AVS1997 Session MM+SS+NS-WeA: Surface Properties/Microanalysis

Wednesday, October 22, 1997 2:00 PM in Room C3/4

Wednesday Afternoon

Start	Invited?	Item
2:00 PM	Invited	MM+SS+NS-WeA-1 Antistiction Processes in Surface Micromachines R. Maboudian (University of California, Berkeley) Adhesion, friction, and wear are prevalent problems in a majority of MEMS devices. Since gravity is negligible at the dimension of most microstructures, understanding of surface interactions in MEMS is of paramount importance for controlling stiction phenomena. This presentation will discuss the use of electrostatically actuated cantilever beam arrays in conjunction with other surface characterization techniques, such as X-ray photoelectron spectroscopy and atomic force microscopy, to measure the surface forces present between polycrystalline silicon surfaces and to manipulate them by utilizing various surface treatments.
2:40 PM		MM+SS+NS-WeA-3 Friction and Durability of Silicon Surfaces Treated with Silane-Based Coupling Agents¹ M.T. Dugger, D.Cowell Senft, G.C. Nelson (Sandia National Laboratories) Reliable production and operation of silicon surface micromachined devices having moving parts requires the ability to maintain the designed separation, and relative motion, between surfaces. We are investigating the use of silane-based coupling agents as molecular scale lubricants for micromachines. The formation of coupling agent films on silicon depends upon reaction with both residual adsorbed water and hydroxyl sites on the silicon surface. The resulting mechanical behavior of the film during sliding contact will depend on the relative amounts of these reactions. We report the use of scanning probe techniques to investigate the molecular scale friction of coupling agent films as a function of substrate surface preparation, structure of the coupling agent molecules, and mechanical contact conditions. The scanning probe data shows that the friction of coupling agent treated surfaces changes in response to sliding contact. Under the conditions investigated in this work, friction is reduced after rubbing the treated surface. The degree of friction change due to rubbing is determined by the applied force between the probe and the surface, as well as the duration of contact. Analysis of silicon samples with patterned ODTS permits direct comparison of the friction response of uncoated versus treated surfaces. We have also investigated coupling agent performance using specially-designed polycrystalline silicon test structures in which sliding contact can be produced under well controlled contact conditions. For both the scanning probe and test structure measurements, time-of-flight SIMS is used to determine film coverage and spatially resolve changes in the film in response to mechanical stress at contact points. This work was supported by the United States Department of Energy under contract DE-AC04-94AL85000. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy.
3:00 PM		MM+SS+NS-WeA-4 Diamond Surface Modification for MEMS P.E. Pehrsson (Naval Research Laboratory); T. Gilbert, D.N. Leonard, M.-S. Chen (GeoCenters, Inc.); J.M. Calvert (Naval Research Laboratory (Currently at Shipley, Inc.)) Si is the subject of most of the current work on surface modification for MEMS applications because of it's physical properties, widespread availability, and the suite of well-understood techniques available for manipulation of it's surface. We are investigating diamond and other materials for MEMS-related applications, in order to take advantage of their often superior material properties. Diamond is the hardest and most thermally conductive material known, has a high lubricity and low thermal expansion coefficient, is radiation hard and a cold cathode field emitter. We have studied the manipulation of diamond surface chemistry using high resolution electron energy loss spectroscopy (HREELS), electron loss spectroscopy (ELS), a Kelvin probe, XPS, and AES. Hydrogenated diamond surfaces were oxidized using both UHV and non-UHV techniques (ex-situ thermal or plasma oxidation). The various oxidation techniques yielded significant differences between the surface oxygen moieties and morphology. Organosilane films were then chemisorbed to the oxidized diamond surfaces, modifying the wettability and other properties. Some organosilanes were then terminated with a Pd catalyst and subjected to electroless plating, in order to produce adherent, patterned metallizations of Ni or Co. Other useful techniques such as etching and thin film separation of CVD films from diamond substrates allow manipulation of diamond for device definition, making it an alternative material for integrated MEMS applications.
3:20 PM		MM+SS+NS-WeA-5 Contact Resistance Performance of Electrostatically Actuated Microswitches S. Majumder, N.E. McGruer, P.M. Zavracky, R.H. Morrison, G.G. Adams, J. Krim (Northeastern University) Electrostatically actuated micromechanical switches have been fabricated and tested. This paper reports results of measurements of lifetime and contact resistance of microswitches with different contact material combinations, and under different operating conditions. Results are compared with a contact resistance model. The fabrication of the microswitch by a surface micromachining process was reported in [1], along with work on an analytical model of the microswitch. Results of preliminary electrical measurements were presented at the AVS 43rd National Symposium [2]. Measured electrical characteristics of the microswitch include a threshold voltage between 30 and 200 V, a current handling capability exceeding 150 mA, and a lifetime exceeding 10⁹ switching cycles. Combinations of gold, ruthenium and rhodium have been used as the contact materials. Switches have been tested in various ambients. Devices have been "hot" switched (closed and opened with a voltage applied across the switch) and "cold" switched (closed and opened without a voltage applied across the switch). In this paper, lifetime and contact resistance measurements are presented and analysed for the above contact materials and operating conditions. Switch lifetime appears to be limited by degradation of the contact resistance, rather than the mechanical behavior of the switch. A model of the contact resistance is presented in this paper. The model includes determination of the contact force as a function of the actuation voltage by numerical analysis; determination of the contact area as a function of the contact force, using profiles of contact surfaces obtained from STM analysis, and a model proposed by Chang, et al. [3]; and determination of the contact resistance as a function of the contact area, based on the work of Holm [4]. Contact resistance measurement results are compared to modeled contact resistance characteristics. [1] P.M. Zavracky, N.E. McGruer, and S. Majumder, "Micromechanical Switches", Journal of Microelectromechanical Systems, vol. 6, pp. 3-9, 1997. [2] S. Majumder, P.M. Zavracky, N.E. McGruer, "Electrostatically Actuated Micromechanical Switches", to appear in Journal of Vacuum Science and Technology. [3] W.R. Chang, I. Etsion, and D.B. Bogy, "An Elastic-Plastic Model for the Contact of Rough Surfaces", Journal of Tribology, vol. 109, pp. 257-263, 1988. [4] R. Holm, Electric Contacts, Springer-Verlac (New York), 1967.
3:40 PM	Invited	MM+SS+NS-WeA-6 Advances in Systems Integration R.C. Anderson (Affymetrix) We have developed a genetic-analysis system that carries out all reactions and processes necessary to analyze tissue samples using a GeneChip oligonucleotide array. Results of fully integrated multistep assays carried out in 10 ul reaction volumes including HIV polymorphism screening with performance equivalent to benchtop assays will be reported. Integrated processes including fluidic manipulation, nucleic-acid extraction, PCR, and hybridization will be described in detail.
4:20 PM	Invited	MM+SS+NS-WeA-8 Electroosmotic and Electrophoretic Integrated Fluidics L. Bousse, A. Kopf-Sill, J.W. Parce (Caliper Technologies Corp.) The concept of integrating a biochemical laboratory on a chip has eceived much attention lately. One widely used method to integrate fluidic perations on a chip is electroosmotic flow. There are good physical reasons that explain the popularity of this type of microfluidics. The ideal channel size for microfluidics is constrained by two basic laws. On the one hand, in channels that are too big, diffusive mixing is very slow. It takes about 200 seconds, for instance, for two flows that are side by side in a 1000 micron wide channel to mix by diffusion. This means channel dimensions must be below about 100 microns. On the other hand, in channels below 10 microns wide, detection becomes difficult because there are simply too few molecules in such a small volume. So the ideal channel size appears to be in the vicinity of 50 by 10 microns, and typical flow rates in such a channel are nanoliters per second. Mechanical pumps that can deliver flow rates in the nanoliter per second range, with accuracies in the picoliter per second range, are not available in a miniaturized form. Electroosmotic pumping , however, is very convenient, and many electrodes can be used in a single microfluidic chip. The advantages of integrated microfluidics will be discussed with two specific examples. One is the achievement of separations with very high theoretical plate numbers in short times and distances, due to the ability to inject well-controlled small sample plugs. The other is the ability to control complex structures containing many parallel separation channels with a small number of electrodes and sample reservoirs.
5:00 PM		MM+SS+NS-WeA-10 A Micromachined Differential Scanning Calorimeter R.E. Cavicchi, J.S. Suehle, N.H. Tea, G.E. Poirier (National Institute of Standards & Technology) We have fabricated and tested a micromachined differential scanning calorimeter. The device consists of a pair of "micro-hotplates" coupled by a thermopile. Each micro-hotplate is made from suspended SiO₂ with an embedded polysilicon heater. The thermopile is a series of buried polysilicon/aluminum junctions that alternate between the two hotplates. The device was fabricated using a CMOS foundry followed by an isotropic silicon etch in XeF₂ to suspend the SiO₂. To produce a calorimetric chemical sensor we have deposited a Pd film on one of the hotplates. Operation consists of scanning the voltage to the heaters in such a way that, in the absence of an analyte, the thermopile is balanced and produces no output voltage. With an analyte present, differences due to the catalytic activity and thermal coupling of the Pd produce a difference signal which depends on temperature and composition and can be either positive or negative. We have operated the device using scans between 20 °C and 350 °C, over a time interval of 160 s, with a temperature resolution of a few millidegrees. Faster or slower operation with a wider or narrower temperature range is possible and will affect the temperature resolution. A second operating mode is to dose the two hotplates, for example by condensing vapor, and then flash-anneal one element prior to scanning so that only one hotplate has analyte adsorbed on its surface. We will present results for ethanol, methanol, acetone, butane, water and other analytes. We anticipate that variants of this device will have broad applications for micro-analysis systems.