belgium, 3 CERMIN, Université catholique de Louvain, Louvain-la-Neuve Belgium, 1 IMMC, Université catholique de Louvain, Louvain-la-Neuve Belgium
Show AbstractThe application of well defined levels of mechanical stress in Si and PolySi is essential to characterize and make use of electromechanical couplings such as piezo-resistivity effects.One of the main limitations for stress driven applications of Si and PolySi, such as for any MEMS, is the inherent defect sensitivity leading to the statistical brittle failure behavior. A versatile on-chip suite of nanomechanical testing units has been developed in order to combine mechanical and electromechanical testing on the same specimen. The concept of actuation is based on the use of internal stress present in one material (tensile LPCVD silicon nitride, named the actuator) to load another specimen, in the present case Si or PolySi beams. Several thousands of tests structures with various lengths and widths, involving electrical connections, are produced on a single wafer, delivering statistically representative fracture data.Here we report tensile tests results on Si nanobeams of various lengths with thickness ranging between 40 to 200 nm and width ranging between 50 and 500 nm patterned by e-beam lithography. The smallest specimens exhibit elastic strains larger than 5%, corresponding to fracture stress on the order of 9 GPa, i.e. about 40% of the theoretical fracture stress of perfect Si. Simple and reliable statistical failure analysis has been demonstrated on both Si and PolySi films. In addition, the surface roughness of released mono-crystalline Si nanobeams has been scanned with AFM for various applied stress values, changing the length of the actuator. The experimental results show a decrease by a factor of 1.5 of the Si surface roughness moving from an unstrained Si nanobeam (0.2 nm) towards a Si nanobeam stressed at around 1 GPa (0.13 nm).Giant piezoresistance effects have been demonstrated in Si nanowires recently under limited applied stress (100 MPa) using a 4-point bending setup. Adding electrical contacts to the on-chip suite of nanostructures used for nanomechanical analyses, the current-voltage characteristics as a function of induced stress (different actuator lengths) can be measured up to values of several GPa. Transient electrical effects due to surface charges and saturation of piezoresistance coefficients in Si nanowires at high level of stress are demonstrated.
5:15 PM - S3.3
Electron Back Scatter Diffraction and Confocal Raman Microscopy in situ Studies of Stress Applied to Polysilicon Grains.
Siddharth Hazra 1 , Ryan Koseski 2 , Frank DelRio 2 , Jack Beuth 1 , Mark Vaudin 2 , Robert Cook 2 , Maarten de Boer 1
1 Mechanical Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania, United States, 2 Ceramics Division, National Institute of Standards and Technology, Gaithersburgh, Maryland, United States
Show AbstractWe have recently developed an in situ, on-chip high throughput microelectromechanical tensile tester for measuring fracture strength of polycrystalline silicon (polysilicon) specimens. This device uses a thermal actuator to mechanically grip and generate tens of milliNewtons of force to fracture polysilicon tensile bars of 70 µm and 2.25×2 µm2 nominal length and cross-sectional area respectively. Stresses resulting from sample-actuator misalignment are negligible and fracture occurs under uniaxial tension. These features, along with its small size, render the device suitable for in situ, small working distance, high-resolution microscopy techniques. Because the measured strength is averaged over a flaw population, the data does not provide direct insight into the mechanics of brittle fracture. It has been reported that fracture in polycrystalline silicon initiates at the side-wall grain boundary flaws. However obtaining quantitative assessments of the stresses acting within such small regions has proven elusive. In this work, we explore confocal Raman microscopy (CRM) and electron back scatter diffraction (EBSD) for in situ local stress mapping on this specimen. CRM provides lower spatial resolution (~100-200 nm), but has the potential for three-dimensional stress characterization. Preliminary Raman results indicate clear stress contours as the fillet necks down to the gage section. EBSD provides greater spatial resolution (~10-20 nm) but is more surface sensitive, and hence detail two-dimensional data can be collected. Preliminary results indicate that individual grains of less than 500 nm diameter are well resolved. Therefore, this tensile tester is amenable to in situ stress studies using these techniques. In this talk we will discuss the feasibility of these techniques in mapping stress within single grains of polysilicon as a function of in situ applied stress.
5:30 PM - S3.4
Characterizing the Effect of Uniaxial Strain on the Surface Roughness of Si Nanowire MEMS-based Microstructures.
Enrique Escobedo-Cousin 1 , Sarah Olsen 1 , Thomas Pardoen 2 , Umesh Bhaskar 2 , Jean-Pierre Raskin 2
1 , Newcastle University, Newcastle upon Tyne United Kingdom, 2 , Universite catholique de Louvain, Louvain-la-Neuve Belgium
Show AbstractThis work uses an original MEMS concept to strain released silicon beams in order to analyze the relationship between on-chip applied strain and nanoscale surface roughness in Si. Roughness affects carrier mobility through surface roughness scattering at high electric fields. The rms surface roughness and correlation length are key parameters to model carrier mobility in MOSFET inversion layers. Simulations indicate that only a reduction in rms roughness compared with bulk Si values can explain the high values of electron mobility observed experimentally in tensile strained silicon devices. However due to the limited characterization techniques available to measure roughness accurately on a nano and sub-nanoscale, to date such assertions have remained largely unconfirmed. An initial observation of strain-induced reduction in surface roughness was recently reported for biaxially strained Si compared with unstrained Si but only one level of strain was investigated [O. Bonno, S. Barraud, D. Mariolle and F. Andrieu, J. Appl. Phys. 103, 63715 (2008)]. Furthermore uniaxial strain was not studied despite offering a wider range of device options and performance benefits compared with biaxial strain. For device modellers investigating transport in uniaxial strain either bulk Si roughness parameters are assumed or limited biaxial data is used. This work addresses the paucity of roughness measurements by reporting on roughness parameters in uniaxial strained Si beams relevant for state of the art MOSFETs, nanowire and MEMS devices, with varying degrees of strain. Roughness is characterized using ultra-high resolution AFM techniques while strain is characterized by Raman spectroscopy. Microstructures comprising a silicon nitride actuator are used to induce a wide range of stress levels in Si beams. The microstructures also allow the comparison of surface evolution in the strain direction (along the Si beam) compared with the unstrained direction (across the Si beam). A gradual reduction in rms roughness amplitude and increase in roughness correlation length in the direction of the applied stress are found for increasing values of strain. In contrast, surface roughness in the direction perpendicular to the applied stress remained largely unchanged from the unstrained initial state. This is the first time that roughness in uniaxially strained structures has been studied and provides unequivocal confirmation that a reduction in rms roughness accompanies increasing tensile strain. Moreover the results suggest that the reduction in rms roughness often assumed by device modellers to generate agreement between measured and modelled data may be overestimated; a proportion of the enhanced carrier mobility observed is likely to instead originate from the increased correlation length. This correlation length parameter has previously been assumed to remain unchanged relative to bulk Si values.
5:45 PM - S3.5
Effects of Plasma-etched Silicon Surfaces on the Strength of Theta-like Micromechanical Test Structures.
Michael Gaither 1 , Frank DelRio 1 , Richard Gates 1 , Robert Cook 1
1 Ceramics Division, National Institute of Standards and Technology, Gaithersburg, Maryland, United States
Show AbstractIn order for the microelectromechanical systems (MEMS) industry to continue to grow and advance, it is critical that methods are developed to determine the mechanical reliability of MEMS devices. Such methods are a crucial element in the “feedback” cycle that optimizes materials and processing selections for device reliability. This is particularly so for advanced devices with contacting, moving components, for which component strength is a key factor in determining reliability. The etching processes used to produce MEMS devices leave residual surface features that typically limit device strength and, consequently, device lifetime and reliability. In order to optimize MEMS device reliability, it is therefore necessary to understand and characterize the effects these etching processes have on MEMS-scale device strengths. At the micro and nano scales, however, conventional strength testing methods cannot be used, and a standardized test method for MEMS-scale strength measurement has yet to be established. The micro-scale NIST theta specimen, shaped like the Greek-letter theta, acts as a tensile test specimen when loaded in compression by generating a uniform tensile stress in the central web of the specimen. Utilizing the theta specimen for strength measurements allows for simple micro-scale strength testing and assessment of etching effects, while removing the difficulties associated with gripping and loading specimens as well as minimizing potential misalignment effects. Three sets of MEMS-scale single crystal silicon theta specimens are fabricated using two deep reactive ion etching (DRIE) recipes and a temperature-controlled cryogenic plasma etching recipe, each set resulting in a different specimen surface quality. DRIE is a silicon etch technique that results in high-aspect-ratio sidewalls and nearly uniform etch steps, or scallops, along these sidewalls. The cryogenic plasma etching technique results in smooth, high-aspect-ratio sidewalls. Each sample is tested by instrumented indentation and finite element analysis (FEA) is used to determine sample strength. Equations developed via FEA translate load-displacement response at the load-point into stress-strain behavior across the theta web region. Strength values for each set of specimens are examined via Weibull statistics. The resulting surface roughness for each etching recipe is determined by atomic force microscopy and sample fragments are examined via field-emission scanning electron microscopy. Surface roughness topography and fracture origins located during fractographic analysis of tested samples are compared with strength-limiting flaw size calculations.
S4: Poster Session: Devices and Fabrication
Session Chairs
Tuesday AM, November 30, 2010
Exhibition Hall D (Hynes)
9:00 PM - S4.1
Hydrothermal Potassium Sodium Niobate Lead-free Piezoelectric Ceramics.
Takafumi Maeda 1 , Norihito Takiguchi 1 , Peter Bornmann 2 , Tobias Hemsel 2 , Takeshi Morita 1
1 Graduate School of Frontier sciences, The University of Tokyo, Kashiwa, Chiba, Japan, 2 Mechatronics and Dynamics, University of Paderborn, Paderborn, North Rhine-Westphalia, Germany
Show AbstractAs a lead-free piezoelectric ceramics, (K,Na)NbO3 is a promising material due to its good piezoelectric properties and high Curie temperature. Usually, to obtain the source powders for these piezoelectric ceramics, the solid-solution method is carried out. However the potassium carbonate K2CO3, which is a potassium source for potassium niobate, is unstable and quite difficult to weigh due to its deliquescence. Moreover, to suppress conductivity, a stoichiometric between potassium and niobium in the ceramic have to be strictly controlled. However, potassium is easily evaporated in the calcination process. Therefore, an additional potassium source must be added to compensate evaporation. To obtain the source powders for these ceramics, we proposed to use the hydrothermal method. Hydrothermal reaction enables to produce high quality powder for ceramics fabrication process.In this study, (K,Na)NbO3 ceramics were sintered from mixed KNbO3 and NaNbO3 powders prepared by the hydrothermal reaction. To obtain the KNbO3 powder, 9.18 g of niobate oxide was put into 140 ml 8.8N KOH solution in a 300ml of pressure vessel, and it was kept at 210deg.C, for 24 hours. For NaNbO3 powders, 37.20g of niobate oxide, 70ml 9N NaOH were put into 125ml of pressure vessel. Other reaction conditions were the same as for KNbO3 powder synthesis. X-ray diffraction pattern of these powders indicated orthorhombic perovskite structure and no secondary impurity. These two powders were mixed with ethanol using a ball milling process, and KNbO3/NaNbO3 molar ratio was controlled to be 0.48/0.52. From this powder, a disk-shaped pellet was obtained using a cold isostatic pressing (CIP) at 200 MPa. After sintering at 1100 deg.C, it was confirmed that the solid solution of this (K0.48Na0.52)NbO3 ceramics was synthesized by XRD measurement. Poling treatments were carried out using a high voltage supply at 2.0 kV/mm in 150 deg.C silicone oil for 1 hour. The measured piezoelectric properties of the (K0.48Na0.52)NbO3 ceramics were as follows: the electromechanical coupling factors k31, k33, the relative free permittivity εT33/ε0, loss tangent δ, the piezoelectric factor d31, d33 and the mechanical quality factor Qm(radial /thickness), were 0.41, 0.53, 2.3%, 474, -77pC/N, 130pC/N, 67(radial mode) and 51(thickness mode) respectively.
9:00 PM - S4.10
Fabrication of an Ultrathin Microfluidic Membrane Bilayer Tissue Engineering Construct.
Alla Epshteyn 1 2 , Jeff Borenstein 1 , John Adams 2 , Steven Maher 2 1 , Angela Holton 1 , Shekhar Bhansali 2 , Amy Taylor 1 , Joseph Cuiffi 1
1 , Draper Laboratory, Tampa, Florida, United States, 2 , University of South Florida, Tampa, Florida, United States
Show AbstractCharacterization of cellular responses using in vitro models is key to solving healthcare problems such as drug discovery and cancer therapeutics. Depending on the organ and the cell type, each cell grows, metabolizes and differentiates in response to its surroundings. The complexity of tissue structure poses a real challenge in creating physiologically relevant in vitro models which closely mimic in vivo cell microenvironments. Microfluidic platforms enhance the ability to control chemical (e.g. nutrient and media delivery), biological (e.g. cell-cell contact and spatial organization), and biophysical factors (e.g. flow-induced shear stress) to create a more mimetic model over standard tissue culture techniques. This paper presents the design and fabrication of a novel ultrathin membrane bilayer cell culture device, specifically tailored as a liver research model for malaria research. This microfluidic platform allows for creation of a cellular bilayer construct while enabling high resolution live-cell imaging on either side of the membrane. Through a unique combination of material choices, surface chemistry modifications and fabrication processes, this high throughput amenable design provides a dynamic research model with high resolution imaging capabilities of live in vitro cell activity. With the unique development of a liver sinusoid construct within the device, our goal is to observe and capture high resolution images of malaria liver invasion.
9:00 PM - S4.11
SU8/ Na-AHA modified MWNT Composite for Piezoresistive Sensor Application.
Prasenjit Ray 1 , V. Seena 1 , Rupesh Khare 2 , Arup Bhattacharyya 2 , Prakash Apte 1 , Ramgopal Rao 1
1 Centre for excellence in Nanoelectronics, Department of Electrical Engineering, Indian Institute of Technology Bombay, Mumbai, Maharashtra, India, 2 Department of Metallurgical Engineering & Materials Science, Indian Institute of Technology Bombay, Mumbai, Maharashtra, India
Show AbstractThere is a huge demand for compliant structural materials such as SU-8 for MEMS applications due to its interesting properties such as lower Young’s modulus, mechanical, thermal stability etc. One of the most interesting and a popular class of MEMS devices is a piezoresitive mcirocantilever which requires a compliant piezoresistive material in order to achieve a very good sensitivity. SU-8 microcantilevers with gold and polysilicon have been reported in the literature with their disadvantages being the lower sensitivity due to the lower gauge factor and the higher young’s modulus with these materials. Ultra-sensitive polymer composite cantilevers made up of SU-8 as a structural layer and Carbon Black as a piezoresistive layer with lower young’s modulus and higher gauge factor have been reported recently by our group. Higher conductivity at lower concentrations of conductive filler is of increased interest. Here we report a novel composite with SU-8 and multi walled carbon nanotube (MWCNT) as a piezoresistor. Purified multiwall carbon nanotubes (MWNT) (NC3100, Nanocyl CA, Belgium, L/D: 100–1000, purity > 95%) were modified with Na-salt of 6-amino hexanoic acid (Na-AHA) in order to achieve debundlled MWNT. MWNT were initially sonicated in distilled water for 20 minutes. Then the required amount of Na- AHA solution was added to the MWNT and again sonicated for 10 minutes. The Na-AHA modified MWNT solution was then subjected to evaporation and the obtained dry powder was left in a vacuum oven at 80oC for 3 h to ensure the complete removal of water. SU-8/CNT composite was spin coated and photolithographically patterned on gold electrodes. The conductivity and temperature dependent piezoresistivity studies of the composite thin film with different concentrations (weight percentage ranging from 0.005 % to 1 %) were carried out. The elastic constant being an important design parameter for MEMS applications, a detailed thin film mechanical characterization using nanoindentation is also conducted for varying concentrations. Finally a polymer mcirocantilever device with integrated SU-8/MWCNT composite has been demonstrated and characterized.
9:00 PM - S4.12
High Yield Polymer MEMS Process for CMOS/MEMS Integration.
Prasenjit Ray 1 , V. Seena 1 , Prakash Apte 1 , Ramgopal Rao 1
1 Centre for Excellence in Nanoelectronics, Department of Electrical Engineering, Indian Institute of Technology Bombay, Mumbai, Maharashtra, India
Show AbstractConventional MEMS devices such as micro-cantilevers and micro-accelerometers are silicon based. However, because of their disadvantages like lower sensitivity, high thermal budget processes and cost of production, MEMS community is trying out better structural materials such as SU-8 that can overcome these disadvantages. The reported suspended structures in SU-8 were fabricated either by using a sacrificial layer etching, electron beam lithography or flip-chip release technique where the entire device chip is released from the dummy substrate. Out of these the first two techniques suffer from stiction and low process yield respectively, while the third method does not support the development of MEMS chip integrated with CMOS. So the development of a high yield low temperature process for suspended structures such as micro-cantilevers, micro-accelerometers fabricated alongside a CMOS wafer would be very promising. Here we report an optical lithography based process for SU-8 based microcantilevers supporting on-chip CMOS integration. The process involves successive steps such as spin coating, softbake, UV exposure and post exposure bake for different layers of SU-8 with a final development step that yields a suspended structure with the anchor attached to the silicon wafer. The process temperature being very low (< 90oC), these SU-8 devices can be fabricated as a post-processing step on an already processed CMOS wafer.
9:00 PM - S4.13
Mechanical and Material Characterization of Bilayer Microcantilever-based IR detectors.
I-Kuan Lin 1 , Ping Du 1 , Yanhang Zhang 1 , Xin Zhang 1
1 Mechanical Engineering, Boston University, Boston, Massachusetts, United States
Show AbstractBilayer microcnatilever-based infrared radiation (IR) detectors have received extensive attention for wide use in military and civilian applications. These detectors can achieve a theoretical noise-equivalent temperature difference (NETD) of below 5 mk. This type of IR detector is based on the bending of bilayer structures upon absorption of IR. The subsequent deformation can be readily determined by using piezoresistive, optical, or capacitive methods. However, the bilayer structures curve significantly after release from a sacrificial layer, largely due to the mismatch of residual stress/strain in the two materials. Therefore, curvature modification is one of the important topics in the post-process assessment of IR detetcors. It is also important to understand the deformation of IR detectors over a significant period of operation time, in order to meet performance and reliability requirements. The inelastic strain behavior (creep) in metal layers results in inelastic deformation in IR detectors. Neglecting the inelastic deformation can lead to misinterpretations of the measurement data from IR detectors and can compromise performance. In this study, the temperature and time -dependent deformations of IR detectors are characterized by using a thermal cycling and isothermal holding testes. First, the thermal cycling technique is employed to flatten as-released IR detectors and characterize the linear thermoelastic behavior. Second, the characterized Power-law creep from is used to develop a numerical model for predicting and simulating the inelastic behavior in long-term operation. The experimental methodologies and theoretical framework developed in this research can be readily applied to study the thermomechanical behavior of various bilayer microcantilever structures, and to improve the fundamental understanding required to design microcantilever-based IR detectors.
9:00 PM - S4.16
Enhanced Piezoelectric Properties of Low Temperature-processed PZT Thick Films by Hybrid Deposition Technique with Chemical Solution Infiltration Process.
Seung-Hyun Kim 1 , Wenyan Jiang 1 , Chang Young Koo 2 , Angus Kingon 1
1 Division of Engineering, Brown University, Providence, Rhode Island, United States, 2 Research Center, INOSTEK Inc., Ansan, Gyeonggi, Korea (the Republic of)
Show AbstractThere is a strong interest in introducing lead zirconate titanate (PZT) thin films for applications in piezoelectric micro-electromechanical systems (MEMS) such as actuators, sensors, and high frequency transducers since they have large piezoelectric coefficients and electromechanical coupling coefficients. However, the stress induced in PZT thin films due to clamping effect by the substrates and other degradation parameters such as low breakdown strength, reduced extrinsic domain wall contribution and insufficient poling have limited these films to be used in commercial MEMS applications. To develop PZT films for MEMS devices, it is necessary to fabricate high quality PZT thick films over 10 µm which can cover the important commercial technological gap between the thin films and the bulk ceramics. However, the preparation of PZT thick films by conventional screen printing method is required high temperature annealing or sintering process above 900 °C. Such an extremely high temperature is unacceptable for Si-based MEMS devices due to severe chemical reaction between electrode and substrate materials, uncontrollable loss of PbO component, porous and rough microstructure, and deteriorated electrical properties. To solve these drawbacks, we introduced a simple process design for high quality PZT thick films using a chemical solution modified hybrid deposition technique, that is, the use of multiple infiltration process with the same composition of PZT solution into the porous screen-printed PZT thick films without any additional sintering aids. With this technique, we successfully lowered the annealing temperature of the PZT thick films to 700 °C without any degradation of piezoelectric performance of the films. To verify the effect of the solution infiltration process on the physical and the electrical properties of the PZT thick films, we compared the microstructure, the ferroelectric and the dielectric properties and piezoelectric coefficients of the hybrid films with those of conventional screen printed films.
9:00 PM - S4.17
The Performance and the Reliability of Micromachined PZT-based Vibration Energy Harvesting Devices.
Jung-Hyun Park 1 , Hosang Ahn 1 , Seon-Bae Kim 1 , Dan Liu 1 , Dong-Joo Kim 1
1 Materials Engineering, Auburn University, Auburn, Alabama, United States
Show AbstractWith higher integration, smaller size, and automated processes, sensors and wireless devices have seen dramatic enhancements to their quality, robustness, and reliability. Recent efforts have been made toward developing autonomous, self-powered remote sensor systems that can offer enhanced applicability and performance with cost savings. The technological challenge of realizing such a system lies in the construction and fabrication of a miniaturized power generator. This work focuses on the development of micromachined piezoelectric energy harvesting devices to achieve maximum efficiency of power conversion. The main factors in this work are focused on the optimally structured materials and device structure from the common design of a MEMS cantilever structure consisting of a Si seismic mass connected to a thin PZT cantilever beam. Factors relating to power improvement and reliability of the device were studied by addressing the effects of the cantilever geometry, electrode design, and other factors such as resonance frequency, output power and voltage, and optimal resistance load. Two types of cantilever geometry include rectangular and triangular shapes, and the triangular design results in more uniform strain in the cantilever beam corresponding to higher output power. In electrode design, transverse (d31) and longitudinal (d33) piezoelectric modes were compared. A longitudinal mode device can provide higher output voltage, but the output power strongly depends on the electrode dimensions. In addition to the output power, the factors related to the reliability such as the degree of poling, the operation temperature, and the number of vibration cycle were systematically investigated. The results can provide a direction for the highly efficient piezoelectric mode for MEMS piezoelectric energy harvesters and can further scientific understanding of the piezoelectric behavior of PZT films.
9:00 PM - S4.18
Parallelized Microfluidic Separations for Large-scale Dewatering of Biofuel Algae.
Kevin Loutherback 1 , Joseph D'Silva 1 , Jost Goettert 3 , Jeff Bargiel 2 , Christopher Lane 2 , Robert Austin 4 , James Sturm 1
1 Electrical Engineering, Princeton University, Princeton, New Jersey, United States, 3 , Lousiana State University, Baton Rouge, Louisiana, United States, 2 , Phycal, LLC, Highland Heights, Ohio, United States, 4 Physics, Princeton University, Princeton, New Jersey, United States
Show AbstractAlgae are a promising candidate for large-scale production of biofuels, an important source of renewable energy [1]. A significant fraction in the cost comes from concentrating (“dewatering”) the algae from the dilute concentrations at which they are cultured to levels necessary for oil extraction. This key step is classically performed by centrifugation, which is costly, energy intensive and must operate in a batch processing mode. In this abstract, we report experiments to replace this dewatering step with deterministic lateral displacement (DLD) arrays, a high resolution, continuous-flow microfluidic particle separation technology. The consists of a channel filled with densely packed vertical posts tilted with respect to the flow direction, so that algae with a diameter above a critical diameter flowing through the channel are displaced to one side of the channel and are thus concentrated [2]. The spacing between posts in these experiments was 10 microns, larger than the critical separation size of 4 microns, so these arrays can be run continuously without clogging, and the devices. The arrays concentrate a dilute algal culture of the species nanochlorella by 25 and 40 times. The concentration is independent of the applied pressure in the range of 1 to 10 psi. Thus, they can be used in the field by the pressure head from a 10 m water tank. Second, we report a packaging method to greatly increase throughput by running many arrays in parallel without increasing the number of external connections. By tiling devices in plane and stacking them vertically with through-device holes lined up, we are able to increase total throughput compared a single device without increasing the applied pressure gradient. This technique is demonstrated in both silicon and soft polymer, materials. In silicon, a gasket layer of soft polymer with punched through holes is necessary to seal individual devices and form connections between layers while soft polymer devices can be sealed to each other without a gasket layer. We expect to demonstrate flow rates on the order of liters/hour.References[1] Y Christi, Biotechnology Advances, 2007, 25, 294-306.[2] LR Huang, EC Cox, RH Austin, JC Sturm, Science, 2004, 304, 987-990.
9:00 PM - S4.19
Development of Ni-Al Superalloys for High Temperature LIGA MEMS Materials.
Devin Burns 1 , Michael Teutsch 2 , Klaus Bade 3 , Jarir Aktaa 2 , Kevin Hemker 1
1 Mechanical Engineering, Johns Hopkins University, Baltimore, Maryland, United States, 2 Institute for Material Research II (IMF II), Karlsruhe Institute of Technology, Karlsruhe Germany, 3 Institute for Microstructure Technology (IMT), Karlsruhe Institute of Technology, Karlsruhe Germany
Show AbstractNickel is one of the most electrodeposited structural LIGA MEMS materials (LIGA is a German acronym for lithography, electroplating and molding). The LIGA technique produces metallic structures with submicron resolution and high aspect ratios. However, LIGA nickel’s mechanical properties degrade when subjected to elevated temperatures. Bulk Ni-based superalloys have shown great success in land and aero based turbine applications. The manufacturing of Ni superalloys microcomponents could therefore increase their application range towards higher temperatures.In this work, we present two approaches for producing LIGA Ni-Al superalloys. The first technique involves the electro codeposition of nickel and aluminum nanoparticles. Results from electrodeposition studies with respect to bath stabilization and particle agglomeration will be discussed. The second approach modifies existing LIGA Ni specimens by vapor phase aluminization. To transform these composite Ni-Al materials into superalloys, different heat treatment cycles are investigated. The effect of heat treatment on the structural and mechanical properties of LIGA superalloys is studied by producing LIGA Ni-Al and pure nickel (for aluminization) microtensile specimens for mechanical testing.
9:00 PM - S4.2
PZT Thick Films for 100 MHz Ultrasonic Transducers Fabricated Using Chemical Solution Deposition Process.
Naoto Kochi 1 2 , Takashi Iijima 2 , Takashi Nakajima 1 , Soichiro Okamura 1
1 , Tokyo University of Science, Shinjuku-ku Japan, 2 , National Institute of Advanced Industrial Science and Technology, Tsukuba Japan
Show AbstractMicromachined ultrasonic transducers (MUT) are one candidate for MEMS device application and especially in demand for medical imaging field. The MUT operating at high frequencies increase the spatial resolution. To observe biological tissue images clearly, ultrasonic waves above 100 MHz are required. Furthermore, the MUT which resonate in the thickness oscillation mode are expected to generate high amplitude ultrasonic waves. This study shows the fabrication and characterization of transducers based on square pillar shaped 10-μm-thick Pb1.1(Zr0.53Ti0.47)O3 films onto a 2 inch Pt/Ti/Al2O3 substrate. The characteristics of the PZT thick films are compared with finite element method (FEM) simulations for a better understanding of the transducer behavior. The PZT thick films were fabricated using chemical solution deposition (CSD) process .The sequence of spin coating and pyrolysis was repeated three times, and then the precursor films were fired. This process was repeated with automatic coating and firing system to increase the film thickness up to 10 μm. Pt top electrode and PZT layer were etched by reactive ion etching (RIE) process and PZT thick film structures were successfully fabricated. The size of the top electrode was varied from 30 X 30 μm2 to 1000 X 1000 μm2. The fabricated PZT thick films exhibited well-saturated P-E hysteresis curves and butterfly-shaped longitudinal displacement curves. The ultrasonic frequency of the PZT thick films generated with a single pulsar-receiver was more than 100 MHz and some reflected ultrasonic waves from the bottom face of the substrate were observed. To clarify the mode of oscillations, electrical impedance properties of the PZT thick films were measured by an impedance analyzer as a function of the electrode length. Lateral oscillations that change by the differences of the electrode length were observed below 70 MHz. Moreover, the PZT thick films showed a number of spurious peaks ranging from 30 MHz to 300 MHz, and thickness oscillation modes were not observed clearly. The FEM simulations agreed with the tendency of the experimental data. These results indicate that the substrate clamping affects the behavior of the spurious peaks. Therefore the optimization of the sample structure that is free from the clamping effects is required to observe the thickness oscillation mode. On the basis of these results, a new structure of PZT thick films onto a silicon substrate was designed in the FEM simulation. The backside of the silicon substrate was etched to eliminate the clamping effects and suppress the reflected ultrasonic waves from the bottom face of the substrate. Consequently, the spurious peaks were reduced and the resonant frequency of the thickness oscillation was observed clearly between 150 MHz and 180 MHz. This FEM simulated structure holds promise for the MUT operating in the thickness oscillation mode above 100 MHz.
9:00 PM - S4.21
Silicon Bonding Using SU-8 and Polyimide.
Aleksander Jonca 1 , Bradley Kaanta 1 , Xin Zhang 1
1 Mechanical Engineering, Boston University College of Engineering, Boston, Massachusetts, United States
Show AbstractWe present two silicon-to-silicon bonding methods using a polymer intermediate layer to create gas tight fluidic channels suitable for high temperature operation. These bonds were created over patterned surfaces as polymer bonding is much more forgiving of surface roughness than fusion or eutectic bonding. Benzocylocbutene (BCB) has often been used for polymer bonding and packaging but has the drawbacks of being very expensive, possessing a short shelf life, and is often used for bonding in a gas-controlled environment. The goal of this work was to use SU-8 and Polyimide as a bonding layer leaving no polymer in the fluidic channels. All bonding was performed in an FC-150 flip chip bonder while exposed to normal atmosphere.To create fluidic channels using SU-8 as the adhesive layer, the photoresist was applied to only one side using contact transfer from a dummy surface. The SU-8 was spun onto the dummy surface at 2μm and then applied via contact transfer to avoid having any photoresist enter the channel. The SU-8 was unexposed to UV light prior to bonding in the FC-150 at a temperature of 90 degrees Celsius. As SU-8 is a negative resist, exposure to light before bonding causes polymer cross linkages, weakening the final bond. Due to this fact, SU-8 is difficult to use as a photo-patternable bonding layer. Finally, the SU-8 was cured at a temperature of 285 degrees Celsius to be able to withstand high temperatures. Following the curing, the device was left to cool slowly before applying nitrogen gas once it reached 150 degrees Celsius.Polyimide was also found to be an effective polymer where a patternable bonding layer is required. The Polyimide was applied to a single side at a thickness of 3.4μm and patterned before bonding. In this case the Polyimide is removed from the center of the channel. The bonding was performed in the FC-150 first at a temperature of 90 degrees, followed by a ramp to 350 degrees Celsius. It was immediately cooled by nitrogen gas following the bond.Both of these bonding methods have advantages over other polymer silicon-to-silicon bonding methods due to SU-8 and Polyimide's relative chemical inertness and the ability to withstand high temperatures. Additionally, these techniques do not require a very smooth surface and can be performed over small raised features, which is difficult with some polymer bonding and not feasible with fusion or eutectic bonding.The bonded seals of the channels were tested by pressurizing the channels with helium and submerging them in a water bath for two minutes without the detection of any air bubbles.References:-Pan, C. T., P. J. Chen, M. F. Chen, and C. K. Yen. "Intermediate Wafer Level Bonding and Interface Behavior." Microelectronics Reliability 45 (2005): 657-63-Niklaus, F., G. Stemme, J. Q. Lu, and R. J. Gutmann. "Adhesive Wafer Bonding." Journal of Applied Physics - Applied Physics Reviews 99.031101 (2006).
9:00 PM - S4.22
A Miniatured Enzymatic Biofuel Cell Based on C-MEMS via Multipoint Covalent Immobilization of Enzyme.
Yin Song 1 , Chunlei Wang 1
1 MME, FIU, Miami, Florida, United States
Show AbstractEnzymatic biofuel cells involving oxidizing biological fuels by enzyme-modified electrodes attracted considerable attention. However, the enzyme stability is always barrier for the practical application of enzymatic biofuel cells miniatured power supply for the portable electronics. In this study, we report a micro biofuel cell based on carbon-microelectromechanical systems (C-MEMS) techniques. A novel design is proposed using multipoint covalent attachment to immobilize enzyme, which can improve the stability and activity of glucose oxidase (GOx) enzyme and increase output potential of enzymatic cells. The experimental results show that the covalent bondings are formed in virtue of the functionlization reaction between the terminal amino groups modified on C-MEMS surface and carboxylic groups of enzyme. This intense multipoint covalence will facilitate the immobilization of enzyme on carbon electrode. It has been found that GOx immobilized onto functionalized carbon surface can enhance its catalytic ability and promote direct electron transfer. In addition, the anodic electrical properties also increase as a result of enzyme immobilization method.
9:00 PM - S4.3
Synthesis and Control of ZnS Nanodots and Nanorods with Different Crystalline Structure from an Identical Raw Material Solution and the Excitonic UV Emission.
Masato Uehara 1 , Satoshi Sasaki 2 , Yusuke Nakamura 2 , Chan-gi Lee 1 , Hiroyuki Nakamura 1 , Hideaki Maeda 1 2 3
1 Measurement Solution Research Center, National Institute of Advanced Industrial Science and Technology (AIST), Tosu, Saga Japan, 2 Interdisciplinary Graduate School of Engineering Sciences, Kyusyu University, Kasuga, Fukuoka Japan, 3 CREST, Japan Science and Technology Agency, Kawaguchi, Saitama Japan
Show Abstract Nanocrystals (NCs) have received much attention due to many attractive properties. The properties of NCs are strongly influenced by the size, shape and crystalline structure. A high temperature thermal synthesis in organic solvent has some advantages for structure control of uniformed NCs. The surfactant molecules play key roles in the synthesis and we can control the size and shape of NC by using adequate surfactants. There are many papers about shape control of NCs via using surfactants, but the type and concentration of surfactants have been selected and optimized in order to obtain the one target shape in most of papers. From the point of industrial view, it would be favorable that the shape of uniformed NCs can be variously controlled from an identical raw material. Christian et al suggested an influence of kinetic condition in the synthesis on the crystalline phase and shape of NCs. We thought that their suggestion implied the possibility of uniformed NC production with the different phase and shape from one identical raw material. In the present work, we tried to control the morphology and crystalline phase of ZnS NC in a thermal synthesis by simple changing of the heating rate. The heating rate was controlled using a micro fluidic system for the rapid heating, in addition to a common heating system which consisted with oil bath and glass vessels. We could successfully synthesis uniformed ZnS NCs and control the shape and crystalline phase from just one identical raw material solution. In the rapid heating, we obtained the uniformed zincblende nanodots, and in the slower rate, the wurtzite nanorods were formed. The length of nanorods was longer as slower heating rate. We concluded that the change in the morphology and phase would be related with the temperature dependence of the surfactants properties, according to the analysis of temporal evolution of NCs by HR-TEM and XRD. Moreover, all products exhibited excitonic emission of UV. This photoluminescence emission peak was narrow without tailing. Since the papers on the excitonic emission without trap-state emission form ZnS NCs are few, the present synthesis of ZnS NCs which exhibit the excitonic emission would be worthwhile.
9:00 PM - S4.4
Mechanism of Hole Inlet Closure in Shape Transformation of Hole Arrays on Si(001).
Reiko Hiruta 1 , Hitoshi Kuribayashi 2 , Ryosuke Shimizu 3 , Koichi Sudoh 4
1 , Fuji Electric Holdings Co. Ltd., Matsumoto Japan, 2 , Fuji Electric Systems Co., Ltd., Matsumoto Japan, 3 , Japan Science and Technology Agency, Tokyo Japan, 4 , The Institute of Scientific and Industrial Research, Osaka University, Osaka Japan
Show AbstractRecently, shape transformation of microstructures fabricated on Si substrates by high temperature hydrogen annealing has been proven to be useful for fabrication processes of three-dimensional structures. When the Si microstructures are annealed in oxygen free ambient, such as in hydrogen gas ambient and in ultrahigh vacuum at high temperatures, they spontaneously change in shape by surface self-diffusion. One of the significant applications of the shape transformation is formation of silicon-on-nothing (SON) structures, in which a vacant space is formed under a thin Si layer, by annealing of an array of deep cylindrical holes [1,2]. For precise control of the spontaneous shape transformation by annealing, detailed understanding of the process is required. In this study we investigate the process of the hole inlet closure, which occurs in the initial stages of the SON structure formation.A periodic square array of holes with the diameter 1.6µm, spacing 1.0µm, and depth 6µm, was fabricated on n-type CZ-Si (100) substrates by anisotropic reactive ion etching (RIE) with a SiO2 mask. The substrates were annealed in 10~60 Torr hydrogen gas ambient at 1150 degree. The structures of the samples were evaluated by scanning electron microscopy (SEM), and atomic force microscopy (AFM). Cross sectional SEM images reveal that the hole inlet is closed by bulging of the surface around the inlet. Top view SEM observations show that the inlet opening shrinks while keeping the circular shape during inlet closure process. Around the inlet openings, we observe complicated morphologies reflecting the four-fold symmetry of the Si(001) surface. The structures are mainly composed of {001}, {111}, and {113} facets with corrugated regions around the holes. AFM observations reveal that the corrugated region is composed of three kinds of microfacets, namely, one {110} and two {113} microfacets. The spontaneous hole inlet closure during annealing is basically understood in terms of morphological evolution by surface diffusion driven by the gradient of chemical potentials along the surface. The observed complicated morphologies are considered to be the structure adapting the anisotropic free energy of the Si surface to the curved geometry during the hole inlet closure caused by surface diffusion.[1] T. Sato et al., Jpn. J. Appl. Phys. 43, 12 (2004).[2] R. Hiruta et al., ICSFS 2008 Proceeding (2008) [3] K. Sudoh et al., J. Appl. Phys. 105, 083536 (2009).
9:00 PM - S4.5
Aspect Ratio Dependence in Evolution of Hole Arrays by Surface Diffusion on Si(001).
Koichi Sudoh 1 , Reiko Hiruta 2 , Hitoshi Kuribayashi 3 , Ryosuke Shimizu 4
1 The Institute of Scientific and Industrial Research, Osaka University, Osaka Japan, 2 , Fuji Electric Holdings Co., Ltd., Nagano Japan, 3 , Fuji Electric Systems Co., Ltd., Nagano Japan, 4 , Japan Science and Technology Agency, Tokyo Japan
Show AbstractRecently, void structure formation by high temperature annealing of hole array patterns on Si substrates has attracted attention as a novel microfabrication technique. Such void formation occurs due to the singular shape transformation of hole arrays by surface self-diffusion [1,2]. In order to obtain desired void structures by this method, proper designing of the initial hole pattern is crucial. One of the important parameters governing the shape transformation is the aspect ratio of the hole. In this work, we study the aspect-ratio dependence of the shape transformation of hole arrays on Si(001) substrates during high temperature annealing. Periodic square arrays of holes with various diameters, depths, and spacings were fabricated on n-type CZ-Si (001) substrates by anisotropic reactive ion etching (RIE) with SiO2 masks. The substrates were annealed in 10~60 Torr hydrogen gas ambient at 1423 K using a ramp furnace. The structures of the samples were observed using scanning electron microscopy (SEM). For samples with aspect ratios of 3.0~7.0, at initial stages of the shape transformation vertically long voids are formed in the bulk Si by hole inlet closure, and subsequently shape change of the individual voids occurs. If the spacing between holes is sufficiently small, a large plate-shaped void is formed by coalescence of the neighboring voids. When the aspect ratio increases up to around 8.0, pinch-off of the vertically long voids occurs during shape change, resulting in formation of two layer structures of voids. We have found that for sufficiently small hole-hole spacing, two-layer plate-shaped voids are formed by coalescence of the neighboring voids. We discuss the aspect ratio dependence of the morphological evolution of hole arrays by surface diffusion, performing numerical simulations using Mullins’ equation [3] assuming an isotropic surface model. The numerical simulation shows that pinch-off of voids occurs for holes with aspect ratios larger than ~ 8.0 in agreement with the experimental results. Based on the results of the numerical simulations, a “phase diagram” showing relationship between the obtained void structure and the structure of the initial hole array is presented. [1] I. Mizushima, T. Sato, S. Taniguchi, and Y. Tsunashima, Appl. Phys. Lett. 77, 3290 (2000).[2] K. Sudoh, H. Iwasaki, R. Hiruta, H. Kuribayashi, and R. Shimizu, J. Appl. Phys. 105, 083536 (2009).[3] W. W. Mullins, J. Appl. Phys. 28, 333 (1957).
9:00 PM - S4.6
Supercritical Fluidic Carbon Dioxide Development of Polymer Nano-springs Fabricated by Two-photon Lithography.
Satoru Shoji 1 , Tomoki Hamano 1 , Satoshi Kawata 1
1 Department of Applied Physics, Osaka University, Suita, Osaka Japan
Show AbstractLaser nanolithography based on two-photon absorption induced polymerization allows us to fabricate three-dimensional freestanding structures of polymer with the spatial resolution of less than 100 nm. Two-photon lithography is one of the powerful tools to shape a variety of polymers into micro/nanoscale three-dimensional structures not only for fabricating micro/nano-devices, but also for studying fundamental properties of polymers in nanoscale. Recently it is reported that such thin nano-sized polymers exhibit several unique behaviors in mechanical, electrical, and thermal properties, which are quite different from those in the bulk state. However, in order to obtain the two-photon lithography-fabricated polymer as dried freestanding micro/nanostructures in air, there is a crucial problem in the developing process. As the shape becomes small, the mechanical rigidity of the polymer structure fails to sustain its original shape against the tensional force from the surrounding solvent when we remove un-polymerized monomers by solvent. This effect becomes significant especially when the structure consists of thin polymer nano-wires with high aspect ratio, such as polymer nano-springs [1-3]. In this presentation, we present a method to develop such fragile polymer micro/nanostructures without severe damage and distortion by means of supercritical fluidic carbon dioxide. After the common routine developing process, i.e. immersing the substrate into the developer, without drying the developer the substrate is exposed to supercritical fluidic carbon dioxide under high pressure. The supercritical fluid smoothly penetrates into the developer, and thereby the developer is finally replaced by the supercritical fluid. Since the viscosity of the supercritical fluid is extremely low compared with the common liquid, the replacement progresses without any tensional force caused by the flow of the fluid. When the exchange is completed the pressure is unloaded so that the supercritical fluid is transformed to carbon dioxide gas. By this method we successfully obtained dried polymer nano-springs consisting of 300 nm thick polymer nano-wires. In the presentation, we show the elasticity of the obtained nano-springs confirmed by means of atomic force microscopy. References : [1]S. Shoji, S. Nakanishi, T. Hamano, and S. Kawata, MRS Proc. 1224, 1224-FF06-05-DD06-05 (2009). [2]S. Nakanishi, S. Shoji, H. Yoshikawa, Z. Sekkat, and S. Kawata, J. Phys. Chem. B 112, 3586 (2008). [3]S. Nakanishi, S. Shoji, S. Kawata, and H.-B. Sun, Appl. Phys. Lett. 91, 063112(2007).
9:00 PM - S4.7
Fabrication and Characterization of MEMS-Based Structures from a Bio-inspired, Chemo-responsive Polymer Nanocomposite.
Allison Hess 1 , Jeffrey Capadona 2 3 4 , Stuart Rowan 3 , Christoph Weder 3 5 , Dustin Tyler 2 4 , Christian Zorman 1
1 Department of Electrical Engineering and Computer Science, Case Western Reserve University, Cleveland, Ohio, United States, 2 , Louis Stokes Veterans Affairs Medical Center, Cleveland, Ohio, United States, 3 Department of Macromolecular Science and Engineering, Case Western Reserve University, Cleveland, Ohio, United States, 4 Department of Biomedical Engineering, Case Western Reserve University, Cleveland, Ohio, United States, 5 Adolphe Merkle Institute, University of Fribourg, Fribourg Switzerland
Show AbstractWe have recently developed a polymer nanocomposite comprised of a poly(vinyl acetate) (PVAc) matrix embedded with stiff cellulose nanofibers that in bulk samples displays a switchable and reversible elastic storage modulus from 5.1 GPa when dry to 12 MPa when exposed to water, mimicking the switchable stiffness of the Cucumaria frondosa dermis. The unique ability to dramatically alter the mechanical properties of a structural material in situ enables new capabilities for MEMS. However, microfabrication of devices utilizing the nanocomposite is challenging due to chemical sensitivities to acids, bases, and organic solvents, as well as to temperatures exceeding 100°C. This paper reports our effort to develop microfabrication processes to facilitate patterning the nanocomposite and its integration with other materials, and to characterize the material for MEMS applications. The nanocomposite was synthesized using a solution-casting and compression-molding technique to form 50 μm-thick, freestanding films. A direct-write CO2 laser was used to pattern structures into this chemical- and temperature-sensitive material. Integration of photolithographically-defined, thin-film Ti/Au (20 nm/200 nm) features onto the nanocomposite was enabled by using a 1 μm-thick parylene film deposited onto one surface of the nanocomposite, which served as a barrier for the wet chemicals involved in metal patterning and protected the metal traces from the moisture absorbed by the nanocomposite. A second 1 μm-thick parylene film was then deposited and patterned over the metal features to provide an insulating capping layer.Microtensile testing using a custom-built apparatus was implemented to determine the stiffness and stimulus response of laser-micromachined, micron-scale nanocomposite structures, both bare and with integrated parylene and metal thin films on one surface. Bare nanocomposite tensile samples with 50 μm-thick, 200 μm-wide, and 3000 μm-long beams were found to have a Young’s modulus of 3.4 GPa in the dry state, and ~20 MPa after soaking in DI water, requiring approximately 3 minutes to achieve the full change in stiffness. The addition of thin film parylene and metal structures to one surface of the nanocomposite did not impede its response to exposure to water. The nanocomposite is dependent upon moisture uptake to display a stimulus response, but exhibits anisotropic swelling with approximately 3 times as much dimensional increase through the film thickness than across film. Adhesion of parylene-encapsulated metal structures on a nanocomposite substrate have been assessed using both tape tests and soak tests in phosphate buffered saline at 37°C for 60 days. After being subjected to both types of tests, the parylene and metal films did not show any evidence of delamination from the nanocomposite. The extended paper will detail the application of this material as a mechanically-dynamic neural probe for cortical interfacing.
9:00 PM - S4.8
Reliability and Stability of Thin-film Amorphous Silicon MEMS on Gass Substrates.
P. Sousa 1 , V. Chu 1 , J. Conde 1 2
1 , INESC Microsistemas e Nanotecnologias (INESC MN) and IN-Institute of Nanoscience and Nanotechnology, 1000-029 Lisboa Portugal, 2 Department of Chemical and Biological Engineering, Instituto Superior Técnico, 1000-049 Lisboa Portugal
Show AbstractMicro-electro mechanical systems (MEMS) are important components for sensor and actuator applications in which miniaturization is a key aspect as they have the potential of enhancing sensitivity and increasing actuation speed and precision. For MEMS applications, the stability and reliability of device performance is critical and knowing how they are related to the changes in mechanical properties upon stress is of great importance in the design of these systems. In this work, we present a reliability and stability study of MEMS resonators based on doped hydrogenated amorphous-silicon (n(+)-a-Si:H) thin-films deposited by radio frequency plasma enhanced chemical vapor deposition (RF-PECVD). We investigate the stress versus life curve (number of cycles to failure) and the material aging by monitoring the structural and electronic material properties due to long-term cyclic loading.Movable parts in MEMS devices under long-term repeated cycling load are subjected to possible failure mechanisms including material fatigue and aging, mechanical fracture, stiction, wear, delamination, residual stress, and environmentally induced failure mechanisms. For a brittle material like silicon, under applied stress fracture occurs at the sites of highest stress concentration which usually are processing-induced. However, micron-scale fatigue has also been reported for single crystal and polycrystalline silicon. The mechanisms that affect MEMS device reliability are still an open question and have not been investigated in the case of thin-film amorphous silicon based MEMS structures.The electrostatic resonators are bridge structures with an underlying gate electrode, fabricated on glass substrates at a maximum processing temperature of 110 °C using CMOS compatible technology. A 1 micrometer thick photoresist layer is used as the sacrificial layer. The micro-resonator structures are electrostatically actuated by applying a voltage between the gate electrode and the bridge at room temperature. The shift in the resonance frequency under long term and repeated cycling load is monitored by with an optical setup and at a pressure of approximately 1E-6 Torr.Failure of the resonating bridges is not observed at room temperature for at least 3E12 cycles with increasing load. It is shown that they can withstand the industry standard of 1E11 cycles at high loading (deflection estimated at 2 nm) with a resonance shift of ~0.3% which occurs primarily in the initial stages of the cycling. Also, no change in the quality factor of the resonant bridges is observed during the entire cycling period. A continuous increase of the electrical conductivity of the thin-film amorphous silicon bridges with the number of cycles was measured (of the order of 250% after 4.6E11 cycles). Reliability of different resonator geometries will be presented as a function of applied loading and scanning electron microscopy imaging will be used to evaluate possible mechanically induced damage.
9:00 PM - S4.9
Improving PZT Thin Film Texture Through Pt Metallization and Seed Layers.
Luz Sanchez 1 2 , Daniel Potrepka 1 , Glen Fox 3 , Ichiro Takeuchi 2 , Ronald Polcawich 1
1 RDRL-SER-L, Army Research Laboratory, Adelphi, Maryland, United States, 2 Materials Science and Engineering, University of Maryland, College Park, Maryland, United States, 3 , Fox Materials Consulting LLC, Colorado Springs, Colorado, United States
Show AbstractLeveraging past research activities in orientation control of lead zirconate titanate (PZT) thin films this work attempts to optimize those research results using the fabrication equipment at the U.S. Army Research Laboratory so as to achieve a high degree of (001) texture and improved piezoelectric properties. The initial experiments examined the influence of Ti/Pt and TiO2/Pt thins films used as the base electrode for the sol-gel PZT thin film growth. In all cases, the starting silicon substrates used a 500nm thermally grown silicon dioxide. The Pt films were sputtered with the TiO2/Pt films using a highly textured titanium dioxide film grown by a thermal oxidation process of a sputtered Ti film. The second objective targeted achieving highly oriented (001) texture in the PZT using a seed layer of PbTiO3 (PT). A comparative study was performed between Ti/Pt and TiOx/Pt bottom electrodes. The results indicate that the use of a highly oriented TiOx led to highly textured (111) Pt which in turn improved both the PT and PZT orientations. Additionally, PZT (52/48) and (45/55) thin films with and without PT seed layers were deposited and examined via x-ray diffraction methods (XRD) as a function of annealing temperature. As expected, the seed layer provides significant improvement in the (001) orientation while suppressing the (111) orientation of the PZT. Improvements in the Lotgering factor (f) was observed from our existing Ti/Pt/PZT process (f=0.66) to samples using the PT seed layer as a template, Ti/Pt/PT/PZT (f=0.87), and finally to films deposited onto the improved Pt electrodes, TiOx/Pt/PT/PZT (f=0.96). 1. N. Ledermann, et.al., Sens. Actuators A 105, 162 (2003). 2. S. Trolier-McKinstry, ARO Final Report 2007.3. D. Potrepka (ARL), G. Fox (self), J. Martin (ARL), R. Polcawich (ARL), unpublished research
Symposium Organizers
Frank W. DelRio National Institute of Standards and Technology
Maarten P. de Boer Carnegie Mellon University
Christoph Eberl Karlsruhe Institute of Technology (KIT)
Evgeni P. Gusev Qualcomm MEMS Technologies
S9: Poster Session: Materials Characterization
Session Chairs
Tuesday PM, November 30, 2010
Exhibition Hall D (Hynes)
S5: Nanocalorimetry and Nanosensors
Session Chairs
Tuesday PM, November 30, 2010
Room 207 (Hynes)
9:30 AM - **S5.1
Nanocalorimetry Measurements of Thermal and Thermodynamic Properties of Thin Film and MEMS Materials.
David LaVan 1 , Parasuraman Swaminathan 2 1 , Feng Yi 1 , Ravi Kummamuru 3 1 , Mark Vaudin 1 , Leslie Allen 3 , Timothy Weihs 2
1 Materials Science and Engineering Laboratory, National Institute of Standards and Technology, Gaithersburg, Maryland, United States, 2 Department of Materials Science and Engineering, Johns Hopkins University, Baltimore, Maryland, United States, 3 Department of Materials Science and Engineering, University of Illinois, Urbana-Champaign, Illinois, United States
Show AbstractNanocalorimetry is a nanoscale technique to measure the thermal and thermodynamic properties of materials such as the melting temperature, solidification temperature, reaction temperature, heat capacity and enthalpy for nanomaterials ranging from discrete nanoparticles to thin films of pure metals and reactions between nanomaterials. Nanocalorimeter chips are similar to conventional calorimeters except they have been scaled down considerably, are produced using silicon micromachining, are much more sensitive (nJ) and respond much faster (10^4 °C/s to 10^6 °C/s is not unusual). A typical nanocalorimeter has a heater and temperature sensor supported on a 50 nm to 100 nm thick silicon nitride membrane; in some instances the temperature is recorded by monitoring the resistance of the heater and comparing that to calibration data for the temperature coefficient of resistance (TCR) for the heater. Experiments may be conducted in air, a controlled gas environment or in vacuum. Details of measurements on a variety of materials including in situ nickel silicide formation, the melting and solidification of pure aluminum and reaction initiation for nickel aluminum multilayers will be presented along with work to improve calibration strategies and techniques for these chips.
10:00 AM - S5.2
Heat Treatment and Thermal Cycling Analysis of Ni-Ti-Zr Shape Memory Alloy Thin Films by Combinatorial Nano-Calorimetry.
Patrick McCluskey 1 , Joost Vlassak 1
1 SEAS, Harvard University, Cambridge, Massachusetts, United States
Show AbstractMicrofabrication techniques have facilitated the development of membrane-based calorimeters with extremely small thermal addendums. Such calorimeters have sensitivities fine enough to measure thermal properties of nanoscale quantities of materials. The recently developed parallel nano-scanning calorimeter (PnSC) combines techniques from nano-calorimetry and combinatorial analysis to create a useful tool for materials research and discovery. The extremely small thermal mass of thin films makes high heating and cooling rates possible. These thermal conditions allow novel phases and microstructures, including the glassy phase in non-optimal glass forming compositions and fine nanocrystalline grain structures. High heating and cooling rates also make serial measurements of large combinatorial sample libraries feasible and reduces times for high-throughput heat treatment and thermal cycling experiments. In fact, thousands of calorimetric measurements can be performed in a mater of hours, a 100-1000 fold improvement over conventional calorimeters. These properties are exploited by the PnSC to study the Ni-Ti-Zr materials system. The PnSC is used to fabricate combinatorial libraries of amorphous Ni-Ti-Zr thin films. These films are crystallized by local heating in a process that lasts just tens of milliseconds. The real-time calorimetric data gives crystallization temperatures and process information through the crystallization peak morphology. Post process data is obtained by x-ray diffraction and resulting phases compare one-to-one with calorimetry peak morphology on a composition basis. The crystallization process determines the phases and microstructure formed; in this case a fine nanocrystalline grain structure and the martensite phase for some compositions. The martensite-austenite phase transformation is studied as a function of composition, heat treatment and thermal cycling, which provides information about the thermo-mechanical coupling, i.e., the shape memory effect, from the thermal perspective. High temperature (900 °C) heat treatments alter the shape memory effect, indicating that the materials are evolving toward an equilibrium state. Heat treatments also reset the thermal fatigue mechanism as determined by low temperature thermal cycling through the shape memory transformation.
10:15 AM - S5.3
Integration of Nanowires onto Microhotplates for the Fabrication of Low Power Consumption and Fast Operated Gas Nanosensors.
Roman Jimenez-Diaz 1 , J.Daniel Prades 1 , Francisco Hernandez-Ramirez 2 3 , Joaquin Santander 4 , Carlos Calaza 4 , Luis Fonseca 4 , Carles Cane 4 , Albert Romano-Rodriguez 1
1 Department of Electronics, University of Barcelona, Barcelona Spain, 2 M-2E, Institut de Recerca en Energia de Catalunya, Barcelona Spain, 3 XaRMAE, University of Barcelona, Barcelona Spain, 4 , Instituto de Microelectrónica de Barcelona, Bellaterra Spain
Show AbstractNanowires have emerged as potential building blocks for future electronic devices [1]. However, significant requirements arise from the use of elements with dimensions in the nanometer scale: large scale synthesis of structures with homogeneous properties and reliable, affordable and fast contact fabrication, among others. In this work, a methodology for the fabrication of gas sensors by integrating individual metal oxide nanowires (NWs) as sensing elements and microhotplates for low power consumption and fast operation is presented. First, monocrystalline CVD-grown SnO2 nanowires were dispersed on ethanol. And later, a droplet of this solution was spread onto over suspended MEMS hotplates which contained integrated microheaters. To guarantee the formation of good electrical contacts between pre-patterned microelectrodes and nanowires, Electron Beam Assisted Deposition and Ion Beam Assisted Deposition processes were performed. These nanowires were electrically contacted using a FEI Strata 235 dual beam instrument equipped with an injector to deposit Pt. The details of this fabrication method were explained in detail elsewhere [2]. Finally, two- and four-probe dc electrical measurements were done using a Keithley 2602 Source Measure Unit, enabling the estimation of the key-parameters of these nanowires. On the other hand, the integration of microheaters expedited the use of these nanowires as gas sensors. Some of them were tested using well-controlled environmental conditions of gas and temperature. This kind of measuring platform enables an optimal control of the working temperature allowing fast and reproducible modulation of the temperature up to 600 K with reduced power consumptions due to the thermal isolation and reduced dimensions of these microhotplates. The obtained results demonstrate the huge potential of nanowires as building-blocks of a new generation of devices with improved performances. For this reason microhotplate-based technologies are a promising approach for the fabrication of nanosensors in a scalable processReferences:[1] E. Comini, Anal. Chim. Acta 568 (2006) 28-40.[2] F. Hernandez-Ramirez et al., Nanotechnology 17 (2006) 5577-5583.Acknowledgments:Work partially funded by the Spanish projects MAGASENS, CROMINA and ISIS. RJ is indebted to the Spanish MEC for the FPU grant. This work has been developed partially using the capacities of ICTS Clean room of micro & nano fabricacion of IMB-CNM (CSIC).
10:30 AM - S5.4
Film Conductivity Controlled Variation of the Amplitude Distribution of High-temperature Resonators.
Denny Richter 1 , Silja Schmidtchen 1 , Han Xia 1 , Holger Fritze 1
1 LaserApplicationCenter, Clausthal University of Technology, Goslar Germany
Show AbstractHigh-temperature stable langasite (La3Ga4SiO14) resonators can be used to determine atmosphere induced changes in the mechanical and electrical parameters of metal oxide films at elevated temperatures. Potential applications include gas sensors for high-temperature applications which show an increased selectivity to different reducing gases like H2 and CO in comparison to conventional conductivity based metal oxide sensors. Alternatively, these transducers can be used to study the electrical and visco-elastic properties of metal oxides under different conditions. In the presented work, the principles of the conductivity induced variation of the effective electrode diameter and its influence on the resonance frequency of piezoelectric resonators are discussed.The electrode layout of the bulk acoustic wave langasite resonators bases on keyhole shaped platinum electrodes of different diameter at front and rear side. The metal oxide film is applied on the smaller electrode. Thereby, the diameter of the film is larger than the electrode diameter. This leads to a sensor film conductivity dependent effective electrode diameter. In this configuration, the frequency shift is not primarily caused by changes in mass load, but rather by changes in the mass sensitivity distribution of a resonator with non-uniform mass load. The local sensitivity of a thickness shear mode resonator shows a Gaussian distribution with a maximum at the center of the electrode. The width of the Gaussian profile depends predominantly on the electrode diameter. Increasing the conductivity of the metal oxide film increases the effective electrode diameter since the film acts as an additional electrode area and broadens the sensitivity distribution. This leads to an increase in sensitivity in the area of the relatively heavy platinum electrode. The mass load of the platinum electrode appears to be larger and causes a decrease in resonance frequency with increasing metal oxide conductivity. Measurements using a laser Doppler vibrometer are performed to determine the shear mode amplitude at high temperatures in order to verify this effect.A model to estimate the distribution of the excitation voltage, the effective electrode diameter and the resulting frequency shift is developed. Thereby, the resonator and electrode geometry as well as the thickness and conductivity of the resonator and film, respectively, are taken into account. Calculated frequency shifts using the presented model are in good agreement with measurements performed with TiO2 and CeO2 coated langasite resonators at at 600 °C and above.
10:45 AM - S5.5
Ultrafine Silicon Nano-wall Hollow Needles and Applications in Inclination Sensor and Gas Transport.
Zeinab Sanaee 1 , Shams Mohajerzadeh 1 , Mahdieh Mehran 1 , Mohammad Araghchini 1
1 Electrical and Computer engineering, University of Tehran, Tehran Iran (the Islamic Republic of)
Show AbstractSmart drug delivery depends the evolution of hollow micro-needles. The transfer of drug through little holes inside such structures allows a replacement for regular injection needles. Apart from biological applications, the cup-like structures can be used as gas and vapor transfer media for mass spectroscopy and ionization sources. We report a method to realize ultrafine hollow needle structures on silicon-based membranes. Such structures were used for liquid and gas transport showing anomalous behavior. The fabrication of ultrafine hollow structures is based on high-resolution deep etching of Si using a modified reactive ion etching method where a combination of H2/O2/SF6 gases is used in a sequential manner to perform passivation and etching steps in fully programmable sub-cycles. Arrays of rods with heights up to 50µm and widths of 1µm are achieved. The fabrication starts by cleaning (100)-Si wafers followed by E-beam deposition of Cr as the mask for future processing. Circular hollow rings with outer diameter of 3-40µm and rim widths less than 200nm are created using precision lithography. Once vertical cup-like structures are created, a slant-angle rotation and deposition method is used to cover the surface of the samples with Cr except for the inner surfaces and especially the bottom of the craters. By continuing the vertical etching, the bottom surface of the cylindrical features is etched away while sidewalls and outer surfaces remain intact. To obtain through-holes, the Si surface needs to be thinned prior to etching, accomplished using backside micromachining in KOH solution. Typical thicknesses of 20-30µm are used. Using this method, Si microneedles with size from 3-40µm are fabricated. High-resolution vertical etching gives the opportunity to have needle wall thicknesses around 60nm for structures with 7µm height, yielding an aspect ratio of 110. These structures have applications for gas transport and as inclination sensors where a cavity based Si microneedle array is realized. Since the backside of the Si substrate has been etched to form membranes, a liquid such as water can be incorporated to realize a capacitive sensor between the backside Cr-plated electrode and another Cr-plated substrate placed just opposite to the needle-holding substrate. The capacitance value depends on the amount of penetration of the liquid (water) with its high permittivity, in the tiny hollow needles (from backside) that depends on the angle of the sensor. This device can serve as high sensitivity inclination sensor. A variation of 0.6 pF/degree of inclination was obtained. Also, such structures are used as media to transfer gas or vapor through their tiny holes. The study of vapor (acetone) transport through these little holes shows an anomalous transient behavior which could be due to surface-assisted properties. Further investigation of the sensor and gas transport properties of such nano-wall structures is being pursued.
11:00 AM - S5: Sensors
BREAK
S6: Microsensors and Nanosensors
Session Chairs
Tuesday PM, November 30, 2010
Room 207 (Hynes)
11:30 AM - **S6.1
Incorporation of Sensitized Materials with CMOS MEMS for Multi-modal Analyte Sensing.
Gary Fedder 1 2 , Nathan Lazarus 2 , Kristen Dorsey 2
1 Institute for Complex Engineered Systems, Carnegie Mellon University, Pittsburgh, Pennsylvania, United States, 2 Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania, United States
Show Abstract(Invited Talk)Over the past eight years, our group has been developing a microsystem for end-of-service-life detectors in cartridges for gas respirators. In this system concept, multi-modal gas sensorsm specifically chemiresistive, gravimetric and chemicapacitive sensors, are combined to provide coverage over a broad variety of volatile organic compounds, while compensating for temperature and humidity effects. CMOS MEMS sensor implementations are sought in order to eventually integrate all devices with interface circuitry onto a common platform.A key materials challenge involves depositing appropriate amounts of sensitized materials onto the MEMS sensors. Inkjet deposition has proven successful in depositing polythiophenes, gold nanocrystals and gold nanoclusters onto gold and platinum electrodes for chemiresistive sensing. Our microcantilever gravimetric sensors require an indirect inkjet deposition where the material dissolved in solvent is wicked into slots micromachined on the cantilever. The distributed evaporation along the length of the beam leads to the material solidifying in a repeatable fashion from tip to base to “dose” the beam. The chemicapacitive sensors have parallel-plate electrodes directly made from the CMOS interconnect layers with a micromachined gap designed to accommodate inkjet delivery as well. The design, process and test results from these devices will be presented.
12:00 PM - S6.2
Metal Insulator Transition-induced Stress Changes in Vanadium Dioxide Thin Films for MEMS Devices.
Viswanath Balakrishnan 1 , Changhyun Ko 1 , Zheng Yang 1 , Shriram Ramanathan 1
1 Harvard School of Engineering and Applied Sciences, Harvard University, Cambridge, Massachusetts, United States
Show AbstractReversible stress change across the first order metal insulator transition of vanadium dioxide (VO2) thin films has been investigated using in-situ wafer curvature technique. The stability of the recoverable stress, hysteresis width and the transition temperature upon thermal cycling has been studied. Kinetics of the stress change across the phase boundary also been studied at various temperatures. Efforts are taken to tune the recoverable stress and phase transition temperature by varying the thickness of vanadium dioxide thin film and film stress. Along with the several orders of magnitude change in electrical resistance and optical transmittance, experimentally measured recoverable stress of 400 MPa, makes vanadium dioxide thin film a potential candidate for actuator/sensor applications. Detailed micro structural studies and electrical transport measurements supporting the stress change phenomenon will be presented to establish the direct structure-property relation in vanadium dioxide thin films.
12:15 PM - S6.3
Development of a Robust Design for Wet Etched Co-integrated Pressure Sensor Systems.
Wolfgang Schreiber-Prillwitz 1 2 , Mikko Saukoski 2 , Gerhard Chmiel 2 , Reinhart Job 1
1 Mathematics and Computer Science, University of Hagen, Hagen Germany, 2 , ELMOS Microsystems, Dortmund Germany
Show AbstractIn a previous report [1], we described and verified a method for the evaluating of electrical signals from MEMS pressure sensor systems in a Wheatstone bridge configuration. In the present work, we have applied this method to optimize the performance of a design of a co-integrated pressure sensor system for 1bar full scale range. We calculated and verified a gain in signal of about 5% by optimizing the position of the piezoresistors on the membrane for the symmetrically target configuration of each resistor having the same distance to the membrane rim. Subsequently, we calculated the influence of potential alignment errors between the backside cavity mask, forming the membrane, and the positions of the piezoresistor on the membrane’s front side. This leads to different distances between resistors and rim. Depending on the amount of such asymmetry, a maximal electrical signal deviation of 1% has been found. Additionally, the influence due to under-etching effects from KOH etching at the backside mask on the electrical signal has been compared for the old and the improved updated design of the pressure sensor system. As this under-etching has a certain range and alters the membrane size, this effect has a significant impact on the sensor performance. The signal variation caused by under-etching could be reduced drastically from 13% to 4%. [1] W. Schreiber-Prillwitz, M. Saukoski, G. Chmiel, R. Job, Design Approach and Realization of Integrated Silicon Piezoresistive Pressure Sensors for Wide Application Ranges, 218th Meeting of the Electrochemical Society, 2010, Symposium J3: "Microfabricated and Nanofabricated Systems for MEMS/NEMS 9", Oct. 10th – 15th, Las Vegas, USA
12:30 PM - S6.4
Fabrication of Blanket Nanoporous Gold Films by Current-controlled Dealloying.
Oya Okman 1 , Jeffrey Kysar 1
1 Department of Mechanical Engineering, Columbia University, New York, New York, United States
Show AbstractNanoporous gold films are fabricated from percursor binary solid solution Ag-Au alloys by selective removal of Ag, which is less noble than Au in a corrosive environment. The final product is a bi-continuous nanoporous gold structure with pore and ligament sizes that may range from a few nanometers to hundreds of nanometers; the process is known as dealloying. Since up to 65% of the atoms in the precursor alloy may be removed during dealloying, the volume of the nanoporous structure is often significantly smaller than the volume of the precursor alloy. In the case of films or membranes constrained to a substrate, the volume contraction causes high film stress and that can lead to cracking. Dealloying can be performed either by free corrosion or in an electrochemical cell, in which the precursor alloy serves as the anode of the cell. For voltage-controlled dealloying processes, the Ag dissolution rate is not uniform throughout the process and the instantaneous stress may achieve high values, increasing the risk of cracking. In this talk, a current-controlled dealloying procedure is suggested to control the rate of dissolution and control the time scale of the surface diffusion of Au for effective stress relief during fabrication. Crack-free blanket films are produced by this method. Incorporation of NPG film as the functional layer on MEMS devices is demonstrated. The relative benefits and disadvantages of the various dealloying methods will then be presented.
12:45 PM - S6.5
Giant Magnetoelectric Cantilever Sensors Based on FeCoSiB/AlN Thin Film Composites.
Henry Greve 1 , Eric Woltermann 1 , Robert Jahns 3 , Stephan Marauska 4 , Bernhard Wagner 4 , Reinhard Knoechel 3 , Manfred Wuttig 2 1 , Eckhard Quandt 1
1 Inorganic Functional Materials, Christian-Albrechts-University of Kiel, Kiel Germany, 3 High Frequency Laboratory, Christian-Albrechts-University of Kiel, Kiel Germany, 4 Microsystems Technology, Fraunhofer Institute of Silicon Technology, Itzehoe Germany, 2 Materials Science and Engineering, University of Maryland, College Park, Maryland, United States
Show AbstractSensors based on resonant Si-cantilevers with magnetoelectric thin film composites as the functional component can be used to measure ultra-low magnetic fields. Such cantilevers are interesting for techniques like magnetoencephalography or magnetocardiography where biomagnetic signals down to the pT range are measured. In general the sensitivity of cantilever sensors is increased as the damping decreases. In this talk we investigate the effects of air damping effects on Si-cantilevers with sputtered FeCoSiB/AlN thin film composites. First we demonstrate the feasibility to adjust the resonance frequency of the cantilevers by suitable design, i.e. including a seismic mass. Additionally we show that air damping which is normally suppressed by evacuation is also substantially reduced by lowering the resonance frequency. Si-cantilevers with FeCoSiB/AlN thin films, that include a seismic mass, feature a resonant magnetoelectric coupling coefficient of 1.8 kV/cmOe at 330 Hz in air which represents the highest achieved value so far.
S7: Process Integration
Session Chairs
Tuesday PM, November 30, 2010
Room 207 (Hynes)
2:30 PM - **S7.1
Design, Process, and Reliability Aspects of MEMS in Display Applications.
Surya Ganti 1
1 Qualcomm MEMS Technologies, Qualcomm, San Jose, California, United States
Show AbstractElectrostatic MEMS are attractive because of their potential to have devices with low power and design voltages. The concept of interference modulators (IMOD) and their application to reflective displays, developed by Qualcomm, will be discussed with particular emphasis on design, reliability and manufacturability requirements of such devices. Each pixel in a MEMS based display is a composite of several MEMS devices and serves as the fundamental unit in color reproduction. In reflective MEMS displays, trade-off between display brightness and color saturation will be discussed. Pixel architectures used to generate primary colors, and procedures used for color reproduction using IMOD design will be discussed.Glass based processes are core to the development of MEMS displays. To lower the cost and improve the scalability of displays, conventional LCD based manufacturing processes are used with customized packaging. Requirements for building MEMS devices on large area glass substrates will be discussed using IMOD display manufacturing as an example.Each display device consists of several thousands of MEMS devices. Reliability of such devices depends on the design, surfaces, and processes employed. The importance of design, packaging and interacting surfaces will be discussed through basic adhesion force models. Relationships between energy release rates and the pixel geometry will be elucidated.
3:00 PM - S7.2
Strain-gated Piezotronic Logic Nanodevices.
Wenzhuo Wu 1 , Yaguang Wei 1 , Zhong Lin Wang 1
1 School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia, United States
Show AbstractConventional CMOS based logic units are electronically triggered and driven by externally applied gate voltages, which are considered as “static” and are separated from the dynamic mechanical actuation units in nano-electromechanical systems (NEMS). We present the first piezoelectric trigged mechanical-electronic logic operation using the piezotronic effect, through which the integrated mechanical actuation and electronic logic computation are achieved using only ZnO nanowires (NWs). By utilizing the piezoelectric potential created in a ZnO NW under externally applied deformation, strain-gated transistors (SGTs) have been fabricated. Using the SGTs as building blocks, universal logic components such as inverters, NAND, NOR and XOR gates have been demonstrated for performing piezotronic logic calculations. In contrast to the conventional CMOS logic units, the SGT based logic units are driven by mechanical agitation and relies only on n-type ZnO NWs without the presence of p-type semiconductor components. The mechanical-electronic logic units can be integrated with NEMS technology to achieve advanced and complex functionalities in nanorobotics, microfluidics and micro/nano-systems. http://www.nanoscience.gatech.edu/zlwang/
3:15 PM - S7.3
Parallel Integration of Single-walled Carbon Nanotube Based Piezoresistive Pressure Sensors.
Brian Burg 1 , Thomas Helbling 1 , Christofer Hierold 1 , Dimos Poulikakos 1
1 Department of Mechanical and Process Engineering, ETH Zurich, Zurich Switzerland
Show AbstractSelective dielectrophoretic integration of individual small band gap semiconducting single-walled carbon nanotubes (SGS-SWNTs) is employed to enable the purely parallel fabrication of SWNT-based piezoresistive pressure sensors. Directed assembly allows precise carbon nanotubes positioning on the designated silicon oxide (SiO2) membrane edges, ultimately the positions of highest strain. Strain components other than from the principal axis are minimized through good alignment. The SWNTs are encapsulated by a protective alumina (Al2O3) coating and can be modulated by a top gate. The pressure sensors have a membrane diameter of 120 µm and thickness of 190 nm. Highest sensitivity of the long-term stable devices is achieved in the off-state of SGS carbon nanotube transistors (CNFETs), reaching values as high as S ~ 0.25 ΔR/R/bar, at a resolution better than 50 mbar, and a power consumption of less than 40 nW. Low contact resistances and high transmission are essential for good sensor resolution. The scale-up of the introduced robust and reliable fabrication process is straight forward and may provide interesting avenues toward SWNT sensor commercialization.
3:30 PM - S7.4
Thin Film Amorphous Silicon Bulk-mode Disk Resonators Fabricated on Glass Substrates.
A. Gualdino 1 , V. Chu 1 , J. Conde 1 2
1 INESC Microsistemas e Nanotecnologias, Institute of Nanoscience and Nanotechnology, Lisbon Portugal, 2 Department of Chemical and Biological Engineering, Instituto Superior Técnico, Lisbon Portugal
Show AbstractSilicon microelectromechanical (MEMS) resonators are promising alternatives to quartz for timing references, can be used as RF filters, and as microbalances in biological and chemical sensor applications. This work reports on the fabrication and characterization of thin-film amorphous silicon surface micromachined bulk-mode disk resonators processed at temperatures below 250C on glass substrates. The ability to fabricate these devices at low temperature enables its potential integration on CMOS chips, its fabrication on large area substrates such as glass or flexible plastic, which can allow the use of MEMS devices in applications that might not have been possible with standard MEMS technologies. Bulk resonators are expected to show higher quality factors, both in vacuum and in dissipative media, and to allow a wider range of tuning of the resonance frequency, than flexural or torsional resonators.The structural layer of the 3-µm-thick microresonators is based on n-type hydrogenated amorphous silicon film (n-a-Si:H) deposited by RF-PECVD. The structures are defined by trench etching using RIE, followed by removal of the aluminum sacrificial layer by wet etch. Resonators in the form of disks with 50 to 300-µm diameter and squares with the same edge lengths were fabricated and studied. All resonators feature four anchored supports. The structures are electrostatically actuated by a capacitive transduction gap of 2 µm. Electrodes are placed around the disk in each of the four quadrants and at the edge of the square resonators. To actuate the wine-glass mode, identical signals are applied on opposing electrodes along one axis. For the extensional modes, the signal is also applied to the electrodes on the orthogonal axis. Displacement resulting from electrostatic actuation is optically monitored. Quality factors up to 5000 are observed in vacuum for the microresonators on glass substrates, while at atmospheric pressure the quality factors go up to 1000. The resonance frequencies are in the MHz range. The dependence of the resonance frequencies and the quality factors on the geometry of resonators, on geometry of the anchor supports, and on the type of electrostatic actuation is studied. The identification of the resonance modes observed is made with the help of finite-element mechanical modeling of the MEMS structures. The influence of the medium on the resonance parameters of the different oscillation modes observed is reported.
3:45 PM - S7.5
Scalable Non-volatile SWNT NEMS Switches for Logic and Memory.
Sivasubramanian Somu 1 , Taehoon Kim 1 , Peter Ryan 1 , Ahmed Busnaina 1 , Luciano Silvestri 1
1 , Northeastern University, Boston, Massachusetts, United States
Show AbstractA single wall carbon nanotube based non-volatile nano-electromechanical bi-stable switch for memory applications is fabricated. This bi-stable switch utilizes one strand of single wall carbon nanotubes as the actuation element. These carbon nanotubes are suspended over two trenches which separated by a distance of 300nm. The trenches are 200nm wide and 20nm deep. Conventional E-beam lithography and microfabrication methods are used to fabricate these switches while modified dynamic electric field assisted directed assembly process is used to assemble the carbon nanotubes on to these fabricated structures. A phase shifter is used in the directed assembly process to enhance the bridging the electrodes with carbon nanotubes. Critical point drying is employed to avoid the stiction of the carbon nanotubes to the bottom of the trenches. After assembly the devices are annealed to reduce the contact resistance between the carbon nanotubes and metal electrodes. The actuation of the switch is achieved at voltages less than 5V. Preliminary results demonstrate that the actuation of ON and OFF states is achieved at the same voltage, compared to several other carbon nanotube based electromechanical devices. Actuation in the ambient conditions and their effect on reliability is also studied. Other parameters governing the switches such as frequency of actuation, number of cycles before failure, reliability, etc are being studied. Further modifications to the existing design would result in Latches, Flip-Flops, registers, etc which are the back bones of a computer processor chip. These devices can also be incorporated with existing CMOS process to fabricate other volatile memory devices.
4:00 PM - S7: Process
BREAK
S8: Packaging and Characterization Methods
Session Chairs
Tuesday PM, November 30, 2010
Room 207 (Hynes)
4:30 PM - S8.1
Microstructure and Mechanical Properties of as Fired Sintered Reaction Bonded Silicon Nitride (SRBSN) Micro Specimens.
Joachim Roegner 1 , Thomas Schwind 1 , Marcus Mueller 2 , Karl-Heinz Lang 1 , Volker Schulze 1
1 Institute of Materials Science and Engineering I, Karlsruhe Institute of Technology, Karlsruhe Germany, 2 Institute for Materials Research III, Karlsruhe Institute of Technology, Karlsruhe Germany
Show AbstractWithin the work of the Collaborative Research Center 499 sintered reaction bonded silicon nitride is investigated with respect to its advantages for the large-scale production of micro mechanical components. Here the investigations on micro three point bending specimens made of SRBSN with varied feedstock composition are described. For the feedstocks the content of sintering additives Y2O3, Al2O3 and MgO as well as the specific surface of Si-powder are varied. The specimens were produced by low pressure injection molding (produced at the Institute for Materials Research III, KIT). Sintered specimens show dimensions of 240 x 240 x 1400 µm3. Mechanical tests were carried out using a self developed micro testing device under monotonic three point bending conditions to detect bending strength. To detect crack growth resistance and R-curve behavior of SRBSN, the specimens were prepared with a crack-like notch by Focused Ion Beam technique (FIB) with a notch lengths between 1 and 15 µm. Macroscopic specimen made of SRBSN show bending strengths of 500 – 1000 MPa while the relative density is about 95 to 100 %. The R-Curve starts at Kth of 1.4 to 2.0 MPam0.5 and shows plateau values of about 5.7 MPam0.5. Characteristic bending strengths of micro specimens were found to be between 540 and 1230 MPa. Due to the size effect a bending strength of about 860 – 1730 MPa was expected (with the validity of a dominating surface effect and a Weibull modulus of 12). The reasons for the lower bending strength were a low relative density of about 80 – 95 % and a large amount of secondary phase which is in part irregular dispersed. The microstructure shows relatively low aspect ratios (1.9 – 3.9) of the needle-like silicon nitride grains. Due to the fact that all micro specimens were tested in the as fired state without a subsequent surface finishing, the origin of fracture can be found on the surface or within a 5 µm thick surface layer. SEM investigations show that fracture initiates at pores with diameters of 1 to 10 µm and at unevenness which results in stress concentration. The results of the bending tests for FIB-notched specimens show increasing fracture resistance with increasing length of the starter notch. With these results an R-curve could be determined using a fitting procedure. For micro specimens (relative density of 90 %) Kth is between 1.4 to 1.9 MPam0.5 and the plateau value seems to be in the same range as for macroscopic specimen. With increasing porosity the R-curve is shifted to lower values of crack growth resistance. This result correlates with the distribution of the specimens with natural defects in the Weibull plots (the more or less dominant R-curve behavior results in a non linear distribution). As conclusion it is to be said that SRBSN qualifies for the production of micro components, but still is challenging because of the large number of factors influencing the development of microstructure and the mechanical properties.
4:45 PM - S8.2
Lead Frame Packaging Utilizing Air-gap Structures for MEMS Devices.
Nathan Fritz 1 , Rajarshi Saha 1 , Paul Kohl 1 , Sue Ann Bidstrup Allen 1
1 Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta, Georgia, United States
Show AbstractAir-gap structures are of interest in a range of microelectronic applications, especially in microelectromechanical systems (MEMS) packaging. In this work, an overcoat material is used to cover a sacrificial polymer, polypropylene carbonate (PPC), which protects the MEMS device during packaging. Once the overcoat is in place, the sacrificial polymer is decomposed freeing the MEMS structure while the overcoat dielectric provides mechanical protection from the environment. An epoxy-based polyhedral oligomeric silsesquioxane (POSS) dielectric was used as a high-selectivity etch mask for the PPC. The dry etching provides improved definition of the encapsulating structure. The POSS provides a rigid overcoat for the sacrificial PPC leading to process improvements. The decomposition process and the overcoat material have been designed for clean removal of the PPC material without damaging the cavity formed during the process. The thermal decomposition of PPC results in low molecular weight products which permeate through the POSS overcoat leaving a clean cavity surrounding the MEMS component. These advancements have demonstrated that air-gaps with rigid overcoats can be used to encapsulate MEMS devices in a wafer-level packaging process. The packaging structures can be designed for a range of MEMS device sizes and operating environments including hermetic and vacuum conditions. However, the air-cavity structures need additional rigidity to withstand chip level packaging conditions, which are dependent on cavity size. Current work is focused on implementing a wafer level air-cavity package into a lead-frame packaging scheme for MEMS devices. Air-gap structures have been studied with regards to metal overcoat for increased rigidity. The rigidity of the overcoat encapsulating the air-clad MEMS structure was measured through nano-indentation using a conospherical diamond tip on 40 µm wide cavities. An increase of 5.6 times in cavity-strength was obtained for a thicker (3X) Al metal film. ANSYS models were used to understand experimental results with regard to cavity deformation as well as provide a useful tool for designing larger structures.
5:00 PM - S8.3
Dynamics of a Supercooled Water Droplet Impact on a Nanostructured Superhydrophobic Surface.
Lidiya Mishchenko 2 , Benjamin Hatton 2 , J. Ashley Taylor 1 , Vaibhav Bahadur 2 , Joanna Aizenberg 2 , Tom Krupenkin 1
2 School of Engineering and Applied Sciences, Harvard University, Cambridge, Massachusetts, United States, 1 Mechanical Engineering, University of Wisconsin-Madison, Madison, Wisconsin, United States
Show AbstractPotential applications of nano and micro-structured superhydrophobic surfaces span an exceptionally broad range, which includes areas as diverse as self-cleaning coatings, surfaces with tunable optical properties, chemical sensors, and hydrodynamic drag reduction surfaces. It has been recently demonstrated that these materials can substantially retard ice formation, sometimes down to temperatures as low as -25°C. In this talk we discuss a theoretical model that describes the mechanism of ice accretion on nano and micro-structured superhydrophobic surfaces. We consider the process of spreading and subsequent retraction of a supercooled water droplet that occurs upon the droplet impact on a superhydrophobic surface and derive a set of conditions required to allow the droplet completely detach from the substrate thus preventing ice accumulation. Dependence of ice formation dynamics on the temperature, details of the surface topography, substrate material, and other factors are investigated. The model is applicable to both regular and random topographies. The predictions of the model are in good quantitative agreement with experimental data. The results of the work can provide new insight into design and optimization of anti-icing structures and coatings.
5:15 PM - S8.4
Characterisation of Hydrophobic Forces for in Liquid Self Assembly of Micron Sized Functional Building Blocks.
Maurizio Gullo 1 , Loic Jacot-Descombes 1 , Laure Aeschimann 2 , Juergen Brugger 1
1 LMIS1 - IMT, EPFL Lausanne, Lausanne Switzerland, 2 , Nanoworld AG, Neuchâtel Switzerland
Show AbstractThe self assembly (SA) of micron sized building blocks has often been performed by using the capillary force (1). However it has proven difficult to achieve guided and selective SA of a defined number of functional building blocks, due to the hardly controllable nature of the capillary force (1). An alternative and more controlled way to self assemble micron sized building blocks would be to use the hydrophobic force. Specific and controlled hydrophobicity can be applied to selected areas of the building blocks by local chemical functionalisation (2, 3). Moreover hydrophobic gradients might be engineered to guide the SA and induce the self repairing of miss aligned assemblies.In order to optimise the design of such functionalised building blocks it is necessary to understand the nature of the hydrophobic force. Especially the force to distance relation of the attraction and the magnitude of binging force is of great interest. We have measured the hydrophobic interaction directly by using a scanning probe microscope. It was possible to quantitatively measure the attractive and binding forces of functionalised surfaces and non functionalised surfaces when immersed in liquids with different pH and ionic strengths. Moreover the change of the binding force in time has been studied. Several theoretical approaches to explain the nature of hydrophobic interactions exist (4). We have chosen a numerical approach and are currently performing molecular dynamic simulation to understand the nature of the hydrophobic interaction as well as its time and environment dependency. References:(1)M. Mastrangeli et al., JOURNAL OF MICROMECHANICS AND MICROENGINEERING, 19, 2009;(2)J. H. Chang et al, MATERIALS LETTERS 64, 2010;(3)L. Feng et al. ADVANCED MATERIALS, 14, 2002;(4)P. Roach et al., SOFT MATTER, 4, 2008;
5:30 PM - S8.5
Depth-resolved Residual Stress Analysis of Thin Coatings by a New FIB-DIC Method.
Marco Sebastiani 1 , Edoardo Bemporad 1 , Christoph Eberl 2 , George Pharr 3 4
1 Mechanical and Industrial Engineering Department , University of Rome "Rome Tre", Rome Italy, 2 izbs, Karlsruhe Institute of Technology, Karlsruhe Germany, 3 Department of Material Science, The University of Tennessee, Knoxville, Tennessee, United States, 4 Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee, United States
Show AbstractA new methodology is presented for the measurement of residual stress profiles through the thickness of thin coatings. The method consists of incremental focused ion beam (FIB) ring-core milling, combined with high-resolution SEM imaging and digital image correlation (DIC) for relaxation strain mapping of the remaining central micro-pillar. The through-thickness profile of the residual stress can be obtained by finite element modeling using Schajer’s integral method, which was originally developed for the hole drilling and extended here to the ring-core geometry.Experimental implementation of the method was carried out on two different coated systems: (i) a chromium nitride (CrN) CAE-PVD 3.0µm coating on steel substrate, and (ii) a gold MS-PVD 1.5µm coating on a Si/SiO2 substrate. In order to protect the sample surface from ion damage during milling (re-deposition, edge smoothening), a thin platinum layer was deposited before milling. A regular pattern of very small Pt dots was also placed on the sample surface before milling with the aim of improving the surface contrast of the SEM images needed for the DIC measurements. Incremental FIB milling was conducted at a current of 28 pA using an optimized milling strategy that produces minimum re-deposition over the sample surface. High-resolution SEM-FEG micrographs were continuously acquired during FIB milling. Digital image correlation (DIC) analysis of the micrographs gave the relaxation strain profile as a function of the milling depth. Results showed an average residual stress of -5.5 GPa in the CrN coating and of +230 MPa in the Au coating. These values are in reasonable agreement with estimates obtained by the XRD-sin2ψ method for the CrN coating and by the curvature method for the Au coating. The analysis revealed an almost constant through thickness stress for the Au coating, but an increasing residual stress from surface to the coating/surface interface in case of the CrN coating. The latter observation is likely related to stress relaxation during grain growth, which was clearly observed in microstructural cross sections. The observation is also in good agreement with models for structure-stress evolution in PVD coatings.
5:45 PM - S8.6
Interface Mechanics of Microtransfer Printing Processes.
Kevin Turner 1 2 , Hyun-Joon Kim-Lee 2 , David Grierson 1
1 Mechanical Engineering, University of Wisconsin-Madison, Madison, Wisconsin, United States, 2 Materials Science Program, University of Wisconsin-Madison, Madison, Wisconsin, United States
Show AbstractMicrotransfer printing is an emerging manufacturing technique to build three-dimensional microstructures. In the process, a soft elastomer stamp is used to retrieve patterned microstructures from a host substrate and subsequently print the structures onto a new substrate or device. Microtransfer printing offers unique opportunities in MEMS fabrication and packaging; for example it can be used to create three-dimensional structures without a release layer, integrate unconventional materials with conventionally fabricated devices, and combine thinned electronic devices with micromechanical devices. In this paper, the interface mechanics of the transfer printing process is investigated theoretically and experimentally to improve transfer efficiency.Successful transfer printing requires a crack to propagate at the interface between the microstructure and substrate during the retrieval step and a crack to propagate between the stamp and microstructure during the printing step. In the present study we examine how the elasticity and geometry of the stamp, microstructure, and substrate affect the strain energy release rate at the stamp-microstructure and microstructure-substrate interfaces. The strain energy release rates are calculated using finite element simulations over a large range of moduli and geometry for different combinations of applied normal and shear loads. Using the strain energy release rates and measured interface toughness values favorable conditions for retrieval and printing are identified and presented in the form of process maps. The modeling results are compared to results from transfer experiments in which 0.5 – 3.0 micrometer thick single crystalline silicon components were transferred.
S9: Poster Session: Materials Characterization
Session Chairs
Wednesday AM, December 01, 2010
Exhibition Hall D (Hynes)
9:00 PM - S9.10
Residual Stress and Corrosion Resistance of Sputtered Silicon Oxide and Silicon Carbide Composite Films.
Ping Du 1 , I-Kuan Lin 1 , Xin Zhang 1
1 Mechanical Engineering, Boston University, Boston, Massachusetts, United States
Show AbstractSilicon carbide (SiC) has received increasing attentions on the integration of micro-electro-mechanical system (MEMS) due to its excellent mechanical and chemical stability at elevated temperatures. However, the deposition process of SiC thin films tends to induce relative large residual stress. In this study, we used RF magnetron sputtering with silicon dioxide (SiO2) and SiC targets to deposit SiOC composite materials whose residual stress can be minimized at appropriate oxygen/carbon ratio. To optimize the material properties with low residual stress, the influence of oxygen/carbon ratio on the residual stress and corrosion resistance of the composite films was studied by Nanoindentation and Reactive Ion Etching (RIE).The SiOC films were deposited on silicon wafers by a Discovery 18 RF magnetron sputter (Denton Vacuum). The oxygen/carbon ratio was controlled by adjusting the RF powers applied to the pure (>99.9% purity) silicon oxide and silicon carbide targets, respectively. Energy dispersive X-ray (EDX) (Oxford Instruments) was employed to characterize the SiOC films in the aspect of stoichiometric composition. The spectra clearly show the peak transition of the element Oxygen and Carbon in the five samples. After that the Young’s modulus and hardness were investigated by nanoindentation. Both the modulus and hardness increase with higher carbon concentrations.The residual stress in the composite films was calculated from the wafer curvature by using Stoney equation. The as-grown films and thermal-annealed (at 400, 600, 800 °C for 10 min) films were used in the experiments. The results show that the residual stress is transformed from compressive to tensile mode as the SiC concentration increasing, and increases with elevated annealing temperatures. It indicates that it is possible to minimize the residual stress in the films by controlling the ratio of SiO2 to SiC and annealing temperature.Nevertheless, the addition of SiO2 will decrease the corrosion resistance of SiC films to some extent. This property was characterized by comparing the etching rate in the RIE process. All the films were etched by 20 sccm CF4 + 4 % O2 for 5 min, and three RF power levels were used, namely, 100, 200 and 300 W. The results show that the etch rate decreases as the SiC concentration increasing, and generally the etch rate increases with higher RF power. The trend of RF power is in agreement with the DC voltage, which represents the ion bombardment energy. It indicates that the etch process is more dominant by the physical process, which agrees the high corrosion resistance of SiC reported in literature.In summary, by choosing the appropriate composition and post processing, a film with relative low residual stress yet good corrosion resistance could be obtained. This flexibility provides more opportunities for SiOC composite films to be integrated into the MEMS.
9:00 PM - S9.11
Seeding Effects on the Structure and Properties of Sr3Bi4Ti6O21 Thin Films Obtained by Chemical Solution Deposition.
Maria Elisabete Costa 1 , Shuo Jin 2 , Isabel Miranda Salvado 3
1 Ceramic and Glass Engineering, CICECO, University of Aveiro, Aveiro Portugal, 2 Ceramic and Glass Engineering, CICECO, University of Aveiro, Aveiro Portugal, 3 Ceramic and Glass Engineering, CICECO, University of Aveiro, Aveiro Portugal
Show AbstractThe awareness of PZT toxicity is prompting the search of new lead free (LF) materials able to respond to an endless list of demanding microelectronic applications including capacitors, ferroelectric memories and microelectromechanical systems. Aurivillius oxides (AO) are a LF family which layered crystal structure is composed by Bi oxide layers (Bi2O2)2+ interleaved by n perovskite-like slabs (An-1BnO3n+1)2- where n may vary from 2 to 8. Due to their fatigue-free properties against repeated polarization switching, AO with n=2 including SrBi2M2O9(M=Ta,Nb) became outstanding candidates for memories applications and thus originated a large amount of studies. Conversely AO with high n(n≥5) still remain almost unexploited being their appropriate synthesis conditions scarcely known and hence their technological potential not fully proved or investigated.The present study is focused on the preparation and properties of AO thin films with n=6, by a chemical solution deposition method combined with a seeding procedure. The selected target composition is Sr3Bi4Ti6O21 (SBTi6), with and without a seeding layer, on a Pt/Ti/SiO2/Si substrate. A sol-gel method starting from alkoxides precursors solution was modified for obtaining single phase SBTi6 thin films. The optimization of experimental variables including chemical reagents concentration and type, spinning parameters and heat treatment schedules allowed to establish the necessary conditions for synthesizing dense films, free of second phases, with uniform microstructure and enhanced ferroelectric properties. The same chemical precursors were also used to precipitate nanosized particles which were then converted by heat treatment into crystalline SBTi6 nanoseeds. In this work the effects of an SBTi6 seeding layer on the crystalline orientation, microstructure and electrical properties of the seeded thin films was also investigated. The properties of SBTi6 thin films, with and without seeds, including structure, dielectric and ferroelectric properties are comparatively discussed so as to identify useful structure-processing-properties relationships applying to AO thin films with high n.
9:00 PM - S9.12
Processing – Structure - Property Relationships in Mn Doped K0.5Na0.5NbO3 for Micro Electromechanical Systems.
Maria Elisabete Costa 1 , Muhammad Asif Rafiq 1 , Paula Vilarinho 1
1 Dep. of Ceramics and Glass Engineering, University of Aveiro, Aveiro Portugal
Show AbstractSmart materials, that include piezoelectric and ferroelectrics, play a crucial role in the development of micro- and nanoelectromechanical systems (MEMS/NEMS), which currently aim at applications as optical displays, acceleration sensing, radio-frequency switching, drug delivery, chemical detection, and power generation and storage. K0.5Na0.5NbO3 (KNN) is one of the leading lead free piezoelectric materials in the perovskite group being considered as a candidate material for MEMS/NEMS. Although, pure KNN has inferior electromechanical properties as compared to the properties of Pb(Zrx,Ti1-x)O3 (PZT), which is currently the most widely used material for electromechanical application, efforts are on going to tailor and improve the properties of KNN by using various dopants, chemical substitutions, or by texturing. Manganese is a versatile dopant, known to serve as sintering aid, dielectric losses suppresser and piezoelectric response improver. Due to its several oxidation states manganese can occupy A- or B-sites of the perovskite lattice depending on its amount, valence state and synthesis conditions. Mn is indeed a commonly used dopant for piezoelectric materials like BaTiO3, PbTiO3. Although also reported in KNN, no systematic study has been reported about the relationship between Mn site occupancy and the electromechanical properties. In this work, the role of Mn-site occupancy on the structural and electromechanical properties of A and B-site doped KNN is presented and the relationships between the structure, properties, processing, and performance of KNN are established.
9:00 PM - S9.13
Characterization of Aluminum Nitride Thin Films for Use in Surface Acoustic Wave Devices Operating at High Temperatures.
J. Justice 1 , L. Rodak 1 , D. Korakakis 1
1 Lane Department of Computer Science and Electrical Engineering, West Virginia University, Morgantown, West Virginia, United States
Show AbstractSurface acoustic wave (SAW) devices have experienced widespread success in the electronics and communications industries as bandpass filters, resonators and delay lines [1]. In recent years, SAW devices have been investigated for their potential use in mechanical, chemical and biological sensors [2]. While there are many materials that exhibit outstanding SAW characteristics at room temperature, such as lead zirconium titanate (PZT), lithium niobate (LiNbO3) and quartz, they fail at high temperatures. Langasite has been in the forefront of research for high temperature materials for use in SAW devices because of its piezoelectric stability up to 1473 °C [3], however, langasite exhibits large acoustic propagation losses at high temperatures that increase with frequency, making it unsuitable for devices operating at high temperatures [4]. Aluminum nitride (AlN) has been considered for use in high temperature SAW devices, but very little experimental research has been conducted to confirm it’s piezoelectric and SAW characteristics at high temperatures [5]. In this study, AlN thin films have been grown via metal organic vapor phase epitaxy (MOVPE) and DC reactive sputtering (DC RS) on silicon and sapphire substrates. AlN samples were annealed at temperatures ranging from 600 °C to 1200 °C in atmospheric and nitrogen environments. AlN thin film quality has been characterized before and after thermal cycling using x-ray diffraction (XRD), x-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), energy dispersive x-ray spectroscopy (EDS) and atomic force microscopy (AFM). AlN thin film piezoelectric properties have been characterized using laser Doppler vibrometry (LDV). SAW interdigital transducers (IDTs) have been fabricated using standard photolithographic processes. SAW devices have not yet been tested, however, a furnace is currently been constructed to allow in-situ measurements of SAW devices on AlN thin films at high temperatures.[1]F.S. Hickernell, 12th Int. Con. on Micro. and Rad., 4, 159 (1998).[2]D.S. Ballantine Jr., R.M. White, S.J. Martin, E.T. Zellers, and H. Wohltjen, Acoustic Wave Sensors: Theory, Design, and Physico-Chemical Applications, 1st ed. (Academic Press, San Diego, 1997) p.70-99.[3]B.V. Mill and Y.V. Pisarevsky, Proc. IEEE/EIA Inter. Freq. Cont. Symp., 10.1109/FREQ.2000.887343, 133 (2000).[4]R. Fachberger, G. Bruckner, R. Hauser, C. Ruppel, J. Biniasch, and L.Reindl, Proc. IEEE Ultrason. Symp. 1-3, 1223 (2004).[5]T. Aubert, O. Elmazria, B. Assouar, L. Bouvot, and M. Oudich, Appl. Phys. Lett., 96, 203503 (2010).
9:00 PM - S9.14
Fabrication and Characterization of A Microscale Cellular Loading Device for Cellular Biomechanical Study.
Qian Wang 1 , Yi Zhao 1
1 Biomedical Engineering, Ohio State University, Columbus, Ohio, United States
Show AbstractMechanical stimuli interfere with cellular behaviors under many physiological conditions. To understand the role of mechanical stimuli, engineered devices are developed to apply mechanical loads to cells in vitro. Despite of their usefulness, these devices are limited since they often lack the capacity of spatial load control, which is essential for intercellular study. Moreover, application of both compressive and tensile loads using a single loading device is challenging. Here, we fabricated and characterized a microdevice for programmable compressive/tensile loading to live cells.The device consists of two PDMS substrates. The top substrate consists of an array of nine thin membranes, which are 500 μm in diameter, 50 μm in thickness and 2500 μm in spacing. Each membrane is connected with a microfluidic channel built in the bottom substrate. Upon actuation, the fluid in the channels deforms the membranes and applies controllable strains to cells cultured on the membranes. In this design, each membrane can be individually controlled to desired strain levels.The top substrate is fabricated using double-side micromolding. Specifically, an array of microdots (5 μm in diameter, single spaced) is patterned on the upper mold; a micropost (500 μm in diameter and 200 μm in height) is patterned on the lower mold. The two molds are aligned face-to-face and separated by a spacer with the height of 250 μm. The PDMS prepolymer (mixed at 1:10 (w/w)) is poured in the gap and cured at room temperature overnight. Afterwards, the upper and lower molds are released, leaving thin PDMS membranes on the top substrate. The microfluidic channels in the bottom substrate are fabricated by soft lithography. After fabrication, the top and bottom substrates are treated by oxygen plasma and bonded. The thin membranes align with the microfluidic channels.The strain generation in the thin PDMS membranes is characterized by mapping the displacement of the dot array. The dot array in an un-deformed membrane is first captured as a reference. Afterwards, liquid is pumped in the microfluidic channel using a precision syringe pump under a small positive pressure overnight. This is to allow the air to escape. Once the channels are fully occupied by liquid, the displacement of the dot array is captured under a systematically varied back pressure. The stain distribution is obtained by comparing with the reference. The result shows that given the aforementioned geometries, the strain at the center of the membrane ranges from -4% (compressive) to about 20% (tensile), agreeing well with finite element analysis. Cell testing is performed using NIH 3T3 fibroblasts. Cells on different membranes are simultaneously loaded with strains of various magnitudes. The morphology of the cells on and between the membranes is examined. The result shows that the cells are viable under both compressive and tensile loads, indicating the device a promising solution for cellular biomechanical study.
9:00 PM - S9.16
Wafer-level Singulation, Release and Post-processing in Surface Micromachining.
Maarten de Boer 1 2 , Robert Anderson 2 , Randy Shul 2 , Peggy Clews 2 , Michael Baker 2
1 Mechanical Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania, United States, 2 , Sandia National Labs, Albuquerque, New Mexico, United States
Show AbstractWe propose and demonstrate a microfabrication processing sequence that enables wafer-scale singulation, release and subsequent processing of microelectromechanical devices. This sequence significantly reduces handling of individual dice compared with more conventional post-processing. Other advantages include faster processing and potentially higher yield. The yield after packaging was 100% on 12 microrelay devices that were processed according to the sequence, including a metallization step after release. The reliability was better than typical compared with previous tests for these devices. This processing sequence is also applicable to nanoelectromechanical devices, enables a direct interface to microfluidic devices, and also makes it possible to manufacture chips of arbitrary shape.
9:00 PM - S9.17
Transport Model for Microfluidic Device for Cell Culture and Tissue Development.
Niraj Inamdar 1 3 , Linda Griffith 1 2 , Jeffrey Borenstein 3
1 Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States, 3 Department of Biomedical Engineering, Charles Stark Draper Laboratory, Cambridge, Massachusetts, United States, 2 Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States
Show AbstractIn recent years, microfluidic devices have emerged as a platform in which to culture tissue for various applications such as drug discovery, toxicity testing, and fundamental investigations of cell-cell interactions in an extracorporeal environment. In this work, we examine the transport phenomena associated with gradients of soluble factors and oxygen in a microfluidic device for co-culture. In particular, we are interested in creating a microfluidic device that can emulate conditions that are known to be important in sustaining a viable culture of cells. Critical parameters include the flow and the resulting shear stresses, the transport of various soluble factors throughout the flow media, and the mechanical arrangement of the cells in the device.Using analytical models derived from first principles, we investigate the transport and flow conditions required to sustain cells in a microfluidic device. A particular device of interest is a bilayer configuration in which critical solutes such as oxygen are delivered through the media into one channel, transported across a nanoporous membrane, and consumed by cells cultured in another. The ability to control the flow conditions in this membrane bilayer device to achieve sufficient oxygenation without shear damage is shown to be superior to the case present in a single channel system. Using the results of these analyses, a set of criteria that characterize the geometric and transport properties of a robust microfluidic device are provided.
9:00 PM - S9.18
Nanolaminates of Cu-Zr Metallic Glass and Nanocrystalline Cu with High Tensile Strength and Ductility.
Ju-Young Kim 1 , Julia Greer 1
1 Materials Science, California Institute of Technology, Pasadena, California, United States
Show AbstractReliable performance of micro-devices and microelectromechanical systems (MEMS) requires mechanical stability of their components – for example, enhanced ductility and tensile strength of metal-based layers. This presentation will describe the results of the in-situ mechanical testing and microstructural characterization of micro-sized Cu-Zr metallic glass/nanocrystalline (nc) Cu nanolaminates subjected to uniaxial tension and compression. It has been shown that – once reduced to nano-scale - some metallic glasses have enhanced tensile ductility, deforming homogeneously rather than forming shear bands. We show that nc Cu layers play a key role in attaining ductility in these composites by preventing shear bands from transferring into the neighboring Cu-Zr layer. The initial films were deposited via RF-sputtering, and dog-bone shaped micro-tensile specimens were patterned using UV lithography, directional ion milling, and undercutting of Si substrate by selective dry etching. By taking into account various materials factors, we report optimum layer thicknesses for attaining maximum combined strength and ductility. We analyze the specific deformation mechanisms leading to the nanolaminates’ amplified mechanical properties by focusing on the non-trivial contribution of interfaces to defect activity during deformation.
9:00 PM - S9.19
Dynamics of Single Ion Transport through a Single Walled Carbon Nanotube Nanopore.
Wonjoon Choi 1 2 , Changyoung Lee 1 , Moonho Ham 1 , Michael Strano 1
1 Department of Chemical engineering, MIT, Cambridge, Massachusetts, United States, 2 Department of Mechanical engineering, MIT, Cambridge, Massachusetts, United States
Show AbstractThe stochastic Coulter effect in nanopores has been demonstrated for whole cells, micro and nanoscale particles, and DNA oligonucleotides, but no system has demonstrated this effect for single small molecule species to date. The reason for this appears to be an inability access the necessary lengths and diameters of the required nanopore. We show that a single walled carbon nanotube (SWNT) can be used as a single molecule conduit demonstrating a small molecule Coulter effect for the first time. In this work, we explore the behavior of such nanopores in comparison to simplfied theory. We indirectly observe the location and operation of electrostatic barriers at the pore mouth of a precise potential threshold. We also show that Markov chain processes describe the dynamics of two SWNT in parallel, and that single ion transport through the interior of the nanotube is linear in electric field as anticipated by theory.
9:00 PM - S9.20
Characterization of Textured PZT Thin Films Prepared by Sol-gel Method onto Stainless Steel Substrates.
Xuelian Zhao 1 , Xufang Yu 1 , Jinrong Cheng 1
1 School of Materials Science and Engineering, Shanghai University, Shanghai China
Show AbstractPbZr0.53Ti0.47O3 (PZT) ferroelectric thin films were deposited on LaNiO3 (LNO) buffered stainless steel substrates (SS) by sol-gel method. The effect of LNO buffer layer thickness on the orientation and electric properties of PZT thin films were studied. Scanning electron microscope (SEM) were applied to characterize the microstructure of the films. SEM images showed that the films were well crystallized. X-ray diffraction (XRD) results indicated that the thickness of LNO buffer layer affected the orientation of the films. Furthermore, the Curie temperature of PZT thin films decreased with the increase of the thickness of LNO buffer layer.
9:00 PM - S9.21
Nanoscale Friction of Uniaxially Stretched Polymer Films.
Yutao Yang 1 , Erik Dunkerley 2 , Jeffrey Haslam 1 , Daniel Schmidt 2 , Emmanuelle Reynaud 3 , Marina Ruths 1
1 Department of Chemistry, University of Massachusetts Lowell, Lowell, Massachusetts, United States, 2 Department of Plastics Engineering, University of Massachusetts Lowell, Lowell, Massachusetts, United States, 3 Department of Mechanical Engineering, University of Massachusetts Lowell, Lowell, Massachusetts, United States
Show AbstractPolymer substrates with a built-in capability for alignment of nanometer-sized objects are of interest for the development, performance, and large-scale production of robust, flexible devices. We have used atomic force microscopy (AFM) in friction mode to investigate the effects of uniaxial stretching on the chain orientation and nanoscale adhesion and friction of glassy polymer substrates. Examples will be shown of the different friction responses of a commercial low-density polyethylene along and across the stretching direction, and how this friction response is altered as the strength of adhesion between the polymer and the AFM tip is deliberately changed.
9:00 PM - S9.23
Whole Field Cell Analysis Based on Diffraction through Microfabricated Polymer Grating Films.
Xiaoyu Zheng 1 , Xin Zhang 1
1 Mechanical Engineering, Boston University, Boston, Massachusetts, United States
Show AbstractConventional cell traction force measurement technique requires individually tracking the sensing units and complex computation efforts for further studying cell contractility. Requirement of high spatial resolution of cell analysis limits the number of cells studied in the whole field of view. This paper presents a robust and high throughput cell contraction analysis system based on interferometric moiré technique. We fabricated micro polydimethylsiloxane (PDMS) grating films and study the diffraction of coherent laser beams through two periodic gratings where the cells were cultured on one of them. The contrast of the fringe pattern was studied by changing the spatial gap between the reference and specimen films. Characterizations of the system were studied by mapping the iso-thermal expansion specimen films prior to cell culture. The strain distribution of the polymer films caused by cell traction forces can be conveniently contoured by the fringe patterns. Multiple cells (up to 20) can be simultaneously studied in the view field due to the magnification effect of interferometric moiré. Cell mechanics study including cell contractile forces, stress and strain distributions during normal and abnormal cell contractions can thus be conveniently analyzed. The distinct signals from moiré patterns in longitudinal and transverse directions revealed abnormal cell mechanical contractility linked to cardiovascular disease.We demonstrated utilizing the transducer to map cardiac myocytes contraction under electric stimulation and vascular smooth muscle cell contractility evolutions triggered by agonist. Given the unique properties of optical moiré techniques (i.e., its automatic displacement and strain contouring, its magnification effect for small displacements), this new approach would be an improvement over existing techniques since it can be integrated with the existing engineered substrates and provide a direct contour of cell forces and fast detection of abnormal cell contractions .
9:00 PM - S9.24
Wetting Interactions on Textured Superhydrophilic Surface.
Hyuk-Min Kwon 1 , Kripa Varanasi 1
1 Department of Mechanical Engineeirng, M.I.T., Cambridge, Massachusetts, United States
Show AbstractTexturing a solid surface is known to profoundly impact its wettability by enhancing its inherent hydrophobicity or hydrophilicity. Interest in the so-called superhydrophobic surfaces has been intense, while relatively few treatments have looked at the wetting properties of superhydrophilic surfaces. In this paper, we study the wetting behavior of water droplets on arrays of lithographically fabricated hydrophilic posts. To understand the effect of texture on droplet wetting state, texture parameters such as feature size, relative spacing, and aspect ratio are systematically varied. Unlike the case of superhydrophobic surfaces, there appears to be no metastable Cassie wetting and drops tend to readily transition to the homogeneous Wenzel wetting state. We show that surface texture can be designed for high negative capillary pressures that can result in effective contact angles as low as zero degrees as well as more rapid spreading rate of liquids. These studies provide a means to design super-wetting surfaces for a variety of heat transfer engineering applications including boiling and evaporation.
9:00 PM - S9.25
Comparison of Fatigue Behavior Between Single and Polycrystalline Silicon Investigated Using a Novel Testing Method.
Masayoshi Ishikawa 1 , Hayato Izumi 1 , Shoji Kamiya 1
1 , Nagoya Institute of Technology, Nagoya Japan
Show AbstractAlsem et al. expected that polycrystalline silicon thin films do not suffer fatigue under high vacuum condition [1]. However their experiment was stopped at 1011. It is essentially difficult in inert environment to survey whether the lifetime is far longer or forever in contrast to humid environment where fatigue lifetime is known to be far shorter. In addition, polycrystalline silicon has the strengthening effect which results in strength higher than initial strength after a certain cycles of stress is applied [2]. From the view point of issues mentioned above, this study is focused on a different approach to the fatigue behavior of single crystalline silicon thin films in inert environment by applying a novel testing method and comparison with the results of polycrystalline silicon thin films obtained in the previous study [3].The experiment was performed under N2 gas environment at 22°C on the specimens fabricated with SOI wafers. A novel "ramping fatigue test" was utilized in this study, where the amplitude of cyclic stress applied to the specimens was gradually increased with respect to the number of cycles. Therefore the effect of fatigue appears as the stress level at the event of fracture, which is reduced by the fatigue damage and therefore smaller than its initial strength. To analyze the effect of fatigue damage, the results are compared to the initial strength. The rate stress increment in the ramping fatigue test is here called ramping rate (Δσ[Pa/cycle]). Fatigue tests were performed with two different ramping rates, Δσ=280 and 80, because it is expected that more significant fatigue damage is introduced before fracture with smaller ramping rate, therefore that the fracture stress becomes smaller.The average initial strength of 2.9GPa was obtained. On the other hand, the average fracture stress levels obtained in ramping fatigue tests were 3.0GPa and 2.6GPa with two ramping rates of Δσ=280 and 80, respectively, which is higher by 3% and lower by 9% than the initial strength. In the previous study with the same environmental conditions [3], Ikeda et.al, reported that polycrystalline silicon specimens broke at 20% higher a stress level than static strength with Δσ=220, and broke at 3% higher a stress level Δσ=70. With Δσ=22, however, breakage at about 7% smaller a stress level was reported.Judging from the results above, it is expected that single crystal silicon also suffers fatigue damage under inert environment of N2 gas and that the strengthening effect is at least far smaller than that observed with polycrystalline silicon. More detailed results with statistical analysis for a quantitative description of fatigue behavior is discussed in the presentation.[1]D.H.Alsem, et al. Applied Physics Letters, 86, 041914, (2005) [2] H. Kahn, et al. Acta Mater . 54 , 667-678, (2006).[3] Y. Ikeda et al. master thesis in Nagoya institute of technology (2010).
9:00 PM - S9.26
Study of Electronic Transport Mechanisms in Au-based MEMS Contact Switches.
Yinxuan Yang 1 , Scott Hoffmann 1 , Doug Irving 2 , Angus Kingon 1
1 School of Engineering, Brown University, Providence, Rhode Island, United States, 2 Dept of Materials Science and Engineering, North Carolina State University, Raleigh, North Carolina, United States
Show AbstractMEMS contact switches are desired for switching applications in mobile devices, as they display low insertion losses and high on-off ratios. However, the desired levels of reliability have not yet been achieved. Contributing to the problem is the fact that failure mechanisms are not fully understood; there are some gaps in the understanding of issues as basic as transport across contacts; and the range of contact materials studied is relatively limited. In this study we focus on the nature of transport in single asperity contacts as the contact force is increased, and the contact undergoes a transition from tunneling to direct contact mode. A comparison is made between the behavior of Au-Au and Au alloy-based contacts. In all cases where there is exposure to the atmosphere the presence of an interfacial layer has a significant impact on the observed phenomenology. The implications for commercial switches is discussed.Funding from the Air Force and from the Office of Naval Research is gratefully acknowledged.
9:00 PM - S9.29
Solid Bridging during Pattern Collapse (Stiction) Studied on Silicon Nanoparticles.
Daniel Peter 1 , Michael Dalmer 1 , Alfred Lechner 2 , Andriy Lotnyk 3 , Lorenz Kienle 3 , Wolfgang Bensch 4
1 , Lam Research AG, Villach, Carinthia, Austria, 2 Microsystems Engineering, University of Applied Sciences Regensburg, Regensburg, Bavaria, Germany, 3 Engineering, Christian-Albrechts-Universität zu Kiel, Kiel, Schleswig-Holstein, Germany, 4 Inorganic Chemistry, Christian-Albrechts-Universität zu Kiel, Kiel, Schleswig-Holstein, Germany
Show AbstractIntroduction: Stiction in MEMS devices and Pattern Collapse in Microelectronics is a damage phenomenon attributed to surface tension forces especially in the drying phase of a wet clean step. Most vulnerable are high aspect ratio structures in close proximity which can be elastically deformed and stuck to the opposite structure by van der Waals (vdW) forces and/or hydrogen bonding. Additionally, after the collapse silica residues can accumulate around the contact site and form a solid bond which is called solid bridging. Corresponding adhesive energies are expected to be larger than hydrogen bonding and vdW forces but no characteristic energy has been reported in the literature due to the inhomogenous deposition of the residues.Nano-scale effects may enhance the solid bridging effect for microelectronic structures e.g. STI. Nanoparticles with their large surface area are used as model material to investigate this phenomenon. Experiments: Silicon nanoparticles with a nominal diameter of 50 nm purchased from Nanostructured & Amorphous Materials Inc. were used in all the experiments in order to have a comparable length scale to STI structures. Stable dispersions were created with de-ionized water (18.2 MΩ) using 10 min of ultrasonic sound (ALLPAX, Type PALSSONIC). The subsequent drying of the particles was done with evaporation and freeze drying techniques. Additionally, fresh particles (A) were aged in laboratory air for six months (B). All particles were characterized by X-ray powder diffraction (XRD) with Rietveld refinement, X-ray photo electron spectroscopy (XPS) and transmission electron microscopy (TEM). Results and Discussion: The XRD measurements of the silicon nanoparticles showed amorphous silicon oxide for the particles B. Results of Rietveld refinement followed by microstructure effect analysis demonstrate that the broadening of the peaks is caused by size effect and not due to strain. The particle size determined by the Debye–Scherrer formula in combination with Rietveld refinement is in the range between 50 and 60 nm and is confirmed by TEM analysis. The core of the particles consists of crystalline silicon with a shell of silicon oxide covering the surface. The average silicon oxide thickness was measured with XPS to be ca. 1.7 nm and more than 3 nm for the particles A and B, respectively. EDX analysis illustrates oxide accumulation at the point of contact of the particles. The amorphous oxide bridge between the particles can be up to 10 nm as measured by high resolution TEM. Thus, solid bridging can be studied with the nanoparticles and is likely for nanostructures.
9:00 PM - S9.3
Measurements of Resonance Frequency of Parylene Microspring Arrays Using Atomic Force Microscopy.
Churamani Gaire 1 , Ming He 1 , Gwo-Ching Wang 1 , Catalin Picu 2 , Toh-Ming Lu 1
1 Physics, Rensselaer Polytechnic Institute, Troy, New York, United States, 2 Mechanical Engineering, Rensselaer Polytechnic Institute, Troy, New York, United States
Show AbstractRecently there have been interests in the use of micro- or nanospring arrays as elements of MEMS (microelectromechanical systems) for optical filters [1] and pressure sensors [2] applications. One of the most important properties of MEMS element is its natural resonance frequency. In this talk we shall report our measurements of the natural resonance frequency of microspring arrays as potential elements for MEMS. Micro-or nanospring arrays can be fabricated using the oblique angle deposition technique. In this technique the spring material is deposited onto a flat Si surface using physical vapor deposition at a large oblique angle with respect to the surface normal while the substrate is rotating around the substrate normal. Due to the shadowing effect and limited mobility of deposited material, isolated islands are formed on the surface and serve as seeds for the growth of micro-or nanosprings. If a capping layer such as a Si wafer is bonded to the top of the springs, one has a mechanical vibration system with micro-or nanosprings sandwiched between two plates. In this work we will report our fabrication of microspring system using Parylene-C. The system contains several millions of Parylene-C microsprings sandwiched between two Si plates. We demonstrated the use of both contact and non-contact atomic force microscope (AFM) to measure the resonance frequency of the system. A PZT (piezoelectric transducer) is used to drive the system from the bottom plate and the AFM is used to record the displacement of the top plate. At the resonance frequency the AFM cantilever gives a large vertical amplitude of displacement. A resonance frequency of over a hundred kilohertz and a Q factor of several hundreds were observed. No difference in resonance frequency is observed between the contact and non-contact AFM measurements, indicating the coupling between the AFM cantilever and the sample is negligible. We shall also report our measurement of the spring constant (k) of the system using nanoindentation. We found that the calculated resonance frequency based on k from nanoindentation is consistent with the measured resonance frequency using AFM. Work partially supported by the NSF 0506738. References:[1] G. D. Dice, M. J. Brett, D. Wang, and J. M. Buriak, Appl. Phys. Lett. 90, 253101 (2007).[2] S. V. Kesapragada, P. Victor, O. Nalamasu, and D. Gall, Nano Letters 6, 854 (2006).
9:00 PM - S9.4
Gold in Flux-less Bonding: Noble or Not Noble.
Marco Balucani 1 , Paolo Nenzi 1 , Fabrizio Palma 1 , Hanna Bandarenka 2 , Leonid Dolgyi 2 , Aliaksandr Shapel 2
1 Electronic Department, University of Rome La Sapienza, Roma Italy, 2 , Belarussian State University of Informatics and Radioelectronics, Minsk Belarus
Show AbstractThe work will highlight the main issues related to the development of a novel compliant contacting technologyapplicable to wafer level probing and to a new electronic packaging technique. The peculiar (essential) processin this technology is a mechanical lifting procedure in which a pulling force is exerted onto the bonded jointsto release contacting wires from the substrate to form free standing vertical structures. The bonding step isobviously the critical for the successful fabrication of the multi probe unit.For this reason the detailed analysis of different bonding techniques was done exploiting solder bumpsbonding in a different ambient to electroplated nickel, gold and preformed solder pad surfaces. Padspassivation employing electroplated gold has been initially preferred and was studied thoroughly.Thermo-compression fluxless flip-chip bonding issues of gold with Sn(63)Pb solder are presented. Thediameter of the investigated electroplated solder bumps fall between 20 and 40 microns. Such values werechosen to keep gold concentration below the 3%wt threshold to prevent embrittlement. It was discovered thateven at these concentrations, gold considerably affects reliability in our application where the joints aresubject to a pulling force. Bonding on gold finishing always gave lower yield (less than 60%) than bonding onbare nickel or on preformed solder independently of the carrier gas used during the flux-less bonding (nitrogenor formic acid).SEM and EDX analysis were performed to understand the reason of such difference in yield on failing joints.All defects were discovered to fall in three categories: solder joint failure without wire release; solder jointfailure after wire release; and silicon cratering with or without wire break. It was proved that the use of gold finish on small bumps (20-40 microns in diameter), must be taken into careful consideration as well as the plating additives used and, if possible, avoided. Gold diffuse readily into molten solder causing joint embrittlement, and the contamination due to additives presented in the platingsolution causes wetting problems. Both contribute to a harmful degradation of solder mechanical characteristics; even at concentrations lower than 3% weight considered safe for SMT applications. Furthermore, it was shown that high yield bonding is possible, in this technology, by removing the SnPb soldering process, moving to flux-less gold-to-gold bonding of pillars to pads. This process also showed issue due to the plating process influencing the final yield. Gold, a noble metal, in bonding processes is shown to lose its nobility and become cause of joint failures. The results of the technique, used to release the contactingwires from the substrate with yield higher than 99,99% will be presented.
9:00 PM - S9.5
Giant Piezoresistive Variation of Metal Particles Dispersed in PDMS Matrix.
Stefano Stassi 1 2 , Giancarlo Canavese 1 , Mariangela Lombardi 1 , Andrea Guerriero 1 3 , Candido Fabrizio Pirri 1 3
1 Centre for Space Human Robotics, IIT-Italian Institute of Technology, Torino Italy, 2 Physics, Politecnico di Torino, Torino Italy, 3 Materials Science and Chemical Engineering, Politecnico di Torino, Torino Italy
Show AbstractThe piezosensitive properties of composites consisting in conductive particles dispersed in a polymeric matrix have been intensively studied during the last decade. Carbon black, graphite flakes, fibers and various metal powders such Zn, Ni, Cu, Ag and Fe used as active fillers in a polymeric matrix have attracted a great interest for a wide range of applications, i.e. electromechanical sensors, micro actuators, tactile sensors for robotics, touchable sensitive screen etc.Investigation of the piezoresistive response of a metal-polymer composite based on nickel conductive filler in a polydimetihylsiloxane (PDMS) insulating matrix for robotic tactile sensor application is presented. A variation of the electric resistance up to nine orders of magnitude was registered when a pressure was applied to the prepared samples. This unexpected and promising behavior is due to the presence of very sharp nanometric spikes on the particle surface. Charges injected in the composite will reside on the fillers, generating very large electric local field at the tips on the particles surface resulting in field assisted Fowler-Nordheim tunneling conduction with a field enhancement factor on the tips that can be as high as 1000. Without any mechanical deformation the composite presents an insulating electric behavior, even above the expected percolation threshold, because the polymer intimately coat the nickel filler, in a way that they do not come in physical contact. When subjected to a compression, the particles come closer, without touching each other, and the resistivity decreases of various orders of magnitude.Samples were prepared dispersing metal particles, having dimension in the range between 3.5 to 7 μm, in silicone rubber. In order to not affect the electrical conduction behavior, ensured by the sharp tips integrity, the composite was gently mixed. The resulting paste was degassed, poured in PMMA molds and then cured in oven at 60°C. Different composite were prepared varying the nickel/PDMS ratio between 3:1 and 5:1 by weight and the copolymer/curing agent ratio between 3:1 and 10:1.All the different composites presented a huge variation (from six to nine orders of magnitude) of electrical resistance when a pressure was applied (from 0 to 350 kPa). An increase in the variation of resistance was registered increasing the content of nickel powder respect to PDMS and enhancing the ratio copolymer/curing agent (i.e. Young’s modulus reduction). Softer materials allow larger displacement of the embedded conductive particles and consequently they assure a larger potential barrier reduction for tunneling conduction when subjected to a mechanical stress if compared with stiffer ones.Cost efficient materials, simplicity of the process, large sensibility, and harsh environment compatibility make this composite a great alternative to the traditional materials used in tactile sensing based on capacitive, magnetic and piezoelectric principles.
9:00 PM - S9.6
Nanoindentation Characterization of PECVD Silicon Nitride on Silicon Subjected to Mechanical Fatigue Loading.
Zhi Kai Huang 1 , Kuang-Shun Ou 1 , Kuo-Shen Chen 1
1 Department of Mechanical Engineering , National Cheng-Kung University, Tainan Taiwan
Show AbstractPECVD nitride films are important and common structural materials used in microsystems applications such as transduction structures, barriers, and mask layers. The mechanical properties of silicon nitride are therefore essential for overall device reliability assessment. In particular, the fatigue and interfacial properties of PECVD nitride coated silicon structure subjected to thermo-mechanical loading would influence the structural longevity of integrated circuits or MEMS actuators. As a result, it is desired to perform characterization to understand the influence of processing and operating parameters such as thermal annealing and loading levels for structural longevity concern.In this work, the fatigue properties of PECVD nitride films are characterized by using indentation and mechanical vibration techniques. PECVD nitride coated silicon cantilevers are served as the test specimens, which are initially micro-indented and then excited by a self-designed high-speed fatigue testing facility. For tests under different stress levels and fatigue cycles, the microscope observed crack propagations are recorded. Fatigue crack growths are observed for stress level above 20 MPa after 3 Million cycles. Brief fracture mechanics analysis is also performed to extract the essential variation in stress intensity factor during stress cycling. Finally, the relationship between the crack propagation rate and the variation of stress intensity factors are correlated and eventually analyzed and presented by using Paris Law for quantifying the fatigue crack growth. Meanwhile, the nanoindentation responses such as apparent hardness, Young’s modulus, and pop-in depth, of silicon nitride after different fatigue loads at different cycles are also characterized. The experimental data shows that the mechanical response varying with different fatigue conditions. For example, the pop-in locations decreases as the fatigue load increases. This implies that the microscopic interfacial structures are changed by fatigue load. A more systematic characterization program is currently performed to evaluate the effects of key operating and processing parameters, such as rapid thermal annealing (RTA) temperature, stress level, stress ratio, and fatigue cycles, on the fatigue and interfacial behaviors of PECVD silicon nitride coated silicon structure. It is believe that this work could make a significant contribution on understanding the thin film fatigue behaviour. In addition, the preliminary characterization results, as well as the established test procedure, should be very useful for related applications such as MEMS and IC reliability assessment.
9:00 PM - S9.7
Multi-scale Simulation of Au-Ni Contacts in RF-MEMS Switches.
Benjamin Gaddy 1 , Douglas Irving 1
1 Materials Science and Engineering, North Carolina State University, Raleigh, North Carolina, United States
Show AbstractOhmic radio-frequency micro-electro-mechanical systems (RF-MEMS) switches have the potential to transform the next generation of high-performance wireless communications devices. Widespread use of MEMS devices is limited by two main types of long-term degradation: stiction and increased contact resistance. Improved rules for materials selection for RF-MEMS devices requires an enhanced understanding of the "in-use" conditions. These conditions are explored with a multi-scale atomistic and continuum simulation method. The results presented will show an improved understanding of the evolution of contacting asperities via modeling the contact of an STM tip with a surface under varying conditions. These conditions include, but are not limited to: varying grain sizes; different Au-Ni alloy compositions (0, 5, 10, and 15 at% Ni) in both random solid solution and Monte-Carlo equilibrated configurations; changing applied voltages and loads. These results will serve as a first step towards the creation of design rules for materials selection for improved RF-MEMS device performance. This work is supported by ONR grant number N00014-10-1-0402.
9:00 PM - S9.8
The Role of Contaminant Layers in the Degradation of Ohmic RF-MEMs Contacts.
Christopher Freeze 1 , Douglas Irving 1
1 Materials Science and Engineering, North Carolina State University, Raleigh, North Carolina, United States
Show AbstractOhmic radio frequency micro-electrical-mechanical systems switches (RF-MEMs) are currently limited by one or two failure mechanisms. The first is stiction and the second is a rapid increase in the contact resistance outside a 1-2 Ohm target regime. Often overlooked in these mechanisms is the presence of environmental contaminant layers. While the hydrocarbons often help prevent stiction, it is unclear what role they play in the rapid rise in contact resistance. This work uses atomistic molecular dynamics and coupled atomistic and continuum simulation to study the dynamics of contact loading in the presence of contaminant layers. These layers have varying strengths of interaction with the underlying substrate. Strongly interacting adlayers composed of alkane thiol self assembled monolayers are compared with weakly physisorbed hydrocarbon contaminants. Varying loads and local thermal conditions caused by Joule heating are explored. This work is supported by ONR grant N00014-10-1-0402.
9:00 PM - S9.9
Effect of Phosphorus Doping on the Young's Modulus and Stress of Polysilicon Thin Films.
Elena Bassiachvili 1 , Patricia Nieva 1
1 , University of Waterloo, Waterloo, Ontario, Canada
Show AbstractPhosphorus doping is often used in order to alleviate the intrinsic compressive stress associated with LPCVD polysilicon and to make the material conductive. However, the effect of the dopant on the Young's modulus of the thin film is not well characterized. In this work, resonant techniques, as well as the M-test are used to extract the Young's modulus of polysilicon thin films which have undergone thermal diffusion from a spin-on source for 30 to 150 minutes. The Young's modulus is extracted using an analytical model which also takes the average stress and stress gradient into account.
Symposium Organizers
Frank W. DelRio National Institute of Standards and Technology
Maarten P. de Boer Carnegie Mellon University
Christoph Eberl Karlsruhe Institute of Technology (KIT)
Evgeni P. Gusev Qualcomm MEMS Technologies
S10: Interfaces
Session Chairs
Wednesday AM, December 01, 2010
Room 207 (Hynes)
9:45 AM - **S10.1
Understanding and Solving Tribology and Lubrication Issues of MEMS.
Seong Kim 1
1 Chemical Engineering, Pennsylvania State University, University Park, Pennsylvania, United States
Show AbstractNo matter how small the mechanical devices are, lubrication of sliding interfaces is needed. In fact, the need for lubricious interfaces becomes even larger as the physical size decreases. This is because the surface area-to-volume ratio of the physical object becomes larger at smaller scales so that surface forces such as adhesion and friction are more significant than body forces such as gravity and momentum. Microelectromechanical systems (MEMS) fall in this category. MEMS devices are usually fabricated from silicon-based materials which have poor tribological properties such as high adhesion, high friction, and low wear-resistance. We have recently demonstrated unprecedented success of MEML lubrication using alcohol vapor. The main difference of alcohol vapor phase lubrication (VPL) from other coating-based approaches is that it allows continuous replenishment of lubricant molecules from the vapor phase, rather than relying on one-time loaded coating layers. This talk will address fundamental insights of nanoscale capillary adhesion in details, demonstrate friction reduction and wear prevention effects of alcohol VPL, and explain mechanisms for lubrication effects.
10:15 AM - S10.2
Understanding Mechanisms of Degradation and Failure in Au-based MEMS Contact Switches.
Angus Kingon 1 , Zhenyin Yang 2 , Scott Hoffmann 1 , Yinxuan Wang 1 , Daniel Lichtenwalner 2 , Doug Irving 2
1 School of Engineering, Brown University, Providence, Rhode Island, United States, 2 Dept of Materials Science and Engineering, North Carolina State University, Raleigh, North Carolina, United States
Show AbstractMEMS contact switches show great promise as for application in RF and microwave circuits, due to the high isolation and low insertion losses. They become more and more competitive against solid state devices as the operational frequency increases. However, widespread adoption has been restricted due to switch degradation and failure issues, in particular where it has been desirable to use generic switch designs in a range of circuits with widely differing operating conditions. This paper synthesizes results and presents implications from an extensive investigation of Au-Au MEMS contact degradation conducted over several years. Multiscale modeling and an experimental study of real switches was first used to understand geometry and topography effects. This led to a detailed investigation of MEMS contact switches using the proxy of single asperity contacts. In this investigation we teased out conditions under which different mechanisms are dominant, including; ductile creep versus plastic deformation (with no current flowing); electronic tunneling versus ballistic transport; and material transport (from one contact surface to another) via adhesion versus field emission mechanisms. The potential for improved contact materials based upon Au alloys is demonstrated, and the importance of controlling the microstructures of these alloys is discussed.Funding from the US Air Force and the Office of Naval Research is acknowledged.
10:30 AM - S10.3
Fabrication and Characterization of Two Compliant Electrical Contacts for MEMS: Gallium Microdroplets and Carbon Nanotube Turfs.
Yoonkap Kim 1 , A. Qiu 1 , J. Reid 1 , R. Johnson 1 , D. Bahr 1
1 Mechanical and Materials Engineering, Washington State University, Pullman, Washington, United States
Show AbstractMicroscale electrical contact switches have been fabricated using Ga and carbon nanotube arrays (CNTs) in place of more traditional solid-solid contact because of the ability to minimize damage from switching and the ability to make good contacts for electrical conductivity. Ga has been electroplated into droplets on the order of 50 microns in radius on single crystal Si to create a MEMS switch that can be annealed to recover its original electrical properties after mechanical damage. CNTs were grown on Si substrates, coated with a thin Au layer, and transferred to other Si or Kapton substrates through thermocompression bonding. In the case of the Ga switch, repeated point contact during switching led to increase the resistance of the contacts due to plastic deformation, but the resistance recovered after a thermal reflow process at 120 °C. Longer term and larger area contacts were used to measure the contact behavior under switching conditions of up to 200 A/cm2. At moderate cycling conditions (on the order of 200 cycles) the adhesion began to significantly degrade the switch. The oxidation behavior of the Ga droplets was characterized for thermal reflow, suggesting a passivating 30 nm oxide forms at 100 °C. The oxide formed by the Ga is thin and fragile as demonstrated by its use in a switch. The Ga droplets were examined with electrical contact resistance nanoindentation and the loads at fracture and the onset of electrical contact were identified. CNT turfs were also tested for making pattered electrical contacts; turfs of lateral dimensions similar to the Ga droplets were tested with ECR and as macroscopic contacts, and shown to be able to carry similar current densities. The results will be compared between the two systems, and benefits and challenges of each will be highlighted for creating compliant electrical switches and contacts.
10:45 AM - S10.4
Atomistic Simulation of the ``in use" Conditions of an Ohmic RF-MEMs Contact.
Benjamin Gaddy 1 , Xiaoyin Ji 1 , Christopher Freeze 1 , Douglas Irving 1
1 Materials Science and Engineering, North Carolina State University, Raleigh, North Carolina, United States
Show AbstractOhmic radio frequency MEMS switches (RF-MEMS) could be important components to next generation wireless communication devices but are currently limited by two degradation mechanisms. The first is stiction and the second is an appreciable rise in the contact resistance outside the targeted 1-2 Ohm regime. In this talk we will present results of our efforts to simulate the dynamics at the contact with atomic resolution by use of a multi-scale methodology that couples numerical solutions of electrical and thermal transport to an underlying MD simulation. Using this method we have begun to explore the dynamics of poly-crystalline Au and Au-Ni alloys of various compositions. Results demonstrating how response changes as a function of load and applied current will be presented. We will also present how the presence of contaminants alters the properties of the contact. Contaminant layers will vary in degree of strength of interaction with the underlying poly-crystalline substrate. For example strongly interacting alkane thiols will be compared to the dynamics of weakly adsorbed hydrocarbon films. These results as a whole begin to provide insight into contact dynamics and assist in the design of new contact materials for these crucial devices. This work is supported by ONR grant N00014-10-1-0402.
11:00 AM - S10: Interf
BREAK
S11: Microfluidics
Session Chairs
Wednesday PM, December 01, 2010
Room 207 (Hynes)
11:30 AM - **S11.1
An Automated System for Performing Multiplexed Immunohistochemistry.
Alex Corwin 1 , Kashan Shaikh 1 , Denise Hollman-Hewgley 2 , Jun Xie 1 , Robert Filkins 4 , Fiona Ginty 3
1 Microsystem and Microfluidics Laboratory, GE Global Research, Niskayuna, New York, United States, 2 Biochemistry and Biological Engineering Laboratory, GE Global Research, Niskayuna, New York, United States, 4 Material Systems Technologies, GE Global Research, Niskayuna, New York, United States, 3 Computational Biology and Biostatistics Laboratory, GE Global Research, Niskayuna, New York, United States
Show AbstractThe future of cancer diagnosis and treatment is held in better understanding the action of disease at the cellular level. Immunohistochemistry, a technique for optically labeling proteins in cells and tissue with fluorescent reporters, is an important tool for investigating disease at this molecular level. Howver, due to spectral overlap, standard fluorescent immunohistochemical methods elimit detection to six proteins, or biomarkers, in a single tissue sample. A novel in situ immunofluorescent multiplexing method for formalin fixed paraffin embedded tissue has been developed at GE that allows multiple cycles consisting of staining, imaging, and chemical inactivation of the fluorphore . Through this multiplexing technique, we can measure over 25 proteins in a single sample, without disruption of tissue architecture. Images are registered using DAPI (which is minimally effected by the inactivation process and imaged in each staining round). Using markers for membrane, cytoplasm and nucleus, tumor and other regions, biomarkers are then quantified on an individual cell basis. The results comprise of a high dimensional data set consisting of spatially resolved biomarker intensity values, which can then be analyzed through a variety of statistical techniques.Key to this multiplexing technique is the need to image a sample, process it (staining or chemical inactivation) over repeated cycles. Automation greatly simplifies this workflow, and allows a sample to be processed in place under the microscope with minimal handling. We have built an automated stainer / imager system based on a microfluidic flow cell. The flow cell is mounted rigidly to the stage of an Olympus IX-81 microscope. Fluidic control is accomplished via a control box containing syringe pumps and fluidic valves. Imaging is performed using with an image based autofocusing routine based on the maximization of the Brenner gradient of an image.Using this system we have taken a sample through more than 8 rounds of staining consisting of 15 biomarkers. Quantitative measures of the data show we have achieved satisfactory staining, while greatly reducing both the time and the effort that would have been needed to perform this same experiment manually.
12:00 PM - S11.2
Biodegradable Microfluidic Scaffolds with Tunable Degradation Properties from Amino Alcohol-based Poly(ester amide) Elastomers.
Jane Wang 1 2 3 , Tatiana Kniazeva 2 , Carly Campbell 2 , Robert Langer 3 4 , Jeffrey Borenstein 2
1 Material Science and Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States, 2 Biomedical Engineering Center, Charles Stark Draper Laboratory, Cambridge, Massachusetts, United States, 3 Program of Polymer Science and Technology, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States, 4 Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States
Show AbstractBiodegradable polymers with high mechanical strength, flexibility and optical transparency, optimal degradation properties and biocompatibility are critical to the success of tissue engineered devices and drug delivery systems. In this work, microfluidic devices have been fabricated from elastomeric scaffolds with tunable degradation properties for applications in tissue engineering and regenerative medicine. Most biodegradable polymers suffer from short half life resulting from rapid and poorly controlled degradation upon implantation, exceedingly high stiffness, and limited compatibility with chemical functionalization. Here we report the first microfluidic devices constructed from a recently developed class of biodegradable elastomeric poly(ester amide)s, poly(1,3-diamino-2-hydroxypropane-co-polyol sebacate)s (APS), showing a much longer and highly tunable in vivo degradation half-life comparing to many other commonly used biodegradable polymers. The device is molded in a similar approach to that reported previously for conventional biodegradable polymers, and the bonded microfluidic channels are shown to be capable of supporting physiologic levels of flow and pressure. The device has been tested for degradation rate and gas permeation properties in order to predict performance in the implantation environment. This device is high resolution and fully biodegradable; the fabrication process is fast, inexpensive, reproducible, and scalable, making the approach ideal for both rapid prototyping and manufacturing of tissue engineering scaffolds and vasculature and tissue and organ replacements.
12:15 PM - S11.3
Microfluidic Methods for Producing Millimeter-size Fuel Capsules for Inertial Fusion.
David Harding 1 3 , Thomas Jones 4 1 , Weiqiang Wang 5 , Zongmin Bei 5 , Matt Moynihan 3 , Shaw Chen 3 1 , Roger Gram 1 , Gregory Randall 2
1 Laboratory for Laser Energetics, University of Rochester, Rochester, New York, United States, 3 Department of Chemical Engineering, University of Rochester, Rochester, New York, United States, 4 Department of Electrical and Computer Engineering, University of Rochester, Rochester, New York, United States, 5 Material Science Department, University of Rochester, Rochester, New York, United States, 2 , General Atomics, San Diego, California, United States
Show AbstractMicrofluidic processes are used to manipulate comparatively large volumes of oil, water, and cryogenic liquids (5 to 30uL) to form fuel capsules for the inertial confinement fusion program. The process involves dispensing separate droplets of oil and water of a precise size using electric fields. These droplets are combined into an emulsion that is, in turn, polymerized to form concentric plastic shells: a critical part of the process is the use of an electric field and the resulting dielectrophoretic force to center the inner droplet within the outer droplet to achieve a uniformly thick-shell wall. The final stage in the process is using the same “lab-on-chip” concept and dielectrophoretic forces at 18 K to manipulate the liquid-deuterium fuel into the plastic shell. We describe the performance of the current microfluidic process for manufacturing polymer shells and include a computational fluid-dynamic model that demonstrates the sensitivity of the process to critical fluid and geometric parameters. The microfluidic scheme used is based upon the “lab-on-a-chip” concept for mass producing fuel capsules for use in an inertial fusion process. The goal is to produce 0.5 million 6-mm-diam foam shells per day to very tight specifications: a diameter variation less than +/-0.05 mm; the 0.35-mm-thick foam wall must be uniformly thick so that the inner surface is concentric within the outer surface to within +/-0.02 mm; and foam is a low-density (0.1 gm/cm3) hydrocarbon with a pore size less than 200nm.This work was supported by the U.S. Department of Energy Office of Inertial Confinement Fusion under Cooperative Agreement No. DE-FC52-08NA28302, the University of Rochester, and the New York State Energy Research and Development Authority. The support of DOE does not constitute an endorsement by DOE of the views expressed in this article.
12:30 PM - S11.4
Traveling Wave-Induced Aerodynamic Propulsive Force using Active Control of the Dynamic Shape of Piezoelectrically-Deformed Plastic Substrates.
Noah Jafferis 1 2 , James Sturm 1 2
1 Electrical Engineering, Princeton University, Princeton, New Jersey, United States, 2 Princeton Institute for the Science and Technology of Materials (PRISM), Princeton University, Princeton, New Jersey, United States
Show AbstractThe local control of the shape of a thin substrate has many applications, including mechanical actuators [1], speaker arrays, adaptive optics, etc. Such substrate deformation is typically achieved by applying voltages across a piezoelectric element, dielectric elastomer [2], or muscular thin film [3]. For accurate control of the shape, the ability to monitor the shape in real time with integrated deformation sensors is desirable. This is especially true for rapidly changing shapes, due to the effects of mechanical resonances, non-linearities, etc.Theoretically, a sheet deformed vs. time in a traveling wave shape has been predicted to produce a force in the direction opposite to the wave propagation. The resulting forward motion of the sheet near the ground should cause it to lift up as it moves – so called “flying paper” [4]. In this work we use integrated piezoelectric actuators and sensors to demonstrate this aerodynamic propulsive force produced by traveling mechanical waves in a thin film substrate in air for the first time.To produce the traveling wave necessary for the “flying paper” effect, actuators were formed along a linear strip of polyvinylidene fluoride (PVDF) sheets (28um thick), and a deformation sensor was integrated next to each actuator. Each sensor voltage signal measures the average curvature in that region, from which the time-varying shape of the sheet may be inferred. Without integrated sensors and feedback, neither the desired traveling wave profile nor any observable propulsive force resulted from the application, to the actuators, of the control signals that had been predicted to produce, under ideal conditions, a traveling wave shape. However, with the sensor data, a feedback matrix was constructed to actively adjust the actual input signals (which can require frequency components up to five times the fundamental). With this approach traveling waves have been produced with amplitude up to 50um in the 6-1600Hz range, confirmed with a high speed camera. For a sheet suspended 1mm above the ground, we observe a propulsive force of ~1uN, corresponding to a displacement of 50-65um, within 20% of that expected by first principles. As expected, the direction of force switches when the direction of the traveling wave is switched. This work demonstrates the advantages of using integrated sensors to control the dynamic shape of thin plastic sheets deformed by piezoelectric elements, and confirms the physical basis for “flying paper.”[1] T. Safakcan et al., Meso-scale Piezoelectric Gripper with High Dexterity. Japanese Journal of Applied Physics 48, 044501 (2009)[2] R. Pelrine et al., High-Speed Electrically Actuated Elastomers with Strain Greater Than 100%. Science 287, 836-839 (2000)[3] A. Feinberg et al., Muscular Thin Films for Building Actuators and Powering Devices. Science 7, 1366-1370 (2007)[4] M. Argentina et al., Settling and Swimming of Flexible Fluid Lubricated Foils. Physical Review Letters 99, 224503 (2007)
12:45 PM - S11.5
Ion-electron Transducing Electrodes with Electrochemically Active Material for Sample-friendly Electroosmotic Pumps.
Per Erlandsson 1 , Nathaniel Robinson 1
1 Physics, Chemistry and Biology, Linköping University, Linköping, Östergötland, Sweden
Show AbstractMicrofluidic LOC devices are in a position to revolutionize the healthcare industry and help solve many of the challenges linked with an aging population in the western world. In order for such systems to be used effectively in point-of-care diagnostics, or in areas with limited healthcare infrastructure, there is a need to miniaturize devices and make them function independently of large external equipment. Improved electroosmotic pumps can increase the potential of many microfluidic technologies such as micro cell culture chambers and lab-on-a-chip (LOC) systems.The pump is one of the critical components to be integrated into microfluidic devices, as it moves liquids between reactors, separators and detectors. Electroosmotic (EO) pumps are a natural choice as they can easily be integrated into chips and generate a controlled plug-like flow. However, electrochemical reactions are required to maintain flow, often hydrolysis, which generates O2 and H2 bubbles and/or H+ and OH- at the electrodes. In large systems the impact of hydrolysis can be mitigated by using large volumes at the electrodes, but when scaling EO pumps and electrophoretic separations down to LOC sizes, these reactions can change the pH which is unacceptable for sensitive biomolecules.Using conductive and electrochemically switchable materials as electrodes, we have demonstrated a hydrolysis-free EO pump that drives aqueous samples without generating gas or changing the solution’s pH. By utilizing switchable electrodes as ion-electron transducers, charge can be transferred at potentials well within the electrochemical stability window of water, making these pumps ideally suited for LOC applications.
S12: Energy Technologies
Session Chairs
Wednesday PM, December 01, 2010
Room 207 (Hynes)
2:30 PM - **S12.1
Energizing and Powering Microsystems.
Gabriel Rincon-Mora 1
1 , Georgia Institute of Technology, Atlanta, Georgia, United States
Show AbstractPortable, lightweight, long-lasting electronics is filling a growing need in military (to lighten and extend reconnaissance mission work, for example), space exploration (for remote sensors), biomedical (for monitoring, prognosis, treatment, etc.), and consumer electronics (such as disposable and rechargeable everyday products). Conforming to microscale dimensions means energy and power supplies, conditioning and processing microelectronics, sensors, wireless transceivers, and other constituent subsystems must synergistically share a common miniaturized platform. Integrating and managing microsources, however, present a myriad of diverse and interdependent mechanical, chemical, and electrical challenges. One pivotal constrain is ultra-small systems cannot store the energy required to sustain practical life times, which is why energy-dense fuel cells and energy harvesters have gained so much attention. The aim of this talk is to illustrate how to energize and power microsystems such as wireless microsensors by reviewing their system requirements, describing the state of the art in miniaturized sources, and presenting fuel-cell and energy-harvesting supply system-on-chip (SoC) and system-in-package (SiP) strategies currently under investigation.
3:00 PM - S12.2
Materials Development for an Energy Harvesting Microsystem Based on a Rankine Microturbine.
Hassan Shahriar 1 , Mokhtar Liamini 2 , Luc Frechette 2 , Srikar Vengallatore 1
1 Mechanical Engineering, McGill University, Montreal, Quebec, Canada, 2 Mechanical Engineering, Universite de Sherbrooke, Sherbrooke, Quebec, Canada
Show AbstractThe Rankine Microturbine is a microelectromechanical system being developed for generating mechanical and electrical power from waste heat, such as from automobile exhaust gases. The goal is to generate 1 W of power using a microsystem that is about 1 cubic centimetre in volume. The design of this device faces the difficult challenges of creating structures that rotate at high-speeds (1 million rpm), sustain large internal pressures (3 MPa) and temperature gradients (100 °C/mm), and micromachining complex geometries to form millimeter-sized components of ceramic and metallic materials with micrometer tolerances. Developing advanced materials to meet these stringent mechanical and thermal loads is essential for designing devices with high performance and sufficient reliability. As a first step towards this goal, performance metrics were identified within the context of the relevant thermodynamic cycle and application. Next, the primary constraints and requirements were defined in terms of the operating conditions (pressure, temperature, and operating speeds), critical shapes and feature sizes, and microfabrication tolerances. Third, a systematic analysis of materials selection was undertaken by developing materials indices, and this exercise revealed a qualitative trade-off between performance, reliability, and manufacturability. Specifically, materials with low thermal conductivity (silica and zirconia) offer the best performance, but low reliability and moderate machinability; titanium alloys and nickel alloys offer better reliability at the cost of reduced performance; and Si and SiC can be readily micromachined but lead to very low performance. Subsequently, this selection was refined by developing detailed thermomechanical models for performance and reliability. Parametric numerical studies were conducted to develop scaling relationships for stresses and deformation at operating conditions. The results of these studies have been used to identify viable device configurations and an optimal set of materials for the various components (evaporator, condenser, pump, rotor, and static insulator). These details, along with approaches for micromachining, integration and packaging, will be presented.
3:15 PM - S12.3
A Picowatt Energy Harvester.
Joe Evans 1 , Johannes Smits 2 , Carl Montross 1 , Gerald Salazar 1
1 , Radiant Technologies, Inc., Albuquerque, New Mexico, United States, 2 , Scaldix, Middleburg Netherlands
Show AbstractThe authors describe an energy harvester circuit fabricated with integrated thin ferroelectric film capacitors on a silicon substrate. The harvesting mechanism is a folded double-beam cantilever with a proof mass at its end point. Interdigitated electrode capacitors are located at the three points on the folded cantilever that are expected to experience maximum bending moment and should produce up to 5V as a function of external vibration. The die has the dimensions of 1.6mm on a side and is designed to be mounted in a TO-18 package transistor-style package. Due to its small size, the self-contained piezoelectric MEMs device should produce 10 picowatts or less in a high vibration environment.
3:30 PM - S12.4
MEMS Based Measurements of Energy Materials Using Electronic Actuation and Detection.
Will Osborn 1
1 Ceramics Division, National Institute of Standards and Technology, Gaithersburg, Maryland, United States
Show AbstractExploiting the mechanical behavior of cantilevers is a staple method for microelectrical systems- (MEMS-)based measurement techniques. We present a MEMS platform currently under development that extends this method to a two-dimensional array of cantilevers designed to make gravimetric measurements of energy materials for two key purposes: hydrogen storage and carbon sequestration. A newly developed all electronic actuation and detection scheme allows the devices to be easily operated in a pressure vessel and to switch between addressable cantilevers at electronic speeds. Given the goal of high-throughput measurements for combinatorial libraries, several engineering challenges will be addressed: (1) design optimization to separate the mechanical behavior of the device structure from the candidate materials, (2) fabrication of a piezoelectric actuator with minimal adhesion layers, and (3) hardware and algorithm development of the electronic actuation and detection system. Modeling results from the structural optimization will be presented along with early experimental results from of the electronic detection technique. Finally, lead zirconate titanate (PZT) deposition without the typical Si/SiO2/Ti/Pt stack will be described.
3:45 PM - S12.5
Piezoelectric and Dielectric Studies of Epitaxial Pb(Zr1-Xtix)O3 Thin Film for Energy Harvesting Devices.
Xin Wan 1 2 , Ruud Steenwelle 1 , Minh Nguyen 1 , Koray Karakaya 2 , Matthijn Dekkers 1 , Dave Blank 1 , Rob Van Schaijk 2 , Guus Rijnders 1
1 MESA+ Institute for Nanotechnology, University of Twente, Enschede Netherlands, 2 , IMEC/Holst Centre, Eindhoven Netherlands
Show AbstractPb(Zr1-xTix)O3 (PZT) piezoelectric thin films offer a number of advantages in energy harvesting systems, due to the high power output density resulting from their high piezoelectric coefficient and relatively low epsilon. In this study, PZT was chosen for its prime piezoelectric properties and the possibilities to obtain well-controlled growth by pulsed laser deposition (PLD). Furthermore, epitaxial PZT films with controlled orientations can be achieved on silicon substrates, enabling fabrication of piezo energy harvesting devices based on MEMS technology. The energy harvesting device performance is given by the figure of merit (e312/ ε), indicating that a high piezoelectric coefficient (e31) combined with a low dielectric constant (ε) is vital to achieve a higher piezoelectric voltage. Pb(Zr1-xTix)O3 thin films, with a composition x=0.8, 0.6, 0.48 and 0.4, were studied and compared with each other. A variety of e31, and ε values was measured, and an optimal composition is found to achieve a large figure of merit value. More important, a shift of morphotropic phase boundary (MPB) is observed, due to the residual strain resulting from the different thermal expansion coefficients between PZT thin films and silicon substrates. These results show that, for epitaxial PZT thin films, the composition of the PZT should be optimized to compensate the strain state. Besides the characterization of material properties, energy harvesting devices have been fabricated and their properties will be discussed.
4:00 PM - S12: Energy
BREAK
S13: Devices and Applications
Session Chairs
Wednesday PM, December 01, 2010
Room 207 (Hynes)
4:30 PM - S13.1
Self-assembled Microscopic Surface Walkers.
Alfredo Alexander-Katz 1
1 Materials Science and Engineering, MIT, Cambridge, Massachusetts, United States
Show AbstractBiological flows at the microscopic scale are important for the transport of nutrients, locomotion, and differentiation. In this respect, nature has devised several mechanisms to transport materials in the microscopic realm, such as flagellar rotation and ciliary beating. Here, we present a novel approach for creating controlled surface-induced flows inspired by a ubiquitous biological system, cilia. Our design is based on a collection of self-assembled colloidal rotors that ``walk'' along surfaces in the presence of a rotating magnetic field. Both experimental and theoretical studies show that the velocity, direction, and particular flow characteristics can be controlled by direct manipulation of the magnetic field that drives the motion of these "surface walking" assemblies. These assemblies are dynamic since they exist only in the presence of an external magnetic field, and their shape is strongly controlled by the magnetic field (e.g. they undergo drag-induced fragmentation transitions at high frequencies). This system represents a completely reversible self-assembled active material that can be used on demand and does not require complex fabrication techniques. Furthermore, it offers a simple and versatile approach for designing novel microfluidic devices as well as for studying fundamental questions in cooperative driven motion and transport at the microscopic level.
4:45 PM - S13.2
Self-folding Micropatterned Polymeric Polyhedra.
Anum Azam 1 , Jatinder Randhawa 1 , Shivendra Pandey 1 , Mustapha Jamal 1 , David Gracias 1
1 Chemical and Biomolecular Engineering, Johns Hopkins University, Baltimore, Maryland, United States
Show AbstractWe demonstrate self-folding of precisely patterned, optically transparent, all-polymeric polyhedra. Polyhedra self-assembled from lithographically patterned templates composed of SU8 panels and polymeric hinges. The strategy allowed for the fabrication of polyhedra with variable shapes, sizes and precisely defined patterns / porosities in all three dimensions. In addition to mechanistic and versatility aspects of the self-folding process, we also describe applications in cell and organisms applications and as building blocks for creating array based aggregates. We provide proof-of-concept for the use of these polymeric containers as encapsulants for beads, chemicals, cells and microorganisms and also compare accelerated hinge degradation rates in alkaline solutions of varying pH. These optically transparent containers resemble 3D micro-Petri dishes and can be utilized to sustain, monitor and deliver living biological components. Precise patterning of these 3D polymeric units with heterogeneous materials are also important in other applications such as enabling periodic 3D electronic and optical devices and materials.
5:00 PM - S13.3
Refractive Index Memory Effect of Ferroelectric Materials by Domain Control.
Kazuhiko Inoue 1 , Takeshi Morita 1
1 Graduate School of Frontier Sciences, The University of Tokyo, 5-1-5 Kashiwa-shi, Chiba Japan
Show AbstractLead lanthanum zirconate titanate (PLZT) has excellent transparency in a wavelength range from visible to infrared, and show controllable optical properties with an applied electric voltage. Until now, using the electro-optical property, the device applications, such as an optical switch, a scanner and a shutter have been studied. In these devices, to keep an optical property, a continuous voltage supply is required. On the contrary, we propose optical properties memory effect with a pulsed voltage. Therefore the continuous voltage supply isn’t required. The optical properties memory effect is based on its hysteresis characteristics in the PLZT. When the sufficiently-large voltage is applied to the PLZT, the relationship between refractive index and voltage becomes the bilaterally shaped symmetric. The refractive index property becomes to be minimum value at the coercive electrical field. With much larger electrical field, by alignment of all of polarization, the refractive index value increased. It is considered there is a correlation between the change of refractive index and domain conditions. Hence, by controlling the polarized and the depolarized condisions, the refractive index memory effect can be realized. The key-point is to control the domain conditions with asymmetric voltage. In the positive direction, the sufficiently-large voltage is applied to align the all of the polarizations, and large refractive index is memorized after removing the electrical field. While in the negative direction, the amplitude of voltage is adjusted to the coercive electrical field for the depolaried conditions, and the minimum refractive index can be kept after removig the electrical field. Therefore, PLZT can have two refractive index state depending on the voltage amplitude.When a laser penetrates PLZT with a prism-shaped top electrode, it goes straight in the absence of an electrical field. While in the presence of an electrical field, the optical path is refracted due to the modification of the refractive index under the top electrode of PLZT. The laser (633 nm) went through PLZT, and voltage was supplied to the prism-shaped top electrode. A two-dimensional position sensor (DSP) was set at 200mm from PLZT to detect the change of the optical path. Measuring the optical path at the DSP to the refractive index change was estimated. The refractive index of PLZT was controlled using the pulse electrical field. The applied electrical field was 14kV/cm in the positive direction, while -6kV/cm in negative direction, and pulse width was 1000ms. Without external electrical field, the refractive index maintained each stable value. As shown here, the principle of refractive index memory effect was proposed, and the pulse operation was demonstrated successfully in this study.
5:15 PM - S13.4
Electrostatic Micromanipulation of a Conductive Particle by a Single Probe with Consideration of an Error in the Evaluated Mass.
Kenji Sawai 1 , Shigeki Saito 1
1 , Tokyo Institute of Technology, Tokyo Japan
Show AbstractRecently, micromanipulation techniques for handling a conductive microparticle have been in demand. Such techniques should also be used in future IC packaging technologies. For instance, on a ball-grid-array (BGA) which is used for high-density IC packaging, the yield ratio of production will be improved if the defects in the soldering can be repaired by manipulating an individual micro solder balls. In micromanipulation, adhesional force between a manipulator and microparticles is dominant since it is proportional to the first power of the object size, whereas the gravitational force is proportional to the third power. Therefore, a repulsive force must be generated to detach the microparticles from the manipulator. Electrostatic micromanipulation with a single probe is a promising technique for such manipulation. While the feasibility of the technique has been proved experimentally, the success rate of the manipulation was 25% and further improvements are required. To enhance the success rate and realize highly reliable electrostatic micromanipulation, this paper proposes an improved design of a voltage sequence which is applied to deposit a microparticle onto a substrate plate. It was found through investigation that the error in the evaluated mass of a microparticle must be considered in order to improve the success rate of the manipulation. Behavior of a microparticle during the electrostatic micromanipulation is calculated by a boundary element method, and the influence of the error is discussed. An improved design of the applied voltage sequence that can tolerate an error in the evaluated mass is described. Moreover, the effectiveness of the newly designed voltage sequence in the electrostatic micromanipulation is experimentally shown.
5:30 PM - S13.5
Deposition of Liquid in Desired Micro-pattern on a Surface Using Vapor-droplet Phase Transition for Mass Alignment of Micro-parts.
Takashi Okamoto 1 , Keita Nakajo 1 , Shigeki Saito 1
1 , Tokyo Institute of Technology, Tokyo Japan
Show AbstractThis study reports a new technique of depositing liquid water on a flat surface in a desired micro-pattern using vapor-droplet phase transition in air. The patterned deposition is demanded for mass alignment of micro-parts by capillary force. The surface of the sample for the deposition is coated with patterned hydrophobic monolayer on hydrophilic substrate. In our experiment, octadecyl-trichloro-silane (OTS) is used for hydrophobic monolayer; silicon dioxide is used for hydrophilic substrate. The monolayer is formed in a grid pattern where squares of 50 micrometer on a side are placed as hydrophilic areas with 50 micrometer spacing. After the water on/above the sample surface is repeatedly condensed and evaporated from/to air by heating and cooling the sample using a Peltier device, the liquid on the surface is successfully obtained in the desired pattern. The demonstration shows the feasibility of this technique for industrial application in batch process of aligning micro-parts such as MEMS parts or solder balls for Ball-Grid-Array (BGA). In addition, the condition for speeding up the process is discussed for further improvement.
5:45 PM - S13.6
3D Photonic Crystal Microcantilever.
Zhenting Dai 1 , XIndi Hu 2 , Huigang Zhang 2 , Taewon Kim 2 , Paul Braun 2 , William King 1 2
1 Mechanical Science and Engineering, University of Illinois Urbana-Champaign, Urbana, Illinois, United States, 2 Materials Science and Engineering, University of Illinois Urbana-Champaign, Urbana, Illinois, United States
Show AbstractWe report the first micromechanical device fabricated from a three-dimensional (3D) photonic crystal (PC), realized as a nickel micro cantilever having an inverse opal structure. Microcantilevers have been extensively used in scanning probe microscopy, chemical sensing, and many other applications. Previous publications have reported two-dimensional (2D) photonic crystals (PCs) integrated into microcantilevers [1]. Here we present a microcantilever made of three-dimensional (3D) PCs. The fabrication process begins with a double-sided polished Si wafer which has 300 nm of SiO2, 10 nm Cr, and 50 nm Au from bottom to top. The microcantilever geometry, 350 micrometers long and 150 micrometers wide, is patterned onto the metal layer using photolithography and etching techniques. Using a self-assembly process, polystyrene spheres with a diameter of 1.8 micrometers are assembled onto the wafer in a cubic lattice. We electroplate nickel through the lattice and then sacrifice the polymer spheres, producing an “inverse opal” nickel foam having a regular pore structure [2]. The porosity can be controlled with electrodeposition or electropolish steps, such that it can be controlled over the range 74% to 95%. The cantilevers are released by inductively-coupled plasma etching of the silicon handle wafer. Fourier transform infrared (FTIR) spectroscopy measurements show that the released cantilevers are very black throughout the mid-infrared spectral region, which is expected for this type of 3D photonic crystal structure [2] but otherwise difficult to achieve in metals. The mechanical properties of these microcantilevers were extensively studied in an Asylum atomic force microscope and Agilent nanoindenter. A typical cantilever has the following parameters: porosity of 78%, overall cantilever thickness of 12 micrometers which corresponds to about 6 lattice periods, cantilever resonant frequency of 45 kHz and a spring constant of 35 N/m. These cantilevers exhibit extremely sharp resonant peaks, with a quality factor up to 700, well more than an order-of-magnitude higher than commercial silicon cantilevers typically used in scanning probe microscopy. The very high quality factor can be attributed to the regular lattice of the inverse opal structure. The cantilever’s surface area is about 20 times greater than a solid cantilever of similar size and shape, making it a good candidate for chemical sensing. While this cantilever is the first micromechanical structure fabricated from a 3D photonic crystal, the fabrication of other 3D PC micromechanical structures is possible and could impact a number of applications, such as photonics or electromagnetics.[1] “Microcantilevers with Nanochannels”, Pyung-Soo Lee et. al., Advanced Materials, 20 (1732-1737), 2008.[2] “Filling Fraction Dependent Properties of Inverse Opal Metallic Photonic Crystals”, Xindi Yu et. al., Advanced Materials, 19 (1689-1672), 2007.