POLYMER MICROELECTROMECHANICAL SYSTEMS: FABRICATION AND APPLICATIONS IN BIOLOGY AND BIOLOGICAL FORCE MEASUREMENTS DISSERTATION Presented in Partial Fulfillment

of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Nicholas J. Ferrell, B.S. *****

The Ohio State University 2008 Dissertation Committee: Approved by Professor Derek J. Hansford, Advisor Professor L. James Lee Professor Allen Yi ___________________________________ Advisor Biomedical Engineering Graduate Program

ABSTRACT

Polymer materials are increasingly being utilized in biomedical micro- and nanotechnolgy applications. This trend has been driven by a several factors ranging from materials compatibility to cost. The manufacturing techniques used to produce these devices are considerably less mature than their silicon-based counterparts. New manufacturing techniques are needed to address unique processing challenges posed by polymer materials. To this end, we have developed a set of soft lithography based micromolding techniques for fabrication of polymer microstructures and devices from a wide range of materials. Materials include common thermoplastic polymers such as poly(methyl methacrylate) (PMMA) and polystyrene as well as functional materials such as conducting polymers. The processing techniques developed through this work are capable of producing a wide range of structures including continuous microstructured films, isolated polymer microstructures, and suspended structures. The nature of the materials and the non-cleanroom based micromolding processes makes these techniques considerably more cost effective with respect to both materials and processing costs. In addition to developing processing techniques, characterization of the processes as well as the materials is a critical step for implementation of polymers in practical device applications. Process characterization was performed by systematically varying process parameters and evaluating the resulting microstructures using common microii

and nanoscale characterization techniques. Scanning electron microscopy, atomic force microscopy, and optical microscopy were all used to evaluate the resulting polymer structures. Nanoindentation techniques were used to characterize the mechanical properties of the materials. Elastic modulus, hardness, creep, scratch resistance, and yield strength of several polymer MEMS materials were evaluated. Application of these techniques for development of functional devices is ultimately the goal. We have used the processing techniques that we have developed to fabricate and test three polymer MEMS devices for biological applications. The first is a microfabricated membrane system for isolation of individual cells or cell clusters. This device could be utilized in a variety of cell biology applications including single cell experimentation, cell cluster biology, and tissue engineering. The other two devices were developed for measuring low magnitude biological forces. A polymer cantilever force sensor was developed for measuring contractile forces produced by fibroblast cells. This device could be used in cell mechanobiology studies, drug evaluation, and cell-based biosensing. The final device is an adapted polymer cantilever sensor for measuring forces produced by protein aggregates known a forisomes. This unique biomaterial could be utilized as a valve or actuator in microdevices.

iii

Dedicated to my parents

iv

ACKNOWLEDGMENTS

I would like to thank my advisor, Derek Hansford, for his support and encouragement dating all the way back to my undergraduate days. I thank Dr. Hansford for his advice and mentorship. Working under his guidance over the years has been a pleasure. I thank all the members of our research group, past and present. Their support, training, mentoring, and patients were indispensable. Specifically, I would like to acknowledge Jay Woodard for his painstaking work with the design and simulation of the cell force sensor. Special thanks to Jingjiao Guan, Joe Kitzmiller, Rob Short, Kelly Larkin, and Jason Sakamoto for showing me the ropes and getting me started on my way down this path. I am grateful to have had the opportunity to work closely with Yang Sun and Randy Butler. I especially thank Daniel Gallego and Natalia Higuita for their enthusiasm and constant willingness to help me out in the lab. I am grateful to Profs. L. James Lee and Allen Yi for serving on my committee. I would like to thank the technical staff and the Nanotech West Laboratory, particularly Derek Ditmer and Paul Steffan, for technical support and training. This research was supported in part by the Air Force Office of Scientific Research MURI award (F49620-03-1-0421). Portions of this work were funded by the National Science Foundation (NSF) Center for Affordable Nanoengineering of Polymer v

Biomedical Devices (EEC-0425626). I also acknowledge the NSF Integrative Graduate Research and Education Traineeship (IGERT) program (0221678) for my graduate fellowship.

vi

VITA September 27, 1979...Born - Van Wert, OH 2003...B.S. Mechanical Engineering, The Ohio State University 2003-present......Graduate Research Fellow, The Ohio State University 2007...Visiting Research Scientist, Instituto de Enginharia Biomdica (INEB), Universidade do Porto PUBLICATIONS Research Publications 1. Nicholas Ferrell, James Woodard, Derek Hansford, Fabrication of Polymer Microstructures for MEMS: Sacrificial Layer Micromolding and Patterned Substrate Micromolding. Biomedical Microdevices, 9, 815-821 (2007). 2. Jingjiao Guan, Nicholas Ferrell, Bo Yu, Derek J. Hansford, L. James Lee, Simultaneous Fabrication of Hybrid Arrays of Nanowires and Micro/Nanoparticles by Dewetting on Micropillars. Soft Matter, 3, 1369-1371 (2007). 3. Manuel Palacio, Bharat Bhushan, Nicholas Ferrell, Derek Hansford, Adhesion Properties of Polymer/Silicon Interfaced for Biological Micro-/nanoelectromechanical Systems Applications. Journal of Vacuum Science and Technology A, 25, 1275-1284 (2007). 4. Nicholas Ferrell, Derek Hansford, Fabrication of Polymer Micro and Nanostructures by Soft Lithography and Spin Dewetting. Macromolecular Rapid Communications, 28, 966-971 (2007). 5. Manuel Palacio, Bharat Bhushan, Nicholas Ferrell, Derek Hansford, Nanomechanical Characterization of Polymer Beam Structures for BioMEMS Applications. Sensors and Actuators A: Physical, 135, 637-650 (2007).

vii

6. Randall T. Butler, Nicholas J. Ferrell, Derek J. Hansford, Spatial and Geometrical Control of Biosilification Using a Patterned Poly-L-Lysine Template. Applied Surface Science, 252, 7337-7342 (2006). 7. Jingjiao Guan, Nicholas Ferrell, L. James Lee, Derek J. Hansford, Fabrication of Polymeric Microparticles for Drug Delivery by Soft Lithography. Biomaterials, 27, 4034-4041 (2006). 8. Guohua Wei, Bharat Bhushan, Nicholas Ferrell, Derek Hansford, Fabrication and Nanomechanical Characterization of Polymer MEMS for Biological Applications. Journal of Vacuum Science and Technology A, 23, 811-819 (2005). 9. Jennifer Lewis, Mark Kotur, Omar Butt, Sumant Kulkarni, Alyssa Riley, Nick Ferrell, Kathryn Sullivan, Mauro Ferrari, Biotechnology Apprenticeship for Secondarylevel Students: Teaching Advanced Cell Culture Techniques for Research. Cell Biology Education, 1, 26-42 (2002).

FIELD OF STUDY Major Field: Biomedical Engineering

viii

TABLE OF CONTENTS Page Abstract ii Dedication... iv Acknowledgmentsv Vita.vii List of Tables ....xiii List of Figures ...xiv Chapters Page

1. Microfabrication and Microelectromechanical Systems (MEMS) ................................. 1 1.1 Introduction............................................................................................................... 1 1.2 Microfabrication and MEMS.................................................................................... 2 1.2.1 Photolithography................................................................................................ 2 1.2.2 Silicon MEMS fabrication ................................................................................. 6 1.2.3 Applications of silicon MEMS ........................................................................ 10 1.2.4 Silicon-based BioMEMS ................................................................................. 10 1.2.5 Microfluidics.................................................................................................... 14 1.2.6 Polymer microfabrication ................................................................................ 17 1.2.7 Applications of polymer MEMS...................................................................... 21 1.3 Conclusion .............................................................................................................. 24 2. Cell Mechanics.............................................................................................................. 31 2.1 Introduction............................................................................................................. 31 2.2 Cell mechanics and cell forces................................................................................ 32 2.2.1 The cytoskeleton .............................................................................................. 32 2.2.2 Forces generated by adherent cells .................................................................. 34 2.2.3 The cytoskeleton and disease........................................................................... 37 2.2.4 Drugs and cell mechanics ................................................................................ 38 2.3 Measuring cell mechanics....................................................................................... 39 2.3.1 Appling forces to cells ..................................................................................... 39 2.3.2 Measuring the mechanical properties of cells.................................................. 40 2.3.3 Measuring forces generated by cells................................................................ 41 2.3.4 BioMEMS for measuring cell forces ............................................................... 43 ix

2.4 Conclusion .............................................................................................................. 46 3. Fabrication of Polymer Microstructures by Double Stamp Micromolding .................. 51 3.1 Introduction............................................................................................................. 51 3.2 Materials and Methods............................................................................................ 53 3.2.1 Double stamp micromolding process............................................................... 53 3.2.2 Process characterization................................................................................... 54 3.3 Results and discussion ............................................................................................ 55 3.4 Conclusion .............................................................................................................. 60 4. Spin Dewetting of Polymers on Polydimethylsiloxane (PDMS) Molds ..................... 62 4.1 Introduction............................................................................................................. 62 4.2 Materials and methods ............................................................................................ 65 4.2.1 Materials .......................................................................................................... 65 4.2.2 PDMS molding ................................................................................................ 65 4.2.3 Spin dewetting and pattern transfer ................................................................. 65 4.2.4 Process characterization................................................................................... 68 4.3 Results and discussion ............................................................................................ 68 4.4 Conclusion .............................................................................................................. 75 5. Lift-Off Processing for Fabricating Micropatterned Sulfonated Polyaniline ............... 78 5.1 Introduction............................................................................................................. 78 5.2 Materials and methods ............................................................................................ 80 5.2.1 SPAN synthesis................................................................................................ 80 5.2.2 Micromolding and Lift-off processing............................................................. 80 5.2.3 Process characterization................................................................................... 82 5.3 Results and discussion ............................................................................................ 83 5.4 Conclusion .............................................................................................................. 87 6. Microfabricated Membranes for Cell Isolation............................................................. 90 6.1 Introduction............................................................................................................. 90 6.2 Materials and methods ............................................................................................ 93 6.2.1 Materials .......................................................................................................... 93 6.2.2 Membrane fabrication...................................................................................... 93 6.2.3 Cell culture and filtration ................................................................................. 95 6.2.4 Characterization ............................................................................................... 97 6.3 Results and discussion ............................................................................................ 97 6.4 Conclusion ............................................................................................................ 104 7. Fabrication of Suspended Polymer Microstructures by Patterned Susbstrate Micromolding and Sacrificial Layer Micromolding....................................................... 107 7.1 Introduction........................................................................................................... 107 x

7.2 Materials and methods .......................................................................................... 109 7.2.1 PDMS mold fabrication ................................................................................. 109 7.2.2 Sacrificial layer micromolding ...................................................................... 110 7.2.3 Patterned substrate micromolding.................................................................. 113 7.3 Results and discussion .......................................................................................... 115 7.4 Conclusion ............................................................................................................ 120 8. Measuring the Mechanical Properties of Polymer Microstructures by Nanoindentation ......................................................................................................................................... 123 8.1 Introduction........................................................................................................... 123 8.2 Materials and methods .......................................................................................... 128 8.2.1 Fabrication of polymer microstructures and thin films.................................. 128 8.2.2 Mechanical characterization .......................................................................... 129 8.3 Results and discussion .......................................................................................... 134 8.3.1 Hardness and elastic modulus by CSM nanoindentaion................................ 134 8.3.2 Creep behavior by CSM nanoindenation....................................................... 135 8.3.3 Scratch resistance from nanindentation scratch test ...................................... 136 8.3.3 Elastic modulus from normal beam bending ................................................. 138 8.3.4 Effects of aqueous environment and temperature.......................................... 141 8.3.5 Yield and breaking strength from normal beam bending .............................. 143 8.3.6 Lateral beam bending..................................................................................... 145 8.4 Conclusion ............................................................................................................ 146 9. Desing, Simulation, and Fabrication of a Polymer Sensor for Measuring Single Cell Forces.............................................................................................................................. 149 9.1 Introduction........................................................................................................... 149 9.2 Materials and methods .......................................................................................... 151 9.2.1 Solid modeling for device visualization ........................................................ 151 9.2.2 Finite element analysis................................................................................... 151 9.2.3 Device development and prototyping process ............................................... 151 9.2.4 Device fabrication.......................................................................................... 152 9.2.5 Device characterization.................................................................................. 154 9.3 Results and discussion .......................................................................................... 154 9.3.1 Design considerations .................................................................................... 154 9.3.2 Device overview ............................................................................................ 155 9.3.3 First generation cell force sensor ................................................................... 157 9.3.4 Second generation cell force sensor............................................................... 162 9.3.5 Third generation cell force sensor.................................................................. 170 9.3.6 Determining the force vectors from displacement data ................................. 177 9.4 Conclusion ............................................................................................................ 181 Biological Testing of the Polymer MEMS Cell Force Sensor........................................ 183 10.1 Introduction......................................................................................................... 183 xi

10.2 Materials and method.......................................................................................... 185 10.2.1 Sample preparation and surface modification.............................................. 185 10.2.2 Experimental setup....................................................................................... 186 10.2.3 Cell culture techniques................................................................................. 188 10.2.4 Cell placement ............................................................................................. 188 10.2.5 Image acquisition and analysis .................................................................... 189 10.3 Results and discussion ........................................................................................ 190 10.3.1 Sample preparation ...................................................................................... 190 10.3.2 Experimental results with first and second generation devices ................... 192 10.3.3 Results with third generation sensors .......................................................... 195 10.3.4 Effects of chemical exposure on cell forces ................................................ 205 10.3.5 Sources of error and measures to minimize error ........................................ 207 10.4 Conclusion .......................................................................................................... 209 11. Measurement of Mechanical Forces Generated by Plant P-Protein Aggregates (Forisomes) ..................................................................................................................... 212 11.1 Introduction......................................................................................................... 212 11.2 Materials and methods ........................................................................................ 213 11.2.1 Design .......................................................................................................... 213 11.2.2 Finite element simulations ........................................................................... 214 11.2.3 Fabrication and Characterization ................................................................. 214 11.2.4 Forisome preparation ................................................................................... 215 11.2.5 Experimental setup....................................................................................... 216 10.3 Results and discussion ........................................................................................ 216 11.4 Conclusion .......................................................................................................... 222 Conclusion and Future Outlook ...................................................................................... 224 Appendices A. Design and Simulation Data for the Cell Force Sensor ............................................ 228 B. Polymer Microfabrication as a Tool for Processing Inorganic Materials .................. 239 List of References ........................................................................................................... 258

xii

LIST OF TABLES Table 1.1 8.1 10.1 10.2 A.1 A.2 A.3 Page Typical etch processes for silicon based materials......9 Mechanical properties of PPMA, PMMA, PS, and PS/Clay ...140 Cell force calculations for second generation cell force sensor. ..194 Cell force calculations for second generation cell force sensor. ..195 Equations for calculating force angle..... 234 Equations for calculating deflection per unit force......237 Spreadsheet for calculating force vectors from displacement data..238

xiii

LIST OF FIGURES Figures Page

Figure 1.1 Schematic diagram of the photolithography process..........................................2 Figure 1.2 SEM micrographs of photoresist patterns ......6 Figure 1.3 Etch profiles for silicon etch processes.......9 Figure 1.4 Silicon MEMS drug delivery device ........12 Figure 1.5 Immunoisolation biocapsule.........12 Figure 1.6 Microcantilever sensor .........14 Figure 1.7 Microscale hot embossing system ...........16 Figure 1.8 Schematic diagram of the PDMS molding process ........18 Figure 1.9 Soft lithography processes ........20 Figure 1.10 Polymer/metal tactile sensor...........22 Figure 1.11 Microfabricated polymer devices for controlled drug delivery ..........23 Figure 1.12 Microfabricated tissue engineering scaffolds .........24 Figure 2.1 Cytoskeletal proteins ........33 Figure 2.2 Molecular components of focal adhesions ...........35 Figure 2.3 Cell motility process .............36 Figure 2.4 Traction force microscopy ........42 Figure 2.5 Silicon MEMS heart cell force sensor ..........44 Figure 2.6 Silicon cantilever fibroblast cell force sensor ..........45 xiv

Figure 2.7 PDMS pillar cell force sensor...........45 Figure 3.1 Schematic diagram of the microtransfer molding process... ........52 Figure 3.2 Schematic diagram of the double stamp micromolding process ..........54 Figure 3.3 Optical micrographs of polystyrene spin coated on PDMS. ........56 Figure 3.4 SEM micrographs of four steps in the double stamp micromolding process ...57 Figure 3.5 Characterization of feature distortion versus polymer concentration.......59 Figure 3.6 Characterization of feature distortion versus transfer pressure ........59 Figure 4.1 Schematic diagram of spin dewetting and pattern transfer. .........67 Figure 4.2 AFM images of polystyrene structures fabricated by spin dewetting on a 2 m diameter pillar mold ...........69 Figure 4.3 Optical micrographs of PPMA dewetting on 20 m diameter pillars.......70 Figure 4.4 Particles size versus (a) solution concentration and (b) mold feature size .......72 Figure 4.5 SEM micrographs at each step in the spin dewetting and pattern transfer process73 Figure 4.6 Dewetting of polystyrene is PDMS wells ........74 Figure 4.7 Nanoscale polystyrene features fabricated by spin dewetting..........74 Figure 5.1 Schematic diagram of the lift-off process for SPAN patterning.. ........81 Figure 5.2 SEM micrographs of sacrificial layers, SPAN deposition of sacrificial layers, and SPAN patterns after lift-off .........83 Figure 5.3 Film thickness versus time for static and dynamic SPAN deposition ..........85 Figure 5.4 Roughness versus time for static and dynamic SPAN deposition............85 Figure 5.5 SEM micrographs of 20 m circular SPAN patterns ...........86 Figure 6.1 Schematic diagram of the cell isolation membrane fabrication process. .....95 Figure 6.2 Experimental setup for cell isolation 96 xv

Figure 6.3 SEM micrographs of cell isolation membranes............98 Figure 6.4 SEM micrographs of NIH 3T3 cells isolated in circular, hexagonal, and square membrane features ...........99 Figure 6.5 SEM micrographs of NIH 3T3 cells after removing the membrane ..........100 Figure 6.6 Fluorescent micrographs of NIH 3T3 cells at low seeding density............101 Figure 6.7 Fluorescent micrographs of NIH 3T3 cells at high seeding density ..101 Figure 6.8 (a) SEM micrograph and (b) fluorescent micrograph of non-adherent THP-1 cells in circular membrane features .............102 Figure 6.9 (a,b) Fluorescent micrographs and (c) SEM micrograph of C3A liver cell clusters in circular membrane features 103 Figure 7.1 Schematic diagram of the sacrificial layer micromolding process.. ...111 Figure 7.2 Schematic diagram of the patterned substrate micromolding process ...114 Figure 7.3 SEM microstructures fabricated by sacrificial layer micromolding...115 Figure 7.4 Polymer cantilevers (a) before and (b) after release of the sacrificial layer. ..116 Figure 7.5 Polymer cantilever deflected with a micropipette......117 Figure 7.6 Polymer beams fabricated by patterned substrate micromolding...118 Figure 7.7 AFM images of features with and without proper selection of processing parameters ........120 Figure 8.1 Schematic diagram of the continuous stiffness measurement loading cycle..125 Figure 8.2 Schematic diagram of the process for fabricating cantilevers for lateral bending .129 Figure 8.3 SEM micrographs of polymer beams for bending and nanoindentation testing...130 Figure 8.4 Nanoindentation experimental setup ......133 Figure 8.5 Hardness and elastic modulus measured using CSM nanoindentation ......135 Figure 8.6 Creep measurements for PPMA, PMMA, PS, and PS/Clay...136 xvi

Figure 8.7 Scratch resistance and corresponding SEM micrographs showing film damage.137 Figure 8.8 Load-displacement plot for beam bending with an uncoated tip....138 Figure 8.9 Load-displacement plots for beam bending with PPMA, PMMA, PS, and PS/Clay beams .........139 Figure 8.10 Effects of aqueous environment and increased temperature on polymer properties..........142 Figure 8.11 Load-displacement plots for beam bending at high loads ....144 Figure 8.12 Lateral bending tests with PS and PS/Clay ......145 Figure 9.1 Schematic diagram of the fabrication process for the cell force sensor.. ...153 Figure 9.2 Circular configuration of the first generation cell force sensor ......156 Figure 9.3 Linear configuration of the first generation cell force sensor ....157 Figure 9.4 L-beam cantilever and direction-angle convention ........158 Figure 9.5 Deflection plot for the first generation sensor ........160 Figure 9.6 SEM micrographs of the first generation cell force sensor ....161 Figure 9.7 Optical micrographs of cantilever bending ........162 Figure 9.8 L-beam line of highest sensitivity ......163 Figure 9.9 Hinge in the second generation beam design .........164 Figure 9.10 CAD images of second generation cell force sensor ........165 Figure 9.11 FEA simulation of second generation cantilever beam ....166 Figure 9.12 Deflection plot for the second generation sensor .........167 Figure 9.13 Comparison of the deflection plots for the first and second generation sensor. ..168 Figure 9.14 SEM micrographs of the second generation sensors made from PPMA and polystyrene .......169 xvii

Figure 9.15 Phase contrast micrographs of the sensor after sacrificial layer release ......170 Figure 9.16 Area of HT 1080 and 3T3 cells before and after spreading .....171 Figure 9.17 Angle convention for third generation device ......172 Figure 9.18 FEA loading conditions ........173 Figure 9.19 Deflection plots for different FEA loading conditions .....174 Figure 9.20 Deflection plots for different measurement locations ......175 Figure 9.21 Comparison of the deflection plots for the first, second, and third generation cell force sensors ......176 Figure 9.22 SEM micrographs of the third generation sensor .........177 Figure 9.23 Plot of force angle versus beam deflection.......178 Figure 9.24 Plot of deflection per unit force versus force angle ......179 Figure 10.1 Experimental setup for force measurements ........187 Figure 10.2 Contact angle for native, tissue culture treated, and O2 plasma modified polystyrene .......191 Figure 10.3 Optical micrographs of a fibroblast on the second generation sensor ......194 Figure 10.4 Optical micrographs of a fibroblast on the third generation sensor .197 Figure 10.5 Force magnitude versus time plot for a WS-1 fibroblast cell.. .199 Figure 10.6 Force direction versus time plot for a WS-1 fibroblast cell .199 Figure 10.7 Force magnitude versus time plot for a WS-1 fibroblast cell.. .201 Figure 10.8 Force direction versus time plot for a WS-1 fibroblast cell .201 Figure 10.9 Force and direction versus time plots for a WS-1 fibroblast cell .203 Figure 10.10 Force and direction versus time plots for a WS-1 fibroblast cell ...204 Figure 10.11 Optical micrographs of a fibroblast cell on the sensor before and after exposure to cytochalasin-D ..206 xviii

Figure 10.12 Force versus time plot for a WS-1 fibroblast cell exposed to cytochalasin-D .207 Figure 11.1 Forisome force sensor...214 Figure 11.2 SEM micrographs of the forisome force sensor ...217 Figure 11.3 Comparison of FEA and analytical solutions for forisome sensor forcedeflection response...219 Figure 11.4 Optical micrographs of forisome before and after actuation ....220 Figure 11.5 SEM micrograph of a forisome on the sensor after testing ..221 Figure 11.6 Radial versus longitudinal forisome forces ..222 Figure A.1 Deflection plot for third generation cell force sensor 229 Figure A.2 Curve fits for calculating force direction ...229 Figure A.3 Curve fits for calculating force per unit deflection234 Figure B.1 Schematic diagram of the silicon wet etch process ...241 Figure B.2 SEM micrographs of silicon after TMAH etching 242 Figure B.3 SEM micrographs of NaCl crystal on PDMS pillars .244 Figure B.4 SEM micrographs of sucrose particles on PDMS pillars ..245 Figure B.5 Plot of crystal/particle size versus NaCl and sucrose concentration .246 Figure B.6 SEM micrographs of ammonium heptamolybdate hydrate particles PDMS pillars246 Figure B.7 SEM micrograph of poly-l-lysine particles on PDMS pillars247 Figure B.8 Schematic diagram of the peptide mediated deposition process.. .249 Figure B.9 SEM micrographs and EDS spectra of peptide mediated deposition of silica .250 Figure B.10 SEM micrographs of peptide mediated gold deposition..250 Figure B.11 Schematic diagram of the sol-gel micromolding process ....252 xix

Figure B.12 SEM micrographs of silica micropatterns fabricated by sol-gel micromolding...252 Figure B.13 SEM micrographs and EDS spectra for sol-gel micromolded silica/HA nanoparticle composites ...253 Figure B.14 SEM micrographs and EDS spectra for selective sol-gel micromolding of silica/HA nanoparticle composites ..255 Figure B.15 Optical micrographs of bone marrow cells grown on patterned and flat silica.256 Figure B.16 SEM micrographs of bone marrow cells grown on micropatterned silica and silica/HA nanoparticles composites .256

xx

CHAPTER 1 MICROFABRICATION AND MICROELECTROMECHANICAL SYSTEMS (MEMS)

1.1 Introduction Micro and nanoscale devices hold great potential in a wide range of technological fields. Micro and nanotechnology concepts have infiltrated almost every scientific discipline including chemistry, physics, engineering, materials science, and medicine. Research efforts in these areas are continuing to add to a growing body of knowledge and push toward applying these concepts in ways that could have a significant impact on society. Micro and nanoscale materials and devices have already found commercially applications ranging from the automotive to the clothing industries. As these research and commercialization efforts continue, no doubt additional progress will be made toward reaching the potential that has been predicted for this field. Traditionally micro and nanoscale devices, often referred to as micro or nanoelectromechanical systems (MEMS and NEMS), have been fabricated from siliconbased materials using processing techniques adapted from the semiconductor industry. Polymer materials are increasingly being utilized in micro and nanoscale biomedical devices. This trend is driven by a number of factors. Many polymer materials are known to be relatively biocompatible, which allows them to be used either as implant materials or for in vitro applications with a minimal detrimental effect on the host tissue or cells. 1

Polymers also have a wide range of properties with respect to bulk chemistry, surface chemistry, and mechanical and physical properties. This broad range of properties provides a large materials selection pool and allows materials to be chosen based on the appropriate behavior in a specific application. In the case of micro and nanodevices, processing costs can become a prohibitive factor due to the specialized materials, equipment, and facilities needed to manufacture such devices. Polymer processing can offer some distinct advantages in minimizing both materials and processing costs. Many polymer materials have significantly lower cost relative to traditional materials used in micro and nanodevices. Much of the processing can also be performed outside of a cleanroom environment without expensive equipment, significantly reducing processing costs. 1.2 Microfabrication and MEMS 1.2.1 Photolithography Microfabrication is a term used to collectively describe a set of techniques used to fabricate miniaturized devices for use in a variety of applications. One of the most widely used techniques in microfabrication is the photolithography process. Photolithography is a process of patterning photosensitive polymer (photoresist) structures. The pattern is created by selectively exposing the photoresist to ultra violet (UV) light. The UV exposure induces a chemical reaction in the photoresist that renders the materials either soluble (positive resist) or insoluble (negative resist) in a chemical developer. The photolithography process schematic for a negative photoresist is shown in Figure 1.1. 2

Figure 1.1 Schematic diagram of the photolithography process for a negative photoresist.

The photolithography process can be broken down into six basic steps: substrate preparation, pre-bake, exposure, post-exposure processing, developing, and cleaning/drying. The first step is preparation of the substrate. Generally single crystal silicon wafers are used, but glass, GaAs, and other materials are also used depending on the application. The wafers are usually cleaned using piranha or the RCA cleaning process.1 Piranha is a mixture of sulfuric acid (H2SO4) and hydrogen peroxide (H2O2) at approximately a 3:1 ratio, respectively. The RCA cleaning procedure is an aqueous 3

mixture of hydrogen peroxide and ammonia (NH4OH) or hydrochloric acid (HCl). For certain photoresists it is necessary to apply an adhesion promoter to improve the adhesion between the wafer surface and the resist. Often this is done by vapor phase deposition of a silane such as hexamethyldisilazane (HMDS). The next step in the process is deposition of the photoresist. This is done by spin coating the resist at speeds ranging from 1000-6000 rpm. The spin coating process can be more complicated than one might expect. In its simplest form, the spin coating process is a balance of the rotational inertial forces and the viscous shear of the fluid. The equilibrium between these two factors leads to a uniform film thickness after the coating process. Of course, several other factors contribute to the process. The use of volatile solvents in the photoresists leads to a time varying viscosity of the fluid. Exhaust flow, edge effects, and equipment variations also need to be considered. A number of mathematical models of varying complexity have been developed to describe the spin coating process.2,3 Process datasheets from the photoresist manufacturer usually contain spin speed versus film thickness curves. These give a good estimate of spin coating properties. However, given the differences in equipment and processing condition, it is usually necessary to create spin speed versus film thickness curves empirically for each photoresist. After spin coating, the photoresist is based to remove any residual solvent. After coating and baking, the wafer is selectively exposed to UV through a photomask. The mask is generally a high quality glass plate that is selectively coated with chrome or another absorbing metal layer. For larger features, a high quality printed mask (transparency mask) can be substituted for the chrome/glass mask.4 The mask is first loaded into the exposure system (mask aligner). The wafer is placed below the mask and 4

the two pieces are either brought close to each other (proximity mode) or brought together (contact mode). Other contact settings can be adjusted to improve the interface between the mask and the wafer. Hard contact mode applies a stronger force to hold the mask and wafer together and vacuum contact mode uses a vacuum to pull the wafer and mask into tighter contact. Improving the contact between the mask and the wafer improves the resolution of the pattern, but harder contact exposure increases wear of the mask. Photoresist can also be transferred from the wafer to the mask, resulting in possible defects in the pattern as well as requiring the mask to be cleaned more often. Post exposure steps are often needed to complete the process. For example, negative tone SU8 photoresists (Microchem Corp.) require a post exposure baking process to complete the cross linking reaction that is initiated by the UV exposure. After exposure and post-exposure processing, the wafer is ready for the development step. This process involves either immersion or spraying with a chemical developer to remove the unwanted portions of the photoresist. For positive resists, the exposed regions are removed, and for negative resist the unexposed regions are removed. The developer chemistry depends on the photoresist. Negative tone resists are generally developed in organic solvents, while positive resists are generally soluble in basic solutions (e.g. NaOH). After development, the wafers are cleaned in deionized (DI) water and dried. Scanning electron micrographs of patterns created by photolithography are shown in Figure 1.2. The images show features fabricated from both negative and positive photoresist. Figure 1.2 (a) is SU8 negative photoresist and (b) is S1813 positive resist positive resist (Shipley). 5

Figure 1.2 SEM micrographs of (a) 5 m diameter SU8 pillars and (b) 5 m diameter S1813 pillars fabricated by photolithography.

1.2.2 Silicon MEMS fabrication The original microfabrication processes were developed for the microelectronics integrated circuit (IC) industry. By combining IC manufacturing techniques with siliconbased deposition and machining techniques, it is possible to integrate electrical and mechanical elements (e.g. sensors, actuators, motors, pumps, etc.) on a single silicon platform. Processing techniques for silicon-based MEMS fabrication vary widely depending on the materials and applications being considered. MEMS processing can broadly be separated into two categories, deposition and machining. Deposition techniques allow different silicon-based materials to be used as either sacrificial or structural materials in MEMS. Some commonly used structural materials are polycrystalline silicon (polysilicon) and silicon nitride (Si3N4). Silicon oxide (SiO2) is generally used as a sacrificial material but can be used as a structural material for certain applications. Other materials, both silicon and non-silicon based, have begun to gain attention for specific applications. For example, significant research has focused on silicon carbide (SiC) 6

MEMS for high temperature, extreme environment applications.5 Other examples include shape memory alloys6 and magnetic materials.7 Deposition of silicon-based materials is usually done via chemical means. Polysilicon and Si3N4 are usually deposited using chemical vapor deposition (CVD) or low pressure chemical vapor deposition (LPCVD). CVD and LPCVD depositions are performed in tube furnaces at high temperature. The reaction of gases introduced to the furnace leads to deposition of a solid material on the silicon wafer surface. Silicon oxide can also be deposited via CVD or LPCVD, but it can also be grown by thermal oxidation of a silicon substrate. Physical methods are also used to deposit materials for MEMS applications. Physical vapor deposition (PVD) processes do not rely on any chemical reaction for the deposition. PVD processes include sputter coating and electron beam (e-beam) evaporation and are often used for metal deposition in MEMS processing. After thin film deposition and photolithography, materials are selectively removed, or etched, to produce the structural or mechanical elements of MEMS. Processes for selectively removing materials are referred to as micromachining techniques. Theses processes are categorized as either surface or bulk micromachining. In the case of bulk micromachining, the wafer material itself is used as the structural material for the device. In surface micromachining, a material deposited on the surface of the wafer is used as the structural material. Methods of micromachining silicon-based materials can be separated into several categories depending on the nature of the etching process and the resulting etch profile. Wet versus dry etching refers to the physical conditions in which the etching is 7

performed. Wet processes use liquid chemical etchants that dissolve the solid material. Dry etches use reactive gases to etch the materials by either chemical (plasma etching) or a combination of physical and chemical processes (reactive ion etching). The resulting etch profile determines if the etch is anisotropic or isotropic. Isotropic etches remove material at an equal rate independent of direction. Anisotropic etches preferentially remove material in a specific direction. The selectivity can either be based on the crystallographic orientation of the substrate or on the directionality of the etch process. For example, potassium hydroxide (KOH) etches silicon around 400 times faster in the <100> crystallographic direction compared to the <111> direction.8 The resulting etch profile for a <100> wafer gives a V-shaped groove with a 54.74 angle between the wafer surface the sidewall of the groove. Directionality in dry etching (plasma and RIE) is not based on the crystallography of the substrate, but is controlled by the etch conditions. Isotropic or anisotropic etching can be achieved depending on the etching parameters. The difference in the etch profiles for isotropic and anisotropic etches are shown in Figure 1.3. Various gases are used for dry etching processes. Some common examples are SF6, CF4, and Cl2. Gas mixtures that incorporate O2, He, and other gases are also used to increase etch rates, control etch directionality, or for cooling.9 Table 1.1 gives some examples of different silicon based etching processes and their characteristics. Note that this table is not a comprehensive list of etch processes, and only provides a few commonly used examples. For a more complete list, see Madous book Fundamentals of Microfabrication.8

Figure 1.3 Typical profiles for silicon based etches.

Etch Chemistry HNA KOH TMAH HF, Buffered HF CF4 CHF3 SF6 Cl2

Etch Material Si Si Si SiO2 SiO2 Si3N4 Si/PolySi Si/PolySi

Mask Material Wet or dry Si3N4 Si3N4/SiO2 SiO2Si3N4 Photoresist Photoresist Photoresist Photoresist Photoresist wet wet wet wet dry dry dry dry

Table 1.1 Typical etch processes for silicon based materials.

Another important step in the micromachining process is selecting a masking material. The mask protects the underlying material during the etching process. A high degree of selectivity is desirable, especially for long or deep etches. The selectivity refers to the relative etch rate of the substrate material compared to the masking material. Some 9

examples of typical masking materials for a given etch are shown in Table 1.1. In the case of dry etches with photoresist etch masks, the photoresists typically have relatively poor selectivity. This requires that the etch depth be rather small or that a thick photoresist layer be used in order to provide sufficient masking for longer etches. 1.2.3 Applications of silicon MEMS Silicon MEMS have found commercial applications in a number of different industries. According to a report by BCC Research, the market for MEMS processing equipment and devices was estimated $5 billion in 2005 and is expected to grow by more than 20% per year to more than $12.5 billion by 2010.10 Another report by Yole Dveloppement estimates a MEMS devices market of around $6 billion in 2006 with an increase to approximately $11 billion dollars in 2011.11 One example of a commercially successful device is the MEMS accelerometer, which is widely used in automotive airbag deployment system. MEMS devices have also found commercial applications in manufacturing of ink jet printer heads, optical devices, pressure sensors, and various radio frequency device (RF MEMS).12 Significant research efforts in industry and academia continue to focus on silicon MEMS device development, processing, materials, and reliability. 1.2.4 Silicon-based BioMEMS Medicine and biology have been significantly impacted, both from a research and commercial perspective, by the development of MEMS technology. Silicon devices have been considered for drug delivery devices, immunoisolation devices, and sensors, just to

10

name a few. Several review articles are available that summarize research efforts in these areas.13-16 Figure 1.4 shows an example of a silicon MEMS device for drug delivery.17 The device consists of an array of drug containing microreservoirs that can be triggered to release a specific quantity or type of drug. The release mechanism is based on electrochemically induced dissolution of metal covers on top of the reservoirs. One of the advantages of this device is that the release can be externally controlled to allow a continuous or time varying release profile. Another advantage of the individually addressable reservoirs is the ability to release multiple different drugs as needed. The device was tested in a dog model and showed steady release of drug over the course of 6 months.18 A more recent version of this device is being commercialized by MicroCHIPS, Inc.

11

Figure 1.4 Silicon MEMS microchip for controlled drug delivery (17).

Figure 1.5 Immunoisolation biocapsule for insulin delivery (19).

The device in Figure 1.5 is designed for immunoisolation of xenografted pancreatic islet cells for the treatment of diabetes.19 The device uses a nanoporous membrane with very small (down to 7 nm) pores to isolate the encapsulated cells from 12

the host immune response while allowing size specific exchange of certain materials. The membrane with 18 nm pores was shown to significantly reduce Immunoglobulin G (IgG) diffusion while maintaining insulin diffusion. The device was tested in a mouse model and effectively provided immunoisolation of the encapsulated cells over a short period of time (1 week).20 Silicon MEMS are also used in a number of biomedical sensors. Several different detection mechanisms are used in bioMEMS sensors. Some of the more common are mechanical, optical and electrochemical detection. As an example, we will consider mechanical microcantilever sensors. Figure 1.6 shows a schematic of the detection mechanism for this type of sensor.21 A target molecule is bound to the surface of the microcantilever. Binding of a receptor molecule creates a stress that causes the cantilever to deflect. The amount of bending is proportional to the amount of bound receptor. This concept has been used to characterize DNA hybridization21 and was sensitive enough to detect a single base pair mismatch between two ssDNA molecules. The same concept has been used to measure receptor-ligand binding. In experiments by Wu et al.22 microcantilevers were used to measure low concentrations (down to 6 ng/ml) of prostate specific antigen (PSA), a key marker for prostate cancer. Microcantilevers have also been used to detect larger particles such as viruses.23 In this case, a change in the resonant frequency of the cantilever was used as the detection mechanism. When virus particles attach to the surface of the cantilever, the mass increases which leads to a decrease in the resonant frequency. This sensor is capable of measuring the attachment of a single virus.

13

Figure 1.6 Detection mechanism for microcantilever sensors. Binding between receptor and target molecules creates a surface stress that bends the cantilever. (21)

1.2.5 Microfluidics One of the most commercially successful and well-researched areas of the MEMS field is microfluidic devices. Microfluidic devices are current being explored for applications that include medical diagnostics,24 DNA and proteomic analysis,25-27 and drug screening.27 The basic concept of microfluidics involves the miniaturization and automation of analytical equipment to allow complete analysis on a single chip with little or no user intervention. Miniaturization of analytical devices allows portability, parallel analysis, rapid response, and improved efficiency, while automation allows devices to be used by non-skilled operators. These devices could potentially allow point of care 14

medical diagnostics in a fraction of the time currently needed for traditional laboratory testing. Microfluidic DNA and proteomics chips have the potential to vastly improve the time and efficiency of DNA and protein sample preparation, separation, and analysis. It is not difficult to imagine the potential improvements in patient care and laboratory technology that could be achieved through these devices. However, the integration of all the necessary control components and sensing mechanisms needed for a functional microsystem continues to be an elusive problem. In order to achieve the automation and miniaturization that are desired in the microfluidic device, a host of different functional components must be simultaneously incorporated into a single, small scale device. These components can include valves, pumps, reaction chambers, separation mechanisms, and detectors. Given the small size of many of the components of microfluidic systems, microfabrication has logically been the method of choice for fabrication of microfluidic systems. A number of different materials and microfabrication techniques have been employed. Silicon and glass were the first materials to be considered due to their common use in microfabrication and MEMS, but for cost considerations and ease of fabrication, polymer based systems have begun to dominate the field. Fabrication methods for polymer microfluidic devices vary significantly from silicon MEMS. Polymer devices are usually fabricated from thermoplastic polymers like poly(methyl methacrylate) (PMMA) or polycarbonate or from the elastomer polydimethylsiloxane (PDMS). Devices made from thermoplastic polymers are commonly fabricated by hot embossing or injection molding.28,29 The first step in the hot embossing process is fabrication of a mold. The mold is usually made from a hard 15

material such as a metal (e.g. nickel) or silicon or from a polymer with a high glass transition temperature relative to the substrate material. The mold is generally fabricated using standard microfabrication processes (photolithography, etching, etc.) sometimes in combination with other processes (e.g. electroplating). The mold is then used as a tool to fabricate multiple polymer devices. The substrate and/or mold are heated above the glass transition temperature of the substrate polymer. Pressure is then applied to the mold to transfer the channel pattern into the polymer. An image of a commercially available microscale hot embossing system (EVG 520, EV Group) is shown in Figure 1.7. The system consists of two heated/cooled plates with computer controlled hydraulics that apply pressure to the plates. The pressure and temperature versus time profiles can both be programmed.

Figure 1.7 (a) Commercially available microscale hot embossing system (EVG 520), (b) close-up of the heated plates.

Another commonly used method for fabrication of polymer microfluidic is soft lithography. Soft lithography covers a wide range of fabrication techniques that involve 16

replication of microstructures in a soft material (PDMS).30 The PDMS microstructures can then be used directly in device applications (such as microfluidics) or as tools for further fabrication processes. The use of PDMS microstructures as microfabrication tools will be discussed in detail in a later section. The Effenhauser group was also one of the first to use PDMS for microfluidics.31 They focused specifically on polydimethylsiloxane (PDMS) as the structural material in a protein separation device. Duffy et al.4 further expanded the use of PDMS as a materials in microfluidic systems. In addition to the reduction in cost resulting form the use of PDMS, the photomask used in the initial fabrication process was printed on a high resolution printer instead of the expensive methods used in traditional photomask fabrication. They also developed a method for covalently sealing PDMS to a number of substrates with a simple exposure to an oxygen plasma. Oxygen plasma also changes the surface chemistry of the PDMS, rendering it hydrophilic, versus its native, highly hydrophobic state. This allows for easier filling of the PDMS microchannels with polar fluids such as water and biological fluids. The charged surface of oxidized PDMS (SiO-) also provides a negatively charged surface that is conducive to electroosmotic flow.32 A significant amount of research has since focused on PDMS microfluidic devices for a wide range of applications. Several review papers have been published in this area.33-35 1.2.6 Polymer microfabrication In addition to microfluidic devices, polymers are being used more frequently in other microfabricated devices. Many polymer microfabrication techniques are based on soft lithography. Figure 1.8 shows the process for fabrication of PDMS molds for soft lithography. A photoresist master is first fabricated using photolithography or another 17

micro/nanofabrication procedure. The PDMS prepolymer is mixed with a curing agent and poured over the master and allowed to cure. The master pattern is transfer to the PDMS creating a negative replica of the original pattern.

Figure 1.8 Schematic diagram for fabricating a PDMS mold from a photoresist master.

Whitesides et al. developed the original soft lithography techniques and subsequently used them to create micropatterns and microstructures from a variety of materials. Microtransfer molding (TM)36 and micromolding in capillaries (MIMIC)37 use microfabricated PDMS molds as templates for further replication of polymer structures. In TM, a prepolymer is applied to the PDMS mold. The excess polymer on top of the mold is removed and the mold is applied to a substrate. The prepolymer is then 18

cured using UV or heat. This process will be discussed in more detail in Chapter 3. For the MIMIC process, the mold is applied to the substrate and a prepolymer solution is applied to one side of the mold. The prepolymer fills the features in the mold either by capillary action or using vacuum. The prepolymer is then cured and the mold is removed. Related soft lithography techniques include replica molding, solvent assisted micromolding,38 and microcontact printing (CP).39 In solvent assisted micromolding, the PDMS is used to imprint a polymer that is dissolved in a solvent. The solvent diffuses through the mold over time, leaving only the polymer in the microfeatures. In CP, the PDMS mold is inked with a specific chemistry, then the mold is brought into contact with a substrate, selectively transferring the chemical pattern to the substrate. This process is generally used for micropatterning self assembled monolayers (SAMs) or proteins.40 Each of these processes are shown in Figure 1.9.

19

Figure 1.9 Schematics for soft lithography processes (38).

Since the development of these initial processing techniques, several variations or related techniques have been developed.41,42 These processes have greatly broadened the versatility of the soft lithographic microfabrication techniques. In addition to processing traditional polymers, significant process has also been made in micropatterning

20

functional materials such as conducting polymers,43-45 piezoelectric polymers,46,47 biodegradable polymers,48,49 and hydrogels.50 While soft lithography processes have found broad applications, they are generally limited to fabrication of two-dimensional (2D) or extruded 2D structures. For fabrication of more complicated three-dimensional shapes or suspended mechanical structures, more complicated techniques are required. In some cases, traditional microfabrication techniques such as photolithography with SU8 photoresist and specialized etching and sacrificial layer techniques can be used to fabricate structural components in polymer MEMS.51,52 In other cases, more complicated techniques such as two-photon lithography53,54 and microstereolithography55 can be used to fabricate 3D polymer structures. However, these processes have additional limitations such as high cost, diminished resolution, and low throughput. 1.2.7 Applications of polymer MEMS Microfabricated polymer structures have found applications in a number of areas, either in the form of all polymer devices or as components in MEMS. Polymers are used in electronic and optical devices. For example, conducting polymers have been used to make all organic field effect transistors (FETs).56,57 Soft lithography has been used to fabricate polymer waveguides.58 Polymers have also been used as components in various sensors (pressure sensors, accelerometers, etc.).59 Figure 1.10 shows a hybrid polymer/metal tactile sensor capable of measuring hardness, thermal conductivity, temperature and surface contour.60

21

Figure 1.10 Hybrid polymer/metal tactile sensor (60).

Polymer microfabrication has also been used extensively in the biomedical field. One of the advantages of microfabrication and MEMS is the ability to interface with cells on their size scale. This ability to interact with, manipulate, and perform measurements on a single cell and biomolecular level has lead to new devices and therapeutic approaches. Some examples include drug delivery devices,61,62 tissue engineering scaffolds,46,47 and cell micromanipulation devices.63 Figure 1.11 shows several examples of microfabricated drug delivery devices. Figure 1.11 (a) is a polymer microreservior system for drug delivery. Biodegradable polymers are used to provide controlled drug release. Figure 1.11 (b-d) shows microfabricated biodegradable polymer microparticles for drug delivery. Microsphere delivery systems are used in several commercial drug delivery systems to give long term, sustained release.64 Microfabrication provides a method to produce monodisperse particles that could potentially provide a more stable release profile than particles produced using current manufacturing methods. Microfabrication also allows fabrication of multiple particle geometries as opposed to 22

only spheres. Flat square or circular particles, high and low aspect ratio cylinders, and many other geometries can be fabricated. Particle geometry could be used to give specific release profiles.

Figure 1.11 (a) Polymer microchip drug delivery device (65), (b) square poly(lactic-coglycolic acid) (PLGA) microparticles fabricated by soft lithography (61), (c) high aspect ratio and (d) low aspect ratio PLGA microparticles fabricated by soft lithography (66). Scale bars equal 50 m, 5 m, and 20 m in (b), (c), and (d), respectively.

Figure 1.12 shows two polymer tissue engineering scaffolds fabricated using soft lithography. Microfabrication is often used in tissue engineering to provide highly controlled cellular environments at the microscale.67 This is an important factor in determining cell behavior. Microfabrication techniques allow precise control of microarchitecture to engineer structure with well-defined pore size and porosity. Figure 23

12 (a) shows a PLGA tissue engineering scaffold fabricated by embossing and CO2 assisted bonding68 and (b) shows a polycaprolactone (PCL) scaffold fabricated using multilayer micromolding.49

Figure 1.12 Microfabricated (a) PLGA (68) and (b) PCL (49) tissue engineering scaffolds.

1.3 Conclusion MEMS have found applications in wide range of industries and continue to be a focus area in the research community. Silicon micromachining techniques including 24

deposition and silicon micromachining are well established and well characterized processes. One of the major focuses of MEMS research and commercialization efforts has been in the biomedical field. Using microdevices to interface with biology on a cellular and molecular level has led to new diagnostic, therapeutic, and sensor applications. As the field of MEMS has progressed, polymers have been increasingly considered as structural and functional components for specific applications. Polymers offer several advantages over silicon-based materials, including biocompatibility, diverse chemical, physical, and mechanical properties, and low cost materials and processing techniques. The push toward incorporating polymers into MEMS devices has led to the development of new microfabrication techniques to meet the unique challenges of polymer processing. Some of the focus areas for polymer MEMS research are microfluidic analytical devices, sensors, drug delivery devices, and tissue engineering. As research in these areas continues and new applications are explored, advances will no doubt be made toward the development of low cost, functional polymer MEMS devices for a wide range of applications in biomedicine and other fields. References W. Kern (ed.), Handbook of Semiconductor Wafer Cleaning Technology: Science, Technology, and Applications, Noyes, Park Ridge, NJ (1993). A.G. Ernslie, F.T. Bonner, L.G. Peck, Flow of a viscous liquid on a rotating disk. Journal of Applied Physics 29 (1958) 858-862.
3 2 1

D. Meyerhofer, Characteristics of resist films produced by spinning. Journal of Applied Physics 49 (1978) 3993-3997. D.C. Duffy, J.C. McDonald, O.J.A. Schueller, G.M. Whitesides, Rapid prototyping of microfluidic systems in poly(dimethylsiloxane). Analytical Chemistry 70 (1998) 49744984. 25

M. Mehregany, C.A. Zorman, SiC MEMS: opportunities and challenges for applications in harsh environments. Thin Solid Films 355-356 (1999) 518-524. Y. Fu, H. Du, W. Huang, S. Zhang, M. Hu, TiNi-based thin films in MEMS applications: a review. Sensors and Actuators A 112 (2004) 395-408.
7 6

D. Niarchos, Magnetic MEMS: key issues and some applications. Sensors and Actuators A 109 (2003) 166-173.

M. Madou, Fundumentals of Microfabrication, CRC Press, Boca Raton, FL (1997) pp.171. H. Janseny, H. Gardeniers, M. de Boer, M. Elwenspoek, J. Fluitman, A survey on the reactive ion etching of silicon in microtechnology. Journal of Micromechanics and Microengineering 6 (1996) 14-28.
10 9

Microelectromechanical Systems (MEMS) Technology: Current And Future Markets, BCC Research (2006).
11

Global MEMS/Microsystems - Markets and Opportunities, Yole Dveloppement (2007).

12

W.H. Ko, Trends and frontiers of MEMS. Sensors and Actuators A 136 (2007) 6267.
13

R.S. Shawgo, A.C. Richards Gradyson, Y.W. Li, M.J. Cima, BioMEMS for drug delivery. Current Opinions in Solid State Materials Science 6 (2002) 329-334.
14

A.C. Richards Grayson, R.S. Shawgo, Y. Li, M.J. Cima, Electronic MEMS for triggered delivery. Advanced Drug Delivery Reviews 56 (2004) 173-184.
15

S.L. Tao, T.A. Desai, Microfabricated drug delivery systems: from particles to pores. Advanced Drug Delivery Reviews 55 (2003) 315-328.
16

R. Bashir, BioMEMS: state-of-the-art in detection, opportunities and prespects. Advanced Drug Delivery Reviews 56 (2004) 1565-1585.
17

J. Santini, M.J. Cima, R.J. Langer, A controlled-release microchip. Nature 397 (1999) 335-338. M. Staples, K. Daniel, M.J. Cima, R. Langer, Application of micro- and nanoelectromechanical devices to drug delivery. Pharmaceutical Research 23 (2006) 847863.
18

26

19

T.A. Desai, D. Hansford, M. Ferrari, Characterization of micromachined silicon membranes for immunoisolation and bioseparation applications. Journal of Membrane Science 159 (1999) 221-231.
20

T.A. Desai, W.H. Chu, G. Rasi, P. Sinibaldi-Vallebona, E. Guarino, Microfabricated biocapsules provide short-term immunoisolation of insulinoma xenografts.Biomedical Microdevices 1 (1999) 131-138.
21

J. Fritz, M.K. Baller, H.P. Land. H. Rothuizen, P. Vettiger, E. Meyer, H.-H. Gntherodt, Ch. Gerber, J.K. Gimzewski, Translating biomolecular recognition into nanomechanics. Science 288 (2000) 316-318.
22

G. Wu, R.H. Datar, K.M. Hansen, T. Thundat, R.J. Cote, A. Majumdar, Bioassy of prostate-specific antigen (PSA) using microcantilevers. Nature Biotechnology 19 (2001) 856-860.
23

A. Gupta, D. Akin, R. Bashir, Single virus particle mass detection using microresonators with nanoscale thickness. Applied Physics Letters 84 (2004) 19761978.
24

T.H. Schulte, R.L. Bardell, B.H. Weigl, Microfluidic technologies in clinical diagnostics. Clinica Chimica Acta 321 (2002) 1-10.
25

G.H.W. Sanders, A. Manz, Chip-based Microsystems for genomic and proteomic analysis. Trends in Analytical Chemistry 19 (2000) 364-378.
26

Yu, X.; Zhang, D.; Li, T.; Hao, L.; Li, X. 3-D microarrys biochip for DNA amplification in polydimethylsiloxane (PDMS) elastomer. Sensors and Actuators A 108 (2003) 103-107.
27

M.T. Cronin, T. Boone, A.P. Sassi, H. Tan, Q. Xue, S.J. Williams, A.J. Ricco, H.H. Hooper. Plastic microfluidic systems for high-throughput genomic analysis and drug screening. Journal of the Association for Laboratory Automation 6 (2001) 71-75.
28

H. Becker, C. Grtner, Polymer microfabrication methods for microfluidic analytical applications. Electrophoresis 21 (2000) 12-26.

H. Becker, C. Grtner, Polymer microfabrication technologies for microfluidic systems. Analytical and Bioanalytical Chemistry 390 (2008) 89-111.
30

29

Y. Xia, G.M. Whitesides, Soft lithography. Angewandte Chemie International Edition 37 (1998) 550-575.

C.S. Effenhauser, G.J.M. Bruin, A. Paulus, M. Ehart, Analytical Chemistry 69 (1997) 3451-3457. 27

31

32

J.S. Buch, P.-C. Wang, D.L. DeVoe, C.S. Lee, Field-effect flow control in a polydimethylsoloxane based microfluidic system. Electrophoresis 22 (2001), 39023907.
33

J.C. McDonald, D.C. Duffy, J.R. Anderson, D.T. Chiu, H. Wu, O.J.A. Scheuller, G.M. Whitesides, Fabrication of microfluidic systems in poly(dimethylsiloxane). Electrophoresis 21 (2000), 27-40.
34

S.K. Sia, G.M. Whitesides, Microfluidic devices fabricated in poly(dimethylsiloxane) for biological studies. Electrophoresis 24 (2003), 3563-3576.
35

J.M.K. Ng, I. Gitlin, A.D. Stroock, G.M. Whitesides, Components for integrated poly(dimethylsiloxane) microfluidic systems. Electrophoresis 23 (2002), 3461-3473.
36

X.-M. Zhoa, Y. Xia, and G.M. Whitesides, Fabrication of three-dimensional microstructures: microtransfer molding. Advanced Materials 8 (1996) 837-840.
37

E. Kim, Y. Xia, and G.M. Whitesides, Polymer microstructures formed by moulding in capillaries. Nature 376 (1995) 581-584.
38

Y. Xia, G.M. Whitesides, Soft lithography. Annual Review of Materials Science 28 (1998) 153-184. A. Kumar, G.M. Whitesides, Features of gold having micrometer to centimeter dimensions can be formed through a combination of stamping with an elastomeric stamp and an alkanethiol ink followed by chemical etching. Applied Physics Letters 61 (1993) 2002-2004.
40 39

R.S. Kane, S. Takayama, E. Ostuni, D.E. Ingber, G.M. Whitesides, Patterning proteins and cells using soft lithography. Biomaterials 20 (1999) 2363-2376. J.A. Rogers, R.G. Nuzzo, Recent progress in soft lithography. Materials Today 8(2) (2005) 50-56.
42 41

M. Wang, H.-G. Braun, T. Kratzmller, E. Meyer, Patterning polymers by microfluid-contact printing. Advanced Materials 13 (2001) 1312-1317.
43

W.S. Beh, I.T. Kim, D. Qin, Y. Xia, G.M. Whitesides, Formation of patterned microstructures of conducting polymers by soft lithography, and applications in microelectronic device fabrication. Advanced Materials 11 (1999) 1038-1041.
44

J.A. Rogers, Z. Boa, V.R. Raju, Nonphotolithographic fabrication of organic transistors with micron feature sizes. Applied Physics Letters 72 (1998) 2716-2718. 28

45

T. Granlund, T. Nyberg, L.S. Roman, M. Svensson, O. Ingans, Patterning of polymer light-emitting diodes by soft lithography. Advanced Materials 12 (2000) 269-273. B. Xu, F. Arias, S.T. Brittain, X.-M. Zhao, B. Grzybowski, S. Torquato, G.M. Whitesides, Making negative Poissons ratio microstructures by soft lithography. Advanced Materials 11 (1999) 1186-1189. D. Gallego, N. Ferrell, D. Hanford, Fabrication of Piezoelectric Polyvinylidene Fluoride (PVDF) Microstructures by Soft Lithography for Tissue Engineering and Cell Biology Applications. in Printing Methods for Electronics, Photonics and Biomaterials, edited by G. Gigli (Mater. Res. Soc. Symp. Proc. Volume 1002E, Warrendale, PA, 2007) 1002-N04-05.
48 47 46

G. Vozzi, C. Flaim, A. Ahluwalia, S. Bhatia, Fabrication of PLGA scaffolds using soft lithography and microsyringe deposition. Biomaterials 24 (2003), 2533-2540.
49

D. Gallego, N. Ferrell, Y. Sun, D.J. Hansford, Multilayer micromolding of degradable polymer tissue engineering scaffolds. Materials Science and Engineering C (in press).
50

J. Guan, H. He, D.J. Hansford, L.J. Lee, Self-folding of three-dimensional hydrogel microstructures. The Journal of Physical Chemistry B 109 (2005) 23134-23137.
51

G. Genolet, J. Brugger, M. Despont, U. Drechsler, P. Vettiger, N.F. de Rooij, D. Anselmetti, Soft, entirely photoplastic probes for scanning force microscopy. Review of Scientific Instruments 70 (1999) 2398-2401.
52

X. Wang, J. Engel, C. Liu, Liquid crystal polymer (LCP) for MEMS: processes and applications. Journal of Micromechanics and Microengineering 13 (2003) 628-633.
53

G. Witzgall, R. Vrijen, E. Yablonovitch, V. Doan, B.J. Schwartz, Single-shot twophoton exposure of commercial photoresist for the production of three-dimensional structures. Optics Letters 23 (1998) 1745-1747.
54

T. Baldacchini, C.N. LaFratta, R.A. Farrer, Acrylic-based resin with favorable properties for three-dimensional two-photon polymerization. Journal of Applied Physics 95 (2004) 6072-6076.
55

A. Bertsch, H. Lorenz, P. Renaud, 3D microfabrication by combining microstereolithography and thick resist UV lithography. Sensors and Actuators A 73 (1999) 14-23.
56

W.S. Beh, I.T. Kim, D. Qin, Y. Xia, G.M. Whitesides, Formation of patterned microstructure of conducing polymer by soft lithography, and applications in microelectronic device fabrication. Advanced Materials 11 (1999) 1038-1041. 29

57

H. Sirringhaus, T. Kawase, R.H. Friend, T. Shimoda, M. Inbasekaran, W. Wu, E.P. Woo, High-resolution inkjet printing of all-polymer transistor circuits. Science 290 (2000) 2123-2126.
58

X.-M. Zhao, S.P. Smith, S.J. Waldman, G.M. Whitesides, M. Prentiss, Demonstration of waveguide couplers fabricated using microtrasnfer molding. Applied Physics Letters 71 (1997) 1017-1019.
59

C. Liu, Recent developments in polymer MEMS. Advanced Materials 19 (2007) 3783-3790.

60

J. Engel, J. Chen, Z. Fan, C. Liu, Polymer micromachined multimodal tactile sensors. Sensors and Actuators A 117 (2005) 50-61.
61

G. Guan, N. Ferrell, L.J. Lee, D.J. Hansford, Fabrication of polymeric microparticles for drug delivery by soft lithography. Biomaterials 27 (2006) 4034-4041.
62

S.L. Tao, T.A. Desai, Microfabrication of multilayer, asymmetric, polymeric devices for drug delivery. Advanced Materials 17 (2005) 1625-1630.
63

N. Chronis, L.P. Lee, Electrothermally activated SU-8 microgripper for single cell manipulation in solution. Journal of Microelectromechanical Systems 14 (2005) 857863.
64

M. Ferrari, A. Lee, L.J. Lee (eds.), Biological and Biomedical Nanotechnology, Springer, Boston, MA (2006).
65

A.C. Richards Grayson, I.S. Choi, B.M. Tyler, P.P. Wang, H. Brem, M.J. Cima, R. Langer, Multi-pulse drug delivery from a resorbable polymeric microchip device. Nature Materials 2 (2003) 767-772. A. Chakrapani, Processing and characterization of polymer microparticles for controlled drug delivery systems. Ph.D. Thesis, Ohio State University (2007).
67 66

S.N. Bhatia, C.S. Chen, Tissue engineering at the microscale. Biomedical Microdevices 2 (1999) 131-144. Y. Yang, S. Basu, D.L. Tomasko, L. James Lee, S.-T. Yang, Fabrication of welldefined PLGA scaffolds using microembossing and carbon dioxide bonding. Biomaterials 26 (2005) 2585-2594.
68

30

CHAPTER 2 CELL MECHANICS

2.1 Introduction Cell mechanics is a broad field that covers a range of mechanical processes, structural elements, and mechanical properties as they relate to cells and cell behavior. Cell processes such as adhesion and motility have a significant mechanical component. Less intuitive and less understood is the role of mechanical forces in regulating other biochemical processes that ultimately control cell behavior. This interrelationship between mechanical and biochemical processes is currently an active area of research often referred to as mechanotrasduction or chemomechanical transduction.1 Mechanical cues act as regulatory signals to control intracellular biochemical processes. External factors like extracellular matrix or substrate stiffness and external shear stress are involved in mediating a host of cell processes including gene expression, differentiation, and proliferation. Cells are also responsible for generating forces that are transmitted to the external environment. In fibroblast cells, for example, intracellular forces are transmitted to the extracellular space and play a critical role in cell movement, adhesion, and signal transduction. The breadth and depth of the field of cell biomechanics makes a comprehensive review of the field impractical. A few specific areas of cell mechanics will be covered as 31

they relate to the development and testing of the polymer MEMS cell force sensor described in Chapters 9 and 10. Topics such as muscle cell mechanics or mechanics of non-adherent cells will not be discussed in detail. Several measurement devices and techniques for evaluating and measuring cell mechanics will be covered. 2.2 Cell mechanics and cell forces 2.2.1 The cytoskeleton The cytoskeleton is set of proteins in the cell that act as the primary load bearing and force generating component of the cell. The cytoskeleton consists of three primary proteins: actin, microtubules, and intermediate filaments. A host of associated proteins and protein complexes are responsible for polymerization and depolymerization of cytoskeletal components, linking cytoskeletal component to each other and to the extracellular space, and regulating cytoskeletal function. Figure 2.1 shows the three molecular components of the cytoskeleton. F-actin, which is the filament form of the monomer G-actin, is the primary structural component of most cell types. Actin filaments consist of twisted strands of actin and have a diameter of 7-9 nm. Actin filaments are polar molecules with a positive barded end where polymerization takes place and a negative pointed end where depolymerization occurs. Actin filaments are responsible in large part for maintaining the structure and morphology of the cells. Polymerization and depolymerization of actin filaments is also the primary mechanism of cell motility in adherent cells such as fibroblast. Microtubules are tubular dimeric proteins consisting of alternating strands of and tubulin. Microtubules are larger than actin filaments, with an outside diameter of 32

about 25 nm. They are also polar, and polymerize faster from the plus end. Microtubules also play a structural role in the cell as well as being involved cell division and transport processes. Microtubules work in connection with the motor proteins dynein and kinesin to provide directional transport of organelles.

Figure 2.1 Cytoskeletal proteins, (A) actin filament, (B) microtubule, and (C) intermediate filament (2). 33

Intermediate filaments are large subset of cytoskeletal proteins. In terms of size, their diameter is between that of actin filaments and microtubules (~10 nm), thus the name intermediate filaments. The assembly process for intermediate filaments has several general steps as shown in Figure 2.1. Polypeptides first assemble into coiled dimers then into tetramer. The individual tetramers assemble into a protofilament. Protofilments then assemble in filaments. Unlike actin filaments and microtubules, intermediate filaments are highly stable. 2.2.2 Forces generated by adherent cells When adherent cells are attached to the extracellular matrix or a tissue culture surface, stresses are generated within the cytoskeleton. Cytoskeletal stresses are responsible for maintaining the structural integrity of the cell and regulating cellular shape and spreading. Molecular linkages between the cytoskeleton and focal adhesion complexes (FACs) transmit forces to the extracellular space. Figure 2.2 shows the molecular mechanism of force transmission.3 Within FACs, actin filaments are connected to transmembrane integrin receptors via protein linkages. Various proteins and protein complexes are involved in linking actin filaments to integrins. Some examples include actinin, talin, paxillin, and viculin.2,3 Integrins are then attached either to the extracellular matrix (in vivo) or to proteins on the cell culture substrate (in vitro).

34

Figure 2.2 Molecular components of a focal adhesion (3).

Recent research has shown that mechanical interactions that occur at the cellextracellular matrix interface are critical in regulating a host of cellular processes. Cell motility requires that cells generate the necessary forces in order to move.4 Figure 2.3 shows the basic processes involved in cell motility. Protrusions, in the form of filopodia or lamellapodia are formed at the lead edge of the cell. These areas are rich in highly dynamic actin filaments. Contractile forces are created in the cytoskeleton. Focal adhesions at the trailing edge of the cell are released, thus propelling the cell in the direction toward the leading edge.

35

Figure 2.3 Cell motility process (2).

Intracellular tension, force transduction across the cell membrane, and associated processes such as cell shape and spreading have been show to affect a number of other cellular processes in addition to cell motility.5,6 A number of interesting experiments have been performed to illustrate this point. By applying mechanical stress directly to integrin receptors via magnetic twisting, Wang et. al7 demonstrated assembly of focal adhesions and stiffening of the cytoskeleton proportional to the applied stress. These experiments are evidence that the integrins are directly involved in transmitting forces between the cytoskeleton and the extracellular matrix. By chemically inhibiting actin filament, microtubule, and intermediate filament function independently, they also showed that all three elements of the cytoskeleton are involved in the mechanical signal transduction process. Chen et al. used micropatterned cell adhesion molecules to control the degree of cell spreading in bovine and human endothelial cells. These experiments showed that restricting the cells ability to spread induced apoptosis.8 This work demonstrated that cells have the ability to use geometrical factors as regulatory cues to determine cell fate. Chrzanowska-Wodnicka and Burridge9 showed that rhoA, a GTP binding protein, 36

increases cell contractility corresponding to the formation of focal adhesions and stress fibers in fibroblast cell. These experiments are just a sampling of the evidence for the bidirectional link between mechanical and biochemical processes in cells. Several review and commentary articles are available with more information on the theory and experimental evidence of the role of cell and cytoskeletal mechanics in regulating cell function.5,6,10-12 2.2.3 The cytoskeleton and disease In addition to normal cell function, a number of pathologies have a direct effect on cytoskeletal function. A prime example is the case of metastatic cancers. Metastasis is largely dependant on the cells ability to migrate from one location to another. As demonstrated previously, the process of cell motility is largely dependant on cytoskeletal dynamics. In a number of cancers, the regulatory mechanisms that control the cytoskeleton do not function properly. In some cases, direct mutations of cytoskeletal actin have been observed. In other cases, mutation or overexpression of regulatory proteins such as members of the WASP family of proteins or Rho GTPases could be responsible for changes in cytoskeletal function.13,14 The cytoskeleton is directly related to a number of other diseases.15 Mutations in keratin intermediate filaments in epithelial cells are responsible for a range of skin disorders.16,17 These disorders are often characterized by fragile skin and blistering. This illustrates the important role of intermediate filament in maintaining the structural integrity of cells under stress. Keratin intermediate filaments have also been implicated in various liver pathologies and could be a factor in intestinal disease.18,19 A lack of the actin 37

regulatory protein WASP has been implicated in the immune disorder Wiskott-Aldrich syndrome.20 2.2.4 Drugs and cell mechanics Given the number of pathologies associated with the cytoskeleton, it is no surprise that drugs have been developed to target specific cytoskeletal components. Various chemotherapeutic agents with cytoskeletal targets have been explored. Primarily these drugs have targeted the microtubules due to their important role in cell division. Paclitaxel (Taxol) is now commonly used to treat various cancers including lung, ovarian, and breast cancer. Paclitaxel stabilizes microtubules to hinder cell mitosis.21-22 Vinca alkaloids23 and colchicine24 also target microtubules and inhibit depolymerization to reduce cell proliferation. Vinca alkaloids are used widely used as chemotherapeutics and colchicine has been explored as an anti-cancer drug as well as being used as a treatment for gout. Cucurbinacin E has shown anti-proliferative effects in prostate cancer cells and been shown to induce significant disruption of the actin cytoskeleton.25 Several other chemicals, while not necessarily being actively explored for therapeutic applications, still have a significant influence on the cytoskeleton and have been indispensable in studying cytoskeletal behavior. Cytochalsins are a group of fungal derived toxins that interfere with actin polymerization primarily by binding the barbed end of actin filaments and decreasing the amount of monomeric actin in the cell.26 Cytochalasins, especially cytochalasin D, have been widely used to study the role of actin in various cellular processes. Phalliodin is another fungal toxin with a significantly different effect on actin. Phalloidin inhibits depolymerization of actin filaments. Phalloidin binds strongly to actin filaments and much less to monomeric actin. This shifts 38

the equilibrium between F-actin filaments and monomeric G-actin toward the formation of filaments.26 Jasplakinolide, a toxin derived from marine sponges, has been shown to induce actin polymerization in vitro27, though some in vivo observations are less straight forward.28 2.3 Measuring cell mechanics A wide array of different techniques and experimental methods have been developed to study cell mechanics. For the purposes of this discussion, the techniques will be broadly grouped to three types of experimental methods: (1) applying forces to cells, (2) measuring the mechanical properties of cells, and (3) measuring forces generated by cells. Each of these is important in its own right and considerable crossover exists between these methods with regards to their importance in understanding cell behavior. 2.3.1 Appling forces to cells Several methods have been developed to deliver different types of mechanical stimuli to various cell types. Externally applied shear stress is an important regulatory cue for many cell types, especially endothelial cells. Shear stress has been shown in induce cytoskeletal remodeling, changes cell morphology and orientation, and affect biological function in endothelial cells.29-31 One of the most common methods to study shear induced endothelial cell response is using a parallel plate flow through systems. This is a simple method for controlling shear stress on cells by controlling the flow rate through a chamber with fixed dimensions. The other commonly used method is the cone plate system. This method uses a rotating cone and a fixed plate to create a controlled shear 39

stress. These studies have been critical in understanding mechanotransduction pathways as well as understanding the specific shear stress induced changes to biochemical processes within endothelial cells. Mechanical forces have also been applied to cells using substrate deformation techniques.32,33 For these experiments, a flexible substrate is mechanically deformed by applying uniaxial tension, biaxial tension, bending, or a range of other mechanical loading conditions. These experiments have been used to study muscle, bone, cartilage, and a variety of other relevant cells. 2.3.2 Measuring the mechanical properties of cells Another active area of cell mechanics involves measuring the mechanical properties of cells and cellular components. Measurements have been carried out on whole cells or on specific regions of the cells. Given that the cytoskeleton is largely responsible for maintaining the structural integrity to the cell, there is a direct link between the elastic properties of the cell and cytoskeletal mechanics. One of the most commonly used tools to study cellular and subcellular mechanical properties is atomic force microscopy (AFM). AFM has been used to create elasticity maps to evaluate the local mechanical properties of several different cell types.34 These studies have been used to show variations in the elastic properties of a cell depending on location. These techniques has also been used to evaluate the effects of different chemicals on the mechanical properties cells.35 A significant decrease was observed in the elastic modulus of cells after exposure to actin disrupting agents (e.g. cytochalasins, jasplakinolide) at sufficient concentrations, but no significant changes to the elastic properties were observed in response to microtubule targeting agents (e.g. cholcicine, taxol). 40

Other methods for measuring the mechanical properties of whole cells include micropipette aspiration and optical tweezers.36 In the case of micropipette aspiration, a tip with an inside diameter smaller than the cell is used and a known pressure is applied to pull the cell into the pipette tip. The deformation of the cell is observed in a microscopy and the elastic and viscous properties of the cell can be calculated.37 Optical tweezers use laser light to apply a known force to a bead. By attaching the bead to a cell, the cell can be deformed and the force and deformation information can be used to calculate the mechanical properties. Optical tweezers have very good force resolution (~1 pN), but a limited range of force (~ 600 pN). Higher forces require higher laser power and local heating can damage the cell. This technique has primarily been applied to study of red blood cells38 and biomolecules.39 2.3.3 Measuring forces generated by cells Some of the seminal work studying forces at the cell-surface interface was done by Harris et al.40 who grew cells on cross-linked silicone rubber substrates and observed the wrinkling of the polymer film. One limitation with this method is the difficulty in quantifying the forces. Since this initial work, improved methods have been developed that incorporate fluorescent microbeads into polyacrylamide gels.41 The displacement of the beads and the stiffness of the gel are used to calculate forces exerted by the cell. This method is commonly referred to as traction force microscopy.42 While calculation of the forces is still not straightforward, this process does allow quantification of spatially resolved forces applied to the substrate. Figure 2.3 shows (a) the bead displacement vectors overlaid on a phase contrast image of the cell and (b) a corresponding colorcoded traction stress map for a NIH 3T3 fibroblast. 41

Figure 2.4 Traction force microscopy for measuring fibroblast forces; (a) Displacement vector field overlaid on a phase contrast image of the cell and (b) corresponding stress map.

Some interesting work has been done looking at the effects of substrate mechanical properties on cell behavior. Pelham and Wang43 used polyacrylamide gels with a range of stiffness to show that the formation of focal adhesions and motility of epithelial (normal rat kidney) and fibroblasts (3T3) cells is dependant on the stiffness of the substrate. Lo et al.44 took this work a step further and showed that 3T3 cells preferentially migrate from the softer regions to the stiffer regions of the substrate. Cells also responded to locally induced stresses created with a microneedle and migrated 42

toward the regions of higher tension and away from region of compression. They used particle tracking methods to measure the forces generated by the cells, and showed that the forces generated by cells on the stiffer substrates were higher than that on the softer materials. These experiments demonstrate several important issues. First, the cells sense the mechanical properties of their surroundings and response accordingly. This work also demonstrates a link between substrate and cytoskeletal mechanics given that the forces exerted on the soft materials were less that those on the stiffer material. 2.3.4 BioMEMS for measuring cell forces Several bioMEMS devices, both silicon and polymer, have been developed for measuring cell mechanics. Figure 2.4 shows an SEM micrograph of a silicon MEMS heart cell force transducer.45 The device consists of a clamp to hold the cell in place. The clamp is connected to a set of thin polysilion beams that are then attached via a slider restraint to the substrate. The heart cell is attached to the clamp using silicone sealant When the cell is stimulated to contract (by varying [Ca2+]), the polysilicon beams are bent toward the center of the device. By measuring the magnitude of the displacement, the force can be calculated. While this device clearly demonstrates measurement of mechanical forces generated by cells, it is designed to measure the relatively large forces generated by cardiomyocytes. The general measurement methodology, namely using cantilever beam deflection to measure cell forces, can be applied more broadly.

43

Figure 2.5 Polysilicon heart cell force sensor (45).

Galbraith and Sheetz developed a silicon-based cantilever device for measuring fibroblast cell forces.46 The device consists of an array of single cantilever beams with a pad at the end of each beam as shown in Figure 2.5. When the cell attaches to the pad and exerts a force, the cantilever is bent proportional to the applied load. Using this device, they were able to measure traction forces exerted by fibroblasts at various locations on the cell (i.e. lead edge, nucleus, and trailing edge) to determine the spatial distribution of cell forces. They were also able to measure the time varying force. One disadvantage of this device is that deflection of the beam can only be measured in the direction orthogonal to the cantilever direction. The force had to be calculated by making the assumption that all of the force was oriented in the direction of the long axis of the cell (i.e. the direction of movement).

44

Figure 2.6 Silicon-based cantilever force sensor for measuring fibroblast traction forces (46).

Figure 2.7 PDMS device for measuring fibroblast contractile forces. The image shows the PDMS pillars and the cell with the force vector for each pillar (47).

45

More recently, microfabricated polymer structures have been developed for measuring cell forces.47 The device in Figure 2.6 shows an array of PDMS micropillars that act as vertical cantilevers. When the cell attaches to the pillars and exerts a force, the pillars are deflected. Using the stiffness of the pillar and the deflection, the force exerted on each PDMS pillar was calculated. They were able to measure the time dependence of the force and by micropatterning adhesion molecules on the surface of the pillars, they showed that the force per pillar is proportional to the amount of cell spreading. 2.4 Conclusion Cell mechanics plays a critical role in regulating cell function. Mechanical processes are involved in cell adhesion, movement, signal transduction, and even play a role regulating cell death. Studying the cytoskeleton both from a biochemical and mechanical perspective has led to a greater understanding of cell mechanics and mechanotransduction. A variety of measurement devices and methods have been developed to investigate these processes. Continued work in this area will lead to a more complete understanding of cell mechanics and how these factors play a larger role in maintaining normal tissue function. Another critical area is understanding the role of cell mechanics in disease and using this information to develop and test drugs and new therapeutic approaches. References A.W. Orr, B.P. Helmke, B.R. Blackman, M.A. Schwartz, Mechanisms of mechanotransduction. Developmental Cell 10 (2006) 11-20.
2 1

M.R.K. Mofrad, R.D. Kamm (eds.), Cytoskeletal Mechanics, Cambridge University Press, Cambridge (2006). 46

A. Nicolas, B. Geiger, S.A. Safran, Cell mechanosensitivity control the anisotropy of focal adhesions. Proceedings of the National Academy of Sciences USA 101 (2004) 12520-12525. M.P. Sheetz, D.P. Felsenfeld, C.G. Galbraith, Cell migration: regulation of force on extracellular matrix-integrin complexes. Trends in Cell Biology 8 (1998) 51-54. M.E. Chicurel, C.S. Chen, D.E. Ingber, Cellular control lies in the balance of forces. Current Opinions in Cell Biology 10 (1998) 232-239. C.G. Galbraith, M.P. Sheetz, Forces on adhesive contacts affect cell function. Current Opinions in Cell Biology 10 (1998) 566-571. N. Wang, J.T. Butler, D.E. Ingber, Mechanotransduction across the cell surface and through the cytoskeleton. Science 260 (1993) 1124-1127. C.S. Chen, M. Mrksich, S. Huang, G.M. Whitesides, D.E. Ingber, Geometric control of cell life and death. Science 276 (1997) 1425-1428. M. Chrzanowska-Wodnicke, K. Burridge, Rho-stimulated contractility drive the formation of stress fibers and focal adhesions. Journal of Cell Biology 133 (1996) 14031415.
10 9 8 7 6 5 4

D.E. Ingber, Cellular tensegrity revisited I: Cell structure and hierarchical systems biology. Journal of Cell Science 116 (2003) 1157-1173.
11

D.E. Ingber, Tensiegrity II. How structural networks influence cellular information processing networks. Journal of Cell Science 116 (2003) 1397-1408.
12

G. Bao, S. Suresh, Cell and molecular mechanics of biological materials. Nature Materials 2 (2003) 715-725.
13

A.E. Cress, R.B. Nagle (eds.), Cell Adhesion and Cytoskeletal Molecules in Metastasis, Springer, Dordrecht, The Netherlands (2006).
14

H. Yamaguchi, J. Condeelis, Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochimica et Beophysica Acta 1773 (2007) 642-652.
15

F.C.S. Ramaekers, F.T. Bosman, The cytoskeleton and disease. Journal of Pathology 204 (2004) 351-354.
16

E.B. Lane, W.H.I. McLean, Keratins and skin disorders. Journal of Pathology 204 (2004) 355-366.

47

17

E. Fuchs, The cytoskeleton and disease: genetic disorder of intermediate filaments. Annual Review of Genetics 30 (1996) 197-231.
18

K. Zatloukal, C. Stumptner, A. Fuchsbichler, P. Fickert, C. Lackner, M. Trauner, H. Denk, The keratin cytoskeleton in liver disease. Journal of Pathology 204 (2004) 367376.
19

D.W. Owens. E.B. Lane, Keratin mutations and intestinal pathology.Journal of Pathology204 (2004) 377-385.
20

Y. Calle, H.-C. Chou, A.J. Thrasher, G.E. Jones. Wiskott-Aldrich syndrome protein and the cytoskeletal dynamics of dendritic cells. Journal of Pathology 204 (2004) 460469.
21

P.B. Schiff, S.B. Horwitz, Taxol stabilizes microtubules in mouse fibroblast cells. Proceedings of the National Academy of Sciences USA 77 (1980) 1561-1565.
22

M.V. Blagosklonny, T. Fojo, Molecular effect of Paclitaxel: myths and reality. International Journal of Cancer 83 (1999) 151-156.
23

V.K. Ngan, K. Bellman, B.T. Hill, L. Wilson, M.A. Jordan, Mechanism of mitotic block and inhibition of cell proliferation by the semisynthetic vinca alkaloids vinorelbine and its newer derivative vinflunine. Molecular Pharmacology 60 (2001) 225-232.
24

G.G. Borisy, E.W. Taylor, The mechanism of action of colchicines. The Journal of Cell Biology 34 (1967) 525-533.
25

K.L.K. Duncan, M.D. Duncan, M.C. Ally, E.A. Sausville, Cucurbitacin E-induced disruption of the actin and vimentin cytoskeleton in prostate carcinoma cells. Biochemical Pharmacology 52 (1996) 1553-1560. J.A. Cooper, Effects of cytochalasin and phalloidin on actin. The Journal of Cell Biology 105 (1987) 1473-1478.
27 26

M.R. Bubb, A.M.J. Senderowiczf, E.A. Sausville, K.L.K. Duncan, E.D. Kern, Jasplakinolide, a cytotoxic natural product, induces actin polymerization and competitively inhibits the binding of phalloidin to F-actin. The Journal of Biological Chemistry 269 (1994) 14869-14871.
28

M.R. Bubb, I. Spector, B.B. Beyer, K.M. Fosen, Effects of Jasplakinolide on the kinetics of actin polymerization. The Journal of Biological Chemistry 275 (2000) 51635170.

48

29

T. Ohashi, M. Sato, Remodeling of vascular endothelial cells exposed to fluid shear stress: experimental and numerical approach. Fluid Dynamics Research 37 (2005) 4059.
30

O. Traub, B.C. Berk, Laminar shear stress: mechanicsms by which endothelial cells transduce an atheroprotective force. Arteriosclerosis, Thrombosis, and Vascular Biology 18 (1998) 677-685.
31

C.R. White, J.A. Frangos, The shear stress of it all: the cell membrane and mechanochemical tranduction. Philosophical Transactions of the Royal Society B 362 (2007) 1459-1467.
32

K.J. Van Vliet, G. Bao, S. Suresh, The biomechanics toolbox: experimental approaches for living cells and biomolecules. Acta Materialia 51 (2003) 5881-5905.
33

T.D. Brown, Techniques for mechanical stimulation of cell in vitro: a review. Journal of Biomechanics 33 (2000) 3-14.
34

T.G. Kuznetsova, M.N. Starodubtseva, N.I. Yegorenkov, S.A. Chizhik, R.I. Zhdanov. Atomic force microscopy probing of cell elasticity. Micron 38 (2007) 824-833.
35

C. Rotsch, M. Radmacher, Drug-induced changes of cytoskeletal structure and mechanics in fibroblasts: an atomic force microscopy study. Biophysical Journal 78 (2000) 520-535.
36

M.R.K. Mofrad, R.D. Kamm (eds.), Cytoskeletal Mechanics: Models and Measurements, Cambridge University Press, Cambridge (2006). R.M. Hochmuth, Micropipette aspiration of living cells. Journal of Biomechanics 33 (2000) 15-22.
38 37

M. Dao, C.T. Lim, S. Suresh, Mechanics of the human red blood cell deformed by optical tweezers. Journal of the Mechanics and Physics of Solids 51 (2003) 2259-2280.
39

A.D. Mehta, M. Rief, J.A. Spudich, D.A. Smith, R.M. Simmons, Single-molecule biomechanics with optical methods. Science 283 (1999) 1689-1695.
40

A.K. Harris, P. Wild, D. Stopak, Silicone rubber substrata: a new wrinkle in the study of cell locomotion. Science 208 (1980) 177-179.
41

M. Dembo, Y.-L. Wang, Stresses at the cell-to-substrate interface during locomotion of fibroblasts. Biophysical Journal 76 (1999) 2307-2316.
42

S. Munevar, Y.-L Wang, M. Dembo, Traction force microscopy of migrating normal and H-ras transformed 3T3 fibroblasts. Biophysical Journal 80 (2001), 1744-1757. 49

43

R.J. Pelham, Y.-L. Wang, Cell locomotion and focal adhesions are regulated by substrate flexibility. Proceedings of the National Academy of Sciences USA 94 (1997) 13661-13665.
44

C.-M. Lo, H.-B. Wang, M. Dembo, Y.-L. Wang, Cell movement is guided by the rigidity of the substrate. Biophysical Journal 79 (2000) 144-152.
45

G. Lin, K.SJ. Pister, K.P. Roos, Surface micromachined polysilicon heart cell force transducer. Journal of Microelectromechanical Systems 9 (2000) 9-17.
46

C.B. Galbraith, M.P. Sheetz, A micromachined device provides a new bend on fibroblast traction forces. Proceeding of the National Academy of Sciences USA 94 (1997) 9114-9118.
47

J.L. Tan, J. Tien, D.M. Pirone, D.S. Gray, K. Bhadriraju, C.S. Chen, Cells lying on a bed of microneedles: An approach to isolate mechanical force. Proceeding of the National Academy of Sciences USA 100 (1993), 1484-1489.

50

CHAPTER 3 FABRICATION OF POLYMER MICROSTRUCTURES BY DOUBLE STAMP MICROMOLDING

3.1 Introduction Micromolding techniques are now widely used for fabrication of microscale polymer components. Hot embossing1 and injection molding2 have been scaled down from their use in macroscale polymer processing to accommodate microscale fabrication. Several of the soft lithography based techniques also allow fabrication of polymer microstructures. Of particular interest to this work is the process of microtransfer molding. Microtransfer molding was introduced by Whitesides et al.3 in 1996. Since then, the process has been used for fabricating micro and nanoscale devices including waveguides,4 Schottky diodes,5 and magnetic separation devices.6 The process of microtransfer molding is shown in Figure 3.1. A polymer solution is applied to the surface of a micropatterned PDMS mold. The mold is then placed in contact with a substrate and the polymer solution is cured and the mold is removed. This process provides a simple and cost effective method to replicate microfeatures over large areas. One limitation of this technique is the residual film that remains between the polymer features after the mold is removed. This film can be removed by additional

51

processing such as oxygen plasma treatment, but the process requires specialized equipment and introduces additional processing costs.

Figure 3.1 Schematic diagram of the microtransfer molding (TM) process (7).

The process of double stamp micromolding is a modified microtransfer molding process that allows fabrication of individual polymer microstructures without the need for additional processing steps. The process adapted from work by Guan et al.8,9 A solution of a thermoplastic polymer is spin coated onto the surface of a PDMS mold. An initial stamping process with low pressure is used to remove the polymer material from the raised portions of the PDMS mold. A higher pressure stamping step is then used to remove the material from the recessed portions of the mold. The double stamp micromolding process is described in detail here, and the process is characterized with respect to the morphology of the features as a function of the polymer solution concentration and the transfer pressure. 52

3.2 Materials and Methods 3.2.1 Double stamp micromolding process Polystyrene was used as the model polymer is this study, but several other thermoplastic polymers have been patterned using minor variations of the same basic process. Other polymers that have been tested include poly(methyl methacrylate) (PMMA), poly(propyl methacrylate) (PPMA), polycaprolactone (PCL), and poly(lacticco-glycolic) acid (PLGA). The process is also flexible with respect to the substrate that can be used for pattern transfer. Rigid substrates such as silicon and glass as well as flexible substrates (polyimide and polycarbonate) have been used. The double stamp micromolding process was performed as described in Figure 3.2. The PDMS mold was spin coated with a polymer solution as shown in Figure 3.2 (b). The polymer on the surface of the mold is then removed using higher temperature and lower pressure. In the case of polystyrene, the temperature for the first stamp was 200 C and the pressure was ~2.5 psi. The polymer that remained in the recess features of the mold was then transferred to the substrate using lower temperature and higher pressure. The lower temperature of the second stamping was needed to maintain the resolution of the features and higher pressure was used to ensure that the polymer in the recesses of the mold made contact with the substrate during the transfer. A transfer temperature of 125 C and transfer pressures between 30 and 200 psi were used in the second stamping. The stamping was performed at varying polymer solution concentrations and pressures to determine if the coating condition or the transfer conditions have a more significant effect on the outcome of the feature morphology. 53

Figure 3.2 Schematic diagram of the double stamp micromolding process. (a) uncoated PDMS mold, (b) mold spin coated with polymer, (c) inverted mold in contact with heated glass to remove surface polymer, (d) polymer removed from the raised portions of the mold, (e) selectively coated mold, (f) inverted mold in brought into contact with the substrate with heat and pressure, (g) polymer structures after transfer to the substrate.

3.2.2 Process characterization Two different PDMS mold geometries were used to characterize the process. For SEM and optical characterization, a 5.3 m deep mold with 5 m wide channels and 45 m spacing between channels was used. The mold was spin coated with 3, 5, 7.5, and 10% (wt/wt) solutions of polystyrene in anisole. The solutions were spin coated onto the mold at 3000 rpm. Optical micrographs were taken to examine the film prior to stamping. For relatively low polymer concentrations, dewetting of the polymer is observed on the mold. The dewetting process and its utility in microfabrication will be 54

discussed in detail in Chapter 4. SEM micrographs were taken of the mold at several steps in the process. Images were taken of the empty mold prior to spin coating, the mold after spin coating with a 7.5% solution of PS, the mold after the first stamping to remove the surface materials, and the substrate after stamping the PS features. For AFM characterization, a 5.3 m deep mold could not be used due to the limited z-range of the AFM. Instead, a 3.3 m deep mold with 5 m wide channels and 5 m spacing was used. AFM scans were run to determine the morphology of the final feature cross section at a given polymer concentration. The features were characterized using tapping mode atomic force microscopy (AFM) with a 300 kHz tip, scan length of 25 m, and scan rate of 0.5 Hz. We define the distortion of the feature according to the sag, or the depression created in the center of the feature during pattern transfer. The sagging was characterized using the equation 1:
S= H max H min 100 H max

(1)

where S is defined as the percent sag, Hmax is the maximum height of the feature cross section, and Hmin is the minimum height of the cross section. 3.3 Results and discussion Figure 3.3 shows the dewetting of the PS on the PDMS mold at concentrations of 1%, 3%, 5%, 7.5%, and 10%. The images show that at 1% the polymer dewets and partially fills the channels in the mold. In addition, a line of polymer forms on the raised features of the mold. As the concentration increases, the wetting of the channels increases while the dewetted line on the raised features become wider. At 5% a small discontinuity exist between the PS in the features and that on the raised portions of the 55

mold. At 7.5% no dewetting occurs (i.e. the polymer film is continuous). Despite the fact that no dewetting occurs, the double stamping process can still be performed. In fact, the yield and feature morphology are more favorable at 7.5% than at any other concentration. The process can also be done with a 10% solution, however, the yield decreases significantly due to removal of material within the channels during the first stamping. At even higher polymer concentrations, the thickness of the film is such that the entire film is removed during the first stamp. This is an effective method for making continuous films of microstructures.

Figure 3.3 Optical micrographs of polystyrene at (a) 1%, (b) 3%, (c) 5%, (d) 7.5%, and (e) 10% PS concentration (wt/wt) in anisole.

Scanning electron micrographs of several steps in the process are shown in Figure 3.4. The empty PDMS mold is shown in (a). Figure 3.4 (b) shows the PDMS mold after spin coating with a 7.5% solution of PS. The images show that the film is continuous but 56

there is a topographical difference in the film on the raised portions of the mold versus the channels. Figure 3.4 (c) shows the mold after the first stamping with material only in the recessed channels. Finally, Figure 3.4 (d) shows the features after removal from the mold. The images show that the features are transferred with little distortion of the original feature morphology.

Figure 3.4 Scanning electron micrographs of four steps in the double stamp micromolding process; (a) uncoated PDMS mold, (b) mold after spin coating, (c) mold after the first stamping with polymer in the recesses of the mold, (d) polymer features after transfer to the substrate.

Figure 3.5 shows the sagging of the PS features after the second stamping as a function of the polymer solution concentration. The plot shows an inverse relationship between the feature sag and the polymer solution concentration. This is likely caused by a combination of two factors. First, at low concentration, the features are not filled 57

completely. The unoccupied space allows the mold to defect during the transfer process, leading to distortion of the features. Second, at low polymer concentration the channels are filled with a more curved profile due to increased retention of material at the wall of the channel versus in the bulk of the channel. As the concentration increases, these effects become less prevalent and the channels fill more completely and more uniformly. Figure 3.6 shows the effect of the transfer pressure on sag of the features for a 7.5% solution spin coated at 3000 rpm. The figure shows that there is a much less significant effect on the polymer features as the pressure is increased and no trend in the sag is observed with increased pressure. This indicates that the choice of polymer solution concentration has a much more significant impact on the feature morphology than transfer pressure used to remove the features from the mold. All scans were performed in triplicate and the error bars are the standard deviation.

58

50 45 40 35 30

3%

% Sag

7.5%

25 20 15 10 5 0 0 2 4 6 8 10 12

10% 5%

Polymer Concentration (%)

Figure 3.5 Plot of the feature distortion (sag) with increasing polymer concentration; the insets show cross sectional AFM scans at each of the four polymer concentrations tested.

50 45 40 35

% Sag

30 25 20 15 10 5 0 0 50 100 150 200 250

Pressure (psi)

Figure 3.6 Plot of the feature distortion (sag) with increasing pressure; a 7.5% PS solution spin coated at 3000 rpm was used to fabricate the structures. 59

3.4 Conclusion SEM, AFM and optical microscopy were used to characterize the double stamping of PS lines. The double stamping process can be used due to topographical differences between the polymer on the raised portions of the mold and polymer in mold recesses. As polymer concentration is further increased, a lower yield of polymer features is obtained due to removal of polymer from the channels during the first stamping process. The results also show that the polymer fills the mold in a predictable manner with increased filling of the features as the polymer solution concentration is increased. The transfer pressure used to remove the materials from the mold has much less effect on the overall feature morphology. This work shows that the double stamping process can be used to fabricate high quality polymer microfeatures. Characterization of the process is critical for obtaining high yield features with quality retention of the original feature morphology. References
1 2

H. Becker, U. Heim, Sensors and Actuators A: Physical 83, 130 (2000).

L.J. Lee, M.J. Madou, K.W. Koelling, S. Daunert, S. Lai, C.G. Koh, Biomedical Microdevices 3, 339 (2001). X.-M. Zhoa, Y. Xia, and G.M. Whitesides, Fabrication of three-dimensional microstructures: microtransfer molding. Advanced Materials 8 (1996) 837-840. X.-M. Zhao, S.P. Smith, S.J. Waldman, G.M. Whitesides, M. Prentiss, Demonstration of waveguide couplers fabricated using microtrasnfer molding. Applied Physics Letters 71 (1997) 1017-1019. J. Hu, T. Deng. R.G. Beck, R.M. Westervelt, G.M. Whitesides, Sensors and Actuators A: Physical 75, 65-69 (1999).
6 5 4 3

T. Deng, M. Prentiss, G.M. Whitesides, Applied Physics Letters 80, 461-463 (2002). 60

Y. Xia, G.M. Whitesides, Soft lithography. Annual Review of Materials Science 28 (1998) 153-184. J. Guan, A. Chakrapani, D.J. Hansford, Polymer microparticles fabricated by soft lithography. Chemistry of Materials 17 (2005) 6227-6229.

J. Guan, N. Ferrell, L.J. Lee, D.J. Hansford, Fabrication of polymeric microparticles for drug delivery by soft lithography. Biomaterials 27 (2006) 4034-4041.

61

CHAPTER 4 SPIN DEWETTING OF POLYMERS ON POLYDIMETHYLSILOXANE (PDMS) MOLDS

4.1 Introduction Dewetting is a process of spatial separation of a material caused by physically, chemically, or thermally induced instability in a thin solid or liquid film. Dewetting is a phenomenon that has been well documented in thin polymer films.1,2 In most cases, this is an undesirable effect that results in formation of holes in polymer films. However, researchers have begun to exploit this process to engineer surfaces that elicit a desired dewetting behavior for creating regularly structured polymer features. Primarily, this has been achieved by locally altering the chemical composition of the surface of interest.3-6 Dewetting of thin polymer films on chemically heterogeneous surfaces has been studied both theoretically and experimentally. The theory of polymer dewetting on chemically non-uniform surfaces has been described by Lenz et al.7 and Konnur et al.8 Microscale differences in the wettability of the substrate lead to instabilities in the film, which in turn leads to formation of a pattern. Experimentally, microcontact printing has been used for inducing specific dewetting behavior. By printing hydrophilic chemistries on hydrophobic substrates, a hydrophobic polymer such as polystyrene will preferentially

62

coat the hydrophobic regions of the substrate, thus creating a pattern.3,4 A similar concept applies to printing of hydrophobic chemistries on hydrophilic substrates.5,6 In addition to chemically induced dewetting, surface topography is also known to be a factor in the dewetting process. The effects of physical, chemical, and physiochemical heterogeneities have been studied theoretically by Konnur et al.9 Their model for physical dewetting was based on a sinusoidally rough surface and predicted polymer dewetting in the depressions of the rough surface. In contrast to chemically induced dewetting, physically induced dewetting is dominated by dynamic interactions between the polymer solution and the topographically modified surface. Bao et al.10 used a combination of chemical patterning and topography for fabricating micro and nanoscale poly(methyl methacrylate) (PMMA) and polycarbonate (PC) features. They coated a topographically patterned SiO2 surface with a higher surface energy silane on the raised surfaces and a lower surface energy silane on the recessed surfaces. During spin coating, PMMA and PC coated the higher surface energy areas. The polymer features were then removed from the raised features using heat and pressure. Here we study the process of spin dewetting on topographically patterned PDMS molds. This combination of soft lithography and physically controlled dewetting allows fabrication of polymer microfeatures in both thin (<100 nm) and thick (>5 micron) polymer films without the need for chemical surface modifications. By spin coating a topographically patterned mold with a polymer solution with the appropriate solvent and polymer content, micro and nanoscale polymer features can be fabricated in a single spin coating process. The spin coating process provides well-controlled coating conditions for 63

patterning of microscale features over a large pattern area (> 5 cm2). While patterning of nanoscale features is less reliable over a large area, the same process shows promise for nanoscale fabrication as well. The lower reliability of replicating nanoscale features is likely due to inconsistencies in the original e-beam fabricated master and due to a relatively small original pattern area. The spin dewetting process allows fabrication of features with either the same or different geometrical properties as the original mold depending on the polymer concentration used. Several levels of dewetting can be observed by altering polymer solution concentration. Complete dewetting takes place at relatively low polymer concentrations, and refers to the complete lack of polymer solution on the raised features of the PDMS molds. For complete dewetting, the polymer remains only in the recessed feature after the spin coating process. Partial dewetting occurs at higher polymer concentrations. In partial dewetting, the polymer separates during the spin process, resulting in filling of the recessed portions of the mold in addition to a second, physically separated polymer pattern that remains on the raised features of the mold. At even higher polymer concentrations, a continuous polymer film is formed on the PDMS mold. Spin dewetting allows precise fabrication of polymer structures at the micro and nanoscale with a single spin process. The dewetting process is initiated purely by surface topography, thus no chemical surface modifications are needed. This process could be applied to a number of applications as a simple method for fabricating physically independent polymer structures at the micro and nanoscale.

64

4.2 Materials and methods 4.2.1 Materials Poly (n-propyl methacrylate) (PPMA, MW ~150,000) (Scientific Polymer Products), poly(methyl methacrylate) (PMMA, MW ~75,000) (Scientific Polymer Products), and polystyrene (PS, melt flow index 4.0) (Aldrich) were used as model polymers. Polymers were dissolved in anisole (Sigma) at concentrations ranging from 0.125% to 20% (wt/wt). From here on, all solution concentrations are given as weight percent 4.2.2 PDMS molding PDMS molds were fabricated using standard soft lithography procedures described previously. Silicon wafer masters were fabricated from three different photoresists depending on the desired mold geometry and thickness. The photoresists used in this study were S1813 positive tone resist (Shipley), SPR 220-7 positive tone resist (Shipley) and SU8 2005 negative tone resist (Microchem Corp.). Nanoscale structures were fabricated using electron beam (e-beam) lithography. Silastic T-2 poly(dimethylsiloxane) (PDMS) (Dow Corning) was mixed at a 10:1 ratio with T-2 curing agent and poured over the patterned wafers. The molds were cured at room temperature for 48 hours before removal from the wafer. 4.2.3 Spin dewetting and pattern transfer The spin dewetting process is illustrated schematically in Figure 4.1. The PDMS mold was spin coated with a polymer solution at 3000 or 4000 rpm for 60 seconds. Spin speed and time were held constant for a given mold geometry to determine the effect of the pattern geometry and solution concentration on the resultant structures. Depending 65

on the concentration of the polymer solution, periodicity of the pattern, and depth of the pattern, four potential coating scenarios are possible: (1) complete dewetting with intermediate separation, (2) complete dewetting with no intermediate separation, (3) partial dewetting, and (4) no dewetting. Complete dewetting with intermediate separation between features is shown on the far left of Figure 4.1(a). In this case, the polymer solution dewets into portions of the recessed PDMS structure, while some recessed areas of the PDMS mold remain uncoated. Complete dewetting with no intermediate separation occurs when the polymer solution coats the entire recessed portion of the mold but does not coat any of the raised portions of the mold. For partial dewetting, the entire recessed portion of the mold is coated and parts of the raise portion are coated. In this case, there is a physical separation between the polymer within the recesses of the mold and the polymer coating the raised portions of the mold. Finally, the case of no dewetting occurs for higher polymer concentrations when the film completely coats the recessed and raised portions of the mold. Each of these scenarios is illustrated in a separate column in Figure 4.1(a).

66

Figure 4.1 Schematic diagram of the spin dewetting and pattern transfer process; (a) four coating scenarios, (b) pattern transfer, and (c) final pattern.

To remove the polymer structures from the mold, the mold is inverted and heat and pressure are applied to remove the polymer from the mold onto the desired substrate. This is illustrated in Figure 4.1(b). For this study, polymer micro- and nanostructures were transferred onto glass or silicon substrates. The substrates were heated to 95 C for PPMA and 170 C for PMMA and polystyrene. Low pressure (~ 2.5 psi) was used to remove features from the raised portion of the mold, and higher pressure (~ 30 psi for S1813 molds and 60 psi for SPR and SU8 molds) was used to remove the features from the recessed portions of the molds. As shown in Figure 4.1(c), the resulting structures can have significantly different geometrical properties from the original mold. This allows features of multiple geometries to be fabricated from a single PDMS mold by varying polymer concentrations. 67

4.2.4 Process characterization Resulting micro and nanostructures were characterized using scanning electron microscopy (SEM) (Hitachi S-3000H), atomic force microscopy (AFM) (Veeco Dimension 3100), and optical microscopy. 4.3 Results and discussion Three dimensional AFM images of polystyrene microstructures fabricated using spin dewetting on 2 m wide, 1.4 m tall pillars are shown in Figure 4.2. The spacing between features is also 2 m. Polymers solutions were spin coated at 3000 rpm for 60 seconds. The images show the evolution of the feature geometry as the polymer solution concentration is increased. Figure 4.2(a) shows complete dewetting of a 0.125 wt.-% polymer solution around the pillars with intermediate separation between the features. The resulting pattern consists of an array of physically separated polymer rings with 2 m holes. At a solution concentration of 1%, the same basic structure is again observed, but the rings are interconnected with each of the adjacent rings. As the solution concentration is increased further, the wetting between the features increases until a continuous film is produced at a solution concentration of 3% [Figure 4.2(c)]. The pattern is a mesh with 2 m through holes. The holes were verified by SEM (data not shown). Between solution concentrations of 3% and 4%, there was a transition from complete dewetting to no dewetting, and the polymer coated the entire surface of the PDMS mold. A solid film with 2 m wells made from a 10 % solution is shown in Figure 4.2(d).

68

Figure 4.2 Three dimensional AFM images and line scans showing the morphological evolution of polystyrene film after removal from the PDMS mold; all scans are 20x20 m; (a) .125 wt.-%, z-scale is 100nm, feature height is ~86 nm, (b) 1 wt.-%, z-scale is 100nm, feature height is ~91 nm, (c) 3 wt.-%, z-scale is 500nm, feature height is ~290 nm, (d) 10 wt.-%, z-scale is 1 m, feature height is ~800 nm.

Dewetting behavior was also studied on a PDMS mold with deeper features. In this case, 7.5 m tall circular features fabricated from SPR220-7 were employed. Solutions of PPMA ranging from 1-20 wt.-% were spin coated at 3000 rpm for 60 69

seconds. The coated molds were then characterized using optical microscopy. Evolution of the film morphology varied slightly from the 2 m pillars due to the differences in both the thickness and lateral dimensions of the features. Figure 3(a) shows an uncoated PDMS mold. Figure 4.3(b-f) shows optical micrographs of spin coated PDMS molds at increasing polymer concentrations. The images show that for a low polymer concentration (1%), the polymer only coated the perimeter of the 20 m pillar and at 3% the polymer coated the perimeter of the pillar and a portion of the area in between the pillars. This behavior is similar to that observed on the 2 m pillars at low polymer concentrations. Partial dewetting was observed at concentrations of 5-15%. Figure 4.3 shows images of the coated mold at (d) 5% and (e) 15% polymer concentrations. The recessed features of the mold were coated with a continuous film while particles formed on the raised features of the mold.

Figure 4.3 Optical micrographs of PPMA dewetting on a PDMS mold with 20 m diameter, 7.5 m tall pillars; (a) Uncoated PDMS mold (b-f) Polymer solution concentrations are (b) 1 wt.-% (c) 3 wt.-% (d) 5 wt.-% (e) 15% wt.-% (f) 20 wt.-%. 70

The diameter of the particles formed on the mold can be well controlled by varying the polymer concentration or the lateral dimension of the features on the mold. Figure 4.4(a) show that the particle diameter increased linearly from 2.4-3.6 m with increased solution concentration. Figure 4.4(b) shows the influence of the lateral dimension of the mold features on the particle diameter. In this case, hexagonal features with lateral dimensions between 5 m and 20 m were used. The particle size increased linearly from 2.6-5.4 m depending on the size of the hexagonal feature that was coated. The solution concentration is held constant at 15%. The ability to form monodisperse microparticles using spin dewetting could provide a simple method for production of microparticles for application such as drug delivery.11,12

71

4.5 4 3.5

Particle Size (m)

3 2.5 2 1.5 1 0.5 0 0 2 4 6 8 10 12 14 16

(a)

Polymer Solution Concentration (%)

Particle Size (m) .

5 4 3 2 1 0 5 10 15 20 25

(b)

Hexagon Size (m)

Figure 4.4 Particle size characterization for (a) varying PPMA solution concentrations on 20 m diameter PDMS pillars and (b) varying hexagonal pillar sizes with a PMMA solution concentration of 15%.

Figure 4.5 shows SEM micrographs of each step of the coating and removal process for a partially dewetted PPMA film on 20 m PDMS features. Figure 4.5(a) is the original PDMS mold. Figure 4.5(b) shows the mold after spin coating. The image shows that the recessed area of the mold is almost completely filled with polymer and physically separate polymer particles form on top of the pillars. Figure 4.5(c) and (d) 72

show the polymer particles that were removed from the top of the pillars using heat and low pressure and the mesh in between the features removed using heat and higher pressure.

Figure 4.5 SEM micrographs of each step in the spin coating and stamping process for partially dewetted features; (a) uncoated PDMS mold, (b) mold after spin coating, (c) substrate with stamped particles, (d) substrate with stamped mesh.

To this point, all of the experiments have described dewetting behavior on pillar structures. In addition to pillars, dewetting can be initiated in well structures. Figure 4.6(a) shows an SEM micrograph of 3% PMMA spin coated on 10m diameter circular wells. The polymer completely dewets and coats only the perimeter of the wells. Figure 4.6(b) shows the polymer rings removed from the wells using the stamping process described previously. 73

Figure 4.6 (a) PDMS mold after spin coating with 3% PMMA, (b) PMMA rings after removal from the mold.

Figure 4.7 SEM micrographs of nanoscale features; (a) no intermediate dewetting between features, (b) intermediate dewetting between features. P represents the primary features from the mold recesses and I indicates the intermediate dewetted features.

The dewetting process was also demonstrated with nanoscale features. A pattern was first created using e-beam lithography and the pattern was transferred to PDMS. The 74

PDMS mold was then spin coated at 4000 rpm with a 1% polystyrene solution. Experiments indicate that in addition to the depth of the pattern, the lateral dimensions and the spacing of the pattern are also important parameters governing dewetting behavior. Figure 4.7 shows SEM micrographs of polystyrene features after removal from the mold. For features with smaller lateral dimensions and a smaller center-to-center distance between features, complete dewetting was observed. For larger features with larger spacing, only partial dewetting was observed. The features in Figure 7(a and b) are approximately 170 nm separated by 300 nm and 250 nm separated by 500 nm, respectively. Note that no residual polymer is observed in (a) and small polystyrene strips approximately 50 nm wide are located between the features in (b) for the same polymer concentration. These results indicate that the original PDMS mold geometry (height, lateral feature dimensions, and spacing) and polymer solution concentration are all important factors in the formation of polymer features by spin dewetting. These features can have significantly different geometries from the original mold. This allows several different types of polymer features to be fabricated from a single PDMS mold. By properly engineering the mold surface, complicated polymer structures can be formed and tailored to a given application. 4.4 Conclusion The process of spin dewetting has potential as a versatile and simple method of fabricating polymer micro and nanostructures. Three different mold geometries were studied here. They can be categorized as thin film, thick film, and nanoscale features. Polystyrene, PMMA, and PPMA were used as model polymers, but the process could 75

potentially be applied to various soluble thermoplastic materials with minor process modifications. The process was demonstrated on both pillar and well structures. In this study, the effect of polymer solution concentration and mold feature geometry on final polymer microstructure was examined. It was found that dewetting occurs on several different types of patterns at differing polymer concentrations. For a given mold geometry, the dewetting transitions can be controlled by varying the concentration of the polymer solution. This results in polymer structures that can have geometries that are significantly different than the original mold. The process was applied to fabrication of microscale polymeric patterns and shows potential for application in polymer nanofabrication. References
1 2

R. Gnter, Dewetting of thin polymer films. Physical Review Letters 68 (1992) 75-78.

R. Xie, A. Karim, J.F. Douglas, C.C. Han, R.A. Weiss, Spinodal dewetting of thin polymer films. Physical Review Letters 81 (1998) 1251-1254. A. Sehgal, V. Ferreiro, J.F. Douglas, E.J. Amis, A. Karim, Pattern-directed dewetting of ultrathin polymer films. Langmuir 18 (2002) 7041-7048. E. Meyer, H. Braun, Controlled dewetting processes on microstructured surfaces- a new procedure for thin film microstructuring. Macromolecular Materials and Engineering 276/277 (2000) 44-50. Z. Zhang, Z. Wang, R. Xing, Y. Han, How to form regular polymer microstructures by surface-patterned-directed dewetting. Surface Science 539 (2003) 129-136.
6 5 4 3

Z. Zhang, Z. Wang, R. Xing, Y. Han, Patterning thin polymer films by surface-directed dewetting and pattern transfer. Polymer 44 (2003) 3737-3743. P. Lenz, R. Lipowsky, Morphological transitions of wetting layers on structured surfaces. Physical Review Letters 80 (1998) 1920-1923. R. Konnur, K. Kargupta, A. Sharma, Instability and morphology of thin liquid films on chemically heterogeneous substrates. Physical Review Letters 84 (2000) 931-934. 76
8 7

K. Kargupta, A. Sharma, Dewetting of thin films on periodic physically and chemically patterned surfaces. Langmuir 18 (2002) 1893-1903.
10

L.-R. Bao, L. Tan, X.D. Huang, Y.P. Kong, L.J. Guo, S.W. Pang, A.F. Yee, Polymer inking as a micro- and nanopatterning technique. Journal of Vacuum Science and Technology B 21 (2003) 2749-2754.
11

J. Guan, A. Chakrapani, D.J. Hansford, Polymer microparticles fabricated by soft lithography. Chemistry of Materials 17 (2005) 6227-6229.
12

J. Guan, N. Ferrell, L.J. Lee, D.J. Hansford, Fabrication of polymeric microparticles for drug delivery by soft lithography. Biomaterials 27 (2006) 4034-4041.

77

CHAPTER 5 LIFT-OFF PROCESSING FOR FABRICATING MICROPATTERNED SULFONATED POLYANILINE

5.1 Introduction Traditional polymeric materials generally have the property of being poor electrical conductors. However, several unique polymer chemistries have been developed to act as organic electrical conductors or conducting polymers. Due to their unique properties, conducting polymers have been considered as potential functional materials in several applications including biosensors,1,2 light emitting devices,3,4 organic transistors,5,6 and actuators.7,8 Sulfonated polyaniline (SPAN) is of particular interest due to the advantages of being self-doped, environmentally stable, and conductive over a wide pH range.9 While sulfonated polyaniline exhibits considerably higher solubility in aqueous solution than native polyaniline,10 solution based processing can still be problematic in certain applications. In situ polymerization of polyaniline on surfaces has been a common route for formation of polyaniline films on flat surfaces.11,12 This method eliminates the need to dissolve the polymer after synthesis. In addition to flat films, in situ polymerization may provide a convenient method for creating micropatterned conducting polymer films.

78

The ability to pattern conducting polymers on the micro and nanoscale is particularly important for many applications. A number of methods have been introduced for patterning conducting polymers.13 By carefully tailoring the conducting polymer chemistry to introduce photosensitivity, photolithography has been used for micropatterning conducting polymers.14,15 The use of photolithography can be used to either locally change the conductivity of the polymer or change the solubility of the polymer in a specific solvent. While photolithographic processing of conducting polymers has been widely studied, these processes require expensive processing equipment and fabrication facilities. In order in reduce the cost and complexity of patterning, soft lithographic techniques have been explored as alternative fabrication routes. Micromolding in capillaries (MIMIC)16-18 and microcontact printing18 are two soft lithography based methods that have been used for the fabrication of conducting polymer microstructures. The method described here combines double stamp micromolding with in situ polymerization in a lift-off technique for microfabrication of SPAN microstructures on both rigid and flexible surfaces. Molding is used to pattern a sacrificial layer from a thermoplastic polymer. SPAN is then polymerized on the surface and the sacrificial layer is dissolved in organic solvent, selectively removing SPAN from the surface. Atomic force microscopy (AFM) was used to measure the film thickness and RMS roughness as a function of the reaction time. Scanning electron microscopy (SEM) was used to image the substrates at various steps in the microfabrication process. This method provides an efficient, low-cost, non-cleanroom method for micropatterning of sulfonated polyaniline

79

on a variety of surfaces. This process may have applications ranging from conducting polymer electronics and sensors to biomedical actuators. 5.2 Materials and methods 5.2.1 SPAN synthesis Sulfonated polyaniline was synthesized in a manner similar to that described previously.19 Analine [AN (Aldrich)] and 3-aminobenzenesulfonic acid [metalinic acid (MA) (Fluka)] were copolymerized in 1 M HCl in an ice bath at 5-15C. Ammonium persulfate [APS (Aldrich)] was used as the oxidant. The molar ratio of AN:MA was 1:1 and the molar ratio of oxidant to each monomer was 1:1. Substrates were first attached to the sidewalls of the reaction vessel. MA was dissolved in HCl and AN was added dropwise to the reaction vessel. APS then dissolved separately in DI water. The reaction beaker was placed in an ice bath and the APS solution was added. The reaction was performed under both static and dynamic conditions. Under static conditions, the reaction components were mixed thoroughly and the reaction was allowed to proceed without agitation. Under dynamic conditions, a magnetic stir plate and stir bar were used to constantly agitate the solution during the reaction. 5.2.2 Micromolding and Lift-off processing To begin the process, a photoresist master was first fabricated using standard photolithographic techniques, and the negative of the photoresist pattern was transferred into PDMS using the molding process described previously. Micropatterning of sulfonated polyaniline was then performed by patterning a sacrificial layer using double

80

stamp micromolding. In situ polymerization on the micropatterned substrate was performed, followed by solvent lift-off. Figure 5.1 shows the process for patterning SPAN. The PDMS mold was first spin coated with a layer of poly(n-propyl methacrylate) (PPMA) (Scientific Polymer Products). PPMA was dissolved in anisole (Aldrich) at various concentrations depending on the geometry of the PDMS mold features. For this study, three different mold geometries were used: a 1.4 m deep mold with 2 m wide channels, a 4.8 m deep mold with a minimum feature size of 5 m, and a 7.5 m deep mold with 20 m circular pillars. The PPMA solution and spin condition for each mold were as follows: 5% PPMA spin coated at 3000 rpm, 10% at 2000 rpm, and 15% at 3000 rpm, respectively.

Figure 5.1 Schematic diagram of sacrificial layer patterning, in situ polymerization, and lift-off; (a) uncoated PDMS mold, (b) PDMS mold spin coated with PPMA, (c) mold is inverted and applied to heated glass to remove surface polymer, (d) PPMA removed from the raised surface of the PDMS mold, (e) PPMA in recessed features of the PDMS mold, (f) mold inverted and heat and pressure applied to remove polymer from recessed features, (g) PPMA features after removal of the mold (h) SPAN polymerized directly onto the surface, (h) PPMA removed in acetone leaving the patterned SPAN features. 81

After spin coating [Figure 5.1(b)], the polymer on the raised surfaces of the mold was removed by bringing the mold into contact with a hotplate at 180 C and applying light pressure (~ 2.5 psi) [Figure 5.1(c)]. Figure 5.1(d) shows the material that was removed from the raised portion of the PDMS mold. The mold was then applied to the substrate at 95 C and pressure was applied to the top of the mold and held for 3-5 seconds. The pressure used for the 1.4 m, 4.8 m, and 7.5 m deep molds were 35, 40, and 50 psi, respectively. This transferred the PPMA from the recessed portions of the mold onto the substrate. Three different substrate materials were used in this study: glass, silicon, and polyimide. The patterned PPMA layer acted as a sacrificial layer for patterning the SPAN. After patterning the PPMA, the SPAN was deposited in situ on the substrates for times ranging from 15 minutes to 2 hours. The substrates were then removed from the reaction and the PPMA layer was removed by sonication in acetone for 10 seconds followed by drying in filtered air. 5.2.3 Process characterization Polymer films were characterized using atomic force microscopy (AFM) (Veeco Dimension 3100) and scanning electron microscopy (SEM) (Hitachi S-3000H). Film thickness and roughness were measured using tapping mode AFM. The film thickness was measured by using the AFM stepheight function. The RMS roughness was measured over a 5 m x 5 m scan area at 1.5 Hz and 512 X 512 pixel resolution. SPAN films deposited on clean silicon substrates were used for thickness and roughness characterization. SEM was used to visualize the substrates at various steps in the fabrication process. 82

5.3 Results and discussion Figure 5.2 shows SEM micrographs of patterned PPMA sacrificial layers (a,b), sacrificial layers after in situ polymerization of SPAN (c,d), and SPAN patterns after sacrificial layer lift-off (e,f). Figure 5.2(c) shows a low magnification image to demonstrate that the process can be used for fabrication of features over a relatively large area. The minimum line width for the pattern shown is nominally 5 m.

Figure 5.2 SEM micrographs of; (a,b) patterned PPMA sacrificial layers, (c,d) PPMA layer with SPAN deposited over the entire surface, and (e,f) SPAN micropatterns after lift-off of the sacrificial layer.

83

SPAN film thickness and roughness were characterized throughout the process in order to optimize the processing conditions to obtain smooth films with well-controlled thicknesses. The SPAN film thickness and RMS roughness versus reaction time are shown in Figure 5.3 and Figure 5.4, respectively. The parameters were measured for both static and dynamic deposition processes to determine the effect of reaction conditions on the film properties. For a static deposition process, the film thickness increased to a steady state thickness of approximately 100 nm after 60 minutes of deposition time. Between 60 and 120 minutes, the film thickness remained relatively constant. Roughness also remained fairly constant between 45 and 90 minutes, but a large increase in the average roughness and standard deviation was observed between 90 and 120 minutes. This is likely due to the formation of larger aggregates of SPAN on the surface. This phenomenon was confirmed qualitatively by SEM observation.

84

180 160 140

Film Thickness (nm)

120 100 80 60 40 20 0 0 20 40 60 80 100 120 140

Reaction Time (min.)

Static

Dynamic

Figure 5.3 SPAN film thickness versus reaction time for static and dynamic reaction processes.

140

120

RMS Roughness (nm)

100

80

60

40

20

0 0 20 40 60 80 100 120 140

Reaction Time (min.) Static Deposition Dynamic Deposition

Figure 5.4 RMS roughness of SPAN films versus reaction time for static and dynamic deposition. 85

Dynamic deposition significantly altered the deposition process, particularly with respect to the film thickness. For dynamic deposition, the film thickness was significantly higher than static deposition for reaction times greater than 60 minutes. A possible explanation for the difference is local depletion of the aniline monomers during static deposition. The higher flux of SPAN to the substrate surface could lead to a larger steady state film thickness. Dynamic deposition did not have a significant effect on the roughness of the films. The same basic trend in the roughness was observed for both deposition processes. An increase in the standard deviation of the roughness at 120 minutes again indicates that larger aggregates were beginning to form on the surface. SEM images of 20 m circular patterns deposited using dynamic deposition are shown in Figure 5.5. Figure 5.5(a) shows a low magnification micrograph of independent features to demonstrate patterning over a large area and a higher magnification image of the pattern is shown in (b).

Figure 5.5 SEM micrographs of discontinuous SPAN patterns at (a) high and (b) low magnifications.

86

5.4 Conclusion The process described here combines double stamp micromolding with in situ polymerization for fabrication of microscale patterns of sulfonated polyaniline. The fabrication method provides a cost effective alternative to conventional cleanroom fabrication processes and allows fabrication on a variety of both rigid and flexible substrates. In situ polymerization eliminates the need for post process dissolution of the polymer and allows polymerization and deposition in a single step. The lift-off process is analogous to conventional lift-off used for fabrication of metal electrodes. The insolubility of SPAN in organic solvent allows lift-off process to be employed as a simple micropatterning technique. The process is capable of producing both continuous (lines) and discontinuous (circles) patterns. This process has potential for simple fabrication of SPAN microstructures over a large area for applications in a variety of conducting polymer applications. References
1

M. Gerard, A. Chaubey, B.D. Malhotra, Application of conducting polymers to biosensors. Biosensors and Bioelectronics 17 (2002) 345-359.
2

S. Geetha, C.R.K. Rao, M. Vijayan, D.C. Trivedi, Biosensing and drug delivery by polypyrrole. Analytica Chimica Acta 568 (2006) 119-125.
3

L. Dai, B. Winkler, L. Dong, L. Tong, A.W.H. Mau, Conjugated polymers for lightemitting applications. Advanced Materials 13 (2001) 915-925.
4 5

D. Braun, Semiconducting polymer LEDs. Materials Today 5(6) (2002) 32-39.

C.J. Drury, C.M.J. Mutsaers, C.M. Hart, M. Matters, D.M. de Leeuw, Low-cost allpolymer integrated circuits. Applied Physics Letters 73 (1998) 108-110.
6

N. Stutzmann, R.H. Friend, H. Sirringhaus, Self-aligned, vertical-channel, polymer field-effect transistors. Science 299 (2003) 1881-1884. 87

E.W.H. Jager, E. Smela, O. Ingans, Microfabricated conjugated polymer actuators. Science 290 (2000) 1540-1545. E. Smela, Microfabrication of PPy microactuators and other conjugated polymer devices. Journal of Micromechanics and Microengineering 9 (1999) 1-18. X.L. Wei, Y.Z. Wang, S.M. Long, C. Bobeczko, A.J. Epstein, Synthesis and physical properties of highly sulfonated polyaniline. Journal of the American Chemical Society 118 (1996) 2545-2555.
10 9 8

J. Yue, Z.H. Wang, K.R. Cromack, A.J. Epstein A.G. MacDiarmid, Effect of sulfonic acid group on polyaniline backbone. Journal of the American Chemical Society. 113 (1991) 2665-2671.
11

J. Stejskal, I. Sapurina, J. Proke, J. Zemek, In-situ polymerized polyaniline films. Synthetic Metals 105 (1999) 195-202. I. Sapurina, A. Riede, J. Stejskal, In-situ polymerized polyaniline films 3. Film formation. Synthetic Metals 123 (2001) 503-507.
13 12

S. Holdcroft, Patterning -conjugated polymers. Advanced Materials 13 (2001) 1753-1765.

D.G. Lidzey, M.A. Pate, M.S. Weaver, T.A. Fisher, D.D.C. Bradley, Photoprocessed and micropatterned conjugated polymer LEDs. Synthetic Metals 82 (1996) 141-148. F.J. Touwslager, N.P. Willard, D.M. de Leeuw, I-line lithography of poly-(3,4ethylenedioxythiophene) electrodes and application in all-polymer integrated circuits. Applied Physics Letters 81 (2002) 4556-4558. W.S. Beh, I.T. Kim, D. Qin, Y. Xia, G.M. Whitesides, Formation of patterned microstructures of conducting polymers by soft lithography, and applications in microelectronic device fabrication. Advanced Materials 11 (1999) 1038-1041. J.A. Rogers, Z. Boa, V.R. Raju, Nonphotolithographic fabrication of organic transistors with micron feature sizes. Applied Physics Letters 72 (1998) 2716-2718. T. Granlund, T. Nyberg, L.S. Roman, M. Svensson, O. Ingans, Patterning of polymer light-emitting diodes by soft lithography. Advanced Materials 12 (2000) 269-273.
19 18 17 16 15

14

I. Mav, M. igon, A. ebenik, Sulfonated polyaniline. Synthetic Metals 101 (1999) 717-718.

88

I. Mav, M. igon, A. ebenik, J. Vohlidal, Sulfonated polyanilines prepared by copolymerization of 3-aminobenzenesulfonic acid and aniline: the effect of reaction conditions on polymer properties. Journal of Polymer Science A 38 (2000) 3390-3398.

20

89

CHAPTER 6 MICROFABRICATED MEMBRANES FOR CELL ISOLATION

6.1 Introduction The ability to control the location and morphology of individual cells is extremely important in several cell biology applications. Isolation of individual cells is critical for biological experiments such as single cell electroporation1,2 and planar patch clamping.3,4 Positioning of cells is also important in cell-based biosensors,5-7 where cells often need to be located on electrodes in order for measurements to be carried out. In addition to cell placement, it is also well established that cell morphology plays a critical role in a variety of cellular processes including proliferation, differentiation, apoptosis, and motility.8-11 Engineering specific cell behaviors by controlling the cellular environment can be particularly important for tissue engineering applications.11 Cells often need to be localized in colonies or clusters in order to function properly. Two examples include liver cells and embryonic stem cells. Isolated hepatocytes are known to quickly lose liver specific function during in vitro culture. Cell aggregates, often in the form of spheroids, are needed in order to maintain proper function.12 Not only the spheroid shape, but also the size of the cluster is important for maintaining cell viability. This is primarily due to nutrient transport issue at the core of the cell spheroid.13 Embryonic stem cells are also known to aggregate into spheroids 90

called embryoid bodies.14 The ability to control spheroid size and the number of cells per spheroid could have significant implications in studying hepatocyte and stem cell biology and in tissue engineering. Microfabrication is an attractive approach for interacting with individual cells because of the ability to fabricate structures and devices on the same size scale as cells. Microcontact printing (CP)15 has been widely used to locally alter surface chemistry for cell patterning. CP is a soft lithography process that uses a poly (dimethylsiloxane) (PDMS) stamp inked with various chemistries including self-assembled monolayers (SAM)16,17. The stamp is then brought into contact with the surface of interest to transfer the chemical pattern. The spatial difference in the surface chemistry alters subsequent protein adsorption, which leads to preferential cell attachment. Cell adhesion proteins such as fibronectin have also been directly patterned on surfaces to alter cell attachment.6,18 In addition to alteration of surface chemistry, a combination of surface topography and chemical modifications have been used to pattern adherent cells.19-20 Several other devices have also been fabricated for isolating individual or groups of cells.21-24 Elastomeric stencils have been used to pattern groups of adherent cells.22 Cells were patterned by culturing them on a PDMS membrane with through holes, and removing the membrane after cell attachment. Other passive systems have been used for cell isolation. Rosenbluth et al.23 used microfabricated SU8 wells for AFM experiments on non-adherent cells. Another passive device was developed by Fukuda and Nakazawa.24 They used microfabricated wells of various diameters to create hepatocyte spheroids of different sizes. Both of these systems rely on passive loading of the cells via sedimentation. 91

Here we present a simple method for active isolation of adherent and nonadherent cells in microwells with well defined geometries. The device consists of a micropatterned membrane fabricated on the surface of a commercially available porous filter. For clarity, it should be noted that from here on the microfabricated structure will be referred to as the membrane and the commercially available portion will be called the filter. The membranes were fabricated from thermoplastic polymers using a spin dewetting process and transfer molding as described previously.25 Several membrane geometries were used including circular, square, and hexagonal patterns to control cell morphology. Individual cells or groups of cells could be isolated within each well by controlling the seeding density and surface area of the pattern. Cells were isolated by placing the device in a filtration setup and using vacuum to pull the cells into the wells. The devices were tested using three different types of cells. NIH 3T3 fibroblasts cells were isolated as either individual cells or multiple cells by altering the seeding density. These cells were also used to demonstrate the ability to control cell morphology by changing the geometrical properties of the microwells. THP-1 lymphocytes were used to show the ability to isolate non-adherent cells. Finally, C3A hepatocytes were isolated as well-defined clusters, with the size of the cluster being dependant on the area (i.e. diameter) of the microwell. The device has several advantages compared with current cell isolation technology: no change in surface chemistry is needed to spatially orient the cells; the device can be used for isolation of adherent or non-adherent cells; and by changing the area of the wells, cell clusters with controlled sizes can by formed. This device could find

92

applications in cell biology studies of individual cells and cell aggregates, cell-based biosensor arrays, and tissue engineering. 6.2 Materials and methods 6.2.1 Materials Membranes were fabricated from either poly(propyl methacrylate) (PPMA, Scientific Polymer Products) or polystyrene (Aldrich). Polymer solutions were made by dissolving the solid polymers in anisole (Aldrich) at various concentrations depending on the desired membrane thickness. Several commercially available filters were used. Track etched polycarbonate filters (Isopore, Millipore) with 200 nm and 400 nm pores were used due to their flexibility and relatively good temperature stability. Anodized alumina filters (Anodisc, Whatman) with 20 nm pores were used. However, cracking of the Anodisc filters during the molding process resulted in low yields. Transparent track etched polystyrene filters (Transwell, Corning) were also used. These filters provided the advantage of being optically transparent, which allows observation using an inverted microscope. However, they can only be used with PPMA membranes due to temperature stability issues during fabrication. 6.2.2 Membrane fabrication Membranes were fabricated using a two-step soft lithographic micromolding process combined with spin dewetting. The geometry of the membrane pores was first defined in a photoresist using standard photolithography. Negative tone SU8 25 (Microchem Corp.) and positive tone SPR 220-7 (Shipley) photoresists were used in the 93

fabrication. Photoresists were spin coated on clean silicon at 3000 rpm and 2000 rpm for the negative and positive photoresists, respectively. The resulting film thickness was 9.2 m for SU8 and 7.5 m for SPR 220-7. The photoresist films were exposed to ultraviolet light through chrome/glass photomasks with various geometries, and films were processed according to the manufacturers suggested parameters. The photolithographically patterned silicon wafers were used as a master for replication of a PDMS mold. The mold was made by first mixing a 10:1 (wt/wt) ratio of PDMS base with curing agent (Silastic T-2, Dow Corning). The base and curing agent were mixed thoroughly and poured over the patterned wafer. The PDMS was degassed in a vacuum dessicator until all bubbles were removed. The PDMS was allowed to cure at room temperature for 48 hours before removing the mold. Figure 6.1 shows the remainder of the membrane fabrication process after making the mold. The mold was spin coated with a solution of PPMA or polystyrene in anisole. Solutions of 15% (wt/wt) were used on the 7.5 m deep molds, and 20% (wt/wt) solutions were used on 9.2 m molds. Solutions were spin coated at 3000 rpm for 1 minute. After spin coating, dewetting of the polymer was observed as the polymer separated to form particles on top of the pillar structures of the PDMS molds. The particles were removed by contacting the polymer with a glass slide at 180C (PPMA) or 200C (PS) using low pressure (~2.5 psi). This is shown in Figure 6.1(c and d). After removal of the surface material, the mold was brought into conformal contact with the filter. The filter was then place on a hotplate at 95C (PPMA) or 125C (PS) and pressure of 60 psi was used to transfer the remaining polymer onto the filter. The resulting structure consisted of through holes in the membrane with access to the underlying filter. 94

Figure 6.1 Schematic diagram of the membrane fabrication process; (a) uncoated PDMS mold, (b) mold spin coated with polymer (particles on the top of the pillars form by spin dewetting), (c,d) the coated mold is brought into contact with heated glass to remove the particles, (e) selectively coated mold, (f) the mold is inverted and brought into contact with the heated filter by applying pressure, (g) the final structure after removing the mold.

6.2.3 Cell culture and filtration NIH 3T3 mouse fibroblasts were cultured in Dulbeccos modified Eagles medium (DMEM) supplemented with 10% calf bovine serum and 1% penicillinstreptomycin. THP-1 monocytes were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. C3A cells were cultured in Eagles minimum essential medium (EMEM) supplemented with 10% FBS and 1% penicillin-streptomycin. All cell lines and culture chemicals were obtained from 95

the American Type Culture Collection (ATCC). Cells were incubated at 37C in a humidified atmosphere with 5% CO2.

Figure 6.2 Experimental setup for cell isolation.

Cells were isolated in the wells using a glass microanalysis vacuum filter setup (Fisher Scientific). The system is shown in Figure 6.2. The membrane/filter was clamped between a glass funnel and the underlying vacuum filter flask (Fisher Scientific). A cell suspension in culture medium, with the desired concentration, was then placed in the funnel on top of the membrane/filter, and vacuum was applied to pull the cell culture medium through the setup. This trapped the cells in the microfabricated wells of the membrane. Additional medium was added if required and the device was placed in an

96

incubator for 30 minutes to allow cell attachment. For non-adherent cells, the membrane was immediately removed from the setup and cells were prepared for examination. 6.2.4 Characterization Microfabricated molds and devices were characterized using scanning electron microscopy (SEM, Hitachi S3000H). Cell isolation was characterized using SEM and fluorescence microscopy (Nikon TS 100). Cells were prepared for SEM by dehydration in graded ethanol solutions (70%, 80%, 90%, and 100%) and hexamethyldisilazane according to the procedure described by Braet et al.26 For fluorescence microscopy, cells were fixed in 70% ethanol for 30 minutes at 20 C. Cells were then stained with propidium iodide RNase (PI RNase, BD Biosciences) for 10 minutes at 4 C. Samples were washed three times in phosphate buffered saline (PBS, ATCC) and placed on glass slides for observation. 6.3 Results and discussion The four membrane geometries used for NIH 3T3 and THP-1 cell isolation are shown in Figure 6.3. The features included 20 m diameter circular features, 20 m hexagons (edge to edge), 30 m squares, and 10 m diameter circles. All of the images were taken using the Isopore polycarbonate filters as the substrate. Figure 6.3(d) clearly shows the track etched pores in the underlying filter. It is critical that access to the pores is maintained throughout the fabrication process in order to provide an open pathway through the entire device to facilitate vacuum filtration of the cells.

97

Figure 6.3 SEM micrographs of cell isolation membranes. Feature geometry and size are (a) 20 m diameter circles. (b) 20 m (side to side) hexagons. (c) 30 m squares. And (d) 10 m diameter circles.

Figure 6.4 presents SEM micrographs of NIH 3T3 cells isolated in the devices. Figure 4(a) shows one NIH 3T3 cell in each of four 20 m circular wells. Figure 4(c) shows cells isolated in two hexagonal wells with two wells empty. Figure 4(e) shows multiple cells isolated in 30 m squares. Multiple cells are observed in the upper left and lower right features, while single cells are isolated in the other two features. Figure 4(b,d,f) shows high magnification images of individual cells isolated in three different membrane configurations. The images show that the filtration process is not detrimental to cell viability, as the cells are still able to attach and spread on the filter material.

98

Figure 6.4 NIH 3T3 cells isolated in three different mold geometries; (a,b) 20 m diameter circles, (c,d) 20 m (side to side) hexagons, (e,f) 30 m squares.

Figure 6.5 shows SEM images of NIH 3T3 cells isolated in (a) 30 m square and (b) 20 m circular features, after removal of the microfabricated membrane from the filter. The images show that the cells adhere well to the filter material and that the cells conform to the geometry of the microfabricated features.

99

Figure 6.5 SEM images of NIH 3T3 cells after removing the polymer membrane from the surface of the filter; (a) 30 m squares, (b) 20 m diameter circles.

Fluorescent microscopy was also used to image the isolated cells. Figure 6.6 shows the three membrane geometries loaded with a relatively low seeding density (2.5 x 104 cells/ml). Primarily individual cells are isolated in the wells. By increasing the seeding density to 5 x 104 cells/ml, multiple cells can be observed in many of the wells as shown in Figure 6.7. The patterns are the same 20 m circles and 30 m squares as Figure 6.7(a and b) with a higher seeding density. This demonstrates the ability to qualitatively control both the efficiency of the isolation and the number of cells isolated in each well for a given well geometry.

100

Figure 6.6 Fluorescent micrographs of NIH 3T3 cells isolated at low density in (a) 30 m square, (b) 20 m hexagonal, and (c) 20 m circular wells.

Figure 6.7 NIH 3T3 cells isolated in (a) 20m circular and (b) 30 m square wells at higher seeding density.

101

Non-adherent THP-1 cells were isolated in 10 m circular wells as show in Figure 6.8. The ability to isolate non-adherent cells is significant, giving the difficulties presented in manipulating these cells. For these cells, the surface chemistry approaches cannot be employed for isolation. Some passive techniques have been used for isolating non-adherent cells, but to our knowledge this is the first active mechanism that has been employed for isolation of multiple adherent and non-adherent cell types.

Figure 6.8 THP-1 cells isolated in 10 m circular wells; (a) SEM micrograph of a single cell and (b) fluorescent micrograph of an array of wells.

Finally, the ability to isolate aggregates of cells was demonstrated with C3A liver cells. Circular wells with 50 m diameter were used. PPMA membranes were patterned on Transwell filters. Cells were seeded at 5x105 cells/ml. Figure 6.9 shows low magnification (a) and higher magnification (b) fluorescent images of groups of hepatocytes isolated within the wells. An SEM micrograph of the cell aggregates after removing the membrane is shown in Figure 6.9(c).

102

Figure 6.9 Images of C3A cell clusters; (a) low and (b) high magnification fluorescent images of cells in 50 m diameter circular wells, (c) SEM micrograph of a cell clusters after removing the patterned membrane.

103

6.4 Conclusion Micromolding of polymer membranes was used for isolation of adherent and nonadherent cell lines. The process provides a simple method for spatial control of cell location as well as control of cell morphology in the case of adherent cells. The efficiency of the isolation as well as the number of cells isolated in each well was controlled by the density of the cell suspension. Adherent cells were able to attach and spread on the filter material. In addition, non-adherent cells were also successfully isolated. These devices could be used in a number of applications including biological studies of cell morphology as well as in cell-based biosensors. These devices are also well suited for single cell experiments where spatial isolation of cells in needed. References
1

Y. Huang, B. Rubinsky, Microfabricated electroporation chip for singe cell membrane permeabilization. Sensors and Actuators A 89 (2001) 242-249.
2

M. Khine, A. Lau, C. Ionescu-Zanetti, J. Seo, L.P. Lee, A single cell electroporation chip. Lab on a Chip 5 (2005) 38-43.
3

K.G. Klemic, J.F. Klemic, M.A. Reed, F.J. Sigworth, Micromolded PDMS planar electrode allows patch clamp electrical recordings from cells. Biosensors and Bioelectronics 17 (2002) 597-604.
4

X. Li, K.G. Klemic, M.A. Reed, F.J. Sigworth, Microfluidic system for planar patch clamp electrode arrays. Nano Letters 6 (2006) 815-819. P. Wang, G. Xu, L. Qui, Y. Xu, Y. Li, R. Li, Cell-based biosensors and its application in biomedicine. Sensors and Actuators B 108 (2005) 567-584. M. Nishizawa, K. Takoh, T. Matsue, Micropatterning of HeLa cells on glass substates and evaluation of respiratory activity using microelectrodes. Langmuir 18 (2002) 36453649. C.S. Chen, M. Mrksich, S. Huang, G.M. Whitesides, D.E. Ingber, Micropatterned surface for control of cell shape, position, and function. Biotechnology Progress 14 (1998) 356-363. 104
7 6 5

C.S. Chen, M. Mrksich, S. Huang, G.M. Whitesides, D.E. Ingber, Geometric control of cell life and death. Science 276 (1997) 1425-1428.
9

Y. Ito, Surface micropatterning to regulate cell function. Biomaterials 20 (1999) 2333-2342.

10

A. Brock, D. Chang, C. Ho, P. LeDuc, X. Jiang, G.M. Whitesides, D.E. Ingber, Geometric determinant of directional cell motility revealed using microcontact printing. Langmuir 19 (2003) 1611-1617.
11

S.N. Bhatia, C.S. Chen, Tissue Engineering at the Micro-scale. Biomedical Microdevices 2 (1999) 131-144.
12

M. Khalil, A. Shariat-Panahi, R. Tootle, T. Ryder, P. McCloskey, E. Roberts, H. Hodgson, C. Selden, Human hepatocyte cell lines proliferating as cohesive spheroid colonies in alginate markedly upregulate both synthetic and detoxificatory liver function. Journal of Hepatology 297 (2001) 68-77.
13

R. Clicklis, J.C. Merchuk, S. Cohen, Modeling mass trasfer in hepatocyte spheroids via cell viability, spheroid size and hepatocellular funtion. Biotechnology and Bioengineering 86 (2004) 672-680.
14

S.M. Dang, M. Kyba, R. Perlingeiro, G.Q. Daley, P.W. Zandstra, Efficiency of embryoid body formation and hematopoietic development from embryonic stem cells in different culture systems. Biotechnology and Bioengineering 78 (2002) 442-453.
15

Y. Xia, G.M. Whitesides, Soft Lithography. Annual Review of Materials Science 28 (1998) 153-184.

16

M. Mrksich, L.E. Dike, J. Tien, D.E. Ingber, G.M. Whitesides, Using microcontact printing to pattern the attachment of mammalian cells to self-assembled monolayers of alkanethiolates on transparent films of gold and silver. Experimental Cell Research 235 (1997) 305-313.
17

R.S. Kane, S.Takayama, E. Ostuni, D.E. Ingber, G.M. Whitesides, Patterning proteins and cells using soft lithography. Biomaterials 20 (1999) 2363-2376.
18

H. Kaji, K. Takoh, M. Nishizawa, T. Matsue, Intracellular Ca2+ imaging for micropatterned cardiac myocytes. Biotechnology and Bioengineering 81 (2003) 748751.
19

K.Y. Suh, J. Seong, A. Khademhosseini, P.E. Laibinis, R. Langer, A simple soft lithographic route to fabrication of poly(ethylene glycol) microstructures for protein and cell patterning. Biomaterials 25 (2004) 557-563. 105

20

M.R. Dusseiller, D. Schlaepfer, M. Koch, R. Kroschewski, M. Textor, An inverted microcontact printing method on topographically structured polystyrene chips for arrayed micro-3-D culturing of single cells. Biomaterials 26 (2005) 5917-5925.
21

E.W.H. Jager, C. Immerstrand, K.H. Peterson, K. Magnusson, I. Lundstrm, O. Ingans, The cell clinic: closable microvials for single cell studies. Biomedical Microdevices 4 (2002) 177-187.
22

A. Folch, B. Jo, O. Hurtado, D.J. Beebe, M. Toner, Microfabricated elastomeric stencils for micropatterning cell cultures. Journal of Biomedical Materials Research 52 (2000) 346-353.
23

M.J. Rosenbluth, W.A. Lam, D.A. Fletcher, Force microscopy of nonadherent cells: a comparison of leukemia cell deformability. Biophysical Journal 90 (2006) 2994-3003.
24

J. Fukuda, K. Nakazawa, Orderly arrangement of hepatocytes spheroids on a microfabricated chip. Tissue Engineering 11 (2005) 1254-1262.
25

N. Ferrell, D. Hansford, Fabrication of micro- and nanoscale polymer structures by soft lithography and spin dewetting. Macromolecular Rapid Communications 28 (2007) 964-967.
26

F. Braet, R. De Zanger, E. Wisse. Drying cells for SEM, AFM and TEM by hexamethyldisilazane: a study on hepatic endothelial cells. Journal of Microscopy 186 (1997) 84-87.

106

CHAPTER 7 FABRICATION OF SUSPENDED POLYMER MICROSTRUCTURE BY PATTERNED SUBSTRATE MICROMOLDING AND SACRIFICIAL LAYER MICROMOLDING

7.1 Introduction MEMS technology has traditionally been focused on silicon based fabrication techniques. In order to broaden the potential applications of MEMS devices, novel materials must be considered for some applications. Polymers are of particular interest due to their unique and diverse material properties and low cost, relative to silicon. Polymer MEMS (P-MEMS) also hold great potential for biological applications due to the inherent biocompatibility of many polymer materials.1 While silicon based fabrication techniques are well established and understood, new processing techniques need to be developed for polymer materials. This chapter introduces two methods for fabrication of freely suspended three dimensional polymer microstructures for bioMEMS applications: sacrificial layer micromolding (SLaM) and patterned substrate micromolding (PSM). Photolithography has been the standard in fabricating polymer microstructures due to its prevalence in microelectronics processing. Recently, several processes have been developed for microfabrication of polymeric microstructures and devices from materials other than standard photosensitive polymers. Microscale hot embossing2 microinjection molding,3 as well as the soft lithography based fabrication methods, 107

including replica molding,4 solvent assisted micromolding4 (SAMIM), microtransfer molding (TM),5 and micromolding in capillaries (MIMIC)6 have all been developed for fabrication of polymer microstructures. These processing methods have found utility in a number of biomedical applications such as microfluidic diagnostic and analytical devices,7,8 tissue engineering,9,10 and drug delivery .11 However, they are generally used for fabrication of planar structures. For suspended structures, other techniques such as etching3 or photolithography using negative photoresist12 must be used. These processes are limited in the materials that can be used in fabrication and require extensive use of expensive cleanroom facilities and equipment. SLaM and PSM provide alternative methods for fabrication of suspended polymer microstructures from a wide range of thermoplastic, biocompatible polymer materials that are not traditionally used in microfabrication but are common in biomaterials applications. In addition, polymer materials are generally inexpensive and a majority of the process can be performed outside of a cleanroom with minimal use of expensive microfabrication equipment, thus reducing the overall cost of the process. The applicability of these processes to a wide range of structural materials also increases the materials selection pool, allowing materials to be selected based on appropriate biological interactions. Both SLaM and PSM involve double stamp molding processes followed by alignment and bonding to a micropatterned substrate. The processes are capable of yielding polymer microstructures from thermoplastic polymers with resolution comparable to the resolution of the photolithography process. In the present work, PPMA, PMMA, and polystyrene were used as model polymers. PPMA was used because 108

of its ease of processing due to its low glass transition temperature (~37C), and PMMA and polystyrene were chosen due to their wide use in biomedical applications. The same basic process could also be extended to a variety of other thermoplastic polymer materials. The processes provide a simple and cost effective method for fabrication of polymer devices for integration into MEMS based sensor and actuator systems. 7.2 Materials and methods 7.2.1 PDMS mold fabrication The device geometries were initially fabricated from photoresist using standard photolithography. A layer of either SU8-5 negative tone photoresist (MicroChem Corp.) or S1813 positive tone photoresist (Shipley) was spin coated on a <100> p-type silicon wafer (WaferNet). The photoresist and spin speed used in the initial process were selected based on the desired final structure geometry and thickness. After coating, the wafers were processed according to the manufacturers suggestion processing parameters. The photoresist features were transferred into a poly(dimethlysiloxane) (PDMS) elastomer for use in fabricating the final polymer structural features. The process of PDMS molding from a photoresist master is described in detail elsewhere.13 Briefly, a 10:1 ratio of T-2 transparent base and curing agent (Dow Corning) was mixed and stirred thoroughly. The mixture was then poured over the patterned silicon wafer and placed in a vacuum dessicator to remove bubbles incorporated during mixing. The sample was removed from the vacuum periodically, and a razor blade was used to remove surface bubbles. After the bubbles were completely removed, the PDMS mold was allowed to

109

cure at room temperature for 48 hours before removing the mold from the wafer. The wafer could be used multiple times for making PDMS molds. 7.2.2 Sacrificial layer micromolding Figure 7.1 shows the fabrication process used in the sacrificial layer micromolding process. Poly(vinyl alcohol) (PVA) (Sigma-Aldrich) was chosen as a sacrificial layer due to its solubility in water, insolubility in organic solvents, and thermal stability. A 10:1 ratio of deionized water and PVA were mixed and heated to 70 C to promote dissolution. After the polymer had completely dissolved, the solution was filtered through high flow rate filter paper to remove impurities. The PVA/water solution was spin coated on silicon wafers for 60 seconds at a spin speed of 1000 rpm. After coating, the wafers were baked at 95 C for five minutes to remove any residual water. The resulting sacrificial layer thickness was ~750 nm. Sacrificial layer thickness was characterized using tapping mode atomic force microscopy (AFM) (Veeco, Dimension 3100). The sacrificial layer thickness can be varied by changing the PVA concentration and spin speed. A thin layer (~400 nm) of poly(methyl methacrylate) (PMMA) (Scientific Polymer Products) was then spin coated on the PVA layer and baked at 115 C for two minutes. The PMMA layer acted to protect the PVA from being dissolved during development in the upcoming photolithography process. A layer of SPR220-7 positive tone photoresist (Shipley) was spin coated on the PMMA layer at 4500 rpm resulting in a photoresist thickness of ~4.5 m. The photoresist was processed according to the manufacturers processing parameters to expose the anchor regions of the underlying PMMA and PVA layers. 110

Figure 7.1 Schematic diagram of the sacrificial layer micromolding method. (a) sacrificial layer patterning process, (a,i) substrate is coated with PVA, PMMA, and photoresist, respectively, and the photoresist is patterned via photolithography, (a,ii) surface is exposed to O2 plasma to etch the PMMA and PVA layers, (a,iii) photoresist and PMMA are removed with acetone, (b) micromolding, (b,i) patterned PDMS mold is uniformly coated with the structural polymer, (b,ii) coated mold is brought into contact with a heated glass plate to remove surface material, (b,iii) glass plate is removed leaving polymer only in the recessed portions of the mold, (c) alignment and bonding, (c,i) sacrificial layer and patterned mold are aligned, (c,ii) sacrificial layer and structural layer are brought into contact and bonded under heat and pressure, (c,iii) mold is removed, (c,iv) sacrificial layer is removed by immersion in water.

The PMMA and PVA layers were then etched with an oxygen plasma in a benchtop reactive ion etcher (Technics, MicroRIE 800). The patterned SPR220-7 layer acted as an etch mask. An oxygen flow rate of 15 sccm, power of 200W, and 163 mTorr pressure were used during the etch process. After the exposed PMMA and PVA were selectively removed and the substrate anchor points were exposed, the wafer was 111

immersed in acetone for two minutes to remove the PMMA and photoresist layers, leaving only the patterned PVA. After removal from the acetone bath, samples were dried with a gentle stream of nitrogen. The structural portion of the devices was fabricated via a modified micro-transfer molding process.14,15 Poly(n-propyl methacrylate) (PPMA) (Scientific Polymer Products), polystyrene (Sigma-Aldrich), and PMMA (Scientific Polymer Products) were chosen as structural materials. Polymers were dissolved in anisole (Sigma-Aldrich) at concentrations ranging from 1-10% (wt/wt). The polymer solution was spin coated onto the patterned PDMS mold at spin speeds from 1000-6000 rpm for 60 seconds. Devices ranged from ~200 nm to >5 m in thickness based on the above parameters. To remove the polymer from the surface (not in the recessed features) of the PDMS, the mold was brought into contact with a glass slide heated to 175 C. The same temperature was used for all three structural materials. The mold was held on the slide for 5-10 seconds under the pressure of its own weight. The mold was immediately removed, with the glass slide still in contact with the heat source. The process was repeated if necessary to completely remove the surface material, leaving the polymer only in the recessed portions of the PDMS mold. The selectively patterned PDMS mold was then aligned with the sacrificial layer under an optical microscope. The substrate was heated to 95 C for PPMA, 120 C for polystyrene, and 175 C for PMMA and a pressure of 0.21 MPa (30 psi) was applied to the backside of the mold and held for 10 seconds. The pressure brought the structural material into contact with the substrate material and the heating process promoted adhesion between the sacrificial and structural layers. The molds were removed, resulting 112

in the final unreleased structures. Resultant device thickness was a function of the PDMS mold depth, polymer concentration, and spin speed. Devices were characterized using a mechanical stylus profilometer (Veeco Dektak, Series 3), atomic force microscopy, and scanning electron microscopy (Hitachi, S-3000H) The devices were released by immersion in DI water, which dissolved the sacrificial PVA layer. Stiction during the drying process prevents the devices from being operated under dry conditions. Our studies have shown that the released devices remain separated from the substrate so long as they are maintained in an aqueous environment. 7.2.3 Patterned substrate micromolding The patterned substrate method is similar to the sacrificial layer method in that the same microtransfer molding process is used for fabrication of the structural portion of the devices. The method differs in choice of the substrate material. For the sacrificial layer method, silicon or glass acted as the substrate material and anchor region for the devices, and a sacrificial material is removed to release the device. For the patterned substrate method, a layer of patterned photoresist acts as the substrate material, and voids in the photoresist layer act as the suspended portions of the devices. Figure 7.2 shows a schematic of the patterned substrate fabrication method. A layer of SU8-25 negative tone photoresist was patterned via standard photolithography processing. The photoresist was coated at 2000 rpm, yielding a film thickness of approximately 25 m. The model substrate geometry for this work consisted of 20-30 m wide channels with 500 m spacing between the channels. Following the photolithography process, the double stamp micromolding process described earlier was performed with a PMDS mold of the desired device geometry. In 113

this case, a mold geometry with 5 m wide channels with 45 m spacing was used, yielding final structures consisting of an array of 5 m wide beams suspended over the 20-30 m channels in the photoresist. This is shown in Figure 7.2(a). The selectively coated mold and photolithographically patterned substrate were then brought into contact with heat and pressure. This transferred the polymer in the PDMS mold onto the photoresist substrate with polymer suspended across the gaps in the patterned substrate.

Figure 7.2 Schematic diagram of the micromolding process used in polymer beam fabrication; (a) selectively coated PDMS mold, (b) mold inverted and aligned with patterned substrate, (c) heat a pressure used to stamp the features, and (d) mold removed.

114

7.3 Results and discussion Scanning electron micrographs of devices resulting from the SLaM process are shown in Figure 7.3. The images show that multiple and complex geometries are achievable and that the integrity of the feature geometry can be maintained throughout the PDMS molding and micromolding processes. The features in Figure 7.3(d) are nominally 750 nm, indicating that the feature resolution of the process reproduces that of the photolithography process. The devices are shown prior to release of the sacrificial layer. Figure 7.4 shows optical micrographs of polymer cantilevers beams before and after release of the sacrificial layer. Figure 7.4(a) shows 10 m wide beams before the sacrificial layer was removed, and 7.4(b) shows the suspended devices after removal of the sacrificial layer.

Figure 7.3 Scanning electron micrographs of devices made by the sacrificial layer micromolding (SLaM) method. Devices are shown prior to release of the sacrificial layer. Arrows indicated the structural PPMA regions, sacrificial PVA layer, fixed anchor regions, and underlying silicon substrate. Structures shown include (a) cantilevers, (b) cross-bridges, and (c) cantilever with through holes. The features in (d) have nominally 750 nm line width and are shown to demonstrate process resolution. 115

Figure 7.4 Optical micrographs of polymer cantilevers fabricated using the sacrificial layer micromolding process (a) before removal of the sacrificial layer and (b) after release.

Figure 7.5(a) shows a phase contrast micrograph of a released PPMA cantilever for low magnitude lateral force measurement applications. The cantilever is 250 m long, 5 m wide, and 3.5 m thick. A micropipette was used to apply a force to the cantilever as shown in Figure 7.5(b). Deflection of the beam was then measure optically and related to the force magnitude using the cantilever beam bending equation:

F=

3EIx L3

(1)

where F is the applied load, x is the measured deflection, L is the distance from the base of the beam to the applied load, E is the Youngs modulus, and I is the area moment of inertia. The load was assumed to be a point load and was applied at 245 m from the base of the cantilever. The modulus was taken to be 0.7 GPa16 and the moment of inertia calculated from the above geometry is 3.65x10-23 m4. The deflection of the beam was measured at 20 m using image analysis software (ImagePro Discovery). The magnitude
116

of the resulting force is 104 nN. The same force applied to a silicon beam (E=190 GPa)17 would result in a lateral deflection of only 74 nm. This deflection would not be detectable using a standard optical microscopy, and a much more sophisticated measurement system would be needed. This illustrates the potential of using polymer MEMS for fabrication of highly sensitive devices that would not be functional if they were fabricated from standard silicon based MEMS materials.

Figure 7.5 Phase contrast micrographs of 5 m wide cantilevers made using the sacrificial layer micromolding process (a) after removal of the sacrificial layer and (b) after application of a 104 nN force using a micropipette tip.

Resulting devices from the patterned substrate micromolding (PSM) process are shown in Figure 7.6. Polystyrene beams are shown in Figure 7.6(a) and PPMA beams in Figure 7.6(b). The images show that the devices remain suspended over the channels with a flat beam profile between the anchor points. The ability to fabricate flat structures over the channels is constrained by the width of the channel (i.e. the portion over which the device is suspended). For the work described here, 30 m was the maximum channel 117

width used for fabrication of 5 m wide structural features. Further work is needed to determine the maximum attainable functional region for this or other feature geometries. The non-uniformity of the channels in the substrate is a result of the use of a printed transparency mask14 as opposed to a chrome/glass mask. Resolution of the substrate channels could be considerably improved by use of a chrome/glass mask, but a significant increase in cost would also be incurred.

Figure 7.6 Scanning electron micrographs of (a) polystyrene and (b) PPMA beams made using the patterned substrate method. The substrate is photolithographically patterned SU8 photoresist.

The proper choice of a polymer MEMS fabrication method is dependant on the given application. Both fabrication methods described here are limited by practical 118

fabrication constrains. The most significant constraint for the sacrificial layer micromolding process at present is the inability to operate devices in dry conditions due to stiction issues. Work is current being done to develop a drying process to minimize this effect, allowing the devices to remain suspended in air. The geometrical constrains of the SLaM technique are minimal. Suspended regions of >200 m can be achieve and remain suspended in aqueous environment. In contrast, the PSM allows devices that can be operated in dry condition, but the length of the suspended regions of the device are limited to those that maintain their structural integrity. Given that both fabrication methods use a soft mold for fabrication, distortion of the features during the final stamping process results in an undesirable loss of feature resolution if the processing parameters for a given device geometry are not appropriately selected and characterized. Figure 7.7 shows AFM section scans of devices of similar geometries with (a) properly selected and (b) poorly selected processing parameters. The parameters that are important for minimizing mold distortion are mold depth, polymer solution concentration, spin speed, and stamping transfer pressure. The processing parameters must be adjusted for any given feature geometry. As an example, the feature in Figure 7.7(a) was processed using a 1.4 m deep mold, 5% polymer solution, 4000 rpm spin speed, and 0.21 MPa (30 psi) transfer pressure. The mold distortion is less critical for the PSM techniques, since the functional regions of the devices are never not brought into contact with the substrate, thus the functional portions maintain their geometrical integrity even if processing parameters are not completely optimized.

119

Figure 7.7 AFM section analysis of features with (a) appropriately selected process parameters and (b) poorly chosen process parameters. Important parameters include mold depth, solution concentration, spin speed, and transfer pressure. The top feature has a height of 875 nm and the bottom feature has a maximum height of 1.12 m and a minimum height of 760 nm.

7.4 Conclusion

Successful fabrication of suspended polymer microstructures was achieved using two different fabrication methods. Both processes are soft lithography based and rely on a two-step micromolding process. The choice of substrate material (i.e. sacrificial layer or patterned substrate) is application dependant, as each method is capable of producing devices of different geometries and for specific operating conditions. The processes are relatively simple and low cost as compared to silicon fabrication techniques, and the

120

transition from silicon to polymer materials could potentially broaden both the functionality and applicability of the devices in a biological setting. References B.D. Ratner, A.S. Hoffman, F.J. Schoen, and J.E. Lemons, Biomaterials Science An Intoduction to Materials in Medicine, Elsevier Academic Press, San Diego, CA (2004) p. 67. H. Becker and U. Heim, Hot embossing as a method for the fabrication of polymer high aspect ratio structures. Sensors and Actuators A 83 (2000) 130-135. L.J. Lee, M.J. Madou, K.W. Koelling, S. Daunert, S. Lai, and C.G. Koh, Design and fabrication of CD-like microfluidic platforms for diagnostics: polymer-based microfabrication. Biomedical Microdevices 3 (2001) 339-351. Y. Xia, G.M. Whitesides, Soft lithography. Annual Review of Materials Science 28 (1998) 153-184. X.-M. Zhoa, Y. Xia, and G.M. Whitesides, Fabrication of three-dimensional microstructures: microtransfer molding. Advanced Materials 8 (1996) 837-840. E. Kim, Y. Xia, and G.M. Whitesides, Polymer microstructures formed by moulding in capillaries. Nature 376 (1995) 581-584. H. Becker, C. Grtner, Polymer microfabrication methods for microfluidic analytical applications. Electrophoresis 21 (2000) 12-26. S.K. Sia, G.M. Whitesides, Microfluidic devices fabricated in poly(dimethylsiloxane) for biological studies. Electrophoresis 24 (2003), 3563-3576. G. Vozzi, C. Flaim, A. Ahluwalia, S. Bhatia, Fabrication of PLGA scaffolds using soft lithography and microsyringe deposition. Biomaterials 24 (2003), 2533-2540.
10 9 8 7 6 5 4 3 2 1

S.N. Bhatia, C.S. Chen, Tissue engineering at the microscale. Biomedical Microdevices 2 (1999) 131-144.

B. Ziaie, A. Baldi, M. Lei, Y. Gu, and R.A. Siegel, Hard and soft micromachining for BioMEMS: a review of techniques. Advanced Drug Delivery Reviews 56 (2004) 145172 G. Genolet, J. Brugger, M. Despont, U. Drechsler, P. Vettiger, N.F. de Rooij, D. Anselmetti, Soft, entirely photoplastic probes for scanning force microscopy. Review of Scientific Instruments 70 (1999) 2398-2401. 121
12

11

D.C. Duffy, J.C. McDonald, O.J.A. Schueller, G.M. Whitesides, Rapid prototyping of microfluidic systems in poly(dimethylsiloxane). Analytical Chemistry 70 (1998) 49744984. J. Guan, A. Chakrapani, D.J. Hansford, Polymer microparticles fabricated by soft lithography. Chemistry of Materials 17 (2005) 6227-6229. J. Guan, H. He, D.J. Hansford, L.J. Lee, Self-folding of three-dimensional hydrogel microstructures. The Journal of Physical Chemistry B 109 (2005) 23134-23137. G. Wei, B. Bhushan, N. Ferrell, D. Hansford, Journal of Vacuum Science and Technology A 23, 811-819 (2005). K.E. Peterson, Silicon as a mechanical material. Proceedings of the IEEE 70 (1982) 420-457.
17 16 15 14

13

122

CHAPTER 8 MEASURING THE MECHANICAL PROPERTIES OF POLYMER MICROSTRUCTURES BY NANOINDENTATION

8.1 Introduction

Mechanical characterization of materials is critical for design, modeling, function, and reliability of MEMS devices. This is particularly important in the case of polymer MEMS, where micro and nanoscale polymer properties may differ significantly from bulk material properties.1-3 In the case of device design and modeling, a good grasp of the materials properties will lead to a better understanding of device behavior prior to fabrication. This in turn leads to fewer iterations in the design, decreased cost, and faster device development. In terms of function, characterization of the material properties is important for ensuring that the device will perform as designed. Finally, characterization of the failure properties of materials will aid in understanding and improving the long term reliability of MEMS. Some of the important properties that are relevant for polymer MEMS development include microscale hardness, elastic modulus, creep, adhesion, scratch resistance, yield strength, and breaking strength. For biomedical applications of polymer MEMS, devices often need to be operated in aqueous condition and/or at elevated

123

temperatures. Therefore, the polymer properties need to be tested under the appropriate environmental conditions to simulate operating conditions. Many well-established techniques are available for bulk macroscale materials characterization. However, these methods are not generally applicable to micro and nanoscale polymer characterization. Specialized characterization techniques are needed to accommodate these high resolution measurements and small sample volumes. Nanoindentation is a powerful tool for micro/nanoscale materials characterization. In nanoindentation, a nanoscale tip with a know geometry is used to penetrate a material. Similar to macroscale indentation tests, this data can be used determine the hardness of the material. In addition, by measuring both load and displacement simultaneously, other materials properties, including elastic modulus, can be determined. Some of the advantages of nanoindentation include the ability to make measurements on extremely thin films. The high resolution of nanoindentation with respect to load and displacement are is also advantageous for measuring properties at very low loads.4 Some disadvantages of nanoindentation are the need to correct for various instrumental factors, including determining the point of the tip-material contact, accounting for the compliance of the load frame, and determining the area function. Additionally, materials issue such as creep, material pile-up around the tip, and substrate effects must be accounted for. While many of these issues can be minimized by proper calibration and data analysis, failure to account for these effects can lead to complication or misinterpretation of data.5 Many nanoindenters are also equipped with the continuous stiffness measurement (CSM) technique.6 The CSM technique allows the nanoindenter tip to be oscillated at high frequency while the load applied by the tip is incrementally increased. A plot of the 124

CSM load cycle is shown in Figure 8.1. Unlike conventional nanoindentation, where data can only be collected at the unloading point, CSM allows continuous measurement of the mechanical properties throughout the loading cycle. In addition, the high frequency tip oscillation minimizes the influence of creep on the hardness and elastic modulus measurements. CSM also allows measurement of the creep properties of a material by applying a constant load and measuring the change in the indentation depth over time.

Figure 8.1 Load-displacement plot for CSM nanoindentation. The inset shows the high frequency oscillation of the tip during the load cycle.

We performed detailed mechanical characterization of four different materials that have been or could potentially be used in polymer MEMS devices. The materials chosen for this study were poly(methyl methacrylate) (PMMA), poly(propyl methacrylate) (PPMA), poly(styrene) (PS) and a polystyrene-nanoclay composite (PS/Clay). PMMA and PS are widely used engineering polymers. Historically, PMMA 125

has been the polymer of choice in ophthalmologic devices due to its high refractive index, hardness and biocompatibility. Its surface can be functionalized with proteins, which promotes bonding of tissues for in vivo implants.7 PMMA is also employed in chips and valve components for immunosensors and other lab-on-a-chip applications.8,9 PPMA has a lower glass transition temperature (Tg) (35-43 oC)10 than PMMA (104106oC)11,12 which allows for easier processing at a lower temperature. Polystyrene is particularly desirable for BioMEMS due to its ubiquitous use in tissue culture applications. Tissue culture treated polystyrene is the most commonly used material for in vitro cell biology studies of adherent cells. Thus, there is a wealth of information about cellular behavior on polystyrene. In addition to tissue culture polystyrene, various surface modification techniques can be employed for functionalizing the PS surface in order to promote cell attachment and proliferation.7 Oxygen plasma modified polystyrene has been shown to improve cell growth, proliferation, and expression of cellular adhesion proteins proportional to the surface oxygen concentration.13 The previous knowledge of cellular interactions with polystyrene makes it a logical choice to use in BioMEMS devices for cellular interactions. Clay nanoparticles in a polymer matrix act to improve the mechanical and thermal properties as compared to the native polymer.14,15 Nanoclay composites can also be used to improve barrier resistance and improve ionic conductivity. The extent to which polymer properties are affected is determined by the clay content, polymer/clay interfacial strength, and dispersion of the clay particles. Improvements in these properties have made polymer nanocomposites candidate materials for use in the automotive and aerospace industries.16,17 In addition, polymer nanocomposites have been 126

considered for diverse applications such as food packaging, flame retardants and batteries.18-20 In polymer MEMS applications, the ability to control material properties over a range of values could significantly improve device design flexibility. The polystyrene/clay nanocomposite is beneficial, as it may allow fine-tuning of the mechanical properties by varying the filler content while maintaining the biocompatibility of the PS matrix. Complete redesign of the device is avoided as the stiffness of the material can be altered as needed. Since the device design and fabrication procedure are not changed, the process becomes more cost-effective. Here, we present microscale mechanical property characterization of PPMA, PMMA, polystyrene, and polystyrene/clay nanocomposites. Hardness, elastic modulus, and creep properties were measured by CSM nanoindentation on supported polymer microstructures. Elastic modulus was also measured by performing force-deflection measurements on suspended microbridges fabricated using patterned substrate micromolding (PSM). Yield and breaking strengths were evaluated by normal beam bending at elevated loads. Scratch tests were performed on polymer films by translating the nanoindenter tip across the surface on the film while ramping the load. The relative scratch resistance of the materials was evaluated using the tip displacement profile and SEM images of the damage. Finally, lateral bending of PS and PS/Clay cantilever beams is demonstrated for the first time. Hardness and elastic modulus experiments were conducted on samples that were soaked in deionized water to assess the effect of the aqueous medium. In addition, the indentation and bending response of the polymer beams

127

was examined at human body temperature (37.5 oC) by fitting a heating assembly into the indenters sample stage.
8.2 Materials and methods

8.2.1 Fabrication of polymer microstructures and thin films Polymer solutions (PPMA, PMMA, and PS) were made by dissolving the materials in anisole at various concentrations depending on the application. To prepare the nanocomposite, the clay was first dispersed in the PS matrix by melt compounding. The composite was then dissolved in anisole and sonicated for at least 8 hours to dissolve the polymer and re-disperse the particles. The resulting concentration of the clay was 10% wt/wt (clay/PS). The clay additive in PS/Clay is Cloisite 20A surface modified natural montmorillonite (Southern Clay Products, Inc.) with thickness of approximately 1 nm and lateral dimensions of 70-150 nm. Polymer microbridges were fabricated using the patterned substrate micromolding technique21 as described in Chapter 7. Polymer cantilevers for lateral force measurements were fabricated by sacrificial layer micromolding with one process variation. The mold for the lateral force cantilevers consisted of 50 m wide channels that were approximately 27 m deep. The PDMS mold was coated with a 15% solution of PS or 10% solution of PS/Clay. The polymers were removed from the mold and transferred to the substrate using the same process that was used for PS in the previous chapter, except that the PS/Clay was removed at a higher temperature (175 C) than PS (125 C). The primary difference in the processes was the addition of a reinforcing structure at the base of the cantilevers as indicated in Figure 8.2(vii). The reinforcement was applied using the same process, with the reinforcing structures oriented perpendicular to the original 128

structures. This provided the necessary reinforcement to ensure that the polymer cantilever was not delaminated from the surface during the test. The polymer films that were used to assess the scratch resistance were made by dissolving the polymers in anisole at various concentrations to achieve a film thickness of approximately 500 nm. The polymer solutions were then spin coated on silicon at 3000 rpm for 1 minute. The films were then baked at 95 C for 2 minutes to remove any residual solvent. The thickness of the film samples was evaluated by scratching the film with a razor blade and measuring the thickness using a stylus profilometer.

Figure 8.2 Schematic diagram of the process for fabrication of polymer cantilevers for lateral force measurements.

8.2.2 Mechanical characterization Figure 8.3 shows an SEM micrograph of a polymer beam with the testing locations for hardness, elastic modulus, and creep. The figure also shows the location of 129

the bending tests that were used to evaluate the elastic modulus, yield strength, and breaking strength. Hardness, elastic modulus, and creep of the supported section of the microstructures were measured using the CSM technique with a NanoIndentor II (MTS Systems Corp) with a diamond Berovich tip. The hardness and elastic modulus were determined continuously between contact and 500 nm using a peak-to-peak load amplitude of 1.2 N at a frequency of 45 Hz. Creep was measured using a 30 N constant load that was held for 600 seconds.

Figure 8.3 SEM micrograph of the microbridges showing the location of the indentation experiment on the substrate supported portion of the structure and bending tests on the suspended region of the structure.

130

Details of the numerical calculations for determining the hardness and elastic modulus from the nanoindentation data can be found in Oliver and Pharr22 or Bhushan and Li.23 Hardness (H) is calculated using equation 1:
H= Pmax A

(1)

where Pmax is the maximum load and A is the projected contact area. The elastic modulus (E) is calculated by first determining what is referred to as the reduced modulus (Er) according to equation 2: Er =

S
2 A

(2)

where S is the stiffness. The elastic modulus for the material can then be calculated from equation 3: E= 1 2 (1 / E r ) ((1 t2 ) / Et ) (3)

where is the Poissons ratio of the material being tested, and t and Et are the Poissons ratio and elastic modulus of the tip material, respectively. Poissons ratio values for PMMA, PS, PS/Clay, and PPMA were taken as 0.35, 0.325, 0.33 and 0.33, respectively. Data for the Poissons ratio of PS/Clay and PPMA were not available, so a value of 0.33 was estimated based on the value for typical thermoplastic polymers.11 For scratch testing, a conical tip with included angle of 90 and radius of curvature of 1 m was translated over the surface of the polymer films as the load was ramped to a maximum of 0.5 mN. The scratch length was 500 m and the velocity was 5 m/s. The coefficient of friction and scratch depth were measured with respect to the

131

normal load. The damage to the polymer film was evaluated using scanning electron microscopy. The beam bending experiments were performed using a conical tip with 1 m diameter and 90 radius of curvature. A schematic diagram of the nanoindenter setup for bending tests is shown in Figure 8.4. The tip was dip coated with 2% PMMA (wt/wt) to avoid damaging the beams with the tip. The bending location was determined using a 1500x microscope objective to apply the load as close as possible to the center of the beam. In order to maintain the proper location, the recalibration of the distance between the indenter and the microscope was repeated frequently according to the manufacturers specifications. This calibration was performed on a soft material, polycaprolactone, to minimize the damage to the PMMA coating on the tip. Samples were also tested after soaking in deionized water for 36 hours to determine the effect of prolonged exposure to aqueous environment. The sample stage was also heated to 37 C to test the material properties at body temperature. Details of the heating stage can be found elsewhere.24 The elastic modulus was calculated from the load-displacement plot according to equations 4 and 5:25 I= bh 3 12 l3 m 192 I (4)

E=

(5)

where I is the area moment of inertia for the rectangular beam cross section, b is the width of the beam and h is the height of the beam. The elastic modulus can then be calculated using equation 5, where l is the length of the beam and m is the slope of the load displacement curve. 132

Yield and breaking strength of the polymer beams were evaluated by normal beam bending at higher loads than those used to evaluate the elastic modulus. The yield and breaking strength () were calculated from equation 6:25

3 Fl 4 bh 2

(6)

where F is the load corresponding to either the onset of yield or total failure, and l, b, and h are length, width, and height of the beam, respectively.

Figure 8.4 Nanoindentation setup for beam bending.

Lateral bending tests were performed on the polymer cantilevers. The cantilevers were first released by immersion in deionized water for 1 hour. The samples were tested 133

in water to avoid stiction of the cantilevers during the drying process. The nanoindenter tip was then position 20 m from the tip of the free end of the cantilever at a distance of approximately 200-300 m away from the edge of the cantilever. A scratch test was performed at a constant load of 400 N. Any additional load was attributed to the lateral force of the beam against the tip.
8.3 Results and discussion

8.3.1 Hardness and elastic modulus by CSM nanoindentaion Typical plots of the hardness and elastic modulus as a function of the penetration depth are shown in Figure 8.5. Each material was tested five times (n=5) and the values were taken at a penetration depth of the 100 nm. The respective hardness and elastic modulus were found to be 390 50 MPa and 5.1 0.4 GPa for PS/Clay, 340 30 MPa and 4.8 0.5 GPa for PMMA, 290 20 MPa and 3.6 0.4 GPa for PS, and 110 30 MPa and 1.7 0.5 GPa for PPMA. The values are given as the average the standard deviation.

134

Figure 8.5 Plots of hardness and elastic modulus as a function of contact depth for CSM nanoindentation measurements.

8.3.2 Creep behavior by CSM nanoindenation The viscoelastic nature of polymer materials makes creep behavior an important consideration for polymer devices. The creep behavior of the materials was tested by holding the load constant at 30 N and monitoring the displacement of the tip over time. Figure 8.6 shows plots of the tip displacement, mean stress, and contact stiffness versus time for each material. The plots show that the displacement increases and the mean stress decreases over time for each of the materials. This behavior is indicative of creep, as the polymer material deforms around the tip. The data shows that all of the materials exhibit some creep behavior. However, the rate of creep is directly related to the hardness and modulus of the materials, with the harder materials with higher modulus showing the

135

slowest creep rate. The addition of a ceramic component in the PS/Clay nanocomposite also helps to decrease the creep.

Figure 8.6 Displacement, mean stress, and contact stiffness as a function of time for a constant applied load of 30 N.

8.3.3 Scratch resistance from nanindentation scratch test Results from scratch testing are shown in Figure 8.7. The plots give the coefficient of friction (left axis) and scratch depth (right axis) as a function of the normal load. SEM images were taken at the beginning (region A), middle (region B), and end (region C) of the scratch to evaluate the damage to the films. PMMA showed the best scratch resistance, as indicated by a low penetration depth and smooth scratch profile. 136

The addition of clay nanoparticles seemed to improve the scratch properties of the PS/Clay material relative to unmodified polystyrene. The spikes in the depth profile for PS/Clay are attributed to roughness in the film. Polystyrene showed signs of film delamination and material pile-up around the tip at approximately 0.45 mN. This is indicated by sudden instability in the depth profile and is confirmed by SEM observations near the end of scratch. PPMA showed the lowest scratch resistance with delamination and pile-up occurring between 0.2-0.25 mN. Again, this was confirmed by SEM imaging of the scratch.

Figure 8.7 Plots of the coefficient of friction and depth profile for nanoindentation scratch tests (left) and SEM images at the beginning (region A), middle (region B), and end (region C) of the scratch. 137

8.3.3 Elastic modulus from normal beam bending Figure 8.8 shows an example of data from a bending experiment performed with an uncoated diamond tip. The plot shows an obvious change in the load/displacement profile between 75-100 nm of displacement and a load of approximately 2 N. This effect results from significant penetration of the tip into the surface of the beam structures. To avoid this effect, the tip was coated with a thin layer of PMMA for all bending test. While the compliance of the PMMA coating is not considered in the calculation of the modulus, it is assumed to be negligible given the small thickness of the coating.

Figure 8.8 Load-displacement plot for PPMA beam bending with an uncoated diamond nanoindenter tip.

All other bending tests were performed with a PMMA coated tip to minimize the damage to the beam. Figure 8.9 shows typical plots of the loading and unloading cycle for each of the four materials tested. The plots show the expected linear behavior for 138

elastic bending. The slope of the load-displacement plot was determined by a best-fit line of the loading portion of the plot. The elastic modulus and standard deviation (n=5) for each material are shown in the figure and are summarized in Table 8.1.

Figure 8.9 Load-displacement plots for beam bending with a maximum load of 10 N. 139

The elastic modulus from the beam bending can be compared with those obtained from CSM nanoindentation as well as the bulk values available in the literature. The bulk elastic modulus for PMMA and polystyrene were found to be 3.1-3.3 GPa11,12 and 3.2-3.4 GPa11, respectively. These values are shown in Table 8.1. The values for PPMA and PS/Clay were not available in the literature. The results show fairly good agreement between the modulus obtained from all three sources. This suggests several things. First, the variation in the data is attributed to differences between the experimental methods and not to variations in the behavior of the materials at the microscale. In other words, with respect to the elastic modulus, the materials seem to behave according to the bulk properties at this size scale. Moreover, the consistency of the beam bending tests seems to validate this method as a legitimate means to characterize the elastic modulus of microscale polymer structures. Some error may be associated with assumptions on the boundary conditions as well as some plastic deformation of the beam surface by the tip. However, these errors do not have a prohibitive effect on the overall measurements.

Polymer Elastic Modulus (GPa) Bulk (literature) PPMA PMMA 3.1-3.311,12 PS 3.2-3.411 PS/Clay Nanoindentation 1.7 4.8 3.6 5.1 Normal beam bending 0.7 1.9 1.9 4.6

Table 8.1 Summary of elastic modulus measured by nanoindentation and microbeam bending and comparison with bulk values.

140

8.3.4 Effects of aqueous environment and temperature The materials were tested after soaking in an aqueous environment and at human body temperature to determine if the materials behave differently under simulated physiological conditions. Any major differences in the behavior have a significant impact on the application of these materials for biological purposes. The materials were tested by both CSM nanoindentation and beam bending. The results are summarized in Figure 8.10. No significant differences were seen in the properties of PMMA, PS, or PS/Clay in response to aqueous environment or elevated temperature. PPMA showed a decrease in elastic modulus in response to both soaking and elevated temperature. This is not unexpected, given the lower glass transition temperature and more open molecular structure of PPMA relative to the other materials. While PPMA may still be a potentially useful material in polymer MEMS devices, the operating conditions must be considered in the design and implementation of PPMA based devices.

141

Figure 8.10 Hardness and elastic modulus of PPMA, PMMA, PS, and PS/Clay in response to soaking in aqueous environment and temperature measured by (a) CSM nanoindentation and (b) beam bending. 142

8.3.5 Yield and breaking strength from normal beam bending Beam bending tests were performed at higher loads (3 mN and 10 mN) to evaluate both the yield strength and breaking strength of the materials. These experiments were also valuable for evaluation of the failure mechanisms of the polymer microstructures. Figure 8.11 shows load-displacement plots for beams tested at 3 mN and 10 mN. Arrows in the plots indicate the yield point and breaking point. The plots at 3 mN show the obvious onset of yield as the load-deflection plot becomes nonlinear. At this load, however, only PS/Clay failed completely as indicated in the figure. At 10 mN, all of the beams broke as indicated by a gap in the load-displacement plot. As expected, when the beam breaks, there is no longer a force being applied to the beam, which in turn leads to a gap in the data collection. The spike in the load-displacement plot after the beams break results from the nanoindenter tip becoming wedged in the channel. The data and SEM images show one significant difference in the failure between the materials: PPMA, PMMA, and PS all exhibited significant plastic deformation prior to total failure, indicating a ductile failure, while PS/Clay, although also exhibiting some plastic deformation, showed a more brittle failure.

143

Figure 8.11 Load-displacement plots and SEM images for bending tests at 3 mN (top) and 10 mN (bottom). 144

8.3.6 Lateral beam bending Results of lateral force bending tests and SEM images of the cantilevers after testing are shown in Figure 8.12. The load-displacement plot is in general agreement with the expected behavior of the materials. A linear region of the plot is followed by the onset of yield. For PS, a decrease in the load following yield indicates cracking of the beam. For PS/Clay, the gap in the data indicates breaking of the beam. While this new method for evaluating materials shows some promise for materials evaluation, the results from the tests do not match well with data obtained via other methods. Therefore, the results from these tests are only considered to be qualitative. However, proof of principle for this method of testing is established, and further refinement of this technique could make it a valuable tool for evaluating materials and devices that are subjected to lateral forces.

Figure 8.12 Load-displacement plots and SEM images of PS and PS/Clay cantilever measured by lateral bending. 145

8.4 Conclusion

Mechanical properties of polymer microstructures and polymer films were evaluated using several different techniques. CSM nanoindentation of supported polymer microstructures was used to measure hardness, elastic modulus, and creep. The scratch resistance of the materials was evaluated with the nanoindenter scratch testing with the aid of SEM imaging. Beam bending experiments were performed on suspended polymer microbridges and the results were used to calculate the elastic modulus, yield strength, and breaking strength of the microstructures. Aside from breaking strength, the property measurements were similar to bulk materials properties obtained from the literature. Hardness and elastic modulus were evaluated under simulated physiological conditions. Only PPMA showed a change in mechanical properties under these conditions. Lateral force bending of the beams was demonstrated as a potential method for evaluating materials, but further refinements to the technique are need in order to obtain meaningful quantitative property measurements. References
1

J.A. Forrest, D. Dalnoki-Veress, The glass transition in thin polymer films. Advances in Colloid and Interface Science 94 (2001) 167-196.

J.L. Keddie, R.A.L. Jones, R.A. Cory, Size-dependent depression of the glass transition temperature in polymer films. Europhysics Letters 27 (1994) 59-64. B. Bhushan, A.V. Kulkarni, W. Bonin, J. Wyrobek, Nano/picoindentation measurement using a capacitance transducer system in atomic force microscopy. Philosophical Magazine A 74 (1996) 1117-1128.
4 5 3

S.J. Bull, Nanoindentation of coating. Journal of Physics D 38 (2005) R393-R413.

A.C. Fischer-Cripps, Critical review of analysis and interpretation of nanoindentation test data. Surface and Coatings Technology 200 (2006) 4153-4165. 146

X. Li, B. Bhushan, A review of nanoindentation continuous stiffness measurement technique and it applications. Materials Characterization 48 (2002) 11-36. I.H. Loh, M.S. Sheu, A.B. Fishcer, Biocompatible polymer surfaces, in Desk Reference of Functional Polymers: Synthesis and Applications (edited by R. Arshady) American Chemical Society, Washington, D.C. (1997). D.B. Holt, P.R. Gauger, A.W. Kusterbeck, F.S. Ligler, Fabrication of a capillary immunosensor in polymethyl methacrylate. Biosensors and Bioelectronics 17 (2002) 95103. A. van der Berg (ed.), Lab-on-a-chip Chemistry in Miniaturized Synthesis and Analysis Systems, Elsevier, Amsterdam (2003). B. Ellis, Polymers: A Properties Database, Available on compact disk, CRC, Boca Raton, FL (2000). J. Brandrup, E.H. Immergut, E.A. Grulke (eds.), Polymer Handbook, Wiley, New York, NY (1999).
12 13 11 10 9 8 7

J.E. Mark, Polymer Data Handbook, Oxford University Press, Oxford (1999).

T.G. van Kooten, H.T. Spijker, H.H. Busscher, Plasma-treated polystyrene surfaces: model surface for studying cell-biomaterial interactions. Biomaterials 25 (2004) 17351747. M. Alexandre, P. Dubois, Polymer-layered silicate nanocomposites: preparation, properties and uses of a new class of materials. Materials Science and Engineering 28 (2000) 1-63. S.S. Ray, M Okamoto, Polymer/layered silicate nanocomposites: a review from preparation to processing. Progress in Polymer Science 28 (2003) 1539-1641. J. Njuguna, K. Pielichowski, Polymer nanocomposites for aerospace applications: properties. Advanced Engineering Materials 5 (2003) 769-778.
17 18 16 15 14

F. Gao, Clay/polymer composites: the story. Materials Today 7 (2004) 50-55.

G. Beyer, Nanocomposites: a new class of flame retardants for polymers. Plastic Additives and Composites 4 (2002) 22-28. M. Kurian, M.E. Galvin, P.E.. Trapa, D.R. Sadoway, A.M. Mayes, Single ion conducting polymer-silicate nanocomposite electrolytes for lithium battery applications. Electrochimica Acta 50 (2005) 2125-2134. 147
19

R.Y. Lockhead, C.R. Haynes. S.R. Jones, V. Smith, The high throughput investigation of polyphenolic coupler in biodegradable packaging materials. Applied Surface Science 252 (2006) 2535-2548. H. Liu, B. Bhushan, Investigation of nanotribological properties of self-assembled monolayerss with alkyl and biphenyl spacer chains. Ultramicroscopy 91 (2002) 185202.
22 21

20

W.C. Oliver, G.M. Pharr, An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. Journal of Materials Research 7 (1992) 1564-1583.

B. Bhushan, X. Li, Nanomechanical characterization of solid surfaces and thin films. International Materials Review 48 (2003) 125-164. H. Liu, B. Bhushan, Investigation of nanotribological properties of self-assembled monolayers with alkyl and biphenyl spacer chains. Ultramicroscopy 91 (2002) 185-202. W.C. Young, R.G. Budynas, Roarks Formulas for Stress and Strain, Wiley, New York, NY (2002).
25 24

23

148

CHAPTER 9 DESIGN, SIMULATION, AND FABRICATION OF A POLYMER SENSOR FOR MEASURING SINGLE CELL FORCES

9.1 Introduction

One particular area where polymer MEMS could be effectively employed is in the field of cellular biomechanics. When cells adhere to the extracellular matrix or to a foreign material, mechanical forces are generated at the adhesive contacts. Following adhesion, cytoskeletal organization occurs, and molecular linkages between the adhesion contacts and cytoskeletal components transmit forces through intracellular stress fibers and adhesion sites to the underlying substrate.1 One of the key barriers to studying cellular mechanics is a lack of sensitive measurement devices with the ability to accurately evaluate static and dynamic cellular forces. A few MEMS and non-MEMS devices have been explored for measuring cellular adhesion2,3, traction forces4,5, cell contractile forces6, and intracellular mechanics7. Polymer MEMS offer several advantages over current devices due to their inherent biocompatibility, high sensitivity based on material properties, and low cost fabrication methods. We set out to design a polymer sensor to measure forces generated by single adherent cells. The device is based on an array of polymer microcantilevers. When cells 149

attach to the cantilevers, forces are transmitted to the beams resulting in deflection. By measuring the beam deflection and calculating the force-deflection response of the beam, the force vector can be calculated. Design and simulation are critical steps in the device development process. Proper design and accurate simulation of the device behavior can be performed prior to fabrication of a prototype device. An adequate understanding of the key design parameters will increase the likelihood that the device will provide the necessary functionality in practice. Well thought out simulations of the device behavior will ensure that the device functions in the expected manner. Both of these steps ultimately lead to fewer design iterations, decreased cost, faster device development, and improved functionality. The design and simulations of the polymer MEMS cell force sensor will be discussed here. A general overview of the device will be given along with a detailed discussion of the parameters that were considered in the design. The force-deflection behavior of the device was modeled using finite element analysis. Improvements in the device design for second and third generation devices will be discussed with respect to improving the force measurement capabilities as well as improving the biological interactions between the cell and the device. The data analysis process for converting experimental beam deflection data into force vector information will also be discussed.

150

9.2 Materials and methods

9.2.1 Solid modeling for device visualization SolidWorks three-dimensional computer aided drafting (CAD) software was used to visualize the device during the design process. This allowed for visualization of different design iterations to consider possible problems with respect to device fabrication and function. Additionally, CAD drawings could be integrated directly into the modeling and fabrication processes. SolidWorks data could be imported into the finite element simulation package and used directly in the photomask fabrication process. 9.2.2 Finite element analysis The response of each prototype design was simulated using ANSYS finite element analysis software. A single probe was isolated and analyzed by applying a range of forces to the end of the cantilever beam and observing the deflection simulations. Several simulation parameters were considered: changes in the simulations with respect to the mesh (i.e. solid versus surface mesh), number of mesh elements, boundary conditions, nature of the applied load, and force application location. Further design modifications were made based on the simulation results to arrive at a set of initial designs. 9.2.3 Device development and prototyping process The process of transitioning from design to prototype began with the photomask preparation process. This involved first developing the layout of the mask in SolidWorks. One important consideration in the layout is the number of different designs that are considered for a single mask. Too few designs can result in a narrow range of operating 151

parameters, and may require new mask production after a short period of time. Too many designs can be impractical given that a limited number of different devices can be tested over a given period of time. Another important consideration is the spacing between the devices. Proper spacing can make device fabrication and the alignment of different device layers significantly easier. In addition, the number of repeats of each design is important. This determines the frequency of fabrication runs and impacts the device fabrication time and cost. After the layout was completed, the SolidWorks file was exported as a .dxf (AutoCAD file extension) file that could be sent directly to the mask manufacturer for processing. 9.2.4 Device fabrication The devices were fabricated using sacrificial layer micromolding (SLaM). The process will be described briefly here, but details of the process can be found in Chapter 7. The fabrication schematic for the sensor is also shown in Figure 9.1. All devices were initially fabricated from SU8 5 or SU 2005 (Microchem Corp.) with minor changes in the photoresist thickness depending on the desired beam thickness. Devices were transferred to PDMS molds using the standard molding process. Polyvinyl alcohol (PVA) sacrificial layers were produced by photolithography followed by reactive ion etching to expose the anchor regions of the substrate. Initial devices were fabricated from PPMA. The polymer was dissolved in anisole (Sigma-Aldrich) at concentrations ranging from 1.2-10% (wt/wt). The polymer solution was spin coated at onto the patterned PDMS mold at spin speeds from 2000-6000 rpm for 60 seconds. Resultant device thickness was a function of the PDMS mold depth, polymer concentration, and spin speed. Devices ranged from 180 nm to 3.7 m based on the above parameters. To remove the surface polymer, the 152

mold was brought into contact with a glass slide heated to 175-185 C. The mold was held on the slide for ten seconds under the pressure of its own weight and then removed. The process was repeated if necessary to completely remove the surface PPMA, leaving the polymer only in the recessed features. The selectively patterned PDMS mold was then manually aligned with the sacrificial layer under an optical microscope at 40x magnification. The substrate was heated to 95 C and a pressure of 30 lb/in2 was applied to the backside of the mold and held for 10 seconds. Later generations of the device were fabricated from polystyrene. In this case, a 7.5% (wt/wt) solution was spin coated on the mold at 3000 rpm. The first stamping was done at 200 C and the temperature of the second stamp was 120 C and the pressure was increased to 70 psi. The polystyrene devices were also annealed at 115 C for 15 minutes to reduce any residual stress in the beams.

Figure 9.1 Fabrication of force sensors by sacrificial layer micromolding; (a) Sacrificial layer patterning process, (b) structural layer molding, and (c) alignment and bonding.

153

9.2.5 Device characterization The thickness of each device was measured individually after testing. The thickness was measured using either atomic force microscopy (AFM) or a stylus profilometer (Dektak, Series 3). Scanning electron microscopy and optical microscopy were used to image the devices.
9.3 Results and discussion

9.3.1 Design considerations Several factors were taken into consideration in the design of the sensor. Factors related to function, fabrication, materials selection, and cost were all considered individually or in combination to determine a set of designs for the initial fabrication run. In term of function, it was necessary to design a device capable of measuring forces with high sensitivity and high resolution. The expected force range for most cells is in the range of hundreds of nanonewtons (nN) and dynamic changes in these forces are often small (on the order of tens of nN) over short periods of time. The device must be capable of measuring forces and changes in force on this scale. Spatial resolution of the force measurements was also considered. A sufficient number of measurement locations on each cell is advantageous in providing information about the spatial distribution of forces. Material selection was a critical parameter in the function, biological interactions, and cost. Polymers offer some particular advantages over traditional MEMS materials. The low elastic moduli of polymers are advantageous for low magnitude force measurements. Cantilevers made from low modulus materials provide larger deflections per unit force, thus increasing the sensitivity of the measurements. Many polymers are 154

also biocompatible and are currently used in a number of implant and in vitro cell biology applications. This is particularly true with polystyrene, which is the most common material for in vitro cell culture. Polymers are generally less expensive than other MEMS materials for both the material and processing costs. Most of the fabrication techniques described here can be performed outside of a cleanroom setting without the expensive processing equipment needed for purely silicon based processing. The beam geometry was considered in the overall cost of device development. One of the most costly steps in the process is mask production. The overall cost of the mask is determined primarily by the smallest feature size. Masks with a minimum feature size of 5 m or larger are significantly less expensive than masks with features smaller than 5 m. This was taken into consideration relative to the width of the beam and the spacing between the cantilevers. A beam width of 5 m was used and a minimum spacing between cantilevers was also 5 m. The design also took into consideration the evaluation of future devices with increased levels of sophistication. It was necessary to use a generalized fabrication protocol that could be easily adapted to different materials, device geometries, and applications. The ability to add higher-level functions to the basic device was also considered. The first several generations of the devices are passive structures that respond to the cell. The ability to incorporate fluidics and actuation mechanisms into an active device could significantly increase functionality. 9.3.2 Device overview The device is based on an array of compound polymer cantilever beams. The device consists of three primary functional regions. These are shown in the first 155

generation device model in Figure 9.2. The first region is the cell active area. This is the portion in the center of the device where the cell is located for the measurements. When the cell attaches and spreads on the cell active area, forces are transmitted to the underlying cantilever beam array. This leads to deflection of the beams proportional to the magnitude of the applied force. The beam deflections are captured over time using a CCD camera attached to an inverted phase contrast microscope. The force vector is then calculated from the deflection data. The second region of the device is the cantilever array. This consists of a set of suspended cantilever beams that span the distance between the cell active area and the outside edge of the device. The cantilevers are configured such that forces can be measured in any direction in the plane of the structure. The shape of the beam is designed to give a specific force-deflection behavior as guided by finite element simulations. The third portion of the device is the anchor region, where the device is fixed to the underlying substrate.

Figure 9.2 Generation 1 cell force sensor with labeled functional regions. 156

9.3.3 First generation cell force sensor The first generation of the device consisted of an array of compound cantilevers with a single bend in the beam (L-beam) that provides the ability to measure force in multiple directions. Two configurations of the first generation device were considered. The circular configuration is shown in Figure 9.2 and the linear configuration is shown in Figure 9.3. In the linear configuration, the beams are fixed at anchor region of the device and converge in the cell active area. For the linear configuration, the beams are fixed in two parallel sets of beams on either side of the device. The circular configuration is well suited for essentially static cells, while the linear configuration was designed to allow measurements on motile cells.

Figure 9.3 Linear configuration of the first generation cell force sensor with labeled functional regions.

The force-deflection behavior of the first generation device was simulated using finite element analysis. Inputs to the finite element simulations included the geometry of 157

the structure and the mechanical properties of the structural material. The geometry of the device was imported directly from the SolidWorks models, with the exception of the beam thickness. A discussion of the method to correct for the thickness of the beam is given in Section 9.3.6. The elastic modulus is also important for simulating the forcedeflection behavior. For PPMA devices, a modulus of 2.0 GPa was used. This value was estimated based on being in the middle range between the values for the elastic modulus obtained from normal beam bending and nanoindentation experiments (see Chapter 8). Force inputs were applied to the beams in 30 increments over 360 as shown in Figure 9.4. Forces at three magnitudes (5 nN, 10 nN, and 20 nN) were applied to a circular area in the center of the pad at the end of the cantilever. The x,y deflection profile for all three forces are shown in Figure 9.5.

Figure 9.4 (a) Image of the L-beam cantilever used in the FEA simulations, (b) close up view of the end of the beam with the direction-angle convention.

158

The deflection profile shows several key features. Ideally the x,y deflection would produce a circular profile with the direction of the deflection being in the same direction as the force. The plot in Figure 9.5 shows a skewed elliptical deflection profile, indicating that the beam responds non-linearly to forces applied in different directions. The consequence is that the device has different resolution and sensitivity depending on the direction of the deflection. For deflections in direction 6 and 12, the device has higher sensitivity and gives a larger deflection for a given force. However, there is lower resolution in determining the direction of the force. The opposite is true for deflection corresponding to directions 3 and 4 (60-90) and directions 9 and 10 (240-270). In these general directions, the device has lower sensitivity and gives smaller deflection for a given force, but the ability to resolve the direction of the force improves. For the direction convention, refer to Figure 9.4. In addition, the plot shows a significant difference between the direction of the deflection and the direction of the force. This complicates determination of the force vector direction from the deflection data. A detailed discussion of determining the force vector magnitude and direction is given in Section 9.3.6. The plot also shows the expected linearity between the magnitude of the force and the magnitude of the deflection, as indicated by the concentricity and spacing of the ellipses.

159

y-deflection (m) -2.0 -1.5 -1.0

x-deflection (m)

Position 7
-3.5 -3.0 -2.5 -2.0 -1.5

Position 4
-1.0 -0.5

-0.5 0.0 0.0 0.5 0.5 1.0 1.5 2.0 2.5 3.0 3.5

Position 10
5 nN
1.0 1.5 2.0

Position 1

15 nN

30 nN

Figure 9.5 Deflection plot for the first generation cell force sensor.

The finite element simulations were also used to determine if the devices behave in the linear elastic regime for the expected range of forces. Estimates of the yield strength for PPMA and polystyrene can be found in chapter 8. The values for maximum stress and strain found in the FEA simulations were well below even more conservative estimates of stress and strain needed to induce yield of the materials. Despite some complications in the L-beam design, this configuration still provides adequate force-deflection characteristics to allow for meaningful measurement to be taken. The first fabrication run was performed with the first generation device in both the circular and linear configurations. Devices with 4, 6, 8, 10, and 12 probe circular configurations and a linear device with 14 cantilevers were chosen for the initial fabrication run. 160

Fabrication results for the first generation sensor are shown in Figure 9.6. The SEM images in Figure 9.6 are shown prior to release of the sacrificial layer. An optical micrograph of the device immersed in phosphate buffered saline after release is shown in Figure 9.7(a). A microprobe attached to a micromanipulator was used to apply a force to the beam to induce a deflection, showing that the beam is free to move and that the beam returns to its initial location after the probe is removed as shown in Figure 9.7(b,c).

Figure 9.6 SEM micrographs of the (a) circular 12 probe device and (b) linear 14 probe device.

161

Figure 9.7 Optical micrographs of the circular 4 probe device (a) before bending with a microprobe, (b) during bending, and (c) after removing micromanipulator tip.

9.3.4 Second generation cell force sensor Given the limitations in the L-beam design and lessons learned from initial fabrication and biological testing with the first generation device, a second generation device was developed to improve the force-deflection response of the beam, make fabrication easier, and improve the cell-sensor interactions. The geometry of the compound cantilever was redesigned to provide a more uniform deflection profile. Several lessons learned from the L-beam were taken into consideration for the new design. Most notably, the highest deflection per unit force was in the direction perpendicular to the line between the fixed base of the beam and the end of the beam as shown in Figure 9.8. While this line provides the highest sensitivity, it also gives the 162

largest offset between the direction of the deflection and the direction of the force. Therefore, by combining two L-beams into a single beam, with specific angles between the fixed base and the ends of the beams, it was possible to decrease the offset between the direction of the deflection and the direction of the force. In addition, this indicates a more homogeneous sensitivity. This was the guiding principle in the development of the second generation beam design.

Figure 9.8 L-beam showing the line of highest sensitivity.

Several other factors were taken into consideration for designing the second generation device. First, geometrical constraints imposed by the circular configuration affects the ability to increase the number of cantilevers on the device without physical interference (contact) between the beams. During the iterative design process, it was also noted that the geometry of the bends in the cantilever affects of the response. This is shown in Figure 9.9. By reducing the width of the beam and the bends, it is possible to produce a hinge-like structure and reduce the non-linearity in the deflection response.

163

However, this is constrained by the width of the beam. It is advantageous to maintain the 5 m minimum feature size to keep the cost of the device low.

Figure 9.9 Image of the beam joint showing the material that was removed to create a hinge-like bend.

With respect to the fabrication and biological aspects of the device, it was noted in the first generation that the width of the anchor region small enough that it was difficult to align the structural and sacrificial regions of the device. The width of the anchor region was increased from 50 m to 130 m to minimize this issue. It was also noted that it was difficult for the cells to attach and spread on the relatively small surface area of the first generation devices. The cell active area of the second generation devices was increased significantly to allow the cell to spread more naturally on the device. The final geometry for the second generation beam is shown in Figure 9.10(a). Figure 9.10(b-e) shows how the new beam design fits into a 4, 6, 8, and 10 probe systems without interference between the beams. The images also show the increase in the area of the anchor region to decrease alignment problems.

164

Figure 9.10 Images of (a) final design of the second generation compound beam (b) 4 probe device (c) 6 probe device, (d) 8 probe device, and (e) 10 probe device.

Figure 9.11 shows an example of the progression of the finite element simulations for the second generation device. The geometry of the beam was first imported from 165

SolidWorks. The boundary conditions were established by placing a fixed support at the end of the beam and the load was applied to top surface of the free end of the beam [Figure 11(a)]. The mesh was then defined as shown in Figure 9.11(b). After imputing the material properties, the simulation was run to solve for the x and y deflections of the beam [Figure 11(c)]. The angle of the force input was then changed in 30 increments and the simulation and analysis was repeated over 360. The deflection of the end of the cantilever was analyzed and used to create the deflection plots.

Figure 9.11 FEA simulations; (a) boundary condition and applied force, (b) mesh, and (c) deflection solution. 166

The x,y deflection plot for the second generation beam is shown in Figure 9.12. The plot shows the profile for 5 nN, 10 nN, and 20 nN forces. The plot shows that there is significantly less offset between the deflection direction and force direction. While the plot is still elliptical, it is much closer to circular than the first generation device. This gives a more uniform sensitivity to forces in any direction. A comparison between the first and second generation beams is shown in Figure 9.13. The plot shows the differences in the deflection profile for a 5 nN force. Improvements to the offset and the uniformity are obvious. Sensitivity of the device is reduced in the direction corresponding to 0 and 180 relative to the sensitivity of the first generation device. The overall uniformity in the sensitivity is improved, thus providing better measurement of forces independent of direction. The deflection profile for the second generation device also improves the ability to resolve the force direction.

y-deflection (m)

-1

-0.5
x-deflection (m)

Pos.4

-1

Pos.7

-0.5 0

0.5
Pos.1

0.5
5 nN
10nN
20 nN

Pos.10

Figure 9.12 Deflection plot for second generation compound beam. 167

y-deflection (m) -0.75

-0.50

-0.25 x-deflection (m) -0.75 -0.50 -0.25 0.00 0.00 0.25 0.50 0.75

0.25

Initial L-beam response

0.50

second generation beam 0.75 response

Figure 9.13 Comparison of the beam deflection response for the L-beam and second generation beam designs.

Fabrication results for the second generation devices are shown in the Figure 9.14. The SEM micrographs show PPMA sensors with (a) 4, (c) 6, and (e) 10 probe configurations. Figure 9.14(b,d,f) shows SEM micrographs of devices fabricated from polystyrene. Polystyrene replaced PPMA as the structural material in order obtain a wellunderstood and biocompatible surface for cell attachment. Figure 9.15 shows phase contrast micrographs of immersed polystyrene sensors after release of the sacrificial layer.

168

Figure 9.14 SEM micrographs of cell force sensors; (a,b) four probe sensors, (c) six probe sensor, (d) eight probe sensor, (e,f) ten probe sensors. The devices in the left column are made from PPMA and the right column are made of polystyrene.

169

Figure 9.15 Phase contrast optical micrographs of (a) 4 probe sensor, (b) 10 probe sensor, and (c) 6 probe sensor after releasing the sacrificial layer in aqueous solution.

9.3.5 Third generation cell force sensor Improvements to the sensor design for the third generation were primarily driven by improving the biological interactions between the cell and the device. Initial experiments with the second generation device revealed that significant improvements could be made to improve the cells ability to attach, spread, and function normally on the devices. Changes were made primarily to the cell active area. First, an adhesion pad was placed in the center of the device and fixed to the substrate in the same manner as the anchor region. This adhesion pad was designed to provide the cell with a surface for initial attachment. The size of the pad was determined based on cell area measurements taken from two cell lines of interest: HT 1080 fibrosarcoma cells and 3T3 mouse 170

fibroblasts. The pad area was designed to provide a surface for cell attachment by also allow the cell to bridge between the pad and the cantilevers during cell spreading. The area of the end of each cantilever was also increased to provide a larger area for the cell to spread. It should be noted that all of the final experiments were performed with WS-1 skin fibroblasts instead of either of the cell lines that were analyzed. However, these measurements gave a good estimate of the area needed to promote healthy cell behavior on the sensor. Cell area measurement for HT 1080 and 3T3 cells are shown in Figure 9.16. The projected cell area for unattached cells and for cells after 12 hours of attachment are given. Based on these measurements, the areas of the adhesion pads that were chosen for the fabrication run were 400 m2 (4 probe sensors-design 1), 420 m2 (6 probe sensor-design 1), and 625 m2 (4 and 6 probe sensors-design 2).

900 800 700 Projected Cell Area (m2) 600 500 400 300 200 100 0 t=0 min. t=720 min. HT1080 3T3

Figure 9.16 Projected cell area for HT1080 and 3T3 cells before attachment (t=0 min.) and after spreading (t=720 min.). 171

Even though the basic geometry of the beam was not changed for the third generation sensors, the increase in the size of the pad led to a decrease in the overall length of the 5 m wide section of the beam. This had a significant impact of the forcedeflection behavior, making it necessary to perform new FEA simulations and produce new force-deflection plots for the third generation device. It should be noted that the angle convention was changed for the simulations of the third generation device. The angle convention is shown in Figure 9.17. The change in the angle convention only changes the appearance of the profile (i.e. the orientation of the profile). It is critical, however, that the same convention is maintained when experimental data is analyzed.

Figure 9.17 Angle convention for third generation device simulations.

Additional FEA simulations were run to determining the effects of the area and location of the applied load, the location on the beam that the deflection is analyzed, and the pad area on the deflection profile. Four different loading scenarios were tested as shown in Figure 9.18. The figure shows the load applied to a (a) 5 m diameter circular area at the center of the pad, (b) 5 m diameter circular area at the end of the pad, (c) line at the end of the pad, and (d) the area of the front half of the pad. The difference in the 172

deflection profile for each of the different loading conditions is shown in Figure 19. The plot does show a small difference in the behavior depending on the loading conditions. It was determined that the load of the front half of the pad area would be used in the analysis. This condition was chosen for two reasons. First, this loading condition seems to fit most appropriately with the actual behavior of a cell on the device. Second, this produces a response in the middle of the two extremes, thus providing a reasonable estimation of the loading.

Figure 9.18 FEA loading conditions; (a) load applied to a 5 m diameter circle at the center of the pad, (b) load applied to a 5 m diameter circle at the center of the pad, (c) load applied to the front edge line of the pad and (d) load applied to the front half area of the pad.

173

0.2 0.15 0.1 0.05 y-deflection 0 -0.2 -0.1 -0.05 0 0.1 0.2

(a) (b) (c) (d)

-0.1 -0.15 -0.2 x-deflection

Figure 9.19 Deflection plot for four different loading conditions; (a),(b),(c), and (d) correspond to the loading conditions in Figure 9.18.

The location on the cantilever where the deflection was measured was also tested to determine if different locations affect the profile. The left and right side of the tip of the cantilever were tested. These locations were chosen because they give the largest displacement for a given force. The locations that were tested are shown in Figure 9.20(a) and the difference in the profile is shown in (b). The plot shows that there is a difference in the profile, with the right side measurement showing more deflection offset. Therefore, the left side displacement was used for simulations and for all experimental measurements.

174

0.2

0.15

0.1

0.05 y-deflection

0 -0.2 -0.15 -0.1 -0.05 -0.05 0 0.05 0.1 0.15 0.2

-0.1

Right Side Left Side

(b)

-0.15

-0.2 x-deflection

Figure 9.20 (a) Convention for location of deflection analysis, (b) deflection plot for deflection analysis on the right and left side of the probe.

The area of the pad was also tested to determine if this had an effect on the profile. This is important because the individual pad area is different for the 4 and 6 probe configurations. The profiles showed that the area of the pad did not have a significant effect on the profile, provided that the length of the 5 m wide section of the beam was maintained constant. A summary in the difference between the deflection plots for the first, second, and third generation devices are shown in Figure 9.21. The angle conventions were transformed into the convention of Figure 9.17 to allow for direct comparison of the profiles. The first generation device showed significant angular offset between the force 175

direction and the displacement direction. The flat elliptical nature of the profile leads to a higher device sensitivity in certain directions relative to others as well as lower resolution in determining the force direction. The second generation device shows a significant improvement in the profile, with better uniformity in the sensitivity for force in any direction, decreased angular offset between the force and displacement direction, and improved resolution in determining the force direction. The third generation device shows that the overall beam is stiffer based on shortening the length of the 5 m wide portion of the beam. The profile shows even less angular offset between the force and displacement directions based on choosing the most appropriate location to analyze the displacement.

0.400

Gen 1 Gen 2 Gen 3

0.200 0.100 0.300

y-deflection

-0.600

-0.400

-0.200

0.000 0.000 -0.100 -0.200 -0.300 -0.400

0.200

0.400

0.600

x-deflection

Figure 9.21 Comparison of the deflection plot for first, second, and third generation cell force sensors. 176

Fabrication results for the third generation device are shown in Figure 9.22. A lower magnification image of the overall device is shown in Figure 9.22(a). A higher magnification SEM image (b) shows the cell active area with the increase in the size of the beam ends and the inclusion of the central adhesion pad.

Figure 9.22 SEM micrographs of the third generation sensor at (a) low magnification and (b) high magnification.

9.3.6 Determining the force vectors from displacement data To this point, the deflection profiles have been given only for a few specific forces. For practical experimental purposes, it is necessary to calculate the force magnitude and force direction from deflection measurements over a wide range of magnitudes and directions. This was done in three basic steps. First, the x,y deflection plot was made from FEA simulations at 10 increments in the force direction in the same manner described previous. The deflection plot is given in Appendix A. The x,y deflection data points were used to calculate the deflection angle for each simulated force direction. The force angle was then plotted versus the beam angle. Parabolic curves were fit to the data over a range of angles that provided a R2 value for the curve fit > 0.998. An 177

example for beam angles between 1.6 and 27.6 is shown in Figure 23. The plots for the remaining angles are given in the Appendix A. This provided a set of equations to transform the experimentally determined beam deflection angle into the force direction angle. The set of equations is also available in Appendix A. The calculated force direction is then used in the next step to calculate the magnitude of the force.

70 60 y = -0.055914x + 3.937841x - 6.182234 R = 0.999931

2 2

Force Angle ()

50 40 30 20 10 0

10

15

20

25

30

Beam Deflection Angle ()

Figure 9.23 Plot of force angle versus beam deflection angle from FEA simulations. The experimentally determined beam angle can then be used to calculate the force angle.

The next step in the process was to calculate the magnitude of the simulated beam deflection by simply taking the sum of the squares of the x,y deflection for a specific force magnitude (i.e. 5 nN). The magnitude of the beam deflection was then plotted versus the force angle for a specific range of force angles. Parabolic curves were fit to this data such that R2 > 0.998 over the force direction range. An example for force angles from 0-60 is shown in Figure 9.24. The remaining plots and equations are given in Appendix A. This gives a set of equations to calculate the expected deflection magnitude 178

per unit force in a given direction. Since the deflection magnitude scales linearly with the force magnitude, the magnitude of the force can be calculated by taking the experimentally determined deflection divided by the simulated deflection per unit force.

0.18 0.16 y = -0.00001578x2 - 0.00026049x + 0.15466464 R2 = 0.99959381

Beam Deflection (m)/5nN

0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 0 10 20 30 40 50 60 70

Force Angle ()

Figure 9.24 Plot of beam deflection per 5 nN of force versus the angle of the force.

The final step in the process was to correct the force magnitude for other variables including the elastic modulus and the beam thickness. Since the force magnitude scales linearly with elastic modulus, it was possible to correct for the elastic modulus by simply multiplying the magnitude of the force by the ratio of the actual modulus to the modulus used in the FEA simulation. The force magnitude does not scale linearly with the beam thickness. To account for this factor, FEA simulations were performed for a range of 179

thicknesses (1-10 m) at a specific force angle (i.e. 0). The magnitude of the deflection was then calculated from the x,y deflection data. The magnitude of the deflection was then normalized to the deflection at a beam thickness of 3 m. The normalized deflection was then plotted again the beam thickness. An exponential curve was fitted to the data. The equation for the curve could be used to calculate a correction factor to account to the beam thickness. The magnitude of the force could then be calculated by dividing the uncorrected force magnitude by the correction factor. In summary, a methodology for calculating the force magnitude and direction from experimental deflection data was given. This is done by first calculating the direction of the force from the experimentally determined beam deflection angle. The force magnitude is then calculated from the previously determine force angle, the experimentally determined beam deflection magnitude, and the expected deflection per unit force at a given force angle. The force magnitude was then corrected for elastic modulus and beam thickness. The beam thickness correction factor was determined by running FEA simulations over a range of beam thicknesses. All of the equations for calculating the force vector and the correction factors were integrated into an Excel spreadsheet. The spreadsheet was configured such that the angle of the probe deflection and the magnitude of the deflection could be input into the spreadsheet. Then for a given material and beam thickness, the angle and magnitude of the force were automatically calculated. The layout of the spreadsheet is given in the Appendix A.

180

9.4 Conclusion

A cantilever sensor for measuring forces generated by single adherent cells was designed with the aid of SolidWorks 3-D CAD software and ANSYS finite element simulations. The design consists of an array of cantilevers with a specific geometry to allow for flexibility in measuring force in multiple directions with high sensitivity. A number of other design consideration were also taken into account during the development process. Some of the other key factors included ease of fabrication, material properties and biocompatibility, and cost. Three generations of the devices were designed with functional and biological improvements being made for each generation. A much improved beam response was achieved for the second generation device by considering the response behavior of the first generations beam (L-beam) and performing an iterative design with the aid of FEA. The third generation device was re-designed with biological improvement being the primary goal. The simulated behavior of each device was discussed in detail along with the methodology for determining force vectors from experiment data. Fabrication of each generation of the force sensor was demonstrated. References C.G. Galbraith, M.P Sheetz, Forces on Adhesive Contacts Affect Cell Function. Current Opinions in Cell Biology 10 (1998) 566-571. A. Yamamoto, S. Mishim, N. Maryuama, M. Sumita, A new technique for direct measurement of the shear force necessary to detach a cell from a material. Biomaterial 19 (1998) 871-879.
3 2 1

K.A. Athanasiou, B.S. Thoma, D.R. Lanctot, D. Shin, C.M. Agrawal, R.G. LeBaron, Development of the cytodetachment technique to quantify mechanical adhesiveness of a single cell. Biomaterial 20 (1999) 2405-2415. 181

C.G. Galbraith, M.P. Sheetz. A micromachined device provides a new bend on fibroblast traction forces. Proceeding of the National Academy of Sciences USA 94 (1997) 9114-9118. J.L. Tan, J. Tien, D.M. Pirone, D.S. Gray, K. Bhadriraju, C.S. Chen, Cells lying on a bed of microneedles: An approach to isolate mechanical force. Proceeding of the National Academy of Sciences USA 100 (2003) 1484-1489. R.M. Hochmuth, Micropipette aspiration of living cells. Journal of Biomechanics 33 (2000) 15-22. D.A. Stenger, G.W. Gross, E.W. Keefer, K.M. Shaffer, J.D. Andreadis, W. Ma, J.J. Pancrazio, Detection of Physiologically Active Compound Using Cell-based Biosensors. Trends in Biotechnology 19 (2001) 304-309.
7 6 5

182

CHAPTER 10 BIOLOGICAL TESTING OF THE POLYMER MEMS CELL FORCE SENSOR

10.1 Introduction

Recent research has shown that cell forces are critical in regulating a host of cellular processes1,2,3. Cytoskeletal components are responsible for regulating cellular shape and spreading. Integrin rich focal adhesions provide a molecular linkage between the intracellular and extracellular spaces that is imperative for establishing bi-directional signal transduction pathways2,6. Adhesion forces at the cell membrane-substrate interface are also responsible for remodeling the extracellular matrix. Transient changes in the adhesion forces and the intracellular stress field are the primary mechanism of cellular motility4,5,7. Cellular adhesion forces also have implication in tissue engineering and biomaterials. The degree and strength with which cells adhere to tissue engineering scaffolds and implant materials can be the limiting factor in their functional effectiveness. By employing a polymer MEMS approach to measure cell forces, a better understanding of the fundamental role of cellular mechanics in regulating cellular processes can be obtained. In addition to normal cell function, abnormalities is the cytoskeleton have been implicated in a wide range of pathologies.8 Cytoskeletal protein abnormalities have been associated with muscular and neurodegenerative disease, 9,10 cancer,11-13 infectious 183

disease14 and pathologies of the immune systems,15 among others. A better understand of the cytoskeleton and cytoskeletal mechanics could lead to improved diagnostic and therapeutic approaches to these diseases. Cytoskeletal proteins, especially actin, also play a critical role in would healing.16,17 There are already a number of drugs and chemicals that directly target cytoskeletal components. Prime examples are chemotherapeutics that target the microtubule network to impede cell proliferation. Paclitaxel (Taxol) is a commonly used drug for treatment of ovarian, breast, and lung cancer. Paclitaxel stabilizes the microtubule network, thus inhibiting cell division.18 Another set of microtubule targeting anticancer drugs are the vinca alkaloids. While they function via a different mechanism of action than Taxol, this class of drugs also inhibits cell mitosis.19 Other chemicals, including the cytochalasins20, phalloidon20, and jasplakinolide21, target the actin network and affect polymerization and depolymerization of actin filaments. The ability to directly test how these and other drugs affect cell mechanics could improve the therapeutic effectiveness of the drugs, lead to new applications of cytoskeletal targeting agents in disease therapy, and potentially provide a new avenue to drug testing and environmental monitoring. Here we describe measurement of fibroblast cell forces using the polymer MEMS sensor described in the previous chapter. Sample preparation, experimental setup, and testing procedure will be described. WS-1 skin fibroblast cells were tested. Forces were measured with the second and third generation devices. Third generation devices were used to measure dynamic force by analyzing time lapse images. The ability to measure cellular response to chemical perturbations was demonstrated by exposing the cell to 184

cytochalasin-D and measuring the force response over time. This work demonstrates the ability to use the polymer MEMS force sensor to measure dynamic cellular forces and measure chemically induced changes in cellular mechanics. This device could find applications in various cell biology studies, chemomechanical transduction studies, and drug and chemical testing.
10.2 Materials and method

10.2.1 Sample preparation and surface modification Initial experiments with PPMA were performed without any surface modification. After changing to polystyrene, the sample surface was modified with an oxygen plasma to promote cell adhesion. The devices were expose to O2 plasma in a benchtop reactive ion etcher (Technics, MicroRIE 800). Samples were exposed for ten seconds with a power of 100 W, O2 flow rate of 30 sccm, and pressure of 187 mTorr. Sample beams were characterized by AFM before and after exposure to O2 plasma to determine if the treatment had a significant effect on the beam thickness or roughness. After O2 plasma treatment, the samples were placed in PDMS coated petri dishes. The dishes were coated with a thin layer of PDMS to fix the samples in place during testing. PDMS was used to avoid using any additional adhesives that could be toxic to the cells. PDMS was prepared by mixing a 10:1 ratio of PDMS T-2 translucent base and T-2 curing agent. The PDMS was mixed thoroughly and a small amount was placed in the dish and spread evenly. The PDMS was allowed to cure for a minimum of 48 hours. The PDMS was sterilized with 70% ethanol for at least 30 minutes before use.

185

The surface modified devices were place in the PDMS coated dishes and cell culture medium was added to the dish within 30 minutes of modification. The O2 plasma surface modification is not stable over time, so devices were modified immediately before testing. The petri dish was placed in the cell culture incubator at 37 C to promote dissolution of the PVA sacrificial layer. After completely dissolving the sacrificial layer, the cell culture was serially diluted three times. A majority of the medium was aspirated from the dish and fresh medium was added. This was done to avoid drying the device completely which would lead to collapse of the beams due to stiction. 10.2.2 Experimental setup The experimental setup for testing the sensor is shown in Figure 10.1. The overall setup is shown in Figure 10.1(a) and a close-up of the sample stage and micromanipulator is given in Figure 10.1(b). The key features of the experimental setup are labeled. An inverted phase contract microscope (Nikon, TS 100) was used for observation. The microscope was placed on a vibration isolation platform [World Precision Instruments (WPI)]. A custom built Plexiglas enclosure was fitted to the microscope. The images in Figure 10.1 are shown with the doors of the enclosure removed to allow the setup to be seen more clearly. Teflon tubing was used to connect a gas cylinder containing a blood gas mixture (5% CO2, 20% O2, 75% N2) to a bulkhead fitting on the enclosure. A heat source was placed in the enclosure to maintain the temperature at approximately 37 C. A thermometer/hygrometer probe was placed in the enclosure to monitor the temperature and humidity throughout the experiment. A three axis manual micromanipulator (WPI, KITE-R) was used to place the cell in the cell active area of the device. The micromanipulator was fixed to the vibration stage using a magnetic holding stand (WPI, 186

M10L). A 2 m inside diameter glass micropipette with a Luer fitting (WPI, TipTM) was prefilled with cell culture medium using a micropipette filling needle (WPI, MicroFilTM). The micropipette was attached to a 10 mL syringe via Tygon tubing. The micropipette was placed in the micropipette holder arm and attached to the micromanipulator. A high resolution (6.6 megapixel) digital camera (Pixelink, PL-A782) was attached to the microscope. The camera was attached to a computer with digital image acquisition software (PixeLINK Capture.OEM).

Figure 10.1 Experimental setup for cell force measurements. 187

10.2.3 Cell culture techniques WS-1 human skin fibroblasts (ATCC) were used for testing. The cells were cultured in T-75 or T-25 tissue culture flasks (Fisher Scientific) in Eagles Minimum Essential Medium supplemented with 10% fetal bovine serum (FBS, ATCC) and 1% penicillin-streptomycin (ATCC). For subculturing, the medium was aspirated from the flask. Cells were detached with 0.25% trypsin-EDTA (ATCC) at 37 C. Fresh medium was then added to the trypsin-EDTA solution and cells were centrifuged at 1200 rpm for 5 minutes. Trypsin containing medium was aspirated and cells were suspended in fresh medium. The cell suspension was then split into the appropriate number of culture flasks (cells were usually split at a 1:4 every 3-5 days). Cells were frozen and thawed as needed. For sensor testing, the cells were detached from the culture flask in the same manner as described previously. A small volume of cell suspension (~0.5 ml) was removed and added to the petri dish with the sensors. The remaining cell suspension was either split or placed back into a single flask depending on the cell density. For studies with cytochalasin-D (Cyto-D, Calbiochem), the material was dissolved in dimethylsulfoxide (DMSO). The cells were allowed to attach and spread on the sensor as normal. Cyto-D was then injected into the medium via a 20 l pipette to a final concentration of 0.1 M. 10.2.4 Cell placement For testing, it was necessary to place a single cell on the cell active area. This was done using the micropipette and three-axis micromanipulator. After adding the cell suspension, the petri dish with sensors was immediately moved to the microscope stage with the enclosure preheated to 37 C and flushed with cell culture gas. The micropipette 188

tip was place in the cell suspension and moved to a position just above the sensors. This allowed the device and the micropipette tip to be imaged simultaneously through the microscope optics. The difference in focal planes between the device and the micropipette tip was used to estimate the distance between them to avoid damaging the device with the tip or breaking the tip. After locating the appropriate cell, positive and negative pressure applied via the syringe was used to move the cell to the center of the device. The micropipette was then removed from the cell culture solution. After cell placement, the enclosure was regularly flushed with cell culture gas to maintain the proper gas concentration. 10.2.5 Image acquisition and analysis After placing the cell on the device, digital images were captured at specific intervals to monitor the deflection of the beams over time as the cell attached and spread on the sensor. Images were captured using the digital image acquisition software. Images were analyzed using ImageJ software. ImageJ is a free software download from the National Institutes of Health and allows pixel mapping among a variety of the other functions. The software can be downloaded from http://rsb.info.nih.gov/ij/. For initial experiments, images were acquired in phase contrast mode. This provided the best contrast between the device and the cell. However, this complicated edge detection due to halos at the edges in the phase contrast images. For most of the experiments with the third generation device, the cell placement was performed in phase contrast mode then the microscope was switched to brightfield mode to capture the images for analysis. Phase contract images were only taken for demonstration purposes (i.e. Figures 10.3 and 10.9). 189

An initial image was acquired prior to cell attachment to determine the initial locations of each of the beams. The images were analyzed by determining the x and y pixel values for the tip of the beams. For the second generation devices, the end of the arrow shaped pad was analyzed, and for the third generation devices the left side of the end of the pad was analyzed (see chapter 9). A fixed point on each device was analyzed to allow correction of any image shift, and a line between fixed points on the images was also analyzed to ensure that there was no angular shift in the images. The subsequent images were analyzed in a similar manner to determine the displacement of the beam relative to the initial position. After correcting for image shift, these values were used to calculate the pixel displacement magnitude and angle of displacement. The pixel displacement magnitude was then converted to distance using the calibration factor for the microscope at a specific magnification. The majority of the images were captured at 100x and the calibration factor was 3.6 pixels/m. After determining the displacement vector for each image, the force direction and magnitude were calculated according to the procedure in Chapter 9.
10.3 Results and discussion

10.3.1 Sample preparation Figure 10.2 shows the static contact angle of polystyrene films before and after plasma surface modification. These values are compared with the static contact angle of tissue culture treated polystyrene. The graph shows that the contact angle of native polystyrene (941) is much higher than that of tissue culture treated polystyrene (462). After modification, contact angle of the plasma treated polystyrene (312) 190

surface had a contact angle even lower than tissue culture treated polystyrene. This surface modification produces a surface similar to that of tissue culture treated polystyrene and should provide a much more suitable surface for cell attachment and spreading.

100 90 80 70 60 50 40 30 20 10 0 Polystyrene Tissue Culture Polystyrene Plasma Treated Polystyrene

Figure 10.2 Static contact angle for native, tissue culture treated, and O2 plasma treated polystyrene.

on the surface morphology or the height of the beam. AFM scans were performed on the same area of the same beam before and after plasma exposure. The stepheight function of the AFM image analysis software was used to measure the height of the beam before and after exposure. In both cases, the beam was measured at 2.3 m. Likely, there is a slight etch due to the oxygen plasma exposure, but given the short exposure time (10 seconds) it 191

Static Contact Angle ()

AFM was used to determine if the plasma etch process had any significant effect

was not enough to produce a significant change in the beam dimensions. The images do not show any noticeable change in the surface morphology. Overall, the plasma exposure did not have a significant detrimental effect on the beam geometry. 10.3.2 Experimental results with first and second generation devices Limitation in the experimental setup and the poor cell interactions with the first generation device prevented any meaningful measurements from being performed. However, lessons learned from the first generation device provided the necessary foundation to improve the beam design, the microscope incubation system, and the image acquisition setup. Initial experiments with the second generation device were performed with unmodified PPMA as the structural material. These experiment were not performed using the stage incubator described previously. Instead, the samples were placed in a cell culture incubator between images. This significantly complicated the image analysis given that large horizontal, vertical and angular image shifts had to be corrected. This procedure also limited the number of images that could be taken and analyzed. However, these experiments did provide a proof of concept for the device functionality, as well as providing further insight into improving the device design, experimental setup, and data acquisition system. In addition, the first meaningful force measurements were demonstrated. Figure 10.3 shows phase contrast optical micrographs of a WS-1 skin fibroblast on the PPMA sensor at times ranging from 0 minutes to 88 minutes. The probes are labeled 1-4 to aid the discussion. The cell was able to attach and spread on three of the four probes. At 49 minutes after cell seeding, the cell is still rounded, and there is little 192

observable deflection in the probes. There is evidence that the cell is beginning to attach to probe 1 and extend some filopodia in various directions as indicated in Figure 10.3(b). By 80 minutes, there is significant deflection of the beams. Probes 1 and 4 have been pulled into contact with one another. There is also significant deflection of probes 1 and 2. By 88 minutes, probe 1 and 2 are also pulled into contact. Probe 3 was never engaged during this experiment. The end of the cantilever was initially out of focus, which indicates that it is not in the same focal plane as the other cantilevers or the cell. This is likely due to damage during the fabrication, release, or cell seeding. Figure 10.4(a) shows an overlay of the images at time=0 min. and time=88 minutes to show the relative displacement of each beam more clearly. The results of the displacement analysis and force calculations are shown in tables 10.1 and 10.2. The resulting forces are on the order of hundreds of nanonewtons (nN).

193

Figure 10.3 Optical micrographs of WS-1 fibroblast on the second generation sensor at times (a) 0 min. (b) 47 min. (c) 80 min. and (d) 88 min.

Cell Force Probe 1 2 3 4 Direction () 261 212 269 Magnitude (nN) 83 70 177

Table 10.1 Cell force calculations for time=80 min. for the experiment corresponding to Figure 10.3. 194

Cell Force Probe 1 2 3 4 Direction () 317 270 270 Magnitude (nN) 159 119 286

Table 10.2 Cell force calculations for time=88 min. for the experiment corresponding to Figure 10.3.

10.3.3 Results with third generation sensors Changes were made in both the design of the sensor and the material used in third generation of the device. All devices were made from plasma modified polystyrene. In addition to improving cell interactions with the device, this material is also much better characterized with respect to its mechanical properties. This decreases the influence of error due to estimation of the elastic modulus of the materials. The area of the cell adhesion region was increased to allow the cell to spread more normally. In addition, considerable process was made in improving the microscope incubation system and image acquisition. This combination of improvements made it possible to make significant process in acquiring meaningful, time resolved cell force measurements. Figure 10.5 shows time lapse optical micrographs of a WS1 skin fibroblast on the third generation device. The image in Figure 10.5(a) shows the cell after initial seeding on the device. The rounded cell is easily visible on the device. At a time of 7.5 minutes, the cell become less visible as it begins to attach and spread. By 15 minutes after seeding, 195

bending of all four of the beams can be observed. The magnitude of the bending increases steadily until time 50 minutes, at which point the beams are in contact with one another or with the adhesion pad.

196

Figure 10.4 Optical micrographs of WS-1 fibroblast on the third generation sensor at times (a) 0 min., (b) 7.5 min., (c) 15 min., (d) 22.5 min., (e) 30 min., (f) 37.5 min., (g) 45 min., and (h) 50 min.

197

The deflection of the beams was analyzed using image analysis. The direction of the force was then calculated according to the FEA simulations. A plot of the magnitude of the force versus time is given in Figure 10.5. The plot shows a steady increase in the force on each beam over time. The plot indicates that the cell attached to the beam relatively quick, with forces applied to three of the four beams 10 minutes after seeding and forces on all four beams at 15 minutes after seeding. Forces were highest in beam 1 and reached a magnitude of 879 nN by 50 minutes. The maximum forces for the beams 2, 3, and 4 were 523 nN, 554 nN, and 761 N, respectively. The maximum force on beams 2, 3, 4 were reach at times of 55, 45, and 50 minutes, respectively. A plot of the force direction versus time is given in Figure 10.6. The plot shows that the forces are concentrated in the direction between 80-120. This indicated that the forces are directed toward the center of the device for all four probes. This further validates that the deflection of the beams is a result of contractile force exerted on the beams by the cells.

198

1000 800

Force (nN)

600 400 200 0 -200 0 10 20 30 40 50 60

Time (min.)
Probe 1 Probe 2 Probe 3 Probe 4

Figure 10.5. Force versus time plot for WS-1 fibroblasts on third generation sensor, plot corresponds to the pictures in Figure 10.4.

300

Force Direction () .

250 200 150 100 50 0 0 10 20 30 40 50 60 70

Time (min.)
Probe 1 Probe 2 Probe 3 Probe 4

Figure 10.6 Force direction plot for WS-1 fibroblast cell corresponding to the force magnitude plot in Figure 10.5. 199

Figures 10.7 and 10.8 show the force direction and magnitude versus time plots for another experiment. The force magnitude plot has several interesting features. First, the force plot shows that the cell did not attached to the probe immediately and the cell attached to the probes at different times. The cell began to exert forces on beam 1 at approximately 30 minutes. The cell attached to probe 4 at around 40 minutes, probe 3 at around 50 minutes. The maximum forces for beams 1, 3, and 4 were 1.00 N, 0.90 N, 0.16 N, respectively. The forces on probes 1 and 3 were considerably higher than those on probe 4. This is likely due primarily to the area of the cell that covered the beam. For example, if the cell were off-center or only attached to the edge of the beam, one would expect the forces on that beam to be much lower than if the cell spread over a large area of the beam surface. The cell did not attach to probe 2. Even though there are no forces exerted on probe 2, this does give a good illustration of the noise in the measurements. The positive and negative values were assigned to give a reference to the direction, with measurements between 0 and 180 (i.e. toward the center of the device) assigned positive values and forces between 180 and 360 (i.e. away from the center of the device) assigned negative values. The plot shows that the force measurements on probe 2 fluctuated between positive and negative values, further indicating that the measurements are due to noise and not to actual forces exerted on the beam. The range of measured values was 42 to +17 nN, so the range of the noise is approximately 60 nN in this case. It should be noted that the images were captured at 100x magnification. If we assume that the noise is primarily due to errors in image analysis, the noise range should decrease considerably with increased image magnification. 200

1200 1000

Force (nN)

800 600 400 200 0 -200 0.00 20.00 40.00 60.00 80.00 100.00 120.00 140.00 160.00

Time (min.)
Probe 1 Probe 2 Probe 3 Probe 4

Figure 10.7 Force magnitude versus time plot for a WS-1 skin fibroblast.

400

Force Direction () .

350 300 250 200 150 100 50 0 0.00 20.00 40.00 60.00 80.00 100.00 120.00 140.00 160.00

Time (min.)
Probe 1 Probe 2 Probe 3 Probe 4

Figure 10.8. Force direction versus time plot corresponding to the force magnitude plot of Figure 10.7.

201

The force direction plot in Figure 10.8 shows that after cell attachment, the forces are concentrated toward the center of the device in the range between 70-110. The plot also shows that the force direction is essentially random for probe 2 and for the rest of the probes prior to cell attachment. This is expected given that these measurements are due to noise in the system.

202

1400 1200 1000

Force (nN)

800 600 400 200 0 0.00

20.00

40.00

60.00

80.00

100.00 120.00 140.00 160.00 180.00 200.00 Probe 3 Probe 4

Time (min.)
(a)
Probe 1 Probe 2

350 300

Force Direction () .

250 200 150 100 50 0 0.00 20.00 40.00 60.00 80.00 100.00 120.00 140.00 160.00 180.00 200.00

Time (min.)
(b)
Probe 1 Probe 2 Probe 3 Probe 4

Figure 10.9 (a) Force and (b) direction versus time plots for a WS-1 fibroblast.

203

1200 1000 800

Force (nN)

600 400 200 0 -200 0 50 100 Probe 1 Probe 2 150 Probe 3 200 Series4 250

Time (min.) (a)

400 350

Force Direction () .

300 250 200 150 100 50 0 0 50 Probe 1 100 Probe 2 150 Probe 3 Probe 4 200 250

Time (min.) (b)

Figure 10.10 (a) Force and (b) direction versus time plots for a WS-1 fibroblast.

Figures 10.9 and 10.10 show force magnitude and direction plots for several other experiments. In each case, similar behavior is observed with respect to the force magnitude and direction plots. The point of cell attachment on each probe can be seen by 204

the increase in the magnitude of the force. After a period of random noise in the angle of the force, the direction stabilizes in the general direction toward the center of the device as expected given the contractile nature of the forces. 10.3.4 Effects of chemical exposure on cell forces One of the potential applications of the cell force sensor is in evaluating the effects of drugs, toxins, and other chemicals on cell mechanics. In order to show that the sensor is capable of measuring chemically induced changes in the cell forces, the cell was exposed to cytochalasin-D (cyto-D). Figure 10.11 shows optical micrographs of a cell that was seeded on the device and allowed to function normally on the device for approximately one hour. The images show that the cell exerted normal contractile forces with the beams being deflected in the general direction toward the center of the device. At 57 minutes after seeding, the cyto-D solution was introduced into the system. After injection of cyto-D, the images show that the beams began to move apart and away from the center of the device, indicating a significant decrease in the force applied to each beam. This is confirmed by image analysis and force calculations as shown in Figure 10.13. The force magnitude plot shows an increase in the forces on each beam over time. The maximum force on each beam was calculated to be 462 nN, 378 nN, 224 nN, and 123 nN for beams 4, 2, 3, and 1, respectively. After injecting cyto-D, the forces on each beam decrease significantly. By 150 minutes, the forces on each beam decreased to 33 nN, 35 nN, 66 nN, and 40 nN for beam 4, 2, 3, and 1, respectively.

205

Figure 10.11 Time lapse optical micrographs of cell forces before and after exposure to cytochalasin-D. The chemical was administered at 57 min. The time point for the images are (a) 0 min., (b) 30 min., (c) 50 min., (d) 70 min., (e) 90 min., (f) 110 min., and (g) 130 min. 206

C y to -D In je c tio n
500 450 400 350

Force (nN)

300 250 200 150 100 50 0 0 20 40 60 80 100 120 140 160

T im e ( m in .)
P rob e 1

P rob e 2

P rob e 3

P rob e 4

Figure 10.12 Cell force magnitude versus time for a cell exposed to Cytochalasin-D. The chemical was injected at 57 minutes after seeding as indicated on the plot.

10.3.5 Sources of error and measures to minimize error Errors in the measurement of the cell forces can be introduced in two main areas. First, errors can be caused by materials and processing related factors. These include the estimation of the elastic modulus and the measurement of the beam thickness. Despite the fact that polystyrene is a fairly well characterized material, variations in the actual elastic modulus of the material versus the value used in the simulations could still lead to some error in the force calculations. The measurement of the beam thickness is also an important factor. Each beam was measured individually and the beams should be quite uniform in thickness given the fabrication processes used, so this factor should only lead to a minimal amount of error. Finally, creep of the polystyrene during the experiment could introduce a small amount of error into the system. These factors are difficult to 207

avoid. A direct calibration of the force-deflection behavior of the beams would eliminate the need to estimate the elastic modulus; however, it is extremely difficult to find a good calibration method for beams at this size scale and force measurement range. Therefore, errors are likely to be introduced by the calibration process itself. The second area that error can be introduced is in the image analysis. Errors in the image analysis are the primary source of noise in the force magnitude plots and result in errors in calculating both the force direction and magnitude. In order for the deflection vector for the beam to be determined, a point at the end of each beam and a fixed point on the device needed to be tracked over time. Ideally, the exact same pixel point on both the end of the probe and the fixed point would be tracked consistently throughout the analysis. However, given that the images were analyzed manually, an element of human error is inevitably introduced into the system. One way to minimize this issue is to use higher magnification images for analysis. For the first and second generation devices, the images could be captured at a maximum magnification of 100x. This was a consequence of not having any stationary points in the center of the device. This required that the images have both the cell active area and the anchor region in the same image to provide a reference point to correct for image shift. The fixed pad in the third generation device not only serves as a point for cell attachment, but also provides a fixed reference point in the image analysis. This allows the images to be collected at higher magnification. As a result, each pixel represents a smaller surface area. Thus, errors in selecting the appropriate pixel lead to less overall error in the measurement. A more sophisticated way to deal with this issue would be to use an automated image analysis system. Not only would this reduce the error in the pixel mapping process, it would also significantly cut 208

down on the time required for analysis. One downside of automated image analysis is the high cost of the software.
10.4 Conclusion

The cell force sensor was tested using WS-1 skin fibroblasts as a model cells. Various improvements were made from one generation of the device to the next in order to improve the response of the beam and the behavior of the cell on the device. In addition, various improvements were made to the experimental setup and the image acquisition and analysis system. By the third generation of the device, time resolved measurements of the cell force magnitude and direction could be made. In addition, the effects of cytochalasin-D on cell force were determined. This is a significant step in the implementation of the devices as a tool for evaluating the effects of drugs, toxins, and other chemicals on cell mechanics. References C.G. Galbraith, M.P. Sheetz, Forces on adhesive contacts affect cell function. Current Opinions in Cell Biology 10 (1998) 566-571.
2 3 1

M.C. Beckerle, Cell Adhesion, Oxford University Press, Oxford (2001).

C.S. Chen, M. Mrksich, S. Huang, G.M. Whitesides, D.E. Ingber, Geometic control of cell life and death. Science 276 (1997) 1425-1428. A.R. Horwitz, J.T. Parsons, Cell migrationmovin on. Science 286 (1999) 11021103. E. Crowley, A. Horwitz, Tyrosine phosphorylation and cytoskeletal tension regulate the release of fibroblast adhesions. Journal of Cell Biology 131 (1997) 525-537. M. Chicurel, R. Singer, C. Meyer, D. Ingber, Integrin binding and mechanical tension induce movement of mRNA and ribosomes to focal adhesions. Nature 392 (1998) 730733. 209
6 5 4

M.P. Sheetz, D.P. Felsenfeld, C.G. Galbraith, Cell migration: regulation of force on extracellular-matrix-integrin complexes. Trends in Cell Biology 8 (1998) 51-54. F.C.S. Ramaekers, F.T. Bosman, The cytoskeleton and disease. Journal of Pathology 204 (2004) 351-354. D. Paulin, A. Huet, L. Khanamyrian, Z. Xue, Desminopothies in muscle disease. Journal of Pathology 204 (2004) 418-427. M. Pekny, M. Pekna, Astrocyte intermediate filaments in CNS pathologies and regeneration. Journal of Pathology 204 (2004) 428-437. H. Yamaguchi, J. Condeelis, Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochimica et Biophysica Acta 1773 (2007) 642-652. Z.Z. Wu, G. Zhang, M. Long, H.B. Wang, G.B. Song, S.X. Cai, Comparison of the viscoelecastic properties of normal hepatocytes and hepatocellular carcinaoma cells under cytoskeletal perturbation. Bioreology 37 (2000) 279-290. M.J. Rosenbluth, W.A. Lam, D.A. Fletcher, Force microscopy of nonadherent cells: a comparison of leukemia cell deformability. Biophysical Journal 90 (2006) 2994-3003. L.H. Miller, D.I. Baruch, K. Marsh, O.K. Doumbo, The pathogenic basis of malaria. Nature 415 (2002) 673-679. Y. Calle, H.C. Chou, A.J. Thrasher, Wiskott-Aldrich syndrome protein and the cytoskeletal dynamics of dendritic cells. Journal of Pathology 204 (2004) 460-469. G.D. Sempowski, M.A. Borrello, T.M. Blieden, K. Barth, R.P. Phipps, Fibroblast heterogeneity in the healing wound. Wound Repair and Regeneration 3 (1995) 120-131. V.C. Sandulache, A. Parekh, H.-S. Li-Korotky, J.E. Dohar, P.A. Hebda, Prostaglandin E2 differentially modulates human fetal and adult dermal fibroblast migration and contraction: implication for would healing. Wound Repair and Regeneration 14 (2006) 633-643. P.B. Schiff, S.B. Horwitz, Taxol stabilizes microtubules in mouse fibroblast cells. Proceedings of the National Academy of Sciences USA 77 (1980) 1561-1565. V.K. Ngan, K. Bellman, B.T. Hill, L. Wilson, M.A. Jordan, Mechanism of mitotic block and inhibition of cell proliferation by the semisyntheic vinca alkaloids Vinorelbine and its newer derivative Vinflunine. Molecular Pharmacology 60 (2001) 225-232. J.A. Cooper, Effects of cytochalasin and phalliodin on actin. The Journal of Cell Biology 105 (1987) 1473-1478. 210
20 19 18 17 16 15 14 13 12 11 10 9 8

M.R. Bubb, I. Spectro, B.B. Beyer, K.M. Fosen, Effects of jasplakinolide on the kinetics of actin polymerization. The Journal of Biological Chemistry 275 (2000) 51635170.

21

211

CHAPTER 11 MEASUREMENT OF MECHANICAL FORCES GENERATED BY PLANT PPROTEIN AGGREGATES (FORISOMES)

11.1 Introduction

Forisomes are small chemomechanically active P-protein aggregates uniquely found in the sieve cells of legumes (Fabaceae). Unlike most other motor proteins, forisomes do not rely on adenosine triphosphate (ATP) to supply the energy for actuation.1 Instead they contract or expand in respond to changes in specific ion concentrations or pH. Given their unique actuation mechanism and ability to produce large conformational changes, forisomes have been suggested for use as micro-actuators, drug delivery devices, or micro-valves.1,2 During forisome contraction, free chemical reaction energy is converted to longitudinal and radial mechanical work and volume change. Upon exposure to Ca(II), Sr(II), and Ba(II) ions or during pH shifts from 7.3 to 4.5 or 7.3 to 11.0, the protein aggregates contract by 10 to 40% of their original length and increase their diameter by as much as 150%.3 The stability of the energy conversion depends on the dissolved oxygen concentration and can be considerably improved by working under anaerobic conditions.

212

The focus of this chapter is the development of a sensor with the ability to measure the forces generated by forisomes in both the lateral and radial directions. Previous work from Knoblauch et al.1 and Schwan et al.4 show that the longitudinal contraction forces are between 50 nN and 120 nN. However, there is a need for a more refined force measurement device for measuring these forces in both the longitudinal and radial directions. Our approach to measuring forisome forces was the development of a microfabricated polymer cantilever sensor. Polymers are being employed more frequently in micro/nanoscale devices. This is often due to their unique physical, chemical, and mechanical properties. In this case, using a polymer with a relatively low elastic modulus provides a cantilever structure that gives relatively large and measurable deflections, even in response to relatively low magnitude forces.
11.2 Materials and methods

11.2.1 Design The layout of the device is shown in Figure 11.1. The structure consists of four cantilever probes, each 250 m in length and 5 m wide. Two of the probes are designed to measure the force in the longitudinal direction of the forisome, and the spacing of the probes is 30 m to match the typical geometry of the forisome in the extended confirmation. Similarly, the other two probes measure force in the radial direction and have a spacing of approximately 3 m. The device is fixed to the substrate at the edges of the structure and the cantilevers are suspended approximately 1 m above the substrate using a water soluble poly(vinyl alcohol) (PVA) sacrificial layer that is dissolved prior to testing. 213

Figure 11.1 (a) Cantilever sensor for measuring forisome forces, (b) close-up of the measurement region, and (c) cross-section of the center region of the sensor.

11.2.2 Finite element simulations The behavior of the device was simulated using finite element analysis (ANSYS). A fixed support was placed at the base of the beam, and forces ranging of 0-600 nN were applied to a 3x3 m surface at the free end of the beam. An elastic modulus of 3.2 GPa and Poissons ratio of 0.325 were used for polystyrene.5 It was assumed that the mechanical properties of the material were similar to the bulk properties as reported in the literature. Previous research on the mechanical properties of polystyrene using nanoindentation indicates that this was a reasonable assumption.6 11.2.3 Fabrication and characterization Devices were fabricated using sacrificial layer micromolding. This process is a soft lithography7 based micromolding technique that is capable of producing suspended 214

polymer structures such as cantilevers. The details of the process are described elsewhere.8 Briefly, the device geometry was defined in SU8 2005 (MicroChem.) photoresist using standard photolithography. The structure was then transferred into poly(dimethylsiloxane) (PDMS). After curing, the PDMS mold was spin coated at 3000 rpm with a 7.5% (wt/wt) solution of polystyrene dissolved in anisole. The resulting film coated both the recesses and raised features of the mold. To remove the material from the raised features of the mold, the surface was brought into contact with a glass slide heated to 200 C. This removed the polymer from the top surface of the mold, resulting in a mold that was selectively filled with polystyrene in the recessed features. The selectively coated mold was then aligned and brought into contact with the patterned sacrificial layer. Heat (125 C) and pressure (~0.35 MPa) were used to transfer the structure onto the sacrificial layer. During testing, a water-based conditioning solution was introduced into the measurement chamber. This solution dissolved the sacrificial layer and released the cantilever beams. 11.2.4 Forisome preparation The forisomes tested in this work were isolated from Vicia faba. They had original lengths between 30 and 50 m and diameters between 2 and 3 m. Forisomes were prepared by mechanically isolating phloem tissue from the stems of plants that were 4 to 6 weeks old. The rind between the first and seventh internodes of the plants was excised and put into a Ca2+-free solution containing 10 mM EDTA, 0.1 M KCl, and 10 mM Tris adjusted to pH 7.3. The separated and pre-dried phloem tissue was powdered

215

under liquid nitrogen and subsequently suspended in a Ca2+-free solution. The suspension was filtered through a nylon sieve with a mesh size of 55 m. 11.2.5 Experimental setup After isolating the forisomes, they were transfer to a storage chamber. The storage chamber was connected by a transfer channel to the measurement chamber. The flow and exchange of conditioning and rinsing solutions were adjusted by computer-controlled piston pumps in the measurement chamber. Two plunger pumps were combined to stock solutions, adjusting the concentrations of Ca2+ ions and EDTA to switch forisomes between the calcium loaded and calcium free states. The force measurement chip was fixed in the measurement chamber. The forisomes were transferred from the storage chamber and attached to the devices using a micropipette attached to a three-axis micromanipulator. With the exception of the pumps, the entire measurement set-up was mounted on an actively damped optical bench. 11.2.6 Data acquisition and analysis During Ca2+ induced actuation of the forisomes, images of the devices were captured on a CCD camera attached to the microscope. The displacement of each probe was analysed using a digital greyscale image correlation program (ARAMIS).
10.3 Results and discussion

A SEM micrograph of the device prior to removal of the sacrificial layer is shown in Figure 11.2. The overall device is shown in Figure 11.2(a). The anchor region can be seen at the outer edge of the device and the beams are suspended over the PVA sacrificial

216

layer. Figure 11.2(b) shows the central region of the device where the forisome is attached. The spacings between the beams are labeled.

Figure 11.2 Scanning electron micrograph of the forisome sensors before release of the sacrificial layer, (a) full view and (b) close-up of the center measurement region.

217

Results from FEA simulations are compared with the analytical solution for beam bending in Figure 11.3. The analytical solution assumes a straight beam 250 m long with a point load applied to the end of the beam. The solution for the area moment of inertia (I) is given in equation 1. Note that since the beam is deflecting laterally, the width of the beam is the squared term in the equation. The force-deflection plot was then determined using the relationship in equation 2:9 I= b3h 12 Fl 3 3EI (1)

d=

(2)

where b is the width of the beam and h is the height of the beam in equation 1 and F is the applied load, l is the beam length, E is the elastic modulus, d is the deflection and I is the moment area of inertia for a rectangular beam cross section in equation 2. The boundary condition for the FEA simulations assumes that the load is applied to a 3x3 m surface at the end of the beam. The differences between the analytical and numerical solution are attributed to differences in the loading conditions and the assumption of a straight beam in the analytical solution.

218

Figure 11.3 Comparison of the finite element and analytical solutions for the forcedeflection behavior of the beam.

Figure 11.4 shows optical micrographs of the forisome attached to the sensor (a) before and (b) after actuation with 10 mM Ca2+. The images show that the forisome contraction resulted in a significant bending of the cantilevers in both the longitudinal and radial directions. The net displacement of both of the beams in a given axis was then used to calculate the corresponding force according to the finite element simulation.

219

Figure 11.4 Optical micrographs of the sensor (a) before and (b) after actuation with Ca2+.

Figure 11.5 shows an SEM micrograph of the sensor with a forisome attached. The forisome was placed with the long axis between the cantilevers with the larger gap spacing. No additional adhesives were used to attach the forisome to the sensor. The natural adhesion of the forisome to polystyrene was sufficient to maintain contact between the forisome and the sensor during the measurements. The SEM image was taken after the device was used and dried. The label A in the image shows damage to the 220

device caused by the micropipette tip. The label B points out that the cantilevers in the images were attached to the substrate. This was a result of stiction of the device during the drying process. The cantilevers remained suspended while in an aqueous environment.

Figure 11.5 SEM micrograph of a forisome on the sensor after testing. The notation A on the image indicates damage to the beam caused by forisome placement and B shows that the beam collapsed during the drying process due to stiction.

The resulting forces for each of seven different forisomes are shown in Figure 11.6. The longitudinal force is plotted versus the radial force. The forces ranged from 84136 nN with an average force of 11719 nN ( standard deviation) in the longitudinal direction of the forisomes. Forces in the radial direction were 22-61 nN with an average force of 4314 nN. The variation in the force is primarily attributed to differences in the geometry and behavior of the individual forisomes. The results for the longitudinal force magnitudes are in qualitative agreement with previous estimations. This is the first time that the lateral forces generated by forisomes have been measured. While the radial forces 221

are smaller relative to the longitudinal force, these forces are sufficiently large to be functionally incorporated into a microscale actuator or valve system.

Figure 11.6 Plot of radial versus longitudinal force for seven experiments.

11.4 Conclusion

In summary, we were able perform parallel measurements of both the radial and longitudinal forces generated by forisomes in response to stimulation by Ca2+. The forces were measured using a unique polymer cantilever sensors. Forces ranged from 84-136 nN in the longitudinal direction and 22-61 nN in the radial direction. The ability to quantify the bi-directional forces generated by forisomes is an important step in applying these materials as functional components in microdevices.

222

References M. Knoblauch, G. A. Noll, T. Mller, D. Prfer, I. Scheider-Hther, A. J. E. van Bel, ATP-independent contractile proteins from plants. Nature Materials 2 (2003) 600-603. M. Knoblauch, W. S. Peters, Biomimetic actuators: where technology and cell biology meet. Cell and Molecular Life Sciences 61 (2004) 2497-2509. S. Schwan, M. Fritzsche, A. Cismak, A. Heilmann, U. Spohn, In vitro investigation of the geometric contraction behavior of chemo-mechanical P-protein aggregates (forisomes). Biophysical Chemisty 125 (2007) 444-452.
4 5 3 2 1

S. Schwan (unpublished data).

J. Brandrup, E. H. Immergut, E. A. Grulke, Polymer Handbook, Wiley, New York, NY (1999). M. Palacio, B. Bhushan, N. Ferrell. D. Hansford, Nanomechanical characterization of polymer beam structures for BioMEMS applications. Sensors and Actuators A 135 (2007) 637-650. Y. Xia, G.M. Whitesides, Soft lithography. Angewandte Chemie International Edition 37 (1998) 550-575. N. Ferrell, J. Woodard, D. Hansford, Polymer microfabrication for MEMS: sacrificial layer micromolding and patterned substrate micromolding. Biomedical Microdevices 9 (2007) 815-821.
9 8 7 6

R. C. Hibbeler, Mechanics of Materials, Prentice Hall, New York, NY, 4th Edition (1997).

223

CHAPTER 12 CONCLUSIONS AND FUTURE OUTLOOK

Polymer MEMS have potential for a wide range of applications. One of the more promising areas is the biomedical field. The biocompatibility, diverse material properties, and low processing costs can all be taken advantage of for development of polymer biomedical devices. Several areas of polymer microdevices have already found commercial success. Specifically, polymer microfluidic devices are widely used for diagnostic and analytical applications. Other areas are less mature, and require new and improved methods of efficient and cost effective fabrication of functional polymer structures and components. In addition to development of new microfabrication techniques, characterization of polymer mechanical properties at the micro and nanoscale is critical for proper implantation of polymers as functional components in MEMS. Finally, there are a wide range of specific biomedical areas where polymer MEMS can be used for practical applications. By directly interfacing with cells and biomolecules on their size scale, new areas of cell biology can be explored. Polymer microfabrication techniques have and will continue to provide a means to develop biomedical tools for cell biology. In addition, polymer microdevices provide an avenue to develop new sensor technologies with unprecedented resolution and sensitivity. MEMS devices have also led

224

to new diagnostic and therapeutic techniques. Continued research in these areas will lead to new and even more exciting developments in biomedical research and clinical practice. The work presented in this thesis focused on three primary areas of polymer MEMS research: (1) development of new polymer microfabrication techniques, (2) characterization of these techniques and of the properties of potential polymer MEMS materials, and (3) development of devices for specific biomedical applications. This led to the characterization of the double stamp micromolding technique and the development and characterization of four other novel polymer microfabrication techniques. Spin dewetting was used to fabricate microstructure from PPMA, PMMA, and polystyrene. The process is capable of fabricating structures with either the same or very different geometry than the original PDMS mold features. Proper design of the mold and selection of polymer solution and spin coating conditions allows tight control over the geometry of the resulting microstructures. Lift-off processing using micromolded PPMA sacrificial layers was used to pattern conducting sulfonated polyaniline. Sacrificial layer micromolding and pattern substrate micromolding were developed to allow fabrication of suspended, mechanically independent polymer microstructures such as cantilevers. A comprehensive characterization of the mechanical properties of potential polymer MEMS materials was conducted using nanoindentation techniques. A new method for conducting microscale beam bending tests was developed and used to measure the elastic modulus, yield strength, and breaking strength of PPMA, PPMA, PS, and PS/clay nanocomposites beams. Conventional nanoindentation techniques were used to measure hardness, elastic modulus, and scratch resistance of the materials. Polymers were also tested at body temperature and after exposure to aqueous environment to 225

determine if these factors have a significant effect on material properties. Results showed that only PPMA had a significant change in properties under these conditions. Three new polymer MEMS devices were developed and tested. The first device is a microfabricated cell isolation devices. The device consists of a microfabricated membrane bonded to a commercially available nanoporous filter. The device allows active manipulation and isolation of adherent cells, non-adherent cells, and cell clusters. In the case of adherent cells, morphology of individual cells can be controlled based on the shape and size of the individual microwells. For cell clusters, the cluster size can be controlled by changing the size of the wells. Two polymer devices were developed for measuring low magnitude biological forces. The cell force sensor was designed and simulated using finite element analysis. The device was fabricated using sacrificial layer micromolding. The cell force sensor was tested using fibroblasts cells and dynamic forces were measured. The effects of chemical disruption of the cytoskeleton on cell forces were also determined. The third device was a modified force sensor for measuring forces generated by plant protein aggregates called forisomes. The forces generated by forisomes in response to changes in calcium concentration were measured for the first time in both the longitudinal and radial directions. The field of polymer MEMS is still in its relative infancy. The fabrication processes developed through this work could find applications in the development of a wide range of polymer devices for biomedical and other applications. The cell isolation device could find applications in liver cell biology, stem cell biology, and tissue engineering just to name a few. The cell force sensor could be used as a tool for cell biologist to study cytoskeletal mechanics in normal and diseased cells. The device could 226

also be used as a drug evaluation and testing method for drugs that target cytoskeletal components. This basic method of force measurement could find broader applications in sensors for other applications.

227

APPENDIX A DESIGN AND SIMULATION DATA FOR THE CELL FORCE SENSOR

228

0.2

0.15

0.1 90 0.05 180 -0.2 -0.15 -0.1 -0.05 -0.05 270 -0.1 0 0 0.05 0.1 0.15 0.2

y-deflection (m)

-0.15

-0.2

x-deflection (m)

Figure A.1 Deflection plot at 10 increments for third generation cell force sensor.

70 60 Force Angle 50 40 30 20 10 0 0 5 10 15 20 25 30 Beam Deflection Angle

Figure A.2 Curve fits for calculating force angle from beam deflection angle. (Continued) 229

Figure A.2 Continued

90 80 70
Force Angle

60 50 40 30 20 10 0 0 10 20 30 40 50 60 70
Beam Deflection Angle

120 100 Force Angle 80 60 40 20 0 0 20 40 60 80 100 120 140 Beam Deflection Angle

(Continued)

230

Figure A.2 Continued

140 120 100 Force Angle 80 60 40 20 0 0 50 100 Beam Deflection Angle 150 200

180 160 140 Force Angle 120 100 80 60 40 20 0 155 160 165 170 175 180 185

Beam Deflection Angle

(Continued)

231

Figure A.2 Continued

300 250 200 150 100 50 0 175

Force Angle

180

185

190

195

200

205

210

Beam Deflection Angle

265 260 Force Angle 255 250 245 240 235 200

210

220

230

240

250

Beam Deflection Angle

(Continued)

232

Figure A.2 Continued

285 280 Force Angle 275 270 265 260 255 0 50 100 150 200 250 300 350 Beam Deflection Angle

315 310 Force Direction 305 300 295 290 285 280 275 300 310 320 330 340 350

Beam Deflection Angle

370 360 Force Angle 350 340 330 320 310 300 340

345

350

355

360

365

Beam Deflection Angle

233

Beam Angle () 1.62-27.62 27.62-62.04 62.04-127.45 127.45-157.03 157.03-179.02 170.02-207.80 207.80-242.04 242.04-307.45 307.45-343.92 343.92-1.62

Equation f= -0.055914b2 + 3.937841b - 6.182234 f = -0.009885b2 + 1.467356b + 27.012935 f = 0.00038811b2 + 0.23221066b + 64.09965349 f = 0.013060b2 - 3.039140b + 275.204737 f = 0.0564224b2 - 16.6828923b + 1348.4376082 f = -0.0550594b2 + 23.7377278b - 2315.1491175 f = -0.00999159b2 + 5.07882517b - 383.93813498 f = 0.0003649b2 + 0.1052494b + 213.1488368 f = 0.017022b2 - 10.275335b + 1830.226010 f = 0.062337b2 - 41.131991b + 7,082.698040 b=beam angle () f=force angle ()

R2 0.999931 1.000000 1.000000 1.000000 0.999936 0.999961 1.000000 1.000000 0.998718 0.999792

Table A.1 Summary of equations for calculating force angle from the beam deflection angle.

Beam Deflection/5nN for Force Angle 0-60

0.18000000 Beam Deflection/5nN 0.16000000 0.14000000 0.12000000 0.10000000 0.08000000 0.06000000 0.04000000 0.02000000 0.00000000 0 10 20 30 40 50 60 70 Force Angle y = -0.00001578x2 - 0.00026049x + 0.15466464 R2 = 0.99959381

Figure A.3 Curve fits for calculating force and from beam deflection angle. (Continued)

234

Figure A.3 Continued

Beam Deflection/5nN for Angle 60-90

0.09000000 0.08000000 0.07000000 0.060000000.00003404x2 - 0.00647351x + 0.34882960 y= 0.05000000 R2 = 0.99773130 0.04000000 0.03000000 0.02000000 0.01000000 0.00000000 0 20 40 60 80

100

Beam Deflection/5nN for Force Angle 90-120

0.1 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 0

y = 0.00002622x2 - 0.00397301x + 0.18712398 R2 = 0.99853986

20

40

60

80

100

120

140

Beam Deflection/5nN for Angle 120-180

0.18000000
2 0.16000000 y = -0.00001731x + 0.00630147x - 0.41971182 2 R = 0.99954678 0.14000000

0.12000000 0.10000000 0.08000000 0.06000000 0.04000000 0.02000000 0.00000000 0 50 100 150 200

(Continued) 235

Figure A.3 Continued

Beam Deflection/5nN for Force Angle 180-240

0.18000000 0.16000000 0.14000000 0.12000000 y = -0.00001638x2 + 0.00568310x - 0.33847476 0.10000000 R2 = 0.99922685 0.08000000 0.06000000 0.04000000 0.02000000 0.00000000 0 50 100 150 200 250 300

Beam Deflection/5nN for Force Angle 240-270

0.09000000 0.08000000 0.07000000 0.06000000 0.05000000 0.04000000 0.03000000 0.02000000 0.01000000 0.00000000 235 y = 0.00003409x2 - 0.01876177x + 2.62208024 R2 = 0.99762382

240

245

250

255

260

265

270

275

Beam Deflection/5nN for Force Angle 270-300

y = 0.00002621x2 - 0.01340333x + 1.75048365 0.10000000 R2 = 0.99854329 0.09000000 0.08000000 0.07000000 0.06000000 0.05000000 0.04000000 0.03000000 0.02000000 0.01000000 0.00000000 265 270 275 280 285 290 295 300 305

(Continued) 236

Figure A.3 Continued

Beam Deflection/5nN for Force Angle 300-360

0.18000000 0.16000000 0.14000000 0.12000000 0.10000000 0.08000000 0.06000000 0.04000000 0.02000000 0.00000000 290 300 310 320 330 340 350 360 370 y = -0.00001723x2 + 0.01249192x - 2.11000316 R2 = 0.99960531

Force Angle 0-60 60-90 90-120 120-180 180-240 240-270 270-300 300-360

Equation R2 DF = -0.00001578f2 - 0.00026049f + 0.15466464 DF = 0.00003404f2 - 0.00647351f + 0.34882960 DF= 0.00002622f2 - 0.00397301f + 0.18712398 DF = -0.00001731f2 + 0.00630147f - 0.41971182 DF = -0.00001638f2 + 0.00568310f - 0.33847476 DF = 0.00003409f2 - 0.01876177f + 2.62208024 DF = 0.00002621f2 - 0.01340333f + 1.75048365 DF = -0.00001723f2 + 0.01249192f - 2.11000316 f=force angle () DF=deflection/5nN force

0.999594 0.997731 0.998540 0.999547 0.999227 0.997624 0.998543 0.999605

Table A.2 Equations for calculating deflection per unit force from beam deflection angle.

237

Input

Beam Direction () 0-27.62 27.62-62.04 62.04-127.45 127.45-157.03 157.03-179.02 179.02-207.8 207.8-242.02 242.04-307.45 307.45-343.02 343.02-360 Cell Force Direction

90 Formula 0 -104.5702 0 79.0104 88.14131 88.14131 0 107.4717 0 303.981 0 -624.7421 0 -7.7813 0 225.57249 0 1043.311 0 3885.772 88.14131 2.5 0.00 0.009146096 293.78 0.042548504 0.00 0.040614885 0.00 0.001198899 0.00 0.035152638 0.00 1.238698981 0.00 0.773029067 0.00 -1.142973022 293.78 1366.703315 293.7823636 307.7689372 10426.23046 355.5920908 10.09123297 16.17015522 -10.93639111

Input

Probe Deflection (m) 0-60 60.01-90 90.01-120 120.01-180 180.01-240 240.01-270 270.01-300 300.01-360 Cell Force Magnitude

Adjustments
Young's Modulus (MPa) Beam Thickness (micron) Beam Width (micron) 3300 2.5 6 Modulus Correction 293.7823636 Thickness Correction 1.21 243.5810949 Width Correction 1.00 243.5323884

Final Results
Cell Force Direction () Cell Force Magnitude (nN) 88.14 243.53

Table A.3 Excel program for calculating the cell force direction and magnitude of experimental beam deflection data. 238

APPENDIX B POLYMER MICROFABRICATION AS A TOOL FOR PROCESSING INORGANIC MATERIALS

239

B.1 Introduction

The techniques described in this thesis have been geared toward fabrication of polymer structures for biological applications. However, several of the basic processing techniques have also been utilized effectively in processes ceramics, ceramic composites, and other inorganic materials. This appendix will describe several processes that were derived from the techniques given in the previous chapters.
B.2 Spin dewetting of PMMA etch masks for silicon wet etching

Chapter 4 describes the process of spin dewetting of polymer materials on PDMS molds. In addition to being used directly in a device application, as was the case with the cell isolation device in Chapter 6, the molded polymer structures can be used etch masks for further silicon processing. PMMA has previously been suggested as a masking material for silicon etching with potassium hydroxide (KOH) and tetramethyl ammonium hydroxide (TMAH).1-3 Spin dewetting provides a simple method for depositing micropatterned PMMA films to act as etch masks. We demonstrated the use of micromolded PMMA as a etch mask for anisotropic etching of <100> silicon in TMAH. Molding of polymer etch masks could significantly reduce both the time and cost of silicon etching as compared to typical masking techniques. The low etch rate of PMMA in TMAH and KOH makes it a strong candidate as a etch mask for silicon processing.3 However, delamination of the PMMA film during etching can be a significant problem. We used baking techniques as well as silicon surface modification with hexamethyldisilazane (HMDS) to increase the adhesion between silicon and PMMA. While PMMA masking time was significantly increased for lower temperature etching (60 C), the masking time was considerable less at higher temperatures (80 C). This 240

study suggests that PMMA patterned using this technique is adequate as an etch mask for lower temperature silicon etching. The etch process is shown in Figure B.1. The PMMA films were fabricated via spin dewetting and pattern transfer with 8% (wt/wt) PMMA solutions in anisole spin coated on the PDMS molds with 30 m square pillars with height of 9 m. The substrates were placed in TMAH with constant stirring. After etching, the mask was removed by sonication in acetone.

Figure B.1 Process for wet etching of silicon in TMAH using a micromolded PMMA etch mask.

241

Figure B.2 SEM micrographs of silicon after etching in TMAH.

Figure B.2 shows silicon after etching in 60 C TMAH for 270 minutes. PMMA showed adequate adhesion for masking of silicon in 60 C TMAH for etch times up to 480 minutes. HMDS deposition and heat treatment were essential to increase the PMMA adhesion. Samples that were not coated with HMDS or annealed showed PMMA delamination after approximately 5 minutes. Samples that were only coated with HMDS or only annealed had masking times of 120 minutes and 40 minutes, respectively. Masking times were taken as the time needed to delaminate the PMMA from the silicon substrate. While PMMA was effective for low temperature etching, as the TMAH bath temperature was increased to 80 C, delamination occurred much more quickly, usually within 30 minutes. These results indicate that PMMA molding using spin dewetting could find applications in specific silicon etch processes, but is not suitable for all applications.
B.2 Spin dewetting of inorganic materials on PDMS Molds

The spin dewetting process is not limited to organically soluble thermoplastic polymers.4 Spin dewetting was also demonstrated using aqueous solutions of NaCl, 242

sucrose, ammonium heptamolybdate tetrahydrate, and poly-l-lysine. The crystal size or particles size for NaCl and sucrose were characterized versus the solution concentration. SEM micrographs of dewetted crystals of NaCl are shown in Figure B.3. A solution of NaCl in DI water was spin coated on the mold at 3000 rpm for one minute. The mold features were 10 m diameter pillars with 10 m spacing between pillars and height of 7.5 m. Sucrose particles were formed using the same mold and spin coating process (Figure B.4). Note the difference in the morphology of the resulting structures. NaCl formed regular single cubic crystals while sucrose formed hemispherical particles. The crystal or particle size versus concentration of the NaCl and sucrose are given in Figure B.5. NaCl crystals as small as 600 nm were formed using low concentration solutions. Figures B.6 and B.7 show particles of ammonium heptamolybdate tetrahydrate and polyl-lysine made using the same process.

243

Figure B.3 SEM micrographs of NaCl crystals formed by spin dewetting at (a) 0.5% NaCl, (b) 10% NaCl, and (c) 25% NaCl.

244

Figure B.4 SEM micrographs of sucrose particles formed by spin dewetting at (a) 0.5% sucrose, (b) 10% sucrose, and (c) 25% sucrose.

245

5 4.5

Crystal/Particle Size (m) .

4 3.5 3 2.5 2 1.5 1 0.5 0 0 5 10 15 20 25 30

Concentration (%)

NaCl

Sucrose

Figure B.5 Crystal/particle size versus concentration (wt/wt) for NaCl and sucrose dewetting on 10 m diameter pillars at 3000 rpm.

Figure B.6 Ammonium heptamolybdate hydrate particles formed from 1% (wt/wt) solution.

246

Figure B.7 Poly-l-lysine particles formed from 0.1% (wt/wt) solution.

B.3 Peptide mediated deposition of silica and gold

In Chapter 5, double stamp micromolding was used to fabricate sacrificial layers that were removed after in situ polymerization of sulfonated polyaniline. The basic process of using micromolded structures as sacrificial layers can be applied to other materials. PPMA sacrificial layers were used for patterned peptide deposition. The patterned peptides were then used to template selective deposition of silica and gold from solution. This process allows deposition of materials at ambient temperature and near neutral pH on a wide range of substrates. Silica deposition has been demonstrated on silicon, polyimide, iron, and nickel. The process schematic is shown in Figure B.8. Sacrificial layers were deposited using double stamp micromolding. Poly-l-lysine, 3X Flag peptide, and bovine serum albumin (BSA) were adsorbed on the surface. The sacrificial layers were removed in organic solvent. Silica and gold were then selectively deposited out of solution on the

247

surface of the peptide. Details of the silica and gold deposition process can be found elsewhere.5,6 Figure B.9 shows an SEM micrograph of patterned silica deposition on polyimide with corresponding energy dispersive spectroscopy (EDS) of the pattern and the substrate. The ESD spectra show that the material on surface is silica as indicated by the Si peak and the increase in the magnitude of the O peak. The spectra also indicate that the deposition is selective to the regions with the patterned peptide, given that no Si peak appears in Scan Area 1. Figure B.10 shows patterned gold nanoparticle deposition using a 3X Flag peptide template.

248

Figure B.8 Schematic diagram of peptide mediated materials deposition process.

249

Figure B.9 SEM of silica deposition on a poly-l-lysine template. EDS spectra show Si and an increased amount of O in the peptide coated region, and no Si is present in the uncoated region.

Figure B.10 Gold nanoparticles deposited on a 3X Flag peptide template.

250

B.4 Sol-gel processing of silica and silica/hydroxyapatite nanoparticle composites

A simple sol-gel micromolding process was developed for fabricating microstructures from silica and silica/hydroxyapatite nanoparticle composites. This could have applications for surface modification of dental implants for improving cell interactions. The process is shown in Figure B.11. A sol (with or without HA nanoparticles) was applied to a substrate and a PDMS mold was applied to the surface and held under pressure. The sol was allowed to gel, and the mold was removed. The substrate was then sintered at 350 C for 30 minutes. Figure B.12 shows 5x5 m line and 5 m diameter pillar patterns made of silica. EDS was used to characterize the chemical makeup of the patterns with HA particles. Figures B.13 shows SEM micrographs and corresponding EDS spectra for silica/HA nanoparticles composites. The EDS spectra indicate that HA particles are present in the line features and in between the features as indicated by the presence of Ca and P peaks. The HA in the pillar features is difficult to detect using EDS. This is likely due to the thickness of the pillar features. If the HA is buried within the feature, it is difficult to detect via EDS.

251

Figure B.11 Schematic diagram of the sol-gel microfabrication process.

Figure B.12 SEM micrographs of (a) 5x5 m lines, (b) 5 m diameter pillars, and (c) cross section view of 5 m diameter pillars.

252

Figure B.13 SEM micrographs and EDS spectra for silica/HA patterns. 253

A process was also developed to give semi-selective deposition of high concentrations of HA particles within the features. In this case, the HA particles were suspended in ethanol. The solution was applied to the patterned PDMS mold and dried. A glass slide was used to spread the particle coating and move the particles into the recessed features of the mold. Tape was used to remove the particles from the surface of the mold. The sol was applied to the substrate and the particle coated mold was applied to the surface. The gelation and sintering processes were repeated as described above. SEM micrographs of line and pillar patterns are shown in Figure B.12 along with EDS spectra of the features and in between the features. The spectra show significant Ca and P peaks on the features but little or no Ca and P in between the features. While this indicates that the process is reasonably selective with regard to the location of the HA particles, it is still possible that some HA is present between the features, but not at high enough concentration to be detected by EDS. Cell culture studies with primary human bone marrow osteoblasts were performed to access the effects of the micropatterns on cell behavior and morphology. Figure B.13 shows optical micrographs of cells grown on a line pattern and flat SiO2. The images show that the lines induce significant changes in cell morphology, with cells aligning and elongating in the direction of the lines. Similar behavior is seen in the SEM images in Figure B.14. All patterns (i.e. line, pillars, and pillars with HA) induced significant changes in cell shape and orientation, with cells exhibiting a long, narrow morphology guided by the direction of the features.

254

Figure B.14 SEM micrographs and EDS spectra for silca/HA composites fabricated using semi-selective deposition. 255

Figure B.15 Optical micrographs of cells grown on (a) micropatterned lines and (b) flat SiO2.

Figure B.16 SEM micrographs of cells grown on SiO2 (a) lines, (b) pillars, and (c) SiO2/HA pillars.

256

References M. Ilie, B. Mrculescu, N. Moldovan, N. Nastase, M. Olteanu, Adhesion between PMMA mask layer and silicon wafer in KOH aqueous solution. in Materials and Device Characterization in Micromachining, edited by C. R. Friedrich and Y. Vladimirsky, (Proc. of SPIE 3512, Santa Clara, CA, 1998) pp. 422-430. D. S. Bodas, S. K. Mahapatra, and S. A. Gangal, Comparative study of spin coated and sputtered PMMA as an etch mask material for silicon micromachining. Sensors and Actuators A 120 (2005) 582-588. D. S. Bodas and S. A. Gangal, Poly(methyl methacrylate) as masking material for microelectromechanical system (MEMS) fabrication. Journal of Applied Polymer Science 102 (2006) 2094-2098. J. Guan, N. Ferrell, B. Yu, D.J. Hansford, L.J. Lee, Simultaneous fabrication of hybrid arrays of nanowires andmicro/nanoparticles by dewetting on micropillars. Soft Matter 3 (2007) 1369-1371. R.T. Butler, N.J. Ferrell, D.J. Hansford, Spatial and geometrical control of silicification using a patterned poly-L-lysine template. Applied Surface Science 252 (2006) 73377342. H. Borteh, N. Ferrell, R. Butler, S. Olesik, D. Hansford, Electroless Deposition of Gold Nanoparticles Over Silicon-based Substrates. in Materials, Processes, Integration and Reliability in Advanced Interconnects for Micro- and Nanoelectronics, edited by Q. Lin, E. T. Ryan, W. Wu, D.Y. Yoon, (Mater. Res. Soc. Symp. Proc. 990, Warrendale, PA, 2007) 0990-B08-17.
6 5 4 3 2 1

257

LIST OF REFEFNCES

1. 2. 3. 4.

W. Kern (ed.), Handbook of Semiconductor Wafer Cleaning Technology: Science, Technology, and Applications, Noyes, Park Ridge, NJ (1993). A.G. Ernslie, F.T. Bonner, L.G. Peck, Flow of a viscous liquid on a rotating disk. Journal of Applied Physics 29 (1958) 858-862. D. Meyerhofer, Characteristics of resist films produced by spinning. Journal of Applied Physics 49 (1978) 3993-3997. D.C. Duffy, J.C. McDonald, O.J.A. Schueller, G.M. Whitesides, Rapid prototyping of microfluidic systems in poly(dimethylsiloxane). Analytical Chemistry 70 (1998) 4974-4984. M. Mehregany, C.A. Zorman, SiC MEMS: opportunities and challenges for applications in harsh environments. Thin Solid Films 355-356 (1999) 518524. Y. Fu, H. Du, W. Huang, S. Zhang, M. Hu, TiNi-based thin films in MEMS applications: a review. Sensors and Actuators A 112 (2004) 395-408. D. Niarchos, Magnetic MEMS: key issues and some applications. Sensors and Actuators A 109 (2003) 166-173. M. Madou, Fundumentals of Microfabrication, CRC Press, Boca Raton, FL (1997) pp.171. H. Janseny, H. Gardeniers, M. de Boer, M. Elwenspoek, J. Fluitman, A survey on the reactive ion etching of silicon in microtechnology. Journal of Micromechanics and Microengineering 6 (1996) 14-28. Microelectromechanical Systems (MEMS) Technology: Current And Future Markets, BCC Research (2006). Global MEMS/Microsystems - Markets and Opportunities, Yole Dveloppement (2007).

5.

6. 7. 8. 9.

10. 11.

258

12. 13.

W.H. Ko, Trends and frontiers of MEMS. Sensors and Actuators A 136 (2007) 62-67. R.S. Shawgo, A.C. Richards Gradyson, Y.W. Li, M.J. Cima, BioMEMS for drug delivery. Current Opinions in Solid State Materials Science 6 (2002) 329-334. A.C. Richards Grayson, R.S. Shawgo, Y. Li, M.J. Cima, Electronic MEMS for triggered delivery. Advanced Drug Delivery Reviews 56 (2004) 173-184. S.L. Tao, T.A. Desai, Microfabricated drug delivery systems: from particles to pores. Advanced Drug Delivery Reviews 55 (2003) 315-328. R. Bashir, BioMEMS: state-of-the-art in detection, opportunities and prespects. Advanced Drug Delivery Reviews 56 (2004) 1565-1585. J. Santini, M.J. Cima, R.J. Langer, A controlled-release microchip. Nature 397 (1999) 335-338. M. Staples, K. Daniel, M.J. Cima, R. Langer, Application of micro- and nano-electromechanical devices to drug delivery. Pharmaceutical Research 23 (2006) 847-863. T.A. Desai, D. Hansford, M. Ferrari, Characterization of micromachined silicon membranes for immunoisolation and bioseparation applications. Journal of Membrane Science 159 (1999) 221-231. T.A. Desai, W.H. Chu, G. Rasi, P. Sinibaldi-Vallebona, E. Guarino, Microfabricated biocapsules provide short-term immunoisolation of insulinoma xenografts.Biomedical Microdevices 1 (1999) 131-138. J. Fritz, M.K. Baller, H.P. Land. H. Rothuizen, P. Vettiger, E. Meyer, H.-H. Gntherodt, Ch. Gerber, J.K. Gimzewski, Translating biomolecular recognition into nanomechanics. Science 288 (2000) 316-318. G. Wu, R.H. Datar, K.M. Hansen, T. Thundat, R.J. Cote, A. Majumdar, Bioassy of prostate-specific antigen (PSA) using microcantilevers. Nature Biotechnology 19 (2001) 856-860. A. Gupta, D. Akin, R. Bashir, Single virus particle mass detection using microresonators with nanoscale thickness. Applied Physics Letters 84 (2004) 1976-1978. T.H. Schulte, R.L. Bardell, B.H. Weigl, Microfluidic technologies in clinical diagnostics. Clinica Chimica Acta 321 (2002) 1-10.

14. 15. 16. 17. 18.

19.

20.

21.

22.

23.

24.

259

25. 26.

G.H.W. Sanders, A. Manz, Chip-based Microsystems for genomic and proteomic analysis. Trends in Analytical Chemistry 19 (2000) 364-378. Yu, X.; Zhang, D.; Li, T.; Hao, L.; Li, X. 3-D microarrys biochip for DNA amplification in polydimethylsiloxane (PDMS) elastomer. Sensors and Actuators A 108 (2003) 103-107. M.T. Cronin, T. Boone, A.P. Sassi, H. Tan, Q. Xue, S.J. Williams, A.J. Ricco, H.H. Hooper. Plastic microfluidic systems for high-throughput genomic analysis and drug screening. Journal of the Association for Laboratory Automation 6 (2001) 71-75. H. Becker, C. Grtner, Polymer microfabrication methods for microfluidic analytical applications. Electrophoresis 21 (2000) 12-26. H. Becker, C. Grtner, Polymer microfabrication technologies for microfluidic systems. Analytical and Bioanalytical Chemistry 390 (2008) 89111. Y. Xia, G.M. Whitesides, Soft lithography. Angewandte Chemie International Edition 37 (1998) 550-575. C.S. Effenhauser, G.J.M. Bruin, A. Paulus, M. Ehart, Analytical Chemistry 69 (1997) 3451-3457. J.S. Buch, P.-C. Wang, D.L. DeVoe, C.S. Lee, Field-effect flow control in a polydimethylsoloxane based microfluidic system. Electrophoresis 22 (2001), 3902-3907. J.C. McDonald, D.C. Duffy, J.R. Anderson, D.T. Chiu, H. Wu, O.J.A. Scheuller, G.M. Whitesides, Fabrication of microfluidic systems in poly(dimethylsiloxane). Electrophoresis 21 (2000), 27-40. S.K. Sia, G.M. Whitesides, Microfluidic devices fabricated in poly(dimethylsiloxane) for biological studies. Electrophoresis 24 (2003), 3563-3576. J.M.K. Ng, I. Gitlin, A.D. Stroock, G.M. Whitesides, Components for integrated poly(dimethylsiloxane) microfluidic systems. Electrophoresis 23 (2002), 3461-3473. X.-M. Zhoa, Y. Xia, and G.M. Whitesides, Fabrication of three-dimensional micro-structures: microtransfer molding. Advanced Materials 8 (1996) 837840. E. Kim, Y. Xia, and G.M. Whitesides, Polymer microstructures formed by moulding in capillaries. Nature 376 (1995) 581-584. 260

27.

28. 29.

30. 31. 32.

33.

34.

35.

36.

37.

38. 39.

Y. Xia, G.M. Whitesides, Soft lithography. Annual Review of Materials Science 28 (1998) 153-184. A. Kumar, G.M. Whitesides, Features of gold having micrometer to centimeter dimensions can be formed through a combination of stamping with an elastomeric stamp and an alkanethiol ink followed by chemical etching. Applied Physics Letters 61 (1993) 2002-2004. R.S. Kane, S. Takayama, E. Ostuni, D.E. Ingber, G.M. Whitesides, Patterning proteins and cells using soft lithography. Biomaterials 20 (1999) 2363-2376. J.A. Rogers, R.G. Nuzzo, Recent progress in soft lithography. Materials Today 8(2) (2005) 50-56. M. Wang, H.-G. Braun, T. Kratzmller, E. Meyer, Patterning polymers by micro-fluid-contact printing. Advanced Materials 13 (2001) 1312-1317. W.S. Beh, I.T. Kim, D. Qin, Y. Xia, G.M. Whitesides, Formation of patterned microstructures of conducting polymers by soft lithography, and applications in microelectronic device fabrication. Advanced Materials 11 (1999) 1038-1041. J.A. Rogers, Z. Boa, V.R. Raju, Nonphotolithographic fabrication of organic transistors with micron feature sizes. Applied Physics Letters 72 (1998) 2716-2718. T. Granlund, T. Nyberg, L.S. Roman, M. Svensson, O. Ingans, Patterning of polymer light-emitting diodes by soft lithography. Advanced Materials 12 (2000) 269-273. B. Xu, F. Arias, S.T. Brittain, X.-M. Zhao, B. Grzybowski, S. Torquato, G.M. Whitesides, Making negative Poissons ratio microstructures by soft lithography. Advanced Materials 11 (1999) 1186-1189. D. Gallego, N. Ferrell, D. Hanford, Fabrication of Piezoelectric Polyvinylidene Fluoride (PVDF) Microstructures by Soft Lithography for Tissue Engineering and Cell Biology Applications. in Printing Methods for Electronics, Photonics and Biomaterials, edited by G. Gigli (Mater. Res. Soc. Symp. Proc. Volume 1002E, Warrendale, PA, 2007) 1002-N04-05. G. Vozzi, C. Flaim, A. Ahluwalia, S. Bhatia, Fabrication of PLGA scaffolds using soft lithography and microsyringe deposition. Biomaterials 24 (2003), 2533-2540.

40.

41. 42. 43.

44.

45.

46.

47.

48.

261

49.

D. Gallego, N. Ferrell, Y. Sun, D.J. Hansford, Multilayer micromolding of degradable polymer tissue engineering scaffolds. Materials Science and Engineering C (in press). J. Guan, H. He, D.J. Hansford, L.J. Lee, Self-folding of three-dimensional hydrogel microstructures. The Journal of Physical Chemistry B 109 (2005) 23134-23137. G. Genolet, J. Brugger, M. Despont, U. Drechsler, P. Vettiger, N.F. de Rooij, D. Anselmetti, Soft, entirely photoplastic probes for scanning force microscopy. Review of Scientific Instruments 70 (1999) 2398-2401. X. Wang, J. Engel, C. Liu, Liquid crystal polymer (LCP) for MEMS: processes and applications. Journal of Micromechanics and Microengineering 13 (2003) 628-633. G. Witzgall, R. Vrijen, E. Yablonovitch, V. Doan, B.J. Schwartz, Single-shot two-photon exposure of commercial photoresist for the production of threedimensional structures. Optics Letters 23 (1998) 1745-1747. T. Baldacchini, C.N. LaFratta, R.A. Farrer, Acrylic-based resin with favorable properties for three-dimensional two-photon polymerization. Journal of Applied Physics 95 (2004) 6072-6076. A. Bertsch, H. Lorenz, P. Renaud, 3D microfabrication by combining microstereolithography and thick resist UV lithography. Sensors and Actuators A 73 (1999) 14-23. H. Sirringhaus, T. Kawase, R.H. Friend, T. Shimoda, M. Inbasekaran, W. Wu, E.P. Woo, High-resolution inkjet printing of all-polymer transistor circuits. Science 290 (2000) 2123-2126. X.-M. Zhao, S.P. Smith, S.J. Waldman, G.M. Whitesides, M. Prentiss, Demonstration of waveguide couplers fabricated using microtrasnfer molding. Applied Physics Letters 71 (1997) 1017-1019. C. Liu, Recent developments in polymer MEMS. Advanced Materials 19 (2007) 3783-3790. J. Engel, J. Chen, Z. Fan, C. Liu, Polymer micromachined multimodal tactile sensors. Sensors and Actuators A 117 (2005) 50-61. G. Guan, N. Ferrell, L.J. Lee, D.J. Hansford, Fabrication of polymeric microparticles for drug delivery by soft lithography. Biomaterials 27 (2006) 4034-4041.

50.

51.

52.

53.

54.

55.

56.

57.

58. 59. 60.

262

61. 62.

S.L. Tao, T.A. Desai, Microfabrication of multilayer, asymmetric, polymeric devices for drug delivery. Advanced Materials 17 (2005) 1625-1630. N. Chronis, L.P. Lee, Electrothermally activated SU-8 microgripper for single cell manipulation in solution. Journal of Microelectromechanical Systems 14 (2005) 857-863. M. Ferrari, A. Lee, L.J. Lee (eds.), Biological and Biomedical Nanotechnology, Springer, Boston, MA (2006). A.C. Richards Grayson, I.S. Choi, B.M. Tyler, P.P. Wang, H. Brem, M.J. Cima, R. Langer, Multi-pulse drug delivery from a resorbable polymeric microchip device. Nature Materials 2 (2003) 767-772. A. Chakrapani, Processing and characterization of polymer microparticles for controlled drug delivery systems. Ph.D. Thesis, Ohio State University (2007). S.N. Bhatia, C.S. Chen, Tissue engineering at the microscale. Biomedical Microdevices 2 (1999) 131-144. Y. Yang, S. Basu, D.L. Tomasko, L. James Lee, S.-T. Yang, Fabrication of well-defined PLGA scaffolds using microembossing and carbon dioxide bonding. Biomaterials 26 (2005) 2585-2594. A.W. Orr, B.P. Helmke, B.R. Blackman, M.A. Schwartz, Mechanisms of mechanotransduction. Developmental Cell 10 (2006) 11-20. M.R.K. Mofrad, R.D. Kamm (eds.), Cytoskeletal Mechanics, Cambridge University Press, Cambridge (2006). A. Nicolas, B. Geiger, S.A. Safran, Cell mechanosensitivity control the anisotropy of focal adhesions. Proceedings of the National Academy of Sciences USA 101 (2004) 12520-12525. M.P. Sheetz, D.P. Felsenfeld, C.G. Galbraith, Cell migration: regulation of force on extracellular matrix-integrin complexes. Trends in Cell Biology 8 (1998) 51-54. M.E. Chicurel, C.S. Chen, D.E. Ingber, Cellular control lies in the balance of forces. Current Opinions in Cell Biology 10 (1998) 232-239. C.G. Galbraith, M.P. Sheetz, Forces on adhesive contacts affect cell function. Current Opinions in Cell Biology 10 (1998) 566-571. N. Wang, J.T. Butler, D.E. Ingber, Mechanotransduction across the cell surface and through the cytoskeleton. Science 260 (1993) 1124-1127. 263

63. 64.

65.

66. 67.

68. 69. 70.

71.

72. 73. 74.

75. 76.

C.S. Chen, M. Mrksich, S. Huang, G.M. Whitesides, D.E. Ingber, Geometric control of cell life and death. Science 276 (1997) 1425-1428. M. Chrzanowska-Wodnicke, K. Burridge, Rho-stimulated contractility drive the formation of stress fibers and focal adhesions. Journal of Cell Biology 133 (1996) 1403-1415. D.E. Ingber, Cellular tensegrity revisited I: Cell structure and hierarchical systems biology. Journal of Cell Science 116 (2003) 1157-1173. D.E. Ingber, Tensiegrity II. How structural networks influence cellular information processing networks. Journal of Cell Science 116 (2003) 13971408. G. Bao, S. Suresh, Cell and molecular mechanics of biological materials. Nature Materials 2 (2003) 715-725. A.E. Cress, R.B. Nagle (eds.), Cell Adhesion and Cytoskeletal Molecules in Metastasis, Springer, Dordrecht, The Netherlands (2006). H. Yamaguchi, J. Condeelis, Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochimica et Beophysica Acta 1773 (2007) 642-652. F.C.S. Ramaekers, F.T. Bosman, The cytoskeleton and disease. Journal of Pathology 204 (2004) 351-354. E.B. Lane, W.H.I. McLean, Keratins and skin disorders. Journal of Pathology 204 (2004) 355-366. E. Fuchs, The cytoskeleton and disease: genetic disorder of intermediate filaments. Annual Review of Genetics 30 (1996) 197-231. K. Zatloukal, C. Stumptner, A. Fuchsbichler, P. Fickert, C. Lackner, M. Trauner, H. Denk, The keratin cytoskeleton in liver disease. Journal of Pathology 204 (2004) 367-376. D.W. Owens. E.B. Lane, Keratin mutations and intestinal pathology. Journal of Pathology204 (2004) 377-385. Y. Calle, H.-C. Chou, A.J. Thrasher, G.E. Jones. Wiskott-Aldrich syndrome protein and the cytoskeletal dynamics of dendritic cells. Journal of Pathology 204 (2004) 460-469. P.B. Schiff, S.B. Horwitz, Taxol stabilizes microtubules in mouse fibroblast cells. Proceedings of the National Academy of Sciences USA 77 (1980) 15611565. 264

77. 78.

79. 80. 81.

82. 83. 84. 85.

86. 87.

88.

89. 90.

M.V. Blagosklonny, T. Fojo, Molecular effect of Paclitaxel: myths and reality. International Journal of Cancer 83 (1999) 151-156. V.K. Ngan, K. Bellman, B.T. Hill, L. Wilson, M.A. Jordan, Mechanism of mitotic block and inhibition of cell proliferation by the semisynthetic vinca alkaloids vinorelbine and its newer derivative vinflunine. Molecular Pharmacology 60 (2001) 225-232. G.G. Borisy, E.W. Taylor, The mechanism of action of colchicines. The Journal of Cell Biology 34 (1967) 525-533. K.L.K. Duncan, M.D. Duncan, M.C. Ally, E.A. Sausville, Cucurbitacin Einduced disruption of the actin and vimentin cytoskeleton in prostate carcinoma cells. Biochemical Pharmacology 52 (1996) 1553-1560. J.A. Cooper, Effects of cytochalasin and phalloidin on actin. The Journal of Cell Biology 105 (1987) 1473-1478. M.R. Bubb, A.M.J. Senderowiczf, E.A. Sausville, K.L.K. Duncan, E.D. Kern, Jasplakinolide, a cytotoxic natural product, induces actin polymerization and competitively inhibits the binding of phalloidin to F-actin. The Journal of Biological Chemistry 269 (1994) 14869-14871. M.R. Bubb, I. Spector, B.B. Beyer, K.M. Fosen, Effects of Jasplakinolide on the kinetics of actin polymerization. The Journal of Biological Chemistry 275 (2000) 5163-5170. T. Ohashi, M. Sato, Remodeling of vascular endothelial cells exposed to fluid shear stress: experimental and numerical approach. Fluid Dynamics Research 37 (2005) 40-59. O. Traub, B.C. Berk, Laminar shear stress: mechanicsms by which endothelial cells transduce an atheroprotective force. Arteriosclerosis, Thrombosis, and Vascular Biology 18 (1998) 677-685. C.R. White, J.A. Frangos, The shear stress of it all: the cell membrane and mechanochemical tranduction. Philosophical Transactions of the Royal Society B 362 (2007) 1459-1467. K.J. Van Vliet, G. Bao, S. Suresh, The biomechanics toolbox: experimental approaches for living cells and biomolecules. Acta Materialia 51 (2003) 5881-5905. T.D. Brown, Techniques for mechanical stimulation of cell in vitro: a review. Journal of Biomechanics 33 (2000) 3-14.

91. 92.

93. 94.

95.

96.

97.

98.

99.

100.

265

101.

T.G. Kuznetsova, M.N. Starodubtseva, N.I. Yegorenkov, S.A. Chizhik, R.I. Zhdanov. Atomic force microscopy probing of cell elasticity. Micron 38 (2007) 824-833. C. Rotsch, M. Radmacher, Drug-induced changes of cytoskeletal structure and mechanics in fibroblasts: an atomic force microscopy study. Biophysical Journal 78 (2000) 520-535. R.M. Hochmuth, Micropipette aspiration of living cells. Journal of Biomechanics 33 (2000) 15-22. M. Dao, C.T. Lim, S. Suresh, Mechanics of the human red blood cell deformed by optical tweezers. Journal of the Mechanics and Physics of Solids 51 (2003) 2259-2280. A.D. Mehta, M. Rief, J.A. Spudich, D.A. Smith, R.M. Simmons, Singlemolecule biomechanics with optical methods. Science 283 (1999) 16891695. A.K. Harris, P. Wild, D. Stopak, Silicone rubber substrata: a new wrinkle in the study of cell locomotion. Science 208 (1980) 177-179. M. Dembo, Y.-L. Wang, Stresses at the cell-to-substrate interface during locomotion of fibroblasts. Biophysical Journal 76 (1999) 2307-2316. S. Munevar, Y.-L Wang, M. Dembo, Traction force microscopy of migrating normal and H-ras transformed 3T3 fibroblasts. Biophysical Journal 80 (2001), 1744-1757. R.J. Pelham, Y.-L. Wang, Cell locomotion and focal adhesions are regulated by substrate flexibility. Proceedings of the National Academy of Sciences USA 94 (1997) 13661-13665. C.-M. Lo, H.-B. Wang, M. Dembo, Y.-L. Wang, Cell movement is guided by the rigidity of the substrate. Biophysical Journal 79 (2000) 144-152. G. Lin, K.SJ. Pister, K.P. Roos, Surface micromachined polysilicon heart cell force transducer. Journal of Microelectromechanical Systems 9 (2000) 9-17. C.B. Galbraith, M.P. Sheetz, A micromachined device provides a new bend on fibroblast traction forces. Proceeding of the National Academy of Sciences USA 94 (1997) 9114-9118. J.L. Tan, J. Tien, D.M. Pirone, D.S. Gray, K. Bhadriraju, C.S. Chen, Cells lying on a bed of microneedles: An approach to isolate mechanical force. Proceeding of the National Academy of Sciences USA 100 (1993), 1484-1489. 266

102.

103. 104.

105.

106. 107. 108.

109.

110. 111. 112.

113.

114. 115. 116. 117.

R. Gnter, Dewetting of thin polymer films. Physical Review Letters 68 (1992) 75-78. R. Xie, A. Karim, J.F. Douglas, C.C. Han, R.A. Weiss, Spinodal dewetting of thin polymer films. Physical Review Letters 81 (1998) 1251-1254. A. Sehgal, V. Ferreiro, J.F. Douglas, E.J. Amis, A. Karim, Pattern-directed dewetting of ultrathin polymer films. Langmuir 18 (2002) 7041-7048. E. Meyer, H. Braun, Controlled dewetting processes on microstructured surfaces- a new procedure for thin film microstructuring. Macromolecular Materials and Engineering 276/277 (2000) 44-50. Z. Zhang, Z. Wang, R. Xing, Y. Han, How to form regular polymer microstructures by surface-patterned-directed dewetting. Surface Science 539 (2003) 129-136. Z. Zhang, Z. Wang, R. Xing, Y. Han, Patterning thin polymer films by surface-directed dewetting and pattern transfer. Polymer 44 (2003) 37373743. P. Lenz, R. Lipowsky, Morphological transitions of wetting layers on structured surfaces. Physical Review Letters 80 (1998) 1920-1923. R. Konnur, K. Kargupta, A. Sharma, Instability and morphology of thin liquid films on chemically heterogeneous substrates. Physical Review Letters 84 (2000) 931-934. K. Kargupta, A. Sharma, Dewetting of thin films on periodic physically and chemically patterned surfaces. Langmuir 18 (2002) 1893-1903. L.-R. Bao, L. Tan, X.D. Huang, Y.P. Kong, L.J. Guo, S.W. Pang, A.F. Yee, Polymer inking as a micro- and nanopatterning technique. Journal of Vacuum Science and Technology B 21 (2003) 2749-2754. J. Guan, A. Chakrapani, D.J. Hansford, Polymer microparticles fabricated by soft lithography. Chemistry of Materials 17 (2005) 6227-6229. M. Gerard, A. Chaubey, B.D. Malhotra, Application of conducting polymers to biosensors. Biosensors and Bioelectronics 17 (2002) 345-359. S. Geetha, C.R.K. Rao, M. Vijayan, D.C. Trivedi, Biosensing and drug delivery by polypyrrole. Analytica Chimica Acta 568 (2006) 119-125. L. Dai, B. Winkler, L. Dong, L. Tong, A.W.H. Mau, Conjugated polymers for light-emitting applications. Advanced Materials 13 (2001) 915-925. 267

118.

119.

120. 121.

122. 123.

124. 125. 126. 127.

128. 129.

D. Braun, Semiconducting polymer LEDs. Materials Today 5(6) (2002) 3239. C.J. Drury, C.M.J. Mutsaers, C.M. Hart, M. Matters, D.M. de Leeuw, Lowcost all-polymer integrated circuits. Applied Physics Letters 73 (1998) 108110. N. Stutzmann, R.H. Friend, H. Sirringhaus, Self-aligned, vertical-channel, polymer field-effect transistors. Science 299 (2003) 1881-1884. E.W.H. Jager, E. Smela, O. Ingans, Microfabricated conjugated polymer actuators. Science 290 (2000) 1540-1545. E. Smela, Microfabrication of PPy microactuators and other conjugated polymer devices. Journal of Micromechanics and Microengineering 9 (1999) 1-18. X.L. Wei, Y.Z. Wang, S.M. Long, C. Bobeczko, A.J. Epstein, Synthesis and physical properties of highly sulfonated polyaniline. Journal of the American Chemical Society 118 (1996) 2545-2555. J. Yue, Z.H. Wang, K.R. Cromack, A.J. Epstein A.G. MacDiarmid, Effect of sulfonic acid group on polyaniline backbone. Journal of the American Chemical Society. 113 (1991) 2665-2671. J. Stejskal, I. Sapurina, J. Proke, J. Zemek, In-situ polymerized polyaniline films. Synthetic Metals 105 (1999) 195-202. I. Sapurina, A. Riede, J. Stejskal, In-situ polymerized polyaniline films 3. Film formation. Synthetic Metals 123 (2001) 503-507. S. Holdcroft, Patterning -conjugated polymers. Advanced Materials 13 (2001) 1753-1765. D.G. Lidzey, M.A. Pate, M.S. Weaver, T.A. Fisher, D.D.C. Bradley, Photoprocessed and micropatterned conjugated polymer LEDs. Synthetic Metals 82 (1996) 141-148. F.J. Touwslager, N.P. Willard, D.M. de Leeuw, I-line lithography of poly(3,4-ethylenedioxythiophene) electrodes and application in all-polymer integrated circuits. Applied Physics Letters 81 (2002) 4556-4558. I. Mav, M. igon, A. ebenik, Sulfonated polyaniline. Synthetic Metals 101 (1999) 717-718. I. Mav, M. igon, A. ebenik, J. Vohlidal, Sulfonated polyanilines prepared by copolymerization of 3-aminobenzenesulfonic acid and aniline: the effect of 268

130. 131. 132.

133.

134.

135. 136. 137. 138.

139.

140. 141.

reaction conditions on polymer properties. Journal of Polymer Science A 38 (2000) 3390-3398. 142. 143. 144. Y. Huang, B. Rubinsky, Microfabricated electroporation chip for singe cell membrane permeabilization. Sensors and Actuators A 89 (2001) 242-249. M. Khine, A. Lau, C. Ionescu-Zanetti, J. Seo, L.P. Lee, A single cell electroporation chip. Lab on a Chip 5 (2005) 38-43. K.G. Klemic, J.F. Klemic, M.A. Reed, F.J. Sigworth, Micromolded PDMS planar electrode allows patch clamp electrical recordings from cells. Biosensors and Bioelectronics 17 (2002) 597-604. X. Li, K.G. Klemic, M.A. Reed, F.J. Sigworth, Microfluidic system for planar patch clamp electrode arrays. Nano Letters 6 (2006) 815-819. P. Wang, G. Xu, L. Qui, Y. Xu, Y. Li, R. Li, Cell-based biosensors and its application in biomedicine. Sensors and Actuators B 108 (2005) 567-584. M. Nishizawa, K. Takoh, T. Matsue, Micropatterning of HeLa cells on glass substates and evaluation of respiratory activity using microelectrodes. Langmuir 18 (2002) 3645-3649. C.S. Chen, M. Mrksich, S. Huang, G.M. Whitesides, D.E. Ingber, Micropatterned surface for control of cell shape, position, and function. Biotechnology Progress 14 (1998) 356-363. Y. Ito, Surface micropatterning to regulate cell function. Biomaterials 20 (1999) 2333-2342. A. Brock, D. Chang, C. Ho, P. LeDuc, X. Jiang, G.M. Whitesides, D.E. Ingber, Geometric determinant of directional cell motility revealed using microcontact printing. Langmuir 19 (2003) 1611-1617. M. Khalil, A. Shariat-Panahi, R. Tootle, T. Ryder, P. McCloskey, E. Roberts, H. Hodgson, C. Selden, Human hepatocyte cell lines proliferating as cohesive spheroid colonies in alginate markedly upregulate both synthetic and detoxificatory liver function. Journal of Hepatology 297 (2001) 68-77. R. Clicklis, J.C. Merchuk, S. Cohen, Modeling mass trasfer in hepatocyte spheroids via cell viability, spheroid size and hepatocellular funtion. Biotechnology and Bioengineering 86 (2004) 672-680. S.M. Dang, M. Kyba, R. Perlingeiro, G.Q. Daley, P.W. Zandstra, Efficiency of embryoid body formation and hematopoietic development from embryonic stem cells in different culture systems. Biotechnology and Bioengineering 78 (2002) 442-453. 269

145. 146. 147.

148.

149. 150.

151.

152.

153.

154.

M. Mrksich, L.E. Dike, J. Tien, D.E. Ingber, G.M. Whitesides, Using microcontact printing to pattern the attachment of mammalian cells to selfassembled monolayers of alkanethiolates on transparent films of gold and silver. Experimental Cell Research 235 (1997) 305-313. H. Kaji, K. Takoh, M. Nishizawa, T. Matsue, Intracellular Ca2+ imaging for micropatterned cardiac myocytes. Biotechnology and Bioengineering 81 (2003) 748-751. K.Y. Suh, J. Seong, A. Khademhosseini, P.E. Laibinis, R. Langer, A simple soft lithographic route to fabrication of poly(ethylene glycol) microstructures for protein and cell patterning. Biomaterials 25 (2004) 557-563. M.R. Dusseiller, D. Schlaepfer, M. Koch, R. Kroschewski, M. Textor, An inverted microcontact printing method on topographically structured polystyrene chips for arrayed micro-3-D culturing of single cells. Biomaterials 26 (2005) 5917-5925. E.W.H. Jager, C. Immerstrand, K.H. Peterson, K. Magnusson, I. Lundstrm, O. Ingans, The cell clinic: closable microvials for single cell studies. Biomedical Microdevices 4 (2002) 177-187. A. Folch, B. Jo, O. Hurtado, D.J. Beebe, M. Toner, Microfabricated elastomeric stencils for micropatterning cell cultures. Journal of Biomedical Materials Research 52 (2000) 346-353. M.J. Rosenbluth, W.A. Lam, D.A. Fletcher, Force microscopy of nonadherent cells: a comparison of leukemia cell deformability. Biophysical Journal 90 (2006) 2994-3003. J. Fukuda, K. Nakazawa, Orderly arrangement of hepatocytes spheroids on a microfabricated chip. Tissue Engineering 11 (2005) 1254-1262. N. Ferrell, D. Hansford, Fabrication of micro- and nanoscale polymer structures by soft lithography and spin dewetting. Macromolecular Rapid Communications 28 (2007) 964-967. F. Braet, R. De Zanger, E. Wisse. Drying cells for SEM, AFM and TEM by hexamethyldisilazane: a study on hepatic endothelial cells. Journal of Microscopy 186 (1997) 84-87. B.D. Ratner, A.S. Hoffman, F.J. Schoen, and J.E. Lemons, Biomaterials Science An Intoduction to Materials in Medicine, Elsevier Academic Press, San Diego, CA (2004) p. 67.

155.

156.

157.

158.

159.

160.

161. 162.

163.

164.

270

165.

H. Becker and U. Heim, Hot embossing as a method for the fabrication of polymer high aspect ratio structures. Sensors and Actuators A 83 (2000) 130135. L.J. Lee, M.J. Madou, K.W. Koelling, S. Daunert, S. Lai, and C.G. Koh, Design and fabrication of CD-like microfluidic platforms for diagnostics: polymer-based microfabrication. Biomedical Microdevices 3 (2001) 339-351. B. Ziaie, A. Baldi, M. Lei, Y. Gu, and R.A. Siegel, Hard and soft micromachining for BioMEMS: a review of techniques. Advanced Drug Delivery Reviews 56 (2004) 145-172. G. Wei, B. Bhushan, N. Ferrell, D. Hansford, Microfabrication and nanomechanical characterization of polymer microelectromechanical systems for biological applications. Journal of Vacuum Science and Technology A 23 (2005) 811-819. K.E. Peterson, Silicon as a mechanical material. Proceedings of the IEEE 70 (1982) 420-457. J.A. Forrest, D. Dalnoki-Veress, The glass transition in thin polymer films. Advances in Colloid and Interface Science 94 (2001) 167-196. J.L. Keddie, R.A.L. Jones, R.A. Cory, Size-dependent depression of the glass transition temperature in polymer films. Europhysics Letters 27 (1994) 5964. B. Bhushan, A.V. Kulkarni, W. Bonin, J. Wyrobek, Nano/picoindentation measurement using a capacitance transducer system in atomic force microscopy. Philosophical Magazine A 74 (1996) 1117-1128. S.J. Bull, Nanoindentation of coating. Journal of Physics D 38 (2005) R393-R413. A.C. Fischer-Cripps, Critical review of analysis and interpretation of nanoindentation test data. Surface and Coatings Technology 200 (2006) 4153-4165. X. Li, B. Bhushan, A review of nanoindentation continuous stiffness measurement technique and it applications. Materials Characterization 48 (2002) 11-36. I.H. Loh, M.S. Sheu, A.B. Fishcer, Biocompatible polymer surfaces, in Desk Reference of Functional Polymers: Synthesis and Applications (edited by R. Arshady) American Chemical Society, Washington, D.C. (1997).

166.

167.

168.

169. 170. 171.

172.

173. 174.

175.

176.

271

177.

D.B. Holt, P.R. Gauger, A.W. Kusterbeck, F.S. Ligler, Fabrication of a capillary immunosensor in polymethyl methacrylate. Biosensors and Bioelectronics 17 (2002) 95-103. A. van der Berg (ed.), Lab-on-a-chip Chemistry in Miniaturized Synthesis and Analysis Systems, Elsevier, Amsterdam (2003). B. Ellis, Polymers: A Properties Database, Available on compact disk, CRC, Boca Raton, FL (2000). J. Brandrup, E.H. Immergut, E.A. Grulke (eds.), Polymer Handbook, Wiley, New York, NY (1999). J.E. Mark, Polymer Data Handbook, Oxford University Press, Oxford (1999). T.G. van Kooten, H.T. Spijker, H.H. Busscher, Plasma-treated polystyrene surfaces: model surface for studying cell-biomaterial interactions. Biomaterials 25 (2004) 1735-1747. M. Alexandre, P. Dubois, Polymer-layered silicate nanocomposites: preparation, properties and uses of a new class of materials. Materials Science and Engineering 28 (2000) 1-63. S.S. Ray, M Okamoto, Polymer/layered silicate nanocomposites: a review from preparation to processing. Progress in Polymer Science 28 (2003) 1539-1641. J. Njuguna, K. Pielichowski, Polymer nanocomposites for aerospace applications: properties. Advanced Engineering Materials 5 (2003) 769-778. F. Gao, Clay/polymer composites: the story. Materials Today 7 (2004) 5055. G. Beyer, Nanocomposites: a new class of flame retardants for polymers. Plastic Additives and Composites 4 (2002) 22-28. M. Kurian, M.E. Galvin, P.E.. Trapa, D.R. Sadoway, A.M. Mayes, Single ion conducting polymer-silicate nanocomposite electrolytes for lithium battery applications. Electrochimica Acta 50 (2005) 2125-2134. R.Y. Lockhead, C.R. Haynes. S.R. Jones, V. Smith, the high throughput investigation of polyphenolic coupler in biodegradable packaging materials. Applied Surface Science 252 (2006) 2535-2548. H. Liu, B. Bhushan, Investigation of nanotribological properties of selfassembled monolayerss with alkyl and biphenyl spacer chains. Ultramicroscopy 91 (2002) 185-202. 272

178. 179. 180. 181. 182.

183.

184.

185. 186. 187. 188.

189.

190.

191.

W.C. Oliver, G.M. Pharr, An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. Journal of Materials Research 7 (1992) 1564-1583. B. Bhushan, X. Li, Nanomechanical characterization of solid surfaces and thin films. International Materials Review 48 (2003) 125-164. H. Liu, B. Bhushan, Investigation of nanotribological properties of selfassembled monolayers with alkyl and biphenyl spacer chains. Ultramicroscopy 91 (2002) 185-202. W.C. Young, R.G. Budynas, Roarks Formulas for Stress and Strain, Wiley, New York, NY (2002). A. Yamamoto, S. Mishim, N. Maryuama, M. Sumita, A new technique for direct measurement of the shear force necessary to detach a cell from a material. Biomaterial 19 (1998) 871-879. K.A. Athanasiou, B.S. Thoma, D.R. Lanctot, D. Shin, C.M. Agrawal, R.G. LeBaron, Development of the cytodetachment technique to quantify mechanical adhesiveness of a single cell. Biomaterial 20 (1999) 2405-2415. D.A. Stenger, G.W. Gross, E.W. Keefer, K.M. Shaffer, J.D. Andreadis, W. Ma, J.J. Pancrazio, Detection of Physiologically Active Compound Using Cell-based Biosensors. Trends in Biotechnology 19 (2001) 304-309. M.C. Beckerle, Cell Adhesion, Oxford University Press, Oxford (2001). C.S. Chen, M. Mrksich, S. Huang, G.M. Whitesides, D.E. Ingber, Geometic control of cell life and death. Science 276 (1997) 1425-1428. A.R. Horwitz, J.T. Parsons, Cell migrationmovin on. Science 286 (1999) 1102-1103. E. Crowley, A. Horwitz, Tyrosine phosphorylation and cytoskeletal tension regulate the release of fibroblast adhesions. Journal of Cell Biology 131 (1997) 525-537. M. Chicurel, R. Singer, C. Meyer, D. Ingber, Integrin binding and mechanical tension induce movement of mRNA and ribosomes to focal adhesions. Nature 392 (1998) 730-733. D. Paulin, A. Huet, L. Khanamyrian, Z. Xue, Desminopothies in muscle disease. Journal of Pathology 204 (2004) 418-427. M. Pekny, M. Pekna, Astrocyte intermediate filaments in CNS pathologies and regeneration. Journal of Pathology 204 (2004) 428-437. 273

192. 193.

194. 195.

196.

197.

198. 199. 200. 201.

202.

203. 204.

205.

H. Yamaguchi, J. Condeelis, Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochimica et Biophysica Acta 1773 (2007) 642652. Z.Z. Wu, G. Zhang, M. Long, H.B. Wang, G.B. Song, S.X. Cai, Comparison of the viscoelecastic properties of normal hepatocytes and hepatocellular carcinaoma cells under cytoskeletal perturbation. Bioreology 37 (2000) 279290. M.J. Rosenbluth, W.A. Lam, D.A. Fletcher, Force microscopy of nonadherent cells: a comparison of leukemia cell deformability. Biophysical Journal 90 (2006) 2994-3003. L.H. Miller, D.I. Baruch, K. Marsh, O.K. Doumbo, The pathogenic basis of malaria. Nature 415 (2002) 673-679. G.D. Sempowski, M.A. Borrello, T.M. Blieden, K. Barth, R.P. Phipps, Fibroblast heterogeneity in the healing wound. Wound Repair and Regeneration 3 (1995) 120-131. V.C. Sandulache, A. Parekh, H.-S. Li-Korotky, J.E. Dohar, P.A. Hebda, Prostaglandin E2 differentially modulates human fetal and adult dermal fibroblast migration and contraction: implication for would healing. Wound Repair and Regeneration 14 (2006) 633-643. J.A. Cooper, Effects of cytochalasin and phalliodin on actin. The Journal of Cell Biology 105 (1987) 1473-1478. M.R. Bubb, I. Spectro, B.B. Beyer, K.M. Fosen, Effects of jasplakinolide on the kinetics of actin polymerization. The Journal of Biological Chemistry 275 (2000) 5163-5170. M. Knoblauch, G. A. Noll, T. Mller, D. Prfer, I. Scheider-Hther, A. J. E. van Bel, ATP-independent contractile proteins from plants. Nature Materials 2 (2003) 600-603. M. Knoblauch, W. S. Peters, Biomimetic actuators: where technology and cell biology meet. Cell and Molecular Life Sciences 61 (2004) 2497-2509. S. Schwan, M. Fritzsche, A. Cismak, A. Heilmann, U. Spohn, In vitro investigation of the geometric contraction behavior of chemo-mechanical Pprotein aggregates (forisomes). Biophysical Chemisty 125 (2007) 444-452. S. Schwann (unpublished data).

206.

207.

208. 209.

210.

211. 212.

213.

214. 215.

216.

274

217.

M. Palacio, B. Bhushan, N. Ferrell. D. Hansford, Nanomechanical characterization of polymer beam structures for BioMEMS applications. Sensors and Actuators A 135 (2007) 637-650. R. C. Hibbeler, Mechanics of Materials, Prentice Hall, New York, NY, 4th Edition (1997). N. Ferrell, J. Woodard, D. Hansford, Fabrication of polymer microstructures for MEMS: sacrificial layer micromolding and patterned substrate micromolding. Biomedical Microdevices 9 (2007) 815-821. M. Ilie, B. Mrculescu, N. Moldovan, N. Nastase, M. Olteanu, Adhesion between PMMA mask layer and silicon wafer in KOH aqueous solution. in Materials andDevice Characterization in Micromachining, edited by C. R. Friedrich and Y.Vladimirsky, (Proc. of SPIE 3512, Santa Clara, CA, 1998) pp. 422-430. D. S. Bodas, S. K. Mahapatra, and S. A. Gangal, Comparative study of spin coated and sputtered PMMA as an etch mask material for silicon micromachining. Sensors and Actuators A 120 (2005) 582-588. D. S. Bodas and S. A. Gangal, Poly(methyl methacrylate) as masking material for microelectromechanical system (MEMS) fabrication. Journal of Applied Polymer Science 102 (2006) 2094-2098. J. Guan, N. Ferrell, B. Yu, D.J. Hansford, L.J. Lee, Simultaneous fabrication of hybrid arrays of nanowires andmicro/nanoparticles by dewetting on micropillars. Soft Matter 3 (2007) 1369-1371. R.T. Butler, N.J. Ferrell, D.J. Hansford, Spatial and geometrical control of silicification using a patterned poly-L-lysine template. Applied Surface Science 252 (2006) 73377342. H. Borteh, N. Ferrell, R. Butler, S. Olesik, D. Hansford, Electroless Deposition of Gold Nanoparticles Over Silicon-based Substrates. in Materials, Processes, Integration and Reliability in Advanced Interconnects for Micro- and Nanoelectronics, edited by Q. Lin, E. T. Ryan, W. Wu, D.Y. Yoon, (Mater. Res. Soc. Symp. Proc. 990, Warrendale, PA, 2007) 0990-B0817.

218. 219.

220.

221.

222.

223.

224.

225.

275