MEMS and Nanotechnology, Volume 2

River Rapids Conference Proceedings of the Society for Experimental Mechanics Series MEMS and Nanotechnology, Volume 2 Tom Proulx Proceedings of the 2010 Annual Conference on Experimental and Applied Mechanics River Publishers

Conference Proceedings of the Society for Experimental Mechanics Series

River Publishers Tom Proulx Editor Proceedings of the 2010 Annual Conference on Experimental and Applied Mechanics MEMS and Nanotechnology, Volume 2

Published, sold and distributed by: River Publishers Broagervej 10 9260 Gistrup Denmark www.riverpublishers.com ISBN 978-87-7004-842-2 (eBook) Conference Proceedings of the Society for Experimental Mechanics An imprint of River Publishers © The Society for Experimental Mechanics, Inc. 2011 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, or reproduction in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Preface MEMS and Nanotechnology, collecting work comprising the 11th International Symposium on MEMS and Nanotechnology, represents one of six tracks of technical papers presented at the Society for Experimental Mechanics Annual Conference & Exposition on Experimental and Applied Mechanics, held at Indianapolis, Indiana, June 7-10, 2010. The full proceedings also include volumes on: Dynamic Behavior of Materials, Role of Experimental Mechanics on Emerging Energy Systems and Materials, Application of Imaging Techniques to Mechanics of Materials and Structures, Experimental and Applied Mechanics, along with the Symposium on Time Dependent Constitutive Behavior and Failure/Fracture Processes. Each collection presents early findings from experimental and computational investigations on an important area within Experimental Mechanics. The current volume on MEMS and Nanotechnology includes studies on: Energy Harvesting and Scaling Effects Metrology and Standards in MEMS and Nanotechnology Carbon Nanotubes Friction and Tribology Meta-materials Optical Methods for MEMS and Nano Adhesion and Stiction Sensors and Actuators Microelectromechanical systems (MEMS) and nanotechnology are revolutionary enabling technologies (ET). These technologies merge the functions of sensing, actuation, and controls with computation and communication to affect the way people and machines interact with the physical world. This is done by integrating advances in various multidisciplinary fields to produce very small devices that use very low power and operate in many different environments. Today, developments in MEMS and nanotechnology are being made at an unprecedented rate, driven by both technology and user requirements. These developments depend on micromechanical and nanomechanical analyses, and characterization of structures comprising nanophase materials.

To provide a forum for an up-to-date account of the advances in the field of MEMS and nanotechnology and to promote an alliance of governmental, industrial, and academic practitioners of ET, SEM initiated a Symposium Series on MEMS and Nanotechnology. The 2010 Symposium is the eleventh in the series and addresses pertinent issues relating to design, analysis, fabrication, testing, optimization, and applications of MEMS and nanotechnology, especially as these issues relate to experimental mechanics of microscale and nanoscale structures. It is with deep gratitude that we thank the organizing committee, session chairs, authors, participants, and SEM staff for making the 11th-ISMAN a valuable and unforgettable experience. The Society would like to thank the organizers of the track, Cosme Furlong, Worcester Polytechnic Institute; Gordon A. Shaw, III, National Institute of Standards and Technology; Barton Prorok, Auburn University; Ryszard J. Pryputniewicz, Worcester Polytechnic Institute for their efforts. vi

Contents 1 Nanomechanical Standards Based on the Intrinsic Mechanics of Molecules and Atoms 1 J.R. Pratt, G.A. Shaw, III, D.T. Smith 2 Magneto-mechanical MEMS Sensors for Bio-detection 9 M. Ramasamy, C. Liang, B.C. Prorok 3 Recent Progress of Piezoelectric MEMS for Energy Harvesting Devices 17 D.-J. Kim, J.-H. Park 4 Performance of Piezoelectric Power Generator in Environmental Conditions 25 5 Comparison of Transduction Efficiency for Energy Harvester Between Piezoelectric Modes 33 J.-H. Park, H. Ahn, S.-B. Kim, S.-H. Kim, D.-J. Kim 6 41 C.M. Byer, K.T. Ramesh 7 Mechanically Probing Time-dependent Mechanics in Metallic MEMS 43 J.P.M. Hoefnagels, L.I.J.C. Bergers, N.K.R. Delhey, M.G.D. Geers 8 Small Mass Measurements for Tuning Fork-based Atomic Force Microscope Cantilever Spring Constant Calibration 49 G.A. Shaw, III, J.R. Pratt, Z.J. Jabbour 9 Linear and Nonlinear Mass Sensing Using Piezoelectrically-actuated Microcantilevers 57 V. Kumar, J.W. Boley, H. Ekowaluyo, J.K. Miller, G.C. Marvin, G.T.-C. Chiu, J.F. Rhoads 10 Calibrating Force and Displacement in the Face of Property Variation 67 J.V. Clark 11 Standardization of Nanoscale Interfacial Experiments Using MEMS 75 T. Ozkan, Q. Chen, I. Chasiotis Size Effects Associated With Microcompression Experiments on Single-Crystal Magnesium S.-B. Kim, J.-H. Park, H. Ahn, D.-J. Kim

12 Arrays of Robust Carbon Nanotube-based NEMS: A Combined Experimental/ Computational Investigation 81 O. Loh, X. Wei, K. Nandy, H.D. Espinosa 13 Electro-mechanical Response of Carbon Nanotube Reinforced Polymer Composites 83 V.K. Vadlamani, V.B. Chalivendra, A. Shukla, S. Yang 14 A Test Platform for Systematic Investigation of Tribology in MEMS 85 N. Ansari, W.R. Ashurst 15 Full Optical Scatter Analysis for Novel Photonic and Infrared Metamaterials 97 T.M. Fitzgerald, M.A. Marciniak 16 Thermal Management and Metamaterials 107 C.T. Roman, R.A. Coutu, Jr., L.A. Starman 17 MEMS Integrated Metamaterial Structure Having Variable Resonance for RF Applications 115 D. Langley, R.A. Coutu, Jr., L.A. Starman, P.J. Collins 18 Characterization and Testing of Adaptive RF Metamaterial Structure Using MEMS 121 C.A. Lundell, P.J. Collins, L.A. Starman, R.A. Coutu, Jr. 19 Design of a Microfluidic Pump, Based on Conducting Polymers 129 20 Micromotor Fabrication by Surface Micromachining Technique 139 D. Barbade, R. Soni, S. Metan 21 Improvement of Piezoresistive Microcantilever Beams for Gas Detection and Sensing 147 N. Wang, B.W. Alphenaar, R.S. Keyton, R.D. Bradshaw 22 Investigation of the Young's Modulus of Fibers in an Electrospun PCL Scaffold Using AFM and its Correlation to Cell Attachment 157 N. Wanasekara, M. Chen, V.B. Chalivendra, S. Bhowmick 23 Recent Progress in E-Beam Lithography for SEM Patterning 163 N. Li, S. Guo, M.A. Sutton 24 167 C.-C. Liao, Y.-L. Lo 25 Experimental Methods for Tensile Testing of Metallic Thin Films at High Temperatures 173 N.J. Karanjgaokar, C.S. Oh, I. Chasiotis 26 Surface Texturing Using Gold Nanoparticles to Reduce Adhesion in MEMS 181 N. Ansari, K.M. Hurst, W.R. Ashurst 27 Design of Microswitch Systems Avoiding Stiction Due to Surface Contact 189 L. Wu, L. Noels, V. Rochus, M. Pustan, J.C. Golinval Analysis of Scattering-type Scanning Near-field Optical Microscopy for Residual-strain Measurements viii K. Kannappan, G. Bogle, J. Travas-Sejdic, D.E. Williams

28 Measurement of Adhesive Force Between two Mica Surfaces With Multiple Beam Interferometry 197 T.Y. Chen, J.C. Jung 29 Performance Studies of a Prototypical MEMS Thermal Actuator 203 W.-Y. Lu, E.J. Garcia, H. Jin, B. Song 30 A New Electrothermal Microactuator With Z-shaped Beams 209 C.H. Guan, Y. Zhu 31 Electrothermal Actuators for Integrated MEMS Safe and Arming Devices 215 R.A. Lake, L.A. Starman, R.A. Coutu, Jr. 32 Contrast Reversal on Surface Plasmon Resonance Reflectivity in Nickel and Nickel Alloy Films 223 A. Horvath, M. Roddy, M. Syed, A. Siahmakoun 33 Effect of Accelerated Ultra Violet and Thermal Exposure on Nano Scale Mechanical Properties of Nylon Fibers 229 34 High Thermal Conductivity Polyurethane-Boron Nitride Nanocomposite Encapsulants 237 J.V. Costa, T. Ramotowski, S. Warner, V.B. Chalivendra 35 Advanced Hard Mask Approach of ICs Copper Interconnects Processes Integration 243 C.-J. Weng 36 Advances in Thin Film Nanoindentation 253 B. Zhou, K. Schwieker, B.C. Prorok 37 Mechanical and Piezoelectric Behavior of Thin Film PZT Composites for MEMS Applications 261 S. Yagnamurthy, I. Chasiotis 38 Fracture Between Self-assembled Monolayers 267 S.R. Na, B. Doynov, A. Hassan, K.M. Liechti, M.J. Krische 39 Control and Quantification of Residual Stresses in Anodically Bonded MEMS Structures 269 R. Inzinga, T. Lin, M. Yadav, H.T. Johnson, G.P. Horn 40 Sub-micron Scale Mechanical Properties of Polypropylene Fibers Exposed to Ultraviolet and Thermal Degradation 275 ix N. Wanasekara, V. Chalivendra, P. Calvert N. Wanasekara, V. Chalivendra, P. Calvert

Nanomechanical standards based on the intrinsic mechanics of molecules and atoms Jon R. Pratt, Gordon A. Shaw, and Douglas T. Smith National Institute of Standards and Technology (NIST) Gaithersburg, MD, USA 20899, jon.pratt@nist.gov ABSTRACT For more than a decade, instruments based on local probes have allowed us to “touch” objects at the nanoscale, making it possible for scientists and engineers to probe the electrical, chemical, and physical behaviors of matter at the level of individual atoms and molecules. In principle, physical interactions on this scale are characterized by fixed, unique values that need only be reliably measured in terms of accurately realized units of force and length to serve as standards. For example, the silicon lattice spacing is often used as a convenient ruler for estimating length in atomic scale images, since this lattice spacing has been independently measured using x-ray interferometry. Recently, the force-induced failure of DNA, often referred to as the overstretch condition, has been proposed as both a standard of force and length in single-molecule bio-physics experiments. Still other nanomechanics researchers have suggested that the rupture force of a single-atom chain is unique to a given metal, and that this intrinsic force can be used to calibrate atomic break junction experiments. In both these examples, a fundamental assumption is that the irreducible nature of nanoscale experimentation, in this case tensile testing, yields consistency befitting a standard. This paper offers context and a condensed overview of recently published results from the NIST Small Force Metrology Laboratory regarding new instruments and capabilities we have developed to examine this fundamental assumption. The reviewed papers describe new test platforms, techniques, and calibration procedures that allow us to bring accurate picoscale measurements of both length and force to bear on the problems of single-molecule and single-atom tensile testing. Keywords: small force, electrostatic force balance, atomic force microscope, colloidal probe calibration. INTRODUCTION An intrinsic standard is an invariant of nature that can be linked to a unit of measure. The triple point of water as a calibration reference in thermometry is a familiar example; the silicon lattice spacing, often used as a convenient ruler for estimating length in atomic scale images, is another. In the realm of piconewton to nanonewton forces, there are a handful of phenomena already referred to as “standards” in biophysical investigations of singlemolecule responses to mechanical deformation and also in similar nanomechanical studies of atomic-size metal contacts. These phenomena, which have been measured in numerous experiments using a variety of techniques, are beginning to be used as convenient reference points for the calibration of the very instruments used in such experimentation. The National Institute of Standards and Technology (NIST) is seeking to securely anchor these reference points to known values by measuring intrinsic force interactions using rigorously calibrated experiments, both physical and computational, that emphasize absolute accuracy through traceability to accepted standards of force and length as defined in the International System of Units (SI). The following paper is drawn from the major published results of these efforts, using the notion of an intrinsic standard to provide a unifying backdrop for the discussion. The paper is organized into four sections; the first two sections deal with specific intrinsic forces, while the latter two sections focus on the more general problems associated with calibrating force and length at the nanoscale. In Proceedings of the SEM Annual Conference June 7-10, 2010 Indianapolis, Indiana USA ©2010 Society for Experimental Mechanics Inc. 1 T. Proulx (ed.), MEMS and Nanotechnology, Volume 2, Conference Proceedings of the Society for Experimental Mechanics Series 2, DOI 10.1007/978-1-4419-8825-6_1, © The Society for Experimental Mechanics, Inc. 2011

general, the sections comprise previously published source material assembled and edited here to support the keynote address delivered at the 2010 Annual Meeting of the Society for Experimental Mechanics. For example, the first section, Au Break Junctions, provides a brief review of Reference [1] where we describe new NIST capabilities developed to study the mechanics and electrical properties of gold nanowires. Gold nanowires apparently neck down under tension to form a single chain of atoms, a physical constriction that leads to the coincidence of quantized force and quantized electrical impedance. It is believed that gold nanowires form in at least three different atomic arrangements. It is not known if all such structures form single-atom chains when stretched to failure and, if so, whether or not the strength of the final bond depends on the initial structure. NIST has been seeking to quantify, to the extent possible, these potential systematic variations. First principles quantum mechanics models along with a new Feedback Stabilized Break Junction (FSBJ) experimental platform have been developed and used at NIST to examine how structure and conductance evolve during elongation, with the ultimate aim to reveal their effect on force at rupture; here, we offer a brief snapshot of this exciting new work. Obviously, the interested reader is encouraged to read the source material for more detail. The intrinsic force associated with the overstretching-induced conformation change of DNA is reported to occur under a tensile force of approximately 65 pN. This force has been measured using both scanning probe microscope (SPM) and optical trapping instruments and has been proposed as both a length and force standard. The quoted force value has known dependencies on environmental factors such as pH, salinity, and temperature. NIST is working to determine how to accurately control these variables to yield reliable data, and we present a small indication of our new single-molecule expertise in the section headed DNA Overstretch. The experimental platforms necessary to investigate these disparate physical phenomena share two things in common. First, both require the accurate measurement of forces at levels at or below a few nanonewtons. Second, because the primary experiments are analogous to macroscale tensile testing, they each require a means of monitoring the elongation of a mechanical specimen with resolution of a few picometers or less. Consequently, Picometer Fiber Interferometry reviews recent developments at NIST regarding the use of fiber interferometry to measure displacements of a nanometer or less. This section provides a highly condensed introduction to the topic; for details the reader should see Reference [2]. Finally, Piconewton Force Calibration summarizes our attempts to improve the calibration of atomic force microscope (AFM) cantilever stiffness and to create known forces at or below the level of a nanonewton. It provides a succinct review of our previous research, and then communicates the key findings from Reference [3] regarding the influence of friction in cantilever-oncantilever calibration, and from References [4-7] regarding our new capabilities to apply electrostatic forces of known magnitudes directly to electrically conducting instrumented indentation and AFM probes. Au Break Junctions We have developed an experimental platform, which we refer to as a feedback-stabilized break junction (FSBJ), to create and deform stable atomic-scale contacts, and have used that platform to probe the phenomenon of quantized electrical conductance in Au nanowires (NWs) and single-atom chains (SACs) [1]. In such a system, electrical conductivity, σ, is known to be quantized in units of G0 = 2e 2/h; that is, σ = nG0 for integer n, with e the charge of the electron and h Plank’s constant [8-10]. The conductivity for the n = 1 state corresponds to a contact resistance of 12.9 kΩ. Quantized conductance has been observed many times, but experimental instabilities typically limit the time a given contact stays in a low-n state to milliseconds [11], and often the presence of quantized states must be inferred from histograms compiled from hundreds or thousands of junction breaks [12]. Because the n = 1 conduction state is believed to occur when there is only one electron conduction channel through the contact [13,14], the ability to maintain that state indefinitely would clearly demonstrate exceptional experimental stability. The NIST FSBJ described in [1] and shown in the photos of Figure 1 is proving to be just such an exceptionally stable experimental platform. By positioning an interferometer cavity directly between an Au surface and probe mount, we have significantly tightened the displacement measurement frame relative to that achieved in prior work. This has allowed us to close a servo loop around the junction separation with long-term, picometer stability, so as to remove thermal drift and low-frequency vibration artifacts and thereby reduce the need for rigorous environmental isolation that is often encountered in these types of experiments. In the near future, the NIST FSBJ will incorporate a stiff elastic force sensor (e.g., an AFM style colloidal probe) coupled with a high resolution fiber interferometer (see Figure 1), so that direct measurements can be made of bond stiffness and breaking force in SACs. 2

Figure 1. Feedback-stabilized break junction platform (A) high-vacuum cryogenic chamber with long standoff microscope (B) chamber lid removed (C) mechanics including probe positioning stage (D) view of tip approaching an AFM colloidal probe as seen using the long standoff microscope. The fiber labeled force records the deflection of the colloidal probe cantilever, while the fiber labeled probe monitors the displacement of the probe stage with respect to their common plane (both fibers are in a single, double bore ferrule that is rigidly attached to the cantilever holder base) for the feedback stabilization of the probe sensor separation. DNA Overstretch Another example of an intrinsic force is the overstretch transition of DNA [15, 16]. During single-molecule force measurements, a transition occurs at approximately 65 piconewtons in which the DNA molecule elongates significantly with very little extra applied force. The force-displacement curve of a single molecule of DNA measured with a force-calibrated AFM is shown in Figure 2, as described previously [17]. This particular molecule was amplified from a segment of the plasmid vector pBR322, and was measured in Tris/sodium chloride buffer. The transformation can be seen as a plateau in the curve. The methods developed for small force measurement at NIST are being used to calibrate the force at which this plateau occurs. The DNA itself then will become a force reference that will allow the calibration of a wide variety of force measuring instruments such as optical, magnetic, and dielectrophoretic tweezers. A B 0.00 0.10 0.20 0.30 0.40 0 50 100 150 200 250 300 d (nm) F (nN) Figure 2. DNA as an intrinsic force standard. (A) Schematic of a single molecule of DNA being stretched by an AFM cantilever. (B) Force displacement curve showing the overstretch transformation of a single DNA molecule measured using an AFM. Arrow indicates the direction in which the DNA was stretched. 3

Picometer Fiber Interferometery Clearly, when dealing with single-atom contact mechanics, such as that realized by the interaction of a probe tip with a flat surface, relative displacement of the tip with respect to the substrate must be monitored with resolution well below the lattice spacing of the atoms (< 0.1 nm). Similarly, the interaction forces must be resolved to within at least a few percent of the rupture force of an atomic bond, which has been observed to be between 1 nN and 2 nN for a pair of gold atoms [18]. Assuming that this force is measured using a spring sufficiently stiff to avoid snap in, a stiffness anticipated to be around 50 N m-1, the detector on the force sensor must resolve relative displacement of the spring to 1.5 pm to achieve even a 5 % uncertainty in determination of the rupture force. The requirements for atomic-scale length metrology in these experiments are clear, and point to the need for an accurate sensor capable of recording full-scale relative displacements on the order of 10 nm with percent-level linearity and a noise floor of picometers, in a bandwidth from DC to hundreds of Hertz. Accurate measurement of displacement requires calibration via comparison to an absolute standard. The unit of length in the International System of Units (SI) is most accurately realized in terms of the wavelength of light; thus, the most accurate determinations of displacements are typically achieved by incorporating measurement tools developed around optical interference, where the displacement is directly compared to the absolute SI standard, minimizing the accrual of uncertainty due to successive calibrations. Simple homodyne interferometers are often employed when the displacement of interest is expected never to exceed the fringe spacing. Robust interferometers based on the low-finesse Fabry-Perot (FP) cavity formed between the end of an optical fiber and the reflective surface of the cantilever spring have been demonstrated for scanning probe microscopes, and we have developed such a fiber-optic interferometer (see schematic in Figure 3) optimized for best performance in the frequency range from DC to 1 kHz, with displacement linearity of 1 % over a range of ± 25 nm, and noiselimited resolution of 2 pm [2]. The interferometer uses a tunable infrared laser source (nominal 1550 nm wavelength) with high amplitude and wavelength stability, low spontaneous self-emission noise, high sideband suppression and a coherence control feature that broadens the laser linewidth and dramatically lowers the lowfrequency noise in the system. The amplitude stability of the source, combined with the use of specially manufactured “bend-insensitive” fiber and all-spliced fiber construction, result in a robust homodyne interferometer system that achieves resolution of 40 fm·Hz-1/2 above 20 Hz and approaches the shot-noise-limit of 20 fm·Hz-1/2 at 1 kHz for an optical power of 10 µW, without the need for differential detection. Figure 3. Schematic diagram of the fiber-optic interferometer components showing the laser, optical isolator (ISO), evanescent wave coupler, single InGaAs detector, and a test interferometer cavity. Note that the forward and return ports of the coupler are terminated using angle polished connectors (APC) to minimize interference in the parasitic cavities along these fiber runs. The light source for our interferometer can be tuned stably and accurately over the wavelength range 1440 nm to 1640 nm. This tuning range enables two very convenient operational features. First, when we either have no control over the macroscopic cavity spacing, as for example in the case of measuring the deflection of a cantilever 4

force sensor, or we choose not to change the cavity out of concern for dimensional stability, we can still operate the interferometer at its maximum sensitivity by tuning λ, so long as the cavity is large enough to contain a quadrature point within the tuning range of the laser. Second, we can make an absolute determination of a fixed cavity length h by sweeping wavelength. Since maxima in reflected intensity occur when 4h/λ = 2m+1, assuming no phase shift at the interfaces (other than standard phase inversion), and minima occur when 4h/λ = 2m, it follows directly that h can be calculated from the measured wavelength values λm and λm+1 of two consecutive fringe maxima or minima: ) 2 ( 1 1 1 m m m m h λ λ λ λ − = + + . For the tuning range of our laser, this means that we can make an absolute determination of cavities as small as h = 6 μm. If a consecutive maximum and minimum are used to calculate h, 3 μm cavities can be measured. Accuracy is limited by our ability to measure λm and λm+1. With curve fitting, wavelengths of maxima and minima can typically be determined within a range of ± 0.1 nm, permitting cavity length determination to better than one part in 1000. Piconewton Force Calibration We have been exploring methods to calibrate microcantilever force sensors, such as those used in atomic force microscope experiments. Our initial work in this field focused on creating a system for accurately realizing or creating known forces of appropriate magnitudes. The NIST Electrostatic Force Balance (EFB), which we created for this purpose [19], is an electromechanical balance system that we have used to measure the force versus displacement response of so-called reference cantilevers [20] for use in cantilever-on-cantilever calibrations, and the force sensitivity of piezoresistive cantilever force sensors that can be similarly used to calibrate the force sensitivity of AFMs [21,22]. This work has been extended recently in two important directions. First, we have found that there are systematic differences between loading and unloading force versus displacement curves obtained using an AFM to probe another cantilever. In Reference [3], the appearance of hysteresis in the slopes of loading curves when probing compliant surfaces with AFM has been explained with an analytical model describing the bending of the test cantilever under the combined influence of a normal contact force and a tangential sliding frictional force. Expressions derived in the paper allowed us to determine coefficients of sliding friction, as well as the stiffness of a sample, provided the spring constant of the AFM cantilever was known. More importantly within the present context, the results were also applied to determine the stiffness of an AFM cantilever using a cantilever-on-cantilever spring constant calibration. A unique aspect of this investigation was that the stiffness of both the AFM cantilever and reference cantilevers was known a priori, based on absolute calibrations performed using the NIST EFB. Making use of corrections described in the paper, it was possible to measure the unknown cantilever stiffness using an accurately calibrated cantilever reference within a relative standard uncertainty of approximately 3 %. Remarkably, similar precision was achieved in the determination of the coefficient of sliding friction. From a practical standpoint, a very useful observation made in the paper is that the accuracy of cantilever-on-cantilever calibration can be markedly improved by simply taking the mean of the slopes recorded during loading and unloading against the reference cantilever. The second important direction of our recent work has been extending the use of calculable electrostatic force to the direct calibration of indentation and colloidal probe force sensors [4-6]. In Reference [4], the capacitance gradient between a spherical indenter probe and a flat electrode was measured. The flat electrode was then used to apply calculable electrostatic forces directly to the spherical indenter, which was mounted on a commercially available force sensor. In essence, the experiment demonstrates the direct application of a force standard to a target instrument without recourse to a transfer artifact, in much the same way calibration has been achieved by hanging known masses on springs in the past. Electrostatic forces ranging from 10 μN to 200 μN were realized with uncertainties estimated less than 1 % (k = 1). Forces were compared in an indirect fashion to deadweights through evaluation of the sensor stiffness, with the relative difference between measured stiffness recorded via the two approaches less than 1 %. In follow up papers [5,6], we have demonstrated calculable electrostatic forces nominally ranging from 320 pN to 100 nanonewtons with uncertainties of a few percent in experiments to perform a direct force calibration of colloidal probe AFM, as illustrated in Figure 4. 5

Figure 4. Schematic of an electrostatic force probe experimental setup as in References [5,6]. Insets show scanning electron microscope micrographs of devices as constructed. CONCLUDING REMARKS A brief overview of NIST’s pioneering work in the field of intrinsic force standards has been presented. The concept of an intrinsic force standard has been introduced, and two candidate physical phenomena have been discussed to describe the current status of this measurement science at NIST. The physical measurements necessary to support the foundational single-molecule and single-atom contact experiments, namely pico-scale force and length metrologies, were reviewed. At present, NIST’s Feedback Stabilized Break Junction platform has demonstrated the stability necessary to perform single-atom contact mechanics experiments with traceable force and length metrology, largely thanks to NIST innovations in the application of fiber interferometery, electrostatic force calibration, and micro-cantilever stiffness calibration. The force curve presented in Figure 2 provides a glimpse of a tantalizing future where accurately determined force transitions that occur in the tensile loading of DNA can serve as readily accessible calibration references for a variety of single-molecule science platforms, including optical and magnetic tweezers. DNA can be manufactured with atomic precision in large quantities, and could provide a very cost-effective reference material for single-molecule nanomechanical experimentation. ACKNOWLEDGEMENTS Thanks to our many collaborators including K-H Chung, L. P. Howard, L. Kumanchik, N. A. Burnham, L. E. Levine, A. M. Chaka, R. S. Gates, M. G. Reitsma, J. A. Kramar, and D. C. Hurley who contributed greatly to various aspects of the research described here. This work has been funded in part by the NIST Innovations in Measurement Science program. REFERENCES 1. Smith, D. T., Pratt, J. R., Tavazza, F., Levine, L. E., and Chaka, A. M., J. Appl. Phys., in press 2010. 2. Smith, D. T., Pratt, J. R., and Howard, L. P., Rev. Sci. Inst., 80:035105, 2009. 3. Pratt, J. R., Shaw, G. A., Kumanchik, L., and Burnham, N. A., J. Appl. Phys., 107:044305, 2010. 4. Chung, K-H, Scholz, S., Shaw, G. A., Kramar, J. A., and Pratt, J. R., Rev. Sci. Inst., 79:095105, 2008. 5. Chung, K-H, Shaw, G. A., and Pratt, J. R., Rev. Sci. Inst., 80:065107, 2009. 6. Chung, K-H, Shaw, G. A., and Pratt, J. R., Proceedings of the IMEKO XIX World Congress, Lisbon, 2009. 7. Chung, K-H, Pratt, J. R., and Reitsma, M. G., Langmuir, 26(2):1386, 2009. 8. Landauer, R., IBM J. Res. Dev. 1, 223 (1957). 9. Büttiker, M., Imry, Y., Landauer, R., and Pinhas, S., Phys. Rev. B, 31:6207, 1985. 10. M. Brandbyge, J. Schiøtz, M. R. Sørensen, P. Stoltze, K. W. Jacobsen and J. K. Nørskov, Phys. Rev. B 52:8499, 1995. 11. V. Rodrigues, T. Fuhrer and D. Ugarte, Phys. Rev. Lett. 85 :4124, 2000. 12. J. L. Costa-Krämer, N. García, P. García-Mochales, P. A. Serena, M. I. Marqués and A. Correia, Phys. Rev. B 55:5416, 1997. 13. T. Frederiksen, M. Paulsson, M. Brandbyge and A.-P. Jauho, Phys. Rev. B 75, 205413, 2007. 6

14. C. Untiedt, M. J. Caturla, M. R. Calvo, J. J. Palacios, R. C. Segers and J. M. van Ruitenbeek, Phys. Rev. Lett. 98 :206801, 2007. 15. Williams, M. C., Rouzina, I., Bloomfield, V. A., Thermodynamics of DNA interactions from single molecule stretching experiments, Acc. Chem. Res., 35:159, 2002. 16. Clausen-Schaumann, H., Rief, M., Tolksdorf, C., Gaub, H. E., Biophys. J., 78:1997, 2000. 17. Shaw, G. A., Pratt, J. R., Proc. SEM Annual Conference, 2009, Albuquerque, NM, June 1-6, 2009. 18. G. Rubio-Bollinger, S.R. Bahn, N. Agraït, K.W. Jacobsen, and S. Vieira, Phys. Rev. Lett., 87:026101, 2001. 19. Pratt, J.R., Kramar, J. A., Newell, D.B., and Smith, D.T., Meas. Sci. Technol., 16:2129, 2005 20. R. S. Gates and J. R. Pratt, Measurement Science and Technology, 17, 2852 (2006) 21. Pratt J R, Smith D T, Newell D B, Kramar J A and Whitenton E., J. Mater. Res. 19:366, 2004. 22. E.D. Langlois, G.A. Shaw, J.A. Kramar, J.R. Pratt and D.C. Hurley, Rev. Sci. Inst., 78 (9):093705, 2007. 7

Magneto-Mechanical MEMS Sensors for Bio-Detection M. Ramasamy, C. Liang and B.C. Prorok Auburn University, Department of Mechanical Engineering 275 Wilmore Labs, Auburn University, AL 36849-5341 Email: prorok@auburn.edu ABSTRACT: Ferromagnetic materials have shown to possess some unique and useful properties, one of which is that they are magnetomechanical transducers, that is, they exhibit a change in dimension when they are subjected to an external magnetic field and vice versa. This magneto-mechanical coupling enables magnetoelastic sensors to be driven to resonance via a modulated magnetic field to detect biological species via frequency shift by mass addition [1,2]. This work details the development of an algorithm to predict the number of captured E. coli cells based only upon the resonance frequency shift. This is an important issue as attaching cells influence resonance based upon their location on the sensor. It is therefore necessary to develop a statistical protocol to predict the concentration of the target agent present. A protocol was developed based upon data from microscale resonators using polystyrene beads as a simulant. The protocol was verified with numerical studies and experiments using E. coli cells. INTRODUCTION: Food safety has become an important issue in the past decade after numerous intentional and unintentional contaminations. With the innovation in science and technology, there has been a remarkable increase in the development of biological weapons throughout the world. Biological agents are considered to be psychologically threatening and therefore provide more appeal to the terrorist. The human pathogenic limit of many engineered biological agents have been reduced in order of magnitude from 1015 to few cells [3]. This has led to more research on developing biological agents’ detection techniques that are very rapid, sensitive and cost effective [4]. Magnetostrictive materials developed into mass-based acoustic wave Sensor operated on longitudinal vibration mode is one such technique. Magnetostrictive materials are soft amorphous ferromagnetic materials that change in magnetic properties when stress is applied or vice versa. These magnetostrictive materials are fashioned into acoustic wave sensors in the form of simple rectangular strips that are actuated in their longitudinal vibration mode when exposed to magnetic field. Due to the applied field the strips resonate at a specific frequency which is dependent on their mass and physical dimensions [1-4]. These devices operate similar to magnetostrictive strips used in stores as a surveillance device to prevent the theft of goods. PRINCIPLE OF OPERATION In this research, the principle of detection of these sensors involved measuring a resonant frequency shift as the target biological species attaches to the sensor; effectively this addition of mass dampens the resonant behavior of the sensor platform [10]. In the case of monitoring harmful biological agents, it is highly desired to detect the presence of a handful of spores/cells since many harmful agents have a very low pathogenic limit in humans, meaning it only takes a few spores/cells to infect them. When biological agents attach to these magnetostrictive sensors in small amounts they represent discrete mass additions, whereas a large number may be more analogous to mass evenly distributed on the strip. It is therefore highly desirable to better understand how spore/cell attachment location influences the resonant frequency of these magnetostrictive strips. The basic sensor structure investigated was a freestanding beam with no fixed ends. The resonance frequency of the first harmonic mode of such a structure can be described as ࢌൌ ૛ ૚ ࡸට࣋ሺ૚ି ࡱ ࣇሻ . (1) Where, L, E, and ρ are the length, Young’s modulus, and density of the sensor strip, respectively. As these freestanding magnetostrictive strips are driven to their first harmonic mode, the entire strip deforms in response to the field. The resulting deformation waves propagate through the strip and reflect back from the free ends and cancel at the strip’s center, which is the zero nodal posit ion for a freestanding beam resonating in its first Proceedings of the SEM Annual Conference June 7-10, 2010 Indianapolis, Indiana USA ©2010 Society for Experimental Mechanics Inc. 9 T. Proulx (ed.), MEMS and Nanotechnology, Volume 2, Conference Proceedings of the Society for Experimental Mechanics Series 2, DOI 10.1007/978-1-4419-8825-6_2, © The Society for Experimental Mechanics, Inc. 2011

harmonic mode [17]. It should be noted that positions further from the center node of the strip move further from the node during deformation due to the accumulated deformations of all positions between it and the node. Thus, the free ends move the furthest which is clearly seen in figure1a portraying the simulation result of a mass sensor developed using Ansys® as the simulation tool. Variation in the shade from black to light shade of gray depicts the level of deformation experienced by the sensor platform due to the longitudinal vibration following the boundary conditions. Figure 1: Deformation of Longitudinally vibrated Sensor and Mass attachment affection the wave speed of the platform When mass becomes attached to the sensor it effectively dampens the speed of the deformation waves propagating in the strip, which reduces its resonant frequency. Figure1b represents the action of deformation waves due to the mass attachment. In the first case E. coli cells are attached to the sensor platform at equal distance from the nodal point [12-16]. The deformation waves subjected to this mass travel with reduced wave speed from that point and reach the central nodal position leading to the resonant frequency shift. Whereas in the second case mass is attached on one side of the sensor leading to shift in the nodal point to balance for the reduced wave speed on one side alone and hence we observe a resonant frequency shift. This paper elaborately discusses detection of biological agent considering the above cases, on the acoustic response of the sensor [11]. The frequency shift as a result of mass attachment for acoustic-based sensors such as these is described the following equation, ∆ࢌൌെࢌ ࢕ ቄࣁ ૛ቅ. (2) Where ߟ ൌ ∆݉ /݉ ௦௘௡ and ∆m is the mass bonded on a sensor, ∆f is the resonant frequency change before and after mass attachment, m0 and f0 is the sensor’s initial mass and resonant frequency, respectively [2-4]. Here, the bonded mass (∆m) on the sensor’s surface is considered as an evenly distributed mass and is considerably small with respect to the sensor’s mass. Accordingly, the change in frequency can be related to the amount of mass bonded on the sensor [8]. However, in the case when a mass is not evenly distributed or is a discrete mass bonded in a particular location this equation is no longer valid. In order to study these effects discrete masses were attached at various positions on the sensor and the response was measured and analyzed. For experimental analysis a schematic setup as shown in figure 2 involving a coil, external magnetic bias and a network analyzer was used to measure the acoustic response of the sensor as discussed in Figure 2: Schematic representation of the measurement rig. 10

previous work by the authors [1-3] DESIGN AND EXPERIMENTAL METHODS: Numerical simulations were carried out using commercially available software Ansys®. Specifically, the simulations involved modal analysis on an undamped, freestanding beam with oscillations in the longitudinal mode both with and without attached mass. The structural physical (engineering) discipline is preferred for the Modal analysis of Magnetostrictive sensors. The selected element type was SOLID186. The sensor size employed here was 250 x 50 x 4 µm and the attached mass were representative of an actual E. Coli O157:H7 cell, size of 1.43 x 0.73 x 0.73 microns and weight of 1 pictogram. The boundary conditions were set to obtain longitudinal vibration mode of the sensor platform. Prediction model was developed involving the factors influencing the resonance behavior such as (a) Mass distribution, (b) Position of the mass distributed and (c) Physical dimension of the sensor platform in compliance with the theoretical equations and simulation results. Mass of E. coli cells were distributed as a single layer for uniform distribution. The later was glued to the sensor platform and subjected to numerical simulation with the application of boundary condition over the whole setup. The density of the layer was modified each time for different amount of mass in uniform distribution case and the corresponding variation in frequency shift was observed. In case of the non uniform distribution the layer was split and concentrated along the free ends gradually moving towards the central nodal line and vice versa as shown in Figure 3. The darker layer in the figure is considered as the mass of E. coli cell distributed over the sensor platform. Glass beads attachment : Sensors with dimensions of 5 mm length and 1 mm width were cut from a 28 μm thick commercially available MetglasTM 2826MB strip. These specimens were prepared, by cleaning and drying, using the identical procedures described by the authors elsewhere. Glass beads with a diameter about 425 μm were employed to simulate the concentrated mass and were carefully loaded on to the sensor surface at prescribed locations and secured with adhesive. The average mass of a sensor and glass bead were 1066 μg and 181.5 μg, respectively. It should be noted that these experiments are aimed as assessing the position of the mass concentrations and not focused on demonstrating minimum sensitivity. Thus, significantly sized beads were employed. The amount of glue employed to affix each bead as well as its position were well controlled to minimize any errors. After a glass bead was loaded on the sensor surface, it was immobilized by drying at Figure 1: Non-Uniform distribution of E. coli cells as layer over the sensor Figure 2: Simulation result of resonant frequency shift due to E. coli cells distribution 11

room temperature for at least two hours. The resonant frequency of the sensor was measured before and after attachment of the glass bead in a manner identical to which is discussed in previous work by the authors. RESULT AND DISCUSSION (a) Resonant Frequency shift due to Mass Distribution: Frequency shift is not the same if the mass is concentrated at a particular location on the sensor platform compared to the uniform distribution of the mass. This is mainly due to reduction in wave speed of the sensor obstructed by the concentrated mass at discrete locations on the platform. Simulation results are shown in the figure 4. From the figure we can notice that uniform distribution shows a linear plot with the increase in frequency shift as the amount of mass increases gradually. The upper and lower plot in the figure explains the distribution of mass as layer as shown in figure 3 from free end to the central nodal point and vice versa. In these 2 cases the plots meet the uniform distribution linear plot at the end since that’s the maximum possible mass that could be attached over the sensor platform. This mass number is based on the assumption that no two E. coli cells accumulate one over the other as demonstrated by Guntupalli, et al. and Wan, et al. in their experimental analysis. In such cases the Maximum resonant frequency shift will be observed at the free ends of the sensor platform which can theoretically determined using equation. ∆ࢌൌെࢌ࢕ቄቀࡷ ૛࣊ ࡸቁെ૚ቅ. (3) Where ࡷൌ ࢼටࡱ ࣋, and β is the angular frequency and using η we can calculate KL value for a free standing beam using the equation, ߟ ሺ ܮܭ ሻൌ tanሺ ܮܭ ሻ Mathematically speaking the frequency shift for certain amount of mass will have a tolerance limit based on their location and distribution. From the above equation we know that the maximum frequency shift for certain amount of mass is observed at the free ends of the sensor platform. This eventually explains the fact that the minimum frequency shift for the same amount of mass would be observed when concentrated at the centre. When mass is distributed close to the nodal point since the deformation experienced is relatively lower compared to the free ends , we observe minimum frequency shift close the centerline of the sensor platform. Figure 5: comparative plot of resonant Frequency shift in theoretical and simulated study. Figure 6: Position sensitive Resonant Frequency shift due to Mass attachment 12

∆ࢌൌെࢌ࢕ ቄࣁ൅ቀࡷ ૛࣊ ࡸቁെ૚ቅ. (4) We were able to formulate an equation for minimum frequency shift due to concentration of mass along the centerline of the sensor platform. These equations were verified with the help of the simulation results obtained by distributing the mass uniformly and at various discrete positions on the sensor platform as shown in Figure 5. From the figure we can also notice that after a certain amount of the mass the movement of the theoretical plot deviate in a linear manner compared to the simulation plot which is mainly due to the assumption made in accordance to the distribution of load along the free ends. but in reality as the mass increases it moves closer to the nodal point and eventually meet the uniform distribution line at a particular amount of mass which is the same in case of minimum frequency shift of the sensor due to mass distribution. (b) Influence of position: Tolerance limit of frequency shift depends on the position of the mass attached irrespective of the amount of mass involved. In this case, the wave speed on the side of the attached mass is less than the opposite side where no mass is attached. During resonance, the wave frequency must match on either side of the sensor. In order to account for the imbalance of wave speeds, the zero nodal point must shift such that the distance the waves travel on each side enables the waves to arrive at the nodal point at the same time to ensure they are in phase. In this case it must shift towards the side with the attached mass, resulting in larger frequency shifts. The simulation and experimental results using glass beads that were attached to the sensor proves this conception. Using these results an equation was developed, to represent the influence of position on the frequency shift of the sensor. ∆ࢌൌି ࢌ࢕૛ ቀ࢓ ∆࢓࢙ࢋ࢔ቁቄ૚െࢉ࢕࢙ቀ࣊ ࡸ ࢞ቁቅ (5) In this case, where the attached mass is a discrete mass bonded in a particular location, the wave speed on the side of the attached mass is less than the opposite side where no mass is attached. During resonance, the wave frequency must match on either side of the sensor. In order to account for the imbalance of wave speeds, the zero nodal point must shift such that the distance the waves travel on each side enables the waves to arrive at the nodal point at the same time to ensure they are in phase. In this case it must shift towards the side with the Figure 7: Plot illustrating the influence of physical dimension on the resonant frequency shift of the sensor platform 13

attached mass. Thus, as the mass is moved further away from the original nodal point, the wave speeds are increasingly imbalanced and the node must shift further to account for it, resulting in larger frequency shifts. Beads were also positioned away from the longitudinal center axis of the sensor to assess the influence of lateral position for the same longitudinal position x. The frequency shift was relatively unaffected indicating that lateral positioning for a discrete mass addition was negligible in comparison to longitudinal positioning. (c) Influence of dimensions on resonant frequency shift: Simulations were carried out to study the effect of variation in resonance behavior due to change in dimensions of the sensor. Figure 5(a) shows that variation in length of the sensor has a dramatic effect on the frequency shift. When the sensor is subjected to Longitudinal vibration, due to the sensor materials magnetic property the deformation occurs in the longitudinal direction along the direction of the applied field. The larger the length of the sensor greater is the sensitivity and hence greater resonant frequency shift for the same amount of mass. In figure 5(b) variation in width of the sensor platform does not have any influence on the acoustic response of the sensor. This is because lateral axis of the sensor does not experience any deformation due to their restricted boundary conditions and hence distribution of mass is considered to uniform along the lateral distance at any point on the sensor platform. This is experimentally proved with the help of polystyrene glass beads that had no influence on the resonant frequency shift when attached along their lateral position of the sensor platform. CONCLUSION A Model developed based on the acoustic response of Magnetostrictive Sensor actuated longitudinally with the modulated magnetic field. Numerical simulations and experimental verifications were carried out with the predicted model considering the factors influencing the resonance behavior of the sensor. The tolerance limit of frequency shift corresponding to the distribution of mass, discrete position of the mass attached and the physical dimensions of the sensor platform were determined in several ways. A Good agreement was found between these results offering a good paradigm for detecting biological agents. References [1] C. Liang and B. C. Prorok, "Measuring the Thin Film Elastic Modulus with a Magnetostrictive Sensor," J. Micromech. Microeng., vol. 17, pp. 709-716, 2007. [2] C. Liang, S. Morshed, and B. C. Prorok, "Correction for Longitudinal Mode Vibration in Thin Slender Beams," Appl. Phys. Lett., vol. 90, pp. 221912, 2007. [3] M. L. Johnson, J. Wan and B. A. Chin et al., “A Wireless Biosensor using Microfabricated Phage-Interfaced Magnetoelastic Properties”,Sens. Act. A 144, 38-47, 2008. [4] W.Shen, R.S.Lakshmanan and B.A.Chin et al., ”Phage Coated Magnetoelastic Micro-Biosensors for realtime Detection of B.anthracis spores”, Sens. Act.B 137,pp.501-506,2009. [5] D. S. Ballantine, R. M. White, S. J. Martin, A. J. Rico, G. C. Frye, E. T. Zellers, and H. Wohltjen, Acoustic Wave Sensors: Theory, Design, and Physical-chemical applications. New York: Academic Press, 1997. [6] R. Raiteri, M. Grattarola, H. J. Butt, and P. Skladal, "Micromechanical cantilever-based biosensors," Sens. Act. B, vol. 79, pp. 115, 2001. [7] P. G. Stoyanov, S. A. Doherty, C. A. Grimes, and W. R. Seitz, "A remotely interrogatable sensor for chemical monitoring," IEEE Trans. Mag., vol. 34, pp. 1315, 1998. [8] C. Mungle, C. A. Grimes, and W. R. Dreschel, "Magnetic field tuning of the frequency-temperature response of a magnetoelastic sensor," Sens. Act. A, vol. 101, pp. 143, 2002. [9] C. Ruan, K. Zeng, O. K. Varghese, and C. A. Grimes, "A staphylococcal enterotoxin B magnetoelastic immunosensor," Biosens. Bioelec., vol. 20, pp. 585, 2004. [10] N. Bouropoulos, D. Kouzoudis, and C. A. Grimes, "The real-time, in situ monitoring of calcium oxalate and brushite precipitation using magnetoelastic sensors," Sens. Act. B, vol. 109, pp. 227, 2005. [11] S. Schmidt and C. A. Grimes, "Characterization of nano-dimensional thin-film elastic moduli using magnetoelastic sensors," Sens. Act. A, vol. 94, pp. 189, 2001. [12] L. S.Q., L. Orona, Z. M. Li, and Z.-Y. Cheng, "Biosensor Based on Magnetostrictive Microcantilevers as a Sensor Platform," Appl. Phys. Lett., vol. 88, pp. 073507, 2006. 14

RkJQdWJsaXNoZXIy MTMzNzEzMQ==