xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001919442
007 cr mnu|||uuuuu
008 071217s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001814
Mier, Alexandro Castellanos.
Poly(N-Isopropylacrylamide) based BioMEMS/NEMS for cell manipulation
h [electronic resource] /
by Alexandro Castellanos Mier.
[Tampa, Fla] :
b University of South Florida,
ABSTRACT: In recent years, BioMEMS/NEMS have been primary elements associated with the research and development efforts in the bioengineering area. International and federal funding has effected an enormous increase in the development of state-of-the-art bioengineering and biomedical technologies. Most of the BioMEMS/NEMS related applications are associated with diagnostics, sensing and detection. Procedures for separation and manipulation of biological components play a paramount role in the function of these bioengineering mechanisms. This research was concerned with the development of a novel BioMEMS device for cell manipulation. The functioning of the device is based on the use of thermally responsive polymer networks, which differs dramatically from existing approaches. This approach is cost effective, requires low power and uses a minimal amount of on-device area, which makes it suitable for personal medical diagnostics and battle field scenarios.^ The device integrates the technologies associated with reversibly binding surfaces and dielectrophoresis, (DEP). The DEP field drives a sample into contact with a binding surface. This surface can be controlled to provide different levels of target selectivity. This system provides a separation strategy that does not suffer from fouling issues. The binding surfaces are fabricated from LCST polymers. The LCST polymers experience hydration-dehydration changes in response to temperature fluctuations. Therefore, separation efficiency can be "dialed in" as a function of temperature to prompt the selection of targets. Furthermore, size-exclusion "trenches" were patterned into the binding surfaces. The trenches permit the passage of the small objects in order to provide size-exclusion separations.^ ^In order to expand the discrimination size range from the micron to the submicron scale, two techniques for submicron patterning of cross-linked reversibly binding surfaces were investigated. The patterning techniques associated with electron-beam lithography and the combination of softlithography and a focused ion beam patterning were found to generate well-defined patterns that retained their thermo-responsiveness. The combination of DEP and reversibly binding surfaces for bio-particle manipulation is a significant contribution to microfluidic based separations in BioMEMS/NEMS. The developments associated with this research provide a novel technology platform that facilitates separations, which would be difficult to achieve by any other existing methods.
Dissertation (Ph.D.)--University of South Florida, 2006.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 89 pages.
Adviser: Wilfrido A. Moreno, Ph.D.
x Electrical Engineering
t USF Electronic Theses and Dissertations.
Poly(N-Isopropylacrylamide) based BioMEMS/NEMS for Cell Manipulation by Alexandro Castellanos Mier A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Electrical Engineering College of Engineering University of South Florida Co-Major Professor: Wilfrido A. Moreno, Ph.D. Co-Major Professor: Ryan Toomey, Ph.D. Stephen E. Saddow, Ph.D. Andrew M. Hoff, Ph.D. James T. Leffew, Ph.D. Peter G. Stroot, Ph.D. Andrs Cardenas-Valencia, Ph.D. Date of Approval: October 17, 2006 Keywords: Dielectrophoresis, Hydrogels, Microfluidics, PDMS, Softlithography Copyright 2006, Alexandro Castellanos Mier
DEDICATION To Viviana, Ana Alexandra, Aitor, Leonardo, Maria Dolores and Family
ACKNOWLEDGMENTS I would like to acknowledge a group of exceptional people who gave me their friendship and advice, which made this disser tation possible. I would like to thank my mentors, Dr. Wilfrido Moreno and Dr. Ryan T oomey for being the best professors and friends. I would also like to than k Dr. Andrew Hoff for allowing me to be his IC processing class TA for 3 years. I would also like to thank Dr. James Leffew for all his help during my four years in the Ph.D. program at USF. I would also like to thank my committee members, Dr. Stephen Saddow, Dr. Peter Stroot, Dr. Andres Cardenas as well as Dr. Carlos Smith fo r always giving me a helping hand when I needed it. There is a list of people who he lped me through all my research: The USFs NNRC staff, Samuel J. DuPont, Matthew R. Cutter, Stefan Cular, Alberto Rodriguez, Sergio Melais, A ugust J. Heim II, Dr. Venkat Bhethanabotla, Dr. Bert Lgel, Dr W. Garrett Matthews, C. Mike Waits, Leonardo Castellanos, The EE department Staff Finally, I would like to esp ecially thank the people from Universidad Veracruzana Dr. Victor Arredondo, Dr. Raul Arias, Emilio Zilli, Enrique Levet, Juan A. Fragoso, Jose S. Bobadilla and all of my friends who gave the most valuable thing in life; their friendship. Thank you all.
i TABLE OF CONTENTS LIST OF TABLES iii LIST OF FIGURES iv ABSTRACT vii CHAPTER 1 INTRODUCTION 1 CHAPTER 2 STIMULI RESPONSIVE POLYMER NETWORKS 9 2.1. Photopolymerization of NIPAAm 12 CHAPTER 3 SYSTEM FABRICATION 16 3.1. Poly(NIPAAm) Processing 18 3.1.1. Soft-lithography 18 184.108.40.206. Standard Photoresist Patterning 19 220.127.116.11. SU-8 Photoresist Patterning 20 3.1.2. E-Beam Lithography 25 18.104.22.168. Resist Preparation 26 22.214.171.124. E-Beam Exposure 27 126.96.36.199. Resist Developer 27 3.1.3. FIB Patterning 30 188.8.131.52. Sample Preparation 30 3.2. Dielectrophoresis Effect 31 3.2.1. Fabrication 38 3.3. Thermal Actuation 39 3.3.1. Heat Transfer 42 3.4. Microfluidics Integration 43 3.4.1. PDMS Microfluidic Ch annel Fabrication Process 44 184.108.40.206. Silicon-PDMS Molding 47 3.4.2. SU-8 Photoresist Microfluid ic Channel Fabrication Process 48
ii CHAPTER 4 EXPERIMENTS AND RESULTS 51 4.1. Poly(NIPAAm) Patterning 51 4.2. High Aspect Ratio Trenches Thermal Responsiveness 53 4.3. Particle Entrapment 54 4.4. Cell Attachment 58 4.5. Flow Velocity 60 4.6. Microfluidic Channel Testing 62 4.7. Device Testing 65 CHAPTER 5 CONCLUSIONS 75 REFERENCES 80 ABOUT THE AUTHOR End Page
iii LIST OF TABLES Table 3.1: DRIE Process Parameters 45 Table 4.1: Data from the Dip Experiment 60
iv LIST OF FIGURES Figure 1.1: Separation Mechanism Using S timuli Responsive Polymer Networks 7 Figure 2.1: Phase Behavior for Swelli ng in Surface Attached Networks 11 Figure 2.2: SU-8 Photoresist Patterns on Substrate 14 Figure 2.3: Detailed SU-8 Photoresist Patterns 14 Figure 2.4: PDMS Molds 15 Figure 2.5: Transferred Poly(NIPAA m) Structures on Substrate 15 Figure 3.1: Integrated BioMEMS Device 17 Figure 3.2: Bright Field Images of the Photoresist Structures 19 Figure 3.3: SU-8 Photoresist Processing for Poly(NIPAAm) Patterning 23 Figure 3.4: SEM Image of the SU -8 Photoresist Structures 24 Figure 3.5: A DAPI Image of the Poly(NIPAAm) Trenches 25 Figure 3.6: PMMA and Poly(NIPAAm) Film Thickness vs. Spin Speed 28 Figure 3.7: Two Dimensional AFM Image of the Poly(NIPAAm) Patterns 28 Figure 3.8: Three Dimensional AFM Image and Section Analysis 29 Figure 3.9: SEM Image of FIB Pattern ed Cross-Linked Poly(NIPAAm) 31 Figure 3.10: Dielectric Particle in a Nonuniform Electric Field 34 Figure 3.11: Simulation of the E ffect of the Electric Field, E, on the Microelectrode Array 37 Figure 3.12: Au Microelectrode Inter-Digitated Array 39
v Figure 3.13: Bright Field Image of the Thermal Actuators 40 Figure 3.14: Nichrome Resistor Structure 42 Figure 3.15: SEM Image of the DRIE Patterned Micro-Channels 46 Figure 3.16: SEM Image of the DRIE Anisotropic Etching 46 Figure 3.17: SEM Image of the DRIE Bosch Process Side Wall Profile 47 Figure 3.18: PDMS Microf luidic Channels 47 Figure 3.19: Microfluidic Channel Fabrication 49 Figure 3.20: Microfluidics Integration 50 Figure 4.1: SEM Images of Soft-Lithography and FIB Patterning 52 Figure 4.2: AFM Measured Section Analysis 52 Figure 4.3: (Top) AFM Image for T = 45 o C; (Bottom) AFM Image for T = 25.6 o C 53 Figure 4.4: Bright Field Images of th e Poly(NIPAAm) Swelling Behavior 54 Figure 4.5: Bright Field Image of the Tren ch Structures Before the Experiment 55 Figure 4.6: Particles Moving Inside the Trenches 55 Figure 4.7: The 20 m Beads Remain on Top of the Poly(NIPAAm) Structures 56 Figure 4.8: Bright Field Image of the Capture Mechanism 56 Figure 4.9: Bright Field Image of the Failure to Release Effect 57 Figure 4.10: E. coli w/GFP on the Membrane Area of a Silicon Substrate 59 Figure 4.11: % E. coli Cells Remaining for Diffe rent Reynolds Numbers 61 Figure 4.12: Plot of Flow Velocity versus % Cells Remaining on Samples 61 Figure 4.13: FITC Image of the E. coli Flowing Inside the PDMS Channels 62 Figure 4.14: Experimental Setup: SU-8 Phot oresist Microfluidic Channel Testing 63
vi Figure 4.15: SU-8 Photoresist Microfluidic Channel 64 Figure 4.16: Experimental Setup for the Device Test 65 Figure 4.17: Inverted Fluorescent Microscope and Syringe Pump 65 Figure 4.18: DUT 66 Figure 4.19: Top Section of th e Microfluidic Channel 67 Figure 4.20: Bottom Section of the Microfluidic Channel 67 Figure 4.21: BioMEMS Device Inte gration Device Top View 68 Figure 4.22: Sequence of Images of the Capture Mechanism 69 Figure 4.23: The Effect of DEP on E. coli 70 Figure 4.24: DEP with 10 l per min, 50 KHz, 8 V pp 71 Figure 4.25: DEP with 10 l per min, 50 KHz, 2 V pp 71 Figure 4.26: Particles after Cold Water Injection 72 Figure 4.27: Particles after Hot Water Injection 73 Figure 4.28: DEP with 20 l per min, 15 KHz, 10 V pp 74
vii POLY(N-ISOPROPYLACRYLAMIDE) BASED BIOMEMS/NEMS FOR CELL MANIPULATION Alexandro Castellanos Mier ABSTRACT In recent years, BioMEMS/NEMS have been primary elements associated with the research and development efforts in th e bioengineering area. International and federal funding has effected an enormous increa se in the development of state-of-the-art bioengineering and biomedical technologies Most of the BioMEMS/NEMS related applications are associated with diagnostic s, sensing and detection. Procedures for separation and manipulation of biological components play a paramount role in the function of these bioengineering mechanisms. This research was concerned with th e development of a novel BioMEMS device for cell manipulation. The functioning of the device is based on the use of thermally responsive polymer networks, which differs dram atically from existing approaches. This approach is cost effective, requires low pow er and uses a minimal amount of on-device area, which makes it suitable for personal medi cal diagnostics and batt le field scenarios. The device integrates the technologies associated with reversibly binding surfaces and dielectrophoresis, (DEP). The DEP field dr ives a sample into contact with a binding surface. This surface can be controlled to prov ide different levels of target selectivity. This system provides a separation strategy th at does not suffer from fouling issues.
viii The binding surfaces are fabricated from LCST polymers. The LCST polymers experience hydration-dehydration changes in response to temperature fluctuations. Therefore, separation efficiency can be diale d in as a function of temperature to prompt the selection of targets. Fu rthermore, size-exclusion trenc hes were patt erned into the binding surfaces. The trenches permit the passage of the small objects in order to provide size-exclusion separations. In order to expand the discrimination size range from the micron to the submicron scale, two technique s for submicron patterning of cross-linked reversibly binding surfaces were investigated. The patterning techniqu es associated with electron-beam lithography and the combina tion of softlithography and a focused ion beam patterning were found to generate well-def ined patterns that re tained their thermoresponsiveness. The combination of DEP and reversib ly binding surfaces for bio-particle manipulation is a significant contribution to microfluidic based separations in BioMEMS/NEMS. The developments associat ed with this research provide a novel technology platform that facil itates separations, which would be difficult to achieve by any other existing methods.
1 CHAPTER 1 INTRODUCTION Bioengineering applications of mi cro and nano technology, (BioMEMS/NEMS), have become progressively more estab lished and have found widespread use in applications that require their interaction with biological system s. Biotechnology has advanced to the level of more than a one hundred billion dollar rese arch and development effort in the last 50 years, . For exam ple, in 2005 the National Institutes of Health, (NIH), increased the funding for the Nati onal Institute of Biomedical Imaging and Bioengineering to $297.6 million, which was targeted toward development of state-ofthe-art instrumentation for bioengineering a nd biomedical technologies. Some of the BioMEMS research areas for diagnostics, sensing and detection applications are: Gene diagnosis and therapy, Cell sorting, manipulat ion, and identification, Bacteria analysis, DNA molecular sizing, Environmental microbial sensing, Tissue Engineering, Drug delivery systems.
2 The largest and most significant research area in BioMEMS is diagnostics. Many BioMEMS devices, for diagnostic applications, differ notably in their designs, fabrication processes, materials and their application area s. Most of these de vices are utilized to detect biological samples such as microorga nisms, cells, proteins, viruses, and DNA. Related to cell identification, the separation, manipulation, sorting of biological particles and sorting of individual cel ls is of paramount importance in microfluidic based diagnostics, [2-6]. In particul ar, there is a growing need fo r portable devices that detect and quantify biological species, isolate sp ecific cell subpopulations from complex matrices and support living cell arrays, [7-9 ]. Moreover, the areas of low-cost personalized medicine, battlefi eld diagnosis and homeland secu rity require inexpensive, low power, reliable devices. In addition, the devices for these areas must be capable of both raw sample processing and detection, whic h does not involve high skills training for their operation. Recent advan ces in BioMEMS/NEMS and mi crofluidics exhibit notable results for sorting and separation mech anisms based on optodynamics, hydrodynamics, electrokinetics and advanced materials, [10-15]. For instance, flow cytometry, (FC), dielectrophoresis, (DEP), optic al tweezers and stimuli sensitive polymers are examples of current approaches, [16-21]. Each of the sorting techniques menti oned exhibits various advantages and drawbacks, which depends on the specific app lication. FC is a technique for examining, counting and sorting microscopic particles su spended in a microfluidic flow. This technique allows for the analys is, usually on the basis of thei r relative fluorescence, of the physical and/or chemical characteristics of single cells. Complex computerized instruments are used to pass a stream of si ngle cells or particle s through the measuring
3 apparatus, which makes it unsuitable for a non sequential process. Normally, the cells are stained with a light-sensitive dye, placed in a fluid and passed in a stream before a laser or other type of light. This is a met hod for measuring the number of particular cells in a sample and the percentage of live cells in a sample. In addition, the method can be used to ascertain certain characteristics of cells such as size, shape and the presence of tumour markers on the cell surface. FC can also be utilized for cell sorting and manipulation by using electrical or mechanical means to divert and collect cells, with one or more measured characteristics, which fall within a range of inte rest. Cells/particles can be selectively charged during a pass and on their exit can be deflected into separate paths of flow. In the MEMS area, FC has been wi dely studied and developed for many important applications, [22-24]. Although th ere has been great pr ogress, within the MEMS field, the majority of the proposed sy stems fall in the categ ory of cell-holding procedures. Therefore, the MEMS devices ar e only designed for microfluidics control and retain the cell sample. The actual sample analysis is performed by a complex out of the chip optical system. The most common method for sample manipulation and attachment in these MEMS devices is DEP a nd most of the optical system, sensors and data analysis reside out side the chip, [25-27]. Recently, some research groups have integrat ed the optical part with the rest of the device, . However, the systems co mplexity is augmented and different issues such as power consumption, portability and operational ease are stil l important obstacles to be overcome. Another important matter is the fact that FC MEMS are mainly intended
4 for the samples optical analysis more than as a sample sorting or a sample preconcentration mechanism, . Optical methods normally require excessive outside the chip infrastructure for proper implementation. Currently, the optical tweezers approach is one of the most exploited optical methods. Optical tweezers ar e widely used to lock in, analyze and manipulate biological samples, [30-32]. Th e operational principle for optical tweezers relies on the optical gradient force, which competes with the radiation pressure. Therefore, very tight focusing is required in order to produ ce a stable optical trap. The trapping area is much smaller than typical cells which results in the force, exerted on the cells, being very non-uniform. In addition, optical tweezers require very high optical power, (1 to 100 mW), to produce a sufficient trapping force, (-1 to 100 pN). The high optical power density could re sult in photodamage of cells, . One possibility for reducing this effect is the use of infrared lasers. Another drawback related to MEMS optical tweezers is that focusing of the la ser for each sample is a time consuming approach. Therefore, the use of optical tw eezers is not a suitable sorting method for substantial sample size. DEP is the phenomenon that a particle, whic h is influenced by an inhomogeneous electric field, experiences. DEP imposes a force proportional to the gradient of the electric field. Due to its functional ease, DEP has been a widely used principle for labon-a-chip MEMS devices. Diel ectrophoretic based particle sorters are very suitable for integration with other system components. Di fferent types of partic le sorters have been presented in the literature, [ 34-37]. The majority of the approaches combine DEP with different techniques such as optical systems. DEP systems suffer from particle adherence
5 to the microelectrode and can cause cell damage due to long time exposure to the electric field, which is a concern. In contrast to the previous two methods the stimuli sensitiv e polymers approach presented in this dissertation exploits the fact that Intelligent thermally responsive polymers swell and contract with changes in temperature, . In order to reduce the complexity of the separation and sorting tasks, there exists a growing interest in these responsive materials for catch and release type applications, [39-40]. Almost two decades have elapsed since the first demonstration of the existence of "Intelligent or Responsive hydrogels, . Presently, thes e materials are being increasingly utilized in the BioMEMS/NEMS area. Environmen tally sensitive hydrogels offer unique opportunities for a wide range of BioMEMS/ NEMS sensor-actuator applications. Environmentally sensitive hydrogels offer excelle nt bio-stability and bio-compatibility. In addition, fabrication methods for envir onmentally sensitive hydrogels are relatively inexpensive, . An example of one thermally responsiv e hydrogel is the cross-linked Poly(NIsopropylacrylamide). Poly(NIPAAm)is normally used to undergo volume transitions by taking advantage of its thermally sensitive hydrophobic interactions within the polymer network, [43-44]. The poly(NIPAAm)hydrogel exhibits a Lower Critical Solution Temperature, (LCST). In addition, poly(NIPAAm) exhibits remarkable hydration-dehydration changes, in aqueous solution, in response to re latively small changes in temperature, . Below the LCST, poly(NIPAAm)chains hydrate to form an expanded structure. Above the LCST, poly(NIPAAm)chains dehydrate to fo rm a collapsed structure. The physical
6 structures and properties of poly(NIPAAm)are readily controlled by simply changing the temperature without changing the chemical structure of the polymer. Therefore, temperature responsive poly(NIPAAm)can be employed in drug delivery systems, cell manipulation, solute separations, concentratio ns of dilute solutions, immobilization of enzymes, coupling of biomolecules and phot osensitive materials, [46-50]. These characteristics made poly(NIPAAm) an ideal candidate for this re search. The most common example for this technology is the use of a thin intelligent polymer network film that integrates sensing and actuation at the material level, . The area of intelligent polyme r networks has developed rapidly in recent years. However, there remain a large number of possibilities for new procedures and improvements to reduce the complexity of sepa ration and sorting tasks. The possibility of reducing the patterning size is due to th e latest nano range l ithography tools. The ability to reduce the patterning size provides the opportunity to interact with an array of different biosamples such as bacteria and vi ruses. One major interest, which drives the use of BioMEMS, is the reduction of the sensor element to the scale of the target species. These sensor elements provide higher sensitivity and the reduction of the time to result, which is due to small volumes a nd higher effective concentrations. This research studied the effects of di fferent topographic pr ofiles and patterning procedures on the behavior of poly(NIPAAm). Additionally, this research studied the impact on the interaction of poly(NIPAAm )with bacteria in order to generate microfluidic based separations for sensor/act uator mechanisms. This research provided a novel technology platform that facilitates se paration processes, which are difficult to achieve by any other method, .
Figure 1.1 presents the sorting and separa tion mechanisms utilized, during this research, as stimuli for responsive polymer networks. Figure 1.1: Separation Mechanism Using Stimuli Responsive Polymer Networks Starting with a mixed biosample population, (A), microfluidics was exploited to allocate the interaction of the samples with a deposited and patterned thin layer of the polymer network, (B). Applicat ion of external stimuli resu lts in physical and chemical property changes to the polymer network that pr oduced an attach or release effect, which was dependent on the external stimuli, (C). The combination of other effects such as DEP and polymer patterning was exploited in or der to take full advantage of the polymer network response. Application of these eff ects maximized the sort ing and separation of the biosamples, (D). This dissertation presents the fundamentals of stimuli responsive polymer networks. The polymerization method employed fo r this research is presented in Chapter two. Chapter three explores the processing st eps, which were applied to fabricate the integrated BioMEMS device. The microfluidic channels integration, which was used for the interaction of the poly( NIPAAm)with biological samples, is explained. The 7
8 collection of methods utilized for poly(N IPAAm) processing is presented. These methods were employed in order to generate 3D structures with different topographic characteristics and a size range down to the su b-micron regime. The integration of all the processing techniques, the study and analysis of the attach and release mechanism between the biomaterial and the biological samp les and other related procedures such as DEP are also presented. Chapter four is de voted to the presentation of the complete approach to the experiments and the analysis of the results. Chapter five presents the conclusions achieved through this resear ch and some possible applications.
9 CHAPTER 2 STIMULI RESPONSIVE POLYMER NETWORKS There have been broad studies for dr ug delivery, separations and diagnostics applications using polymer systems that undergo phase transitions in response to external stimuli. These Intelligent polymers are know n to respond to a variety of stimuli such as pH, temperature and the electric fiel d, [53-56]. One of the main temperatureresponsive categories is the alkyl acrylamide polymers. An example of this category is poly(N-isopropylacrylamide) or poly(NIPAAm), which undergoes a sharp phase separation at its lower critical solution temper ature, (LCST), when submerged in water. The existence of this reaction in water is th e reason why they are called hydrogels. This endothermic process is driven by a gain in en tropy, which is associat ed with the release of hydrophobically bound water molecules, . Poly(NIPAAm) has a LCST, in water, when heated above 32 o C, (89.6 o F), . A large quantity of information is available for these materials since they have been employe d extensively as biomaterials. However, Poly(NIPAAm) gels are only responsive in the presence of aqueous solutions with relatively low ionic strength, wh ich significantly limits the type s of fluids that can be manipulated, . The hydrogels temperature and pH responses have been completely characterized. The hydrogel exhibits tunable phase transitions, whic h are sharp and may be exploited in many BioMEMS applications.
10 The LCST of these thermally sensitive polymers can be tuned to a desired temperature range by copolymerization. Copolymerization with a more hydrophilic comonomer raises the LCST while use of a more hydrophobic comonomer lowers the LCST, . Carboxylic acid m onomers such as acrylic aci d, (AA), or methacrylic acid, (MAA), which are readily available, have been copolymerized with NIPAAm. These copolymerizations, which utilize traditional free radical polymerization techniques, are employed to form random copolymers with desired temperature and pH-responsive properties, . The degree of the physical reaction of swelling increases dramatically with the incorporation of i onizable comonomers. Moreover, the transition temperature increases as the ionized groups make the pol ymer more hydrophilic, . The swelling behavior of the hydrogel permits their use in sensor-actuator mechanisms where the volume phase transition rate is known to scal e with the square of the gel dimension. Therefore, swelling should increase dramatical ly with microscaling, . However, the time scale is also a function of gel compos ition, microstructure and surface properties . There is also a difference in the swel ling response to increasing and decreasing temperature. The collapse of the gel network is significantly faster than the swelling process, . The Flory-Rehner theory ha s been used to study and interpret the anisotropic swelling of hydrogels. The theory qualitatively captur es the experimental features observed in poly(NIPAAm) based ne tworks. The Flory-Rehner theory has been widely used to analyze the phase transition of networks, .
The osmotic pressure, ( ), of the network can be described by six variables: = ( d, o f, N c ) (1) The parameter d is the dimension in which the network can swell. The volume fraction of the swollen gel is given by the parameter. The volume fraction of the preparation state is given by the, o parameter. The degree of ionization is given by the, f parameter. The number of segments between cross-links is given by the, N c parameter. The Flory-Huggins interaction parameter is a function of the te mperature, T, . Figure 2.1 presents a plot of the equilibrium polymer volume fraction of the network as a function of the reduced temperature with 1-2 Figure 2.1: Phase Behavior for Swe lling in Surface Attached Networks In this phase diagram, the isobars illustrate how the materials discontinuous volume transition depends on different factors such as the ion content, the state of preparation, the cross-link dens ity and its chemical compositi on. In the plot, the volume 11
12 fraction in the preparation state, o, was set to 1.0, the segment between cross links, N c was set at 500 and different degrees of ionization, f were applied. Figure 2.1 reveals the capability to tune-up the networks precise thermal response depending on the specific applications parameters. One of the main objectives of research with hydrogels is the evaluation of their responses as building materials for BioMEMS using several criteria. A few of these criteria are the ease of fabric ation of the hydrogel structures the kinetics of the volume phase transition as a function of gel size and composition, their ability for sensor-actuator mechanisms, anisotropic swelling of the hydr ogel, modeling of the swelling geometry, and the response to different stimuli. 2.1. Photopolymerization of NIPAAm In order to create the poly(NIPAAm) st ructures, a technique known as Softlithography is applied. This technique is de scribed in detail in Chapter 3. Softlithography generates poly(dimet hylsiloxane) or PDMS mold s, which are applied to produce the polymer features. To obtain the polymer biomater ial, NIPAAm, (97 % Aldrich), was purchased, . N, N-Met hylenenbisacrylamide, (BisAAm, Chemzymes Ultra Pure), was obtained from Po lysciences Inc. and used as a cross linker. An initiator, 2,2-dimethoxy-2-phenyl-acetophenone, (DMPA, 99% Aldrich), was used. Acrylic acid was purchased from Fisher, (99.0%), N, NDimethylformamide, (DMF), was purchased from Aldrich. All the chemicals were used without further purification. Poly(NIPAAm) was prepared on a Pyrex glass surface. The pr eparation of the substr ates included, in the given order, the actions:
13 Submerged in sulfuric acid for 10 minutes, Washed with isopropanol, Rinsed in deionized water, Dried with dry nitrogen gas, Cleaned for 15 minutes in a plasma cleaner, Soaked in a photosensitive sila ne coupling agent for 15 minutes, Heated in an oven at 100C for 10 minutes, Rinsed with deionized water, Dried with dry nitrogen gas. The photopolymerization process began w ith 0.1% (w/w) of initiator, [DMPA], 2.5% (w/w) of cross-linker, [Bis AAm], in a test tube with 1ml of Acetone. The test-tube was stirred for 5 minutes and then purged with nitrogen to remove the dissolved oxygen. The mix was injected into the soft -lithography PDMS mo lds using capillary forces. UV light was applied, for 8 minutes, to photo-initiate the grafting of NIPAAm while nitrogen was used to purge the polymerization chamber of any oxygen. The thiocarbamate group of the sensitizer was bonded to the glass surface by coupling the silane agent with the hydroxyl groups on the glass surface. Figures 2.2-2.5 present the processing steps for the poly(NIPAAm) structures.
Figure 2.2: SU-8 Photores ist Patterns on Substrate In Figure 2.2, the final thickne ss of the SU-8 photoresist was 40 m. The trench structures were attached to the two lateral s quare features, which we re used as the inlet and outlet ports for the capi llary injection of the mix. Figure 2.3: Detailed SU-8 PhotoresistPatterns Figure 2.3 presents the details of the SU -8 trenches where the PDMS was applied to generate the molds. 14
Figure 2.4: PDMS Molds In Figure 2.4, the PDMS was applied to obtain the master molds and the poly(NIPAAm) was In-situ photo-pol ymerized onto the substrate. Figure 2.5: Transferred Poly(NIP AAm) Structures on Substrate Figure 2.5 displays the excellent pattern fidelity that was achieved using softlithography. The final Poly(NIPAAm) thickness is 40 m. 15
16 CHAPTER 3 SYSTEM FABRICATION The processing methods, which were developed and applied to fabricate the proposed BioMEMS device, are presented in this chapter. The device integrated different technologies to achieve particle manipulati on and sorting. The device was based on the main concepts embodied in Intelligent reactive polymer network layers, Dielectrophoresis, (DEP), the effect of Joule heating and microfluidics. Thermal responsive high aspect ratio polymer trenches were fabricated in order to exploit their swelling/collapse behavior in order to separate biological particles with micron and submicron dimensions. The swelling/collapse behavior of the polymer was described in Chapter 2. When the trenches ar e open, objects smaller than their width can be driven down into the cha nnel through the use of DEP. On ce the objects are driven into the trenches they are captured by closing of th e trenches. Objects th at do not fit into the trench are not retained when the trenches are closed and pass over the trenches. The opening and closing of the trenches provides the mechanism for size-selection separation. This novel approach represents a key technology platform that facilitates separation processes that are difficult to ach ieve by any other so rting or manipulation method.
Figure 3.1 presents the basic concept associat ed with this sorting mechanism. It displays the four main building blocks of the system: DEP, microfluidics, heating element, and the Poly(NIPAAm) structures. Figure 3.1: Integrated BioMEMS Device The fluid flow enters a micro chamber th ru microfluidic channels. In the micro chamber the particles are driven down by ne gative dielectrophoresis forces. The openclose effect of the poly(NIPAAm) captures the particles and effects size sorting. 17
18 3.1. Poly(NIPAAm) Processing 3.1.1. Soft-Lithography Soft-lithographic methods use an elastomeri c stamp or mold, which is prepared by casting the liquid prepolymer of an elastomer against a master that has a patterned relief structure, in order to create structures su ch as microfluidic channels, . Photolithography is used for fabrication of the masters. Most of the research associated with soft-lithography has used poly(dimethylsiloxane) or PDMS as the elas tomer. PDMS has several properties that make it well suited for soft-lithography, . PDMS is biocompatible, permeable to gases and can be used for cell culture(s). PDMS is optically tran sparent above 300 nm. Since PDMS is elastomeric, it can contact non-planar su rfaces conformally, . Two different mold or stamp fabricat ion methods were applied during this research. One method was based on the use of positive, standard microelectronic, photoresist. Use of this method allowed fa brication of poly(NIPAAm) structures up to 1.5 m in height. Another method, which involve d the use of SU-8 photoresist, allowed fabrication of poly(NIPAAm) structures up to 120 m in height. The two methods are described in detail in the following sections.
220.127.116.11. Standard Photoresist Patterning In this approach the masters were made of positive patterned photoresist and used as molds for PDMS embossing, (Shipley 1813) Silicon wafers were used as the substrate. The substrates were RCA cleaned, DI water rinsed a nd dehydration baked on a hot plate at 200 o C for 10 minutes, . Preparation of the substrates continued with HMDS and photoresist spin-coating prior to UV exposure at the mask aligner. Once the desired photoresist structures were deve loped, the photoresist was hard baked at 155 o C for 2 hours. A mix of PDMS, (Dow Corning) and curing agent was prepared in a 10/1 ratio and allowed to outgas for 1 hour insi de a laboratory hood. The PDMS mix was placed on top of a 2 inch <100> n-type silicon wafer and placed on a hot plate at 90 o C for 90 minutes. Afterwards, the PDMS mix was allowed to cool down for 2 hours in order to obtain the final, desired, patterns. Openings at the edge of the PDMS patterns were constructed in order to apply capillary force effects to the poly(NIPAAm). After the poly(NIPAAm) accepted the shape of the PD MS patterns, it was photopolymerized, Insitu, as described in Chapter 2. Figure 3.2 presents the final photoresist structures, which were obtained with this soft-lithography method. Figure 3.2: Bright Fi eld Images of the Photoresist Structures 19
20 18.104.22.168. SU-8 Photoresist Patterning An alternate method to generate high asp ect ratio structures encorporated the use of SU-8 photoresist, (PR). The SU-8 photores ist is a thick, negative, high aspect ratio and epoxy-photoplastic resist. SU-8 photoresist is utilized primarily as a substitute, in several MEMS applications for the LI GA, (lithographie, galvanoformung und abformung), process. The LIGA process, whic h the use of the SU-8 photoresist replaces include: Fabrication of plastic micro-molds or metal micro-molds by electroplating, , Microfluidics for SU-8 microchannels , Fabrication of photoplastic structur es such as micro-gears, , Micro-coil fabrication, , Rapid prototyping using laser machining Bonding material for optical MEMS. Very high aspect-ratio micro-structures with vertical sidewalls and high resolution have been reported using the LIGA technique, [ 77]. However, the need for a synchrotron radiation source limits the application of th e LIGA technique. In addition, the process is inherently time consuming and expensive. Th e SU-8 photoresist can be spin coated in a standard spinner with th icknesses ranging from 1 m to 300 m and thicknesses up to 2 mm have been reported with multilayer coatings. Combined with standard lithographic processes, the use of SU-8 photoresist offers excelle nt opportunities in MEMS applications and packaging, which makes it s use very popular in the MEMS community.
21 The processing of SU-8 photoresist has been described by various groups, . SU-8 photoresist has very suitable properties of thickness and chemical stability. Additionally, SU-8 photoresist possesses good mechanical and optical properties. The disadvantages associated with the use of SU-8 photoresist are concerned with stress, adhesion selectivity and resi st stripping. The adhesion SU-8 photoresist is good with materials such as silicon and gold, . However, on materials such as glass, nitrides, oxides and other metals the adhesion is poor The SU-8 photoresist can easily be removed from the surfaces of such materials during the development process. On many surfaces such as silicon or glass, which are suitable for spinning the SU-8 photoresist, the thermal expansion coefficient mismatch is la rge. This mismatch causes large amounts of stress at the material interface, which is due to shrinkage of the SU-8 photoresist while cross-linking during curing. The stress eff ect is pronounced in la rge SU-8 photoresist structures and, if poor adhesion is obtained during processing, the photoresist delaminates easily. The challenges for fabricating an ul tra-thick micro-structure with SU-8 photoresist include coating, baking, the de veloping efficiency and the tremendous residual stress formed afte r curing. Different groups have reported that a micro-structure with a height of 1.2 mm can be formed by double coating the SU-8 photoresist layers, , . However, multicoat ing is a time consuming process. The surface flatness is also a critical issue. The solvent content in subsequent coating layers may be different, which results in a much more complicated lit hography process. Single coating of an SU8 photoresist film, thicker than 1mm, is not pr actical using a conventional spin coater due to the high viscosity of the photoresist. As a photoplastic material, the SU-8 photoresist
22 is chemically stable and resistant to most acids and other solvents Consequently, after cross-linking it is difficult to remove and suitable methods of stripping, which are compatible with other materials in the structur e, are often not effectiv e or their use is not desirable. The process was initiated, by applying RCA cleaning to an n-type <100> SingleSide-Polished 2 silicon wafe r. Different MicroChem Co rporation SU-8 photoresists, (SU-8 2005, 2025 and 100), were used to spin co at the sample, . Initially, ramping spinning at 500 rpm for 5 s was employed. Depending on the selected SU-8 photoresist and the desired thickness, a second rampi ng speed was employed. For example, the second speed was set to 1500 rpm for 25 sec onds, which resulted in a coating thickness of 120 m, when SU-8 100 photoresist was used. After spin coating, the sample was allowed to relax for one hour in order to allow reflow to complete and to prevent problems with step coverage on the silicon. A pre-exposure bake was performed using temperature ramping to reduce stress. The sample was baked on a hotplate at 65 C for 5 minutes. Immediately afterward the temperat ure was ramped and baking continued at 95 C for 40 minutes. The pre-exposure bake was deliberately extended to allow for maximum solvent evaporation. A Karl Suss Mask Aligner UV light source was used for exposure. A mask, which defined the locations of the SU-8 photoresist on the wafer, was utilized and exposure persisted for 35 sec onds. The SU-8 photoresist structures on the cantilevers formed the trenches. The trenches possessed a width of 15 m and an opening space of the same dimension. The post exposure bake was carried out at 65 o C for 3 minutes and 95 o C for 12 minutes. Afterwards, the samples were left at room temperature for 20 minutes in order to avoi d thermal stress. The substrates were
developed using a MicroChem SU-8 photoresis t developer. The development process consisted of gentle agitation for 20 mi nutes followed by a hard bake at 150 o C for 2 hours. The poly(NIPAAm) patterning process is presented in Figure 3.3 Figure 3.3: SU-8 Photoresist Pro cessing for Poly(NIPAAm) Patterning Figure 3.3 A-E presents the poly(NIPAAm) patterning steps. A silicon or glass substrate is cleaned for the SU-8 patter ning, Figure 3.3A. Figure 3.3 B displays the patterned SU-8 structures. The SU-8 photores ist structures were us ed as molds for the PDMS, Figure 3.3C. The PDMS mix was placed on top of the patterned SU-8 photoresist and placed on a hot plate at 90 o C for 90 minutes. Afterwards, the sample was allowed to cool down for 2 hour s in order to obtain the de sired final patterns. Figure 3.3D, represents the process where the PDMS membrane wa s peeled-off and placed on top of a new substrate. The ne w substrate could be either sili con or glass, which has been preprocessed with the silane binding agent. Openings at the edge of the PDMS patterns 23
were constructed in order to apply capillar y force effects on the poly(NIPAAm). Figure 3.3E, the poly(NIPAAm) took the shap e of the PDMS patterns and was photopolymerized In-situ as described in Chapter 2. A SEM was used to characterize the height SU-8 structures, which were found to be 120 m. Figure 3.4 presents a SEM image of the SU-8 photoresist structures. Figure 3.4: SEM Image of the SU-8 Photoresist Structures Responsive, high aspect ratio, poly(NIPAAm) trenches were fabricated in order to exploit their, temperature dependent, swelling/ contraction behavior to separate submicron biological particles from real samples. When the trenches are open, objects smaller than the width can be pulled into the channel and held there upon closing the trench. Objects that do not fit into the trench are not retain ed upon closing the channels, and are passed over the trench. This mechanism offers si ze-selection separation. Figure 3.5 presents a DAPI, (4', 6-diamidino-2-phenylindole), im age of the final poly(NIPAAm) structure, which was synthesized by this method. 24
Figure 3.5: A DAPI Image of the Poly(NIPAAm) Trenches 3.1.2. E-Beam Lithography Recently, most of the hydrogel related research has focused on the fabrication process of structures using micro-fabrication techniques, [ 86]. One of the most common fabrication approaches is the use of uvlight photolithography in order to pattern structures, . Different techniques for fabr icating sub-micron length scales that span a wide range are available. This section pres ents a fabrication process using an e-Beam Lithography, (EBL), system within a Scanning Electron Mi croscope, (SEM), which was enhanced with a Nanometer Pattern Generati on System, (NPGS),  General purpose scanning electron microscopes compare well, in terms of controllability and stability, with dedicated e-Beam Direct Writers and dedicated CD SEMs. These tools provide medium throughput, high quality and lithography capability for nanotechnology devices. Extensive development of commercial EBL systems has provided various modes of use for positive and negative resists. Howe ver, work still continues to be performed with the common polymer PMMA resist on conv erted SEMs, . PMMA is one of the most common positive e-beam resists employe d today since it possesses the highest resolution, . There are many other negativ e resists available su ch as COP, SAL and 25
26 P, (SI-CMS), which work by cross-linking the po lymer chains together. Cross-linking of the polymer chains renders them le ss soluble in the developer, . A novel technique for e-Beam lithography has been developed using poly(NIPAAm) as a resist to pattern a silic on substrate. The topographic profiles of the structures achieved were characterized and analyzed using an Atomic Force Microscope, (AFM). The results showed that the poly(NI PAAm) behaves as a negative resist for eBeam lithography. Therefore, it can be exploi ted in a direct patterning procedure for a thermally responsive hydrogel in orde r to achieve nano scale structures. 22.214.171.124. Resist Preparation The hydrogel utilized during th is research as a resist for e-Beam lithography was poly(NIPAAm). This polymer was dissolved using cyclohexanone. Afterwards, the solution consisting of the polymer and the so lvent was mixed and stirred for 24 hours at room temperature. The poly(NIPAAm) concentr ation was set at 5%. It is important to note that the solvent normally used for P MMA, which is Anisol, did not dissolve the poly(NIPAAm) at this concentration level. The cyclohexanone and the poly(NIPAAm) were acquired from Aldrich, . The resist was spin-coated at different sp eeds onto Silicon substrates in order to analyze the speed effect on the resists final th ickness. The speed values were selected in accordance with the PMMA vendor suggestions and employed for 45 seconds each, . Following the spin coating the resi st was baked on a hot plate at 140 o C for 75 seconds.
27 126.96.36.199. E-Beam Exposure E-Beam lithography was performed, for each machined pattern structure, using a JOEL SEM Model 840, which was enhanced with the Nabity NPGS system. The exposure parameters, which were used for resolution optimization, were 30 kV for the acceleration voltage, 24 pA for the beam current and an e-beam dose range between 300 C/cm 2 and 720 C/cm 2 188.8.131.52. Resist Developer Once the e-Beam lithography process was performed, the resist was submerged into a MIBK/IPA developer. A MIBK/IPA developer ratio of 1:3, which is used for PMMA, was applied for 70 seconds. Application of the developer was followed by 20 seconds in IPA and a rinse for 20 seconds in DI water. In order to complete the development process the samples were rinsed and dried with no post or hard baking step performed on the samples. The topographic profile of the fabricated structures was characterized and analyzed using an Atomic Force Microscope, (AFM). Figure 3.6 presents a comparison of the effect of different spin speeds and th e final film thickness. The thicknesses are in angstroms, (), for poly(NIPAAm) and PMMA. The data was obtained using a profilometer, (Tencor Alpha Step 200). The results exhibit a similar behavior, am ong the two resist polymers, for the spin speed effect on the final film thickness.
Figure 3.6: PMMA and Poly(NIPAAm) Film Thickness vs. Spin Speed The two dimensional image of the patt erns acquired using an Atomic Force Microscope, (AFM), is presented in Figure 3.7. Figure 3.7: Two Dimensional AFM Im age of the Poly(NIPAAm) Patterns The scan size was 40 m at a scan rate of 0.5 Hz. In Figure 3.7, the smallest feature sizes obtained are in the nano meter scale range as illustrated. 28
Figure 3.8 displays the three dimensional image of the poly(NIPAAm) patterns. The different section analysis test points appear as inverse triangles. Figure 3.8: Three Dimensional AFM Image and Section Analysis 29 In addition, Figure 3.8 presents the section analysis of the patterns. The average final thickness value for the patterns was 140 nm The achieved side walls profile is also displayed in the image.
30 3.1.3. FIB Patterning A possible alternative for poly(NIPAAm) patterning is the use of focused ion beam, (FIB), systems combined with soft-lithography. FIB systems have proven to be very useful in micro and nano processing. The utility of FIB systems is primarily due to their ability to select ively remove and deposit material without the use of a patterned resist mask, . Applications for FIB in the fabrication of MEMS/NEMS have been described with emphasis on the processing. FIB techniques can also be very useful during prototyping of microstructures, . This patterning technique demonstrated an interesting approach to the achievement of nano scale structures of poly(NIPAAm). 184.108.40.206. Sample Preparation Silicon samples were prepared and supplied with poly(NIPAAm) using softlithography. These samples were rinsed with DI water and allowed to dry. The samples were taken to the FIB system, which was located at USFs NNRC, (Quanta 3D USF), where they were patterned. The SEM/FIB was configured and optimized for PMMA patterning.
Figure 3.9 presents the structures obtain ed by combining soft-lithography and FIB patterning. The large features were obtained with soft-lit hography and the trenches were obtained by FIB patterning. Figure 3.9: SEM Image of FIB Patte rned Cross-Linked Poly(NIPAAm) 3.2 Dielectrophoresis Effect Current approaches to cell sorting, most frequently, exploit differences in cell density, specific immunological or receptor-ligand in teractions to isol ate target cells. These techniques are often inadequate. Theref ore, sorting devices cap able of identifying and selectively manipulating cells through nove l physical properties are desirable. Electrical-field-induced forces on polarizable particles have b een the subject of extensive studies in recent years because of their potent ial applications for pa rticle characterization 31
32 and manipulation. Particles, including biol ogical cells, become electrically polarized under the influence of applied fields. The induc ed polarizations can, in turn, interact with applied fields, which results in net electrical forces on the particle s. Dielectrophoresis, (DEP), is an important phenomenon for the manipulation and distinction of various dielectric particles, . DEP possesses the ab ility to influence cells or particles strictly as a function of their intrinsic dielectric differences without any modification to them. Therefore, DEP has become one of the most attractive manipulation techniques. In recent studies, many research groups have reve aled the dielectric properties of various biological and separated polystyrene micr o-beads such as human peripheral blood mononuclear cells, yeast, leukocytes, erythroc ytes, neuronal cells and malaria, . During the past few decades, the application of dielectrophoresis for collecting, positioning and separating particles suspended in liquids has advanced tremendously due to improvements in micro-fabrication techniques. Two approaches, which depend on migr ation and retention, have been demonstrated for DEP separation. DEP migr ation exploits the opposing polarities of DEP forces exerted on different particle types. One type is attracted toward high-field regions by positive dielectrophoresis while the other type(s) are repelled by negative dielectrophoresis. Therefore, different particle types are focu sed at different regions of a microelectrode structure and sp atial separation is achieved, [9 7]. DEP retention exploits the competition between DEP and fluid-flow forces. Particles experiencing a weaker DEP force, either negative or a small posit ive, are eluted by fluid flow. However, particles experiencing strong pos itive DEP forces are trapped at electrode edges against the drag of the fluid flow.
There has been great interest in a family of techniques collectively termed fieldflow fractionation, (FFF). FFF techniques exploit the velocity gradient of a hydrodynamic flow profile to achieve particle fractionation, . In these techniques, a laminar fluid profile is established in a thin chamber such that fluid moves faster with increasing distance from the chamber walls. A force field is applied perpendicular to the flow direction and across the flow profile. Th e force field causes particles to be forced toward a so-called accumulation wall of the chamber, which is typically the bottom wall. In normal-mode FFF, particles are sufficiently small that the diffusion and accumulation forces counteract and an equilibrium concen tration profile results along the force-field direction and hence across the fl uid velocity profile. Thereby particles are carried by the fluid at velocities that de pend on their equilibrium profile Dissimilar particle types exhibit different characteris tic concentration profiles. The different concentration profiles are carried at different velocities, which causes the different particle types take different times to traverse the separation cham ber. Therefore, they can be separated by collecting them at appropriate times at the chamber exit. This principle was exploited during this research in order to deflect th e particles down to the poly(NIPAAm) trenches where the catch and release mechanism was executed to manipulate the samples. The movement of dielectric particles insi de a microfluidic channel is dominated by the dielectrophoretic velocity induced by a non-uniform electric field and the hydrodynamic velocity imparted by the fluid fl ow. The total velocity is given by: Hyd DEP TotalVVV (3.1) 33
The dielectrophoretic velocity result s from the equilibrium between the dielectrophoretic force and the viscous dr ag force, which is represented by: r F f F VDEP Drag DEP DEP6 (3.2) The DEP force acts on neutral bodies in nonhomogeneous electrical fields. It differs with respect to the aspe ct of neutrally from the electrical force, which acts only on charged bodies in an electrical field, E In the presence of the field, E the body becomes polarized. The polarization of the body makes it e quivalent to an electric dipole. In an electric dipole there exists an excess of positive charges area and an excess of negative charges in another area and the two areas ar e separated by some distance. This dipole establishment for an initially neutral dielectric particle, which is subjected to a nonhomogeneous field is illu strated in Figure 3.10). Figure 3.10: Dielectric Particle in a Nonuniform Electric Field Since the field is nonhomogeneous, the fo rces acting on the two ends are not equal and the total net for ce is nonzero. The force, F is given by: 34
rEQdrEQF (3.3) where Q + and Q represent the excesses of positive and negative charges on the body. If the distance associated with the induced dipole, d, is small compared to the distances with respect to which the field varies appreciably, then Q + Q (3.4) and E ( r+d ) E ( r ) + (d ) E (3.5) Substitution of the approximations given by equations 3.4 and 3.5 yields: EpF (3.6) where p = Q d (3.7) is the induced dipole moment of the body. 35
The time-averaged dielectrophoretic force that represents the interaction force between the electric field and the induced dipole moment is given by: 2 3)Re(2 EfrFCM m DEP (3.8) and * **2 )(m p mp CMwf (3.9) where r is the radius of the particles, m *and p are the complex permittivities of the suspending medium and particle respectively, E is the local electric field and Re( f CM ) is the real part of the polarization factor ,( ClausiusMossotti factor), which is dependent on the applied field frequency a nd the conductivity medium. Re( f CM ) can have values between +1 and -0.5. Substitution of equation (3.8) in equati on (3.2) and simplifying shows that the dielectrophoretic velocity is pr oportional to the cube of the pa rticle radius and the real component of the polarizati on factor with a fixed medi um conductivity and a fixed electric field. Consequently, the DEP force can be used to separate, position or deflect particles with different sizes and dielectrophore tic affinities under the appropriate electric field distribution. With the electric field gradient of the microelectrode along the perpendicular direction to the fluid flow, particles with differe nt dielectrophoretic velocities can be selectively fractionated in a microfluidic channel. 36
The Electric field, E can be numerically simulated through the use of finite element analysis. Figure 3.11 presents a simulation for the array of inter-digitated microelectrodes. The simulation was perfor med with the ANSOFT HFSS software, . Figure 3.11: Simulation of the Effect of the Electric Field, E on the Microelectrode Array In Figure 3.11(A), a plane is defined at the center of one finger section of the microelectrode array. The gap A-A distance is 50 m. The plane is perpendicular to the flow inside the microfluidic channel. In Figure 3.11(B) an AC signal is applied to the microelectrode array. The applied signal produces an electric field in a plane perpendicular to the surface plane of the mi croelectrodes. The vectors indicate the direction of the field, which wi ll produce a DEP force perpendicular to the flow inside the 37
38 microfluidic channel. The strength of th e field is given by equation 3.8. Figure 3.11(C) displays the field intensity, whic h, as expected, is larger near the array and reduces in the z axis direction. Figure 3.11(D) presents a detailed view of the electric field, E close to the A-A cross section between to electrodes of the array. In the final device, the DEP force is necessary to deflect the particles into the microfluidic channels, which exhibit laminar flow behavior as e xplained in Section 3.4. The induced DEP force could be, either negative or positive. If the DEP fo rce is negative the particles would move away from the microelectrode arra y. If the DEP force is posi tive the particles would move toward the microelectrode array. The dire ction of the DEP force depends on the final value of the ClausiusMossotti factor, which is indicated in equation 3.8. In order to bind the samples and the poly(NIPAAm), the DE P effect must be applied inside the microfluidic channel. 3.2.1. Fabrication A microelectrode array was fabricated on top of a glass substrate with, Futurrex NR9-1500PY, negative photoresist, (PR), using conventional photolithography. The initial processing of the glass involved several steps, whic h consisted of the substrate being: Cleaned with a sulfuric acid solution for 5 minutes, Dipped in 50:1 HF for 30 s to improve adherence, PR was spin coated at 3000 rpm for 30 s, Soft baked, on a hot plate, for 2 minutes at 150 o C, Exposed to UV light for 35 s,
Baked, post exposure, for 1 minute at 100 o C, Developed for 20 s with Futurrex RD6 developer. After the initial proce ssing, 1800 of Au was deposited on the substrate via sputtering. Lift-off was performed using acetone and isopropanol. The device was DI water rinsed and nitrogen dried. The cross se ction and bright field image of the final inter-digitated microele ctrode array are presented in Figure 3.12. Figure 3.12: Au Microelectr ode Inter-Digitated Array 3.3 Thermal Actuation In order to modify the thermally se nsitive hydrophobic interactions, within the polymer network, micro-machined thermal actua tors were installed in the final device. A micro-heater to thermally affect the hydrogel was fabricated using conventional techniques, . A therma l insulation layer of SiO 2 was thermally grown up to 3000 The SiO 2 was patterned by employing standard photolithography a nd diffusion openings 39
were defined. Boron diffusion was applied on th e substrate in order to obtain the desired doping profile. Finally, aluminum contacts were patterned to connect the diffused resistor. The final temperature obtained was, according to the Joul e effect, proportional to the geometric properties and the current a nd voltage applied to the resistor, . Figure 3.13 presents two diffused re sistor prototypes, which were fabricated as part of the research. Figure 3.13: Bright Field Image of the Thermal Actuators Another proposed method to generate the heating element in the system was to use a nichrome resistor heater Using the Joule heating eff ect, which is determined by: RIQR 2 Watts (3.10) where I is the current applied to the resistor in amperes and R is the resistors final resistance value in ohms. 40
41 Thin film heaters can be made of differ ent materials on a variety of substrates. The thin film circuit substrate was selected to ensure the performance and reliability of the circuit. The main requirement for the substrate, for this research, was the electrical insulation properties. A number of thin film resistor materials such as: tantalum nitride, polysilicon, tungsten, and platinum can be used for different applications. However, the most popular material is the nickel-chromium alloy called ni chrome, which is 80% nickel and 20% chrome. Nichrome was selected as th e material for fabricat ion of the resistor film for this research. Thin film resistors, (thickness le ss than 24,000 or 0.1 mil), offer the advantage of a low temperature coeffici ent of resistance, (TCR), and finer line definition capabilities over their thick film e quivalents. Nichrome was selected since it can be easily e-beam deposited on glass substr ates and allows standard processing for patterning. Nichrome resistors are corrosion -resistant and, with proper annealing, offer a very low TCR, which results in precise and thermally stable resistors. The resistor fabrication process starte d by cleaning the gla ss substrates in a sulfuric acid solution for 10 minutes, rinsi ng with DI water and desiccating, on the hot plate, at 200 o C for 10 minutes. Next, th e resistor network was pa tterned using, Futurrex NR7-3000 PY, negative PR on the wafer, (Futu rrex, Inc., Franklin, NJ). The nichrome resistor was deposited in an e-beam evapor ator for a total thickness of 1800 Typical sheet resistance values for nichrome range from 100 to 200 Ohms-per-Square, with a tolerance of 10% of the nominal value and a TCR of 0 50 ppm per o C. Figure 3.14 presents the cross section and the bright field image of the final structure fabricated, which yielded a final resistance value of 1.9 K
Figure 3.14: Nichrome Resistor Structure 3.3.1 Heat Transfer In order to design the heating element, it was necessary to estimate the heat transfer applied to the microflu idic channel. The total amount of heat necessary to effect a change in temperature, from room temp erature and above the LCST, is given by: TfcQw (3.11) where f is the fluid flow, c is the specific heat and T is the change in temperature. To obtain a ch ange in temperature of 20 o C with a fluid flow of 10 l per min the total heat, Q w required was 6 x 10 -3 Kcal per hour. 42
The heat loss, due to transfer by c onvection and conduction through the channels side walls and top part, can be calculated using Fouriers la w of heat conduction, which is given by: TAQf (3.11) where is the thermal conductivity, A is the area and T is the change in temperature. The total amount of heat, Q f applied to the cha nnel was 0.012 Kcal per hour, which is equivalent to 0.014 W per hour, in electric units. This value was used to calculate the total current required for th e nichrome heating device, which was in accordance with the final m easured resistance value. 3.4 Microfluidics Integration In the final device the poly(NIPAAm) struct ures were placed inside microfluidic channels to interact with biological samples. Microfluidic systems constitute a steadily growing sector of the MEMS field. Flow ch annels for fluids ar e fundamental building blocks of microfluidic systems. Flow channels in microfluidic systems are analogous to wires or thin film electrica l interconnects in c onventional integrated circuits, . There is a clear trend to reinvent microfluid ic components, as dis posables, by replacing classic MEMS/NEMS materials such as silicon or pyrex with inexpensive polymer materials such as PDMS, parylene or SU-8 photoresist, . Two different microfluidic channel integration methods were developed during this research. One was based on the use of PDMS which was bonded to the subs trate. The other method utilized SU-8 43
44 photoresist as an intermediate layer between two surfaces in or der to create microfluidic channels. PDMS has had a large impact on research and development. The PMDS impact is due to its ease of processing, in standard laboratory conditions, and its biocompatibility predestines, which make it suitable for rapid prototyping in microfluidics, . This elastomer is widely used in microfluidic applications to form components such as channels, valves, and di aphragms, . PDMS is a low cost material that offers many advantages. PMDS can be processed by easily molding. PMDS is elastic and can form fluid seals effectively. PDMS is commonly used as a bulk material, . The predominant fabricati on process associated with PDMS is bulk molding. The principle of the PDMS pattern ing process used during this research is discussed in the following section. 3.4.1 PDMS Microfluidic Channel Fabrication Process The initial phase for the Silicon-PDMS molding, used during this research, was bulk silicon processing using an induction coupled plasma (ICP), a deep reactive ion etcher, (DRIE), which employed the advanced silicon etch, (ASE), process, . The ASE process was patented by Bosch GmhH. The basic idea behind the anisotropic reactive ion etching efforts, in use today, is to find a bala nce between trench side wall passivation and trench bottom etching. Tr ench bottom etching is activated by the bombardment of ions from the plasma discharge. The method known as the Bosch Process is a room temperature process based on continuous cyc ling of subsequent passivation and etching steps. The Bosch Process achieves high aspect ratio microstructures. Aspect ratios over 20 and etching depths up to 500 m can be achieved.
A typical etching rate is 2 m per min, the selectivity to resist is 75:1 and the selectivity to silicon dioxide is 250:1, . Dry etching and molding is a fast and flexible production process for polymer microstructure products based on a sequence of process steps, one, of which, is a DRIE step, . The high aspect ratios, directional freedom low roughness, high etching rates and high selectivity with respect to th e mask material of DRIE allow a versatile fabrication process of micro-moulds for subsequent molding and embossing. The feasibility of this process has been demonstrated for feature sizes of several micrometers up to tens of micrometers. The layouts of the mould inserts are etched by DRIE into silicon. The etched structures are used for PDMS embossing. Silicon wafers <100> 4 inch, n-type were DRIE processes using a Unaxis, PlasmaTherm 770 ICP system to produce microfluidic channel molds on Silicon for the final application during this research. The DRIE conditions used in order to obtain the structures on the Silicon wafe rs are presented in Table 3.1. Table 3.1: DRIE Process Parameters 45
The structures obtained by DRIE, during this research, are presented in Figure 3.15. The final depth was 8 m. Figure 3.15: SEM Image of the DR IE Patterned Micro-Channels Figure 3.16 displays the cha nnels topographic profile. Figure 3.16: SEM Image of the DRIE Anisotropic Etching 46
A typical Bosch Process side wall effect is outlined in Figure 3.17. The passivation and etching step s produced this profile. Figure 3.17: SEM Image of the DRIE Bosch Process Side Wall Profile 220.127.116.11. Silicon-PDMS Molding PDMS was poured into the Silicon molds and cured for 1 hour at 65 o C using a 10:1 ratio of elastomer and curing agent. After curing the PMDS was peeled from the mold revealing the microfluidic channels. The channels are presented in Figure 3.18. Figure 3.18: PDMS Microfluidic Channels 47
48 3.4.2. SU-8 Photoresist Microfluidi c Channel Fabrication Process Another method applied to fabricate micr ofluidic channels, during this research, was based on SU-8 photoresist patterning. Polymerized SU-8 photoresist possesses excellent chemical and mechanical properties. The properties of the polymerized SU-8 photoresist have made possible many novel microf luidic devices whose capabilities have been demonstrated, . However, the fabr ication of high-aspect ratio microfluidic channels is still under extensive investigation. During this research, a novel method was developed for the fabrication of microfluidic structures. SU-8 photoresist was used as an intermediate layer between two substrates. One of the substrates contained the DEP microelectrode array and the other contai ned the poly(NIPAAm) trenches. Polished, 1.1mm thick, microscopic glass substrates were cut to fit the design of the device. Prior to the lithography process, the glass substrates were cleaned in a boiling Piranha solution, (H 2 SO 4 :H 2 O 2 = 3:1), for 10 min. The photolithography masks were gene rated using layout design software, (AutoCAD). The SU-8 photoresist process was the same as described in Section 18.104.22.168 in order to obtain the high aspect ra tio structures. The bonding method between substrates utilized a highly transparent UVsensitive adhesive and a UV light source. PDMS with a 1:1 ratio was used for sealing the microfluidic devices and the capillary tubing. The process for fabricating the SU -8 photoresist microfluidic devices is presented in Figures 3.19 and 3.20.
Figure 3.19: Microfluidic Channel Fabrication Figure 3.19 presents the microfluidic ch annel cross-section where the upper substrate, (Black), contains the DEP microel ectrode array and the lower substrate, (Gray), contains the polymer trenches. The SU-8 photoresist was patterned to form the microfluidic channels. Finally, both substrates were bonded and the channels inlet and outlet ports, (I/O), were attach ed and sealed using PDMS. 49
Figure 3.20: Microflu idics Integration Figure 3.20 presents the complete microfluid ics integration process. In step A, the SU-8 photoresist is patterned according to the channel design. In step B the channel is aligned to the DEP microel ectrode array and the poly(NIPA Am) trenches. In step C, the substrates are bonded and the channel formed using the SU-8 photoresist as an intermediate layer. Figure 3.20D presents a bright field image of the final channel structure. Figure 3.20E displays the DEP Microelectrode inside the microfluidic channel. 50
51 CHAPTER 4 EXPERIMENTS AND RESULTS 4.1. Poly(NIPAAm) Patterning The topographic profile of the Soft-lit hography-FIB patterned structures, which were fabricated, was characterized and an alyzed using an AFM. AFM imaging was performed in the FIB exposed area using an Asylum Research's MFP-3D system and uMasch silicon cantilevers w ith nominal spring constants of 1nN under ambient air and deionized water. The sample was submerged in 50 o C water and maintained at that temperature by recycling the water droplet for 15 minutes. The effects of temperature, on the gel, were obtained by systematically sca nning a small area of 10 scan lines as the fluid equilibrated with the su rrounding air temperature. Se ctional data was taken on the same position of the infrastructure for comparison. The AFM results illustrated the swelling behavior of the poly(NIPAAm) due to a change in water temperature from 45 o C to 25 o C. The swelling resulted in an 80% volume transition in the vertical plane. The FIB patterning demonstrated nano-s cale range patterning capabilities.
An SEM image of the patterns generated is presented in Figure 4.1. The lighter colored areas are the Poly(NIPAAm) structures. Figure 4.1: SEM Images of So ft-Lithography and FIB Patterning Figure 4.1 presents a structure with di mensions less than 500 nm. Figure 4.2 presents the AFM section analysis fo r the FIB patterned structures test. Figure 4.2: AFM Measured Section Analysis Figure 4.3 presents the AFM scans for temperatures of 45 o C and 25.6 o C. The swelling properties of the H ydrogel with changes in temper ature are clearly illustrated. 52
Figure 4.3: (Top) AFM Image for T = 45 o C; (Bottom) AFM Image for T = 25.6 o C 4.2. High Aspect Ratio Trenches Thermal Responsiveness The high aspect ratio trenches, which were constructed of the UV polymerized In-situ poly(NIPAAm) using soft-lithography, were tested to confirm their temperature responsiveness. The collapse/swelling beha vior expected was obt ained by applying a 20 o C temperature change, ( T). DI water was poured on the samples at different temperatures above and below the LCST of the poly(NIPAAm). 53
It was observed that at 40 o C, (104 o F), the trenches were open. However, using DI water at room temperature causes the poly( NIPAAM) to swell and induces a closing effect on the trenches. This effect is a key technology platform that facilitates separation processes that are difficult to achieve by any other met hod. Figure 4.4 presents the poly(NIPAAm) swelling behavior due to temperature stimuli. Figure 4.4: Bright Field Images of the Poly(NIPAAm) Swelling Behavior 4.3. Particle Entrapment A series of experiments were performed in order to analyze the particles catch and release effect inside the high aspect ratio tr enches. Fluorescent polystyrene microbeads, (PolyScience PA), were added to a mix cons isting of DI water, 20m diameter beads and 6m diameter beads. This mix was applied to the top of the trenches at different temperatures and different beads concentration values in order to examine their behavior. Figure 4.5 presents the trench structures, submerged in water, with a temperature above the LCST of the poly(NIPAAm). There is no microbeads mix present in Figure 4.5. 54
Figure 4.5: Bright Field Im age of the Trench Structures Before the Experiment The BioMEMS device exploits the opening and closing effect of the fabricated trenches to entrain particles inside them in order to generate a sorting mechanism. Figure 4.6 presents an image with the microbeads mix present. The 6 m beads move inside the trenches, while the 20 m beads stay on top of the structures. Figure 4.6: Particles Movi ng Inside the Trenches 55
Figure 4.7 illustrates the behavior of th e beads mix diagrammatically. The size sorting mechanism is applied to 20 and 6 m diameter microbeads. Figure 4.7: The 20 m Beads Remain on Top of the Poly(NIPAAm) Structures After the surface was scoured with cold water, the thermal responsive polymer swelled closing the trenches and capturing the particles inside. This closing and capture behavior is displayed in the sequence presented in Figure 4.8. Figure 4.8: Bright Field Imag e of the Capture Mechanism The areas inside the circled point displa y the captured particles and the closed trenches. In Figure 4.8A, cold water was applied and the particles moved inside the trench. In figure 4.8 B, the trenches start to close. In Figure 4.8C, the particles are captured. The experiment was repeated using differe nt microbeads concentrations. One set of experiments contained 6 m and 20 m microbeads. Other sets of experiments 56
contained only the 6 m microbeads. In th e experiments with a mix containing the two different diameter size microbeads, the majo rity of the 20 m beads were removed. However, some of them remained on top of the trenches and a significant amount of particles were not tra pped in the trenches. It appears that there is so me interaction that produces a sticking effect between these beads and the polymeric surface, which produces a failure to release of the particles. After the trenches were scoured again with hot water, very few, if any, of the particles that were trapped were able to be removed. This situation is presented in Figure 4.9. Figure 4.9: Bright Fiel d Image of the Failure to Release Effect Although some applications such as particle entrapment for analysis dont require the release of the particles fr om the trenches, the scope of the application is reduced by the lack of a release mechanism. The fi nal BioMEMS device must be capable to integrate both the capture and release mechanisms. 57
58 4.4. Cell Attachment An experiment was developed to anal yze the interaction behavior of the poly(NIPAAm) and Escherichia coli bacteria. Two silicon substrate samples were prepared and supplied with Poly(N-isopropylacrylamide) on each. These samples were rinsed with RO/DI water and allowed to dry. In preparation for the experiment, a hybridization oven and a water bath were bot h set to 38C. One of the samples was placed in the oven while the other was left out at room temperature, (22C). A 50 ml conical tube of rinse water, (RO/DI), was placed in the water bath and another was left at room temperature. Two cultures of E. coli w/green fluorescent protein, (GFP), were prepared and allowed to equilibrate at the two operating temperatures. The cultures were prepared as follows: A single colony of E. coli w/GFP was added to 1 ml of minimal media and resuspended by vortexing, The culture, (0.5 ml), was added to tw o tubes containing 1.5 ml of minimal Media, Each of the two cultures were mixed by vortex and allowed to equilibrate at the two defined temperatures. Once the components were equilibrated at the defined temperatures, the cultures were transferred by pipette to the poly( NIPAAm) membrane spots on each silicon substrate. One spot was left dry on each ch ip for comparison purposes. Each sample was hybridized for five minutes and then rinsed in their respectiv e rinse water tubes. Each sample was submerged into the tube for ten seconds and then removed for ten seconds. The water velocity, relative to the sample, was 0.56 cm/s, (Re = 1,288, which is laminar
flow). Following the rinse, the samples were given a quick shake to remove the excess surface moisture. Images presented in Figure 4.10 were co llected at a magnification of 250x using an epifluorescence microscope with a FITC filter cube. Figure 4.10: E. coli w/GFP on the Membrane Area of a Silicon Substrate In Figure 4.10A, the background of the si licon substrate was hybridized for 5 minutes and rinsed at 22C. In Figure 4.10B, the poly(NIPAAm) membrane area was hybridized for 5 minutes and rinsed at 22 C. In Figure 4.10C, the background of the silicon substrate was hybridized for 5 minutes and rinsed at 38C. In Figure 4.10D, the poly(NIPAAm) membrane area was hybridized for 5 minutes and rinsed. The results demonstrate that the E. coli was attached to the poly(NIPAAm) membrane at 38 o C. 59
60 4.5. Flow Velocity The Flow Velocity experiment was carried out in order to determine if water velocity has an impact on whether or not the E. coli cells, (w/GFP), are washed off the membrane. The temperature was held constant All samples were kept at a temperature of 38C, placed in a hybridization oven a nd allowed to equilibrate for 10 minutes. Approximately 200 l of resuspended E. coli, (w/GFP), was placed on each spot on the samples. The samples were allowed to hybridize for another 5 minutes. Once hybridization was complete, the samples were individually lowered in to RO water, (at 38C), for the times presented in Table 4.1. Af ter the soak time the samples were pulled up in the water for the same amount of time. These, down then up, cycles were repeated fives times for each slide at 38C. Table 4.1: Data from the Dip Experiment Slide Number Dip time (seconds) Total time in water (sec) Estimated Velocity (cm/s) Reynolds Number Estimated cells remaining (%) 1 10 100 0.56 cm/s 1,288 100 2 5 50 1.12 cm/s 2,576 10 3 2 20 2.8 cm/s 4,293 3 4 1 10 3.73 cm/s 8,587 <1 After each sample was dipped, they were allowed to dry in the hybridization oven for 5 minutes and subjected to image analysis. The images are presented in Figure 4.11.
Figure 4.11: % E. coli Cells Remaining for Different Reynolds Numbers Based upon the images displayed in Figur e 4.11, it is apparent that the water velocity produced an effect on the ability of the membrane to hold the cells. As the values of the Reynolds number increased the percentage adhesion of the bacteria decreased dramatically. Figure 4.12 presents a plot of the velocity vs. cells remaining. Figure 4.12: Plot of Flow Velocity versus % Cells Remaining on Samples 61
It is obvious that there is a rapid sloughing effect as the flow velocity increases from 0.56 cm/s. This velocity represents a Reynolds Number, (Re), of 1,288, which is well into the laminar range. Based upon this pr eliminary data, it appear s that the cells are able to stay attached to the membrane for laminar flow conditions. However, the situation reverses for turbulen t flow conditions. The next stage of the experiment should be performed with flow conditions corre sponding to Reynolds numbers of 1000 and 2000. Results from this experiment should reveal whether or not the ve locity of the flow or the laminar/turbulent condi tions determines cell adhesion. 4.6. Microfluidic Channel Testing The PDMS microfluidic channels were a ttached to a microscope cover slip and observed through the cover slip. The bacteria, E. coli w/ green fluorescent protein, (GFP), was introduced at the edge of the mi crofluidic channel and allowed to move through the microfluidic cha nnels by capillary forces. Figure 4.13 displays the microfluidic channels present on th e other side of the cover slip. 62 Figure 4.13: FITC Image of the E. coli Flowing Inside the PDMS Channels
By adjusting the focal plane, the E. coli w/ green fluorescent protein, (GFP), was visible when observed through the FITC filter cube. The E. coli bacteria are clearly in the channels. The SU-8 photoresist fabricated microfluidic channels were also tested using the experimental setup presented in Figure 4.14. Figure 4.14: Experimental Setup: SU-8 Photoresist Microfluidic Channel Testing The syringe pump, (Harvard Apparatus), was programmed to generate an input flow from 10 l/min to 600 l/min. Fluorescent polystyrene beads, (PolyScience PA), of 6 m nominal size were added to the DI wate r solution with a final concentration of 1x 10 6 per ml as visual targets to be observed through the microscope FITC filter cube. The volume flow rate relationship fo r a given geometry is given by: AvF in m 3 per sec (4.1) 63 where F is the volume flow rate, A is the cross-sectional area of the flow channel in m 2 and v is the average velocity of the flui d in m/s. A volume flow rate of 10 l/min,
(1x10 -8 m 3 /min), produced an average velocity, v for the DUT of 0.013 m/s. To calculate the Reynolds number of the flow, which is re quired to determine the relative turbulence of the flow stream, equation 4.2 was applied: vL Re (4.2) where is the fluids density, v is the fluids velocity, L is the channel length and is the fluids viscosity. As expected, the Reynolds number in the microfluidic channels was very low, which is indicative of a lami nar flow behavior. The low Reynolds number value, (0.2), essentially elim inates the presence of any turbulence. In the absence of turbulence certain effects such as mixing are very hard to achieve. These results demonstrate the necessity of implementi ng a strategy for sample and poly(NIPAAm) interaction in the presence of laminar flow. Figure 4.15 presents an image of the flow at a rate of 10 l/min. Figure 4.15: SU-8 Photoresist Microfluidic Channel Faced with these results it was conclude d that the use of the SU-8 photoresist channels was appropriate for the final device. 64
4.7. Device Testing The BioMEMS device was fully tested to verify its functionality. The experimental setup is presented in Figure 4.16. Figure 4.16: Experimental Setup for the Device Test An inverted fluorescent microscope, (LEICA DMI400B), with a CHROMA 31001 cube was used to capture video and still images. The flow, inside the microfluidic channel, was controlled by a Harvard Apparatus PHD 2000 syringe pump, operating in infusion mode. The microscope and sy ringe pump are pictured in Figure 4.17. Figure 4.17: Inverted Fluorescent Microscope and Syringe Pump 65
The Device-Under-Test, (DUT), was inst alled inside a petri dish, which was bonded and leveled using an epoxy layer. The DUT is pictured in Figure 4.18. Figure 4.18: DUT The capillary tubing was attached to the different syringes using a plastic connector which was attached to the capilla ry tubing. The DEP signal was connected to the gold pads in the device. The DUT contains: the microfluidic channel, the DEP microelectrode array loca ted at the top of the channel, the poly(NIPAAm) trenches located on the bottom substrate, the inlet and outlet ports, the capillary tubing, which was at tached to the inlet port, the syringe pump. The video and still images were processe d with IMAGE Pro Plus software, which was installed in an INTEL P4, 1G-RAM, PC. The AC signal for the DEP microelectrode array was generated using a SONY/Te ktronix AFG320 Function Generator. 66
Figure 4.19 presents an image of the top sec tion of the microfluidic channel. The microfluidic channel inlet port is on the left side of the image. The DEP microelectrode array is inside the SU-8 microfluidic channel. Figure 4.19: Top Section of the Microfluidic Channel Figure 4.20 presents an image of the bottom part of the microfluidic channel. The microfluidic channel outlet port is on the right side of the image. Figure 4.20: Bottom Section of the Microfluidic Channel 67 The inlet and outlet ports were drilled in the glass using a diamond coated drill. After drilling, the DEP microelectrode and the poly(NIPAAM) trenches were fabricated next to each port to avoid dama ge caused by the drilling process.
An inverted fluorescent microscopes imag e, of the DUT, is presented in Figure 4.21. The DEP array is on top. The trenches can be seen located at the bottom part of the channel. Figure 4.21: BioMEMS Device In tegration Device Top View The experiment was initiated by programm ing the syringe pump to generate an input flow of 10 l/min. Fluorescent polystyrene beads, (Polyscience PA), of 6 m and 20 m nominal size were added to the DI water solution. The final concentration was set at 1 x 10 6 per ml at 60 o C. At this temperature and flow values, the trenches were open. However, the particle density attached to the trenches was almost zero. This was expected due to the laminar flow region inside the microfluidic channel. A negative DEP force was applied perpendicular to the flow in order to defl ect the particles down to the trenches. Figure 4.22 presents a sequence of images of the particles path. 68
Figure 4.22: Sequence of Images of the Capture Mechanism In Figure 4.22A and Figure 4.22B, the flow contains 6 m and 20 m particles at 60 o C. The attached particle density is zer o. In Figure 4.22C, DEP force is applied and the particles start to be deflected to the tren ches, which increases their density. In Figure 4.22D, Figure 4.22E and Figure 4.22F the capture process is presented. During the capture process the 6 m particle density in the tren ches is increased and the 20 m particles are swept away. In Figure 4.22F just one of the 20 m bead remains attached to the top of the trenches. As designed, the sorting mechanism was confirmed by the presence of the 6 m beads, which are displayed inside the trenches. Previous work was performed related to DEP The E. coli bacteria were tested using another microelectrode array to verify the DEP effect on them. Figure 4.23 displays the E. coli aligned to the microelectrode array when the DEP was on and scattered six seconds after the DEP was turned off. 69
Figure 4.23: The Effect of DEP on E. coli DEP was applied while changing three main parameters: Flow rate, Frequency, AC Amplitude. Flow rate was varied from 5 l/min to 20 l/min. The frequency was varied from 1 to 50 KHz. The amplitude was varied from 1 to 10 V pp The image sequences were analyzed using a particle c ounting routine, from MATLAB, in order to calculate the particle density inside the trenches. Figure 4.24 presents the effect of the DEP on the trench density with a flow rate of 10 l per min, a frequency of 50 KHz and AC amplitude of 8 V pp The data was taken every 3s along the x axis. 70
Figure 4.24: DEP with 10 l per min, 50 KHz, 8 V pp Figure 4.25 presents the effect of the DEP on the trench density with a flow rate of 10 l per min, a frequency of 50 KHz and AC amplitude of 2 V pp The data was taken every 3s along the x axis. Figure 4.25: DEP with 10 l per min, 50 KHz, 2 V pp The plots presented in Figures 4.25 and 4.25 clearly illustrate the increase in density due to the use of DEP. The experiment demonstrated that the DEP effect on the particles was as proposed. The experiment also demonstrat ed that the DEP and the flow value are correlated with the particle density. Although not demonstrated in the Figures of 4.24 and 4.25, the highest particle concentrat ion areas coincided with the highest E Field 71
produced by the microelectrode array. Th is result corresponded to the simulation performed, which was presented in Chapter 3. The next stage of the experiment consis ted of injecting cold water into the microfluidic channel. Figure 4.26 presents the image sequence of the particles inside the trenches. Figure 4.26: Particles afte r Cold Water Injection In Figure 4.26A, the particle s clearly occupy a larger ar ea inside the trench. In Figure4.26B, the flow temperature starts to close the trenches and align the particles while reducing the space. In Figure 4.26C, the trenches are closed and the particles are maintained inside. The sequence of images, presented in Figure 4.26, clearly demonstrates the capture mechanism. The final experimental stage was focused on the release mechanism. Hot water, at 40 o C, was injected into the channel, which caused the trenches to open. The particles demonstrated the same sticking effect described in Section 4.3, which did not allow the particles to be released fr om the trenches. As proposed, DEP was applied and the frequency range was changed from 10-60 KHz to 10-20 MHz, which shifted the ClausiusMossotti factor. Application of a positive DEP combined with an increase in 72
the flow value to 20 l/min generated particle displa cement inside the trenches and started the release mechanism. Figure 4.27 presents the image seque nce of the release mechanism. Figure 4.27: Particles after Hot Water Injection In Figure 4.27A the particles are inside and attached to the trenches. In Figure 4.27B, Figure 4.27C and Figure 4.27D, the positive DEP force effect on the particles is demonstrated. The particles start to be released, Figure 4.27C, and the particle density inside the trenches is decreased, Figure 4.27D. 73
Figure 4.28 illustrates the decreasing part icle density due to the positive DEP force. As in the negative DEP case, the experiment demonstrated that the DEP and the flow value are correlated with the particle density. Figure 4.28: DEP with 20 l per min, 15 KHz, 10 V pp The experiment proved the devices f unctionality was in accordance with the proposed mechanisms. The release mechanism is more dependant on the flow value and the voltage peak-to-peak amplitude than the attach mechanism. The release mechanism takes less time. 74
75 CHAPTER 5 CONCLUSIONS A cell sorting integrated BioMEMS devi ce based on the use of Intelligent thermally responsive polymers ha s been described, developed and fabricated. This novel sorting and manipulation strate gy hasnt been previously exploited or reported. Therefore, it can be considered to be an important contribution to the BioMEMS field. Compared to the alternate existing methods the proposed approach accomplishes the desired characteristics of portability, low comple xity, cost effectiveness, power efficiency and ease of operation. As illust rated in this dissertation, th e complete integration of different technologies to achieve the desired results has displayed promising results related to this technology platform. This dissertation has described differe nt procedures for the prototyping and fabrication of responsive poly(NIPAAm) hydr ogel topographies usi ng soft-lithography, e-beam lithography and soft-lithography combined with focused ion beam patterning. Soft-lithography, together with the use of innovative proce ss materials such as SU-8 photoresist and PDMS, presente d the capability of obtaining high aspect ratio polymer structures. Soft-lithography ha s also proven to provide an excellent approach since its In-situ polymerization capability provides significant flexibility to the hydrogels property tuning. Although e-beam lithography doesnt offer such flexibility, it can also
76 be considered as an alternate patterning solu tion in order to achieve submicron scale size ranges. The behavior of the poly(NIPAAm) pol ymer as a negative e-beam resist unveils the possibility of using this technique as a direct patterning procedure for hydrogel based structures. Future experiments will account for relevant factors such as film thickness, ebeam dose, process step temperatures and othe r issues that need to be optimized for high resolution nano scale range pattern genera tion. The latter ap proach, through the combined use of the soft-lithography and FIB patterning also demonstrated a nanometer structure size range, which exploits the advantages of soft-lithography. The AFM structural and functional analysis and resu lts have demonstrated that the hydrogels thermal response is maintained after the FI B exposure. In addition, the AFM results demonstrated that the thermal response was ma intained while reducing the patterning size and increasing the BioMEMS devi ce interaction range. The increase in the interaction range extends the usefulness of the device to different Biosamples su ch as Bacteria and viruses, provides a higher sensitivity a nd a reduction of the device response time. In the micron size range, the thermally responsive polymer trenches exhibited the desired open and close behaviour due to temperature changes. Fluorescent microscopy results demonstrated the polymers thermal re sponse consistency for temperature changes ranging from above to below its LCST. The opportunity to create smart trenches to discriminate and capture targets for simple and quick sample processing could be a key technology platform. The existence of such a platform facilitate s separation processes that would be difficult to achieve by th e other methods mentioned in Chapter 1.
77 A major challenge related to the propos ed approach was the particle-polymer layer release mechanism. Even though th e capture mechanism was easily assured, the experiments conducted with the E. coli bacteria and the microbeads displayed some attractive sticking to the polymer structures in laminar flow regimes. This activity was probably due to Van der Waals force effect. While some specific applications wont be affected, the use of a release mechanism based on the reverse dielec trophoresis effect was proposed to overcome these results. The re sults obtained proved that the proposed approach was correct and the re lease mechanism was achieved. With respect to the therma l actuation mechanism, the proposed nichrome resistor approach proved to be the best suited for the integration of the final device. The fabrication process, for the nichrome resistor, illustrated that it was highly compatible with the rest of the device fabrication processes. Nichrome resistors have been widely studied and reported in the literature. Future work related to the actuators effectiveness and its processing optimization must be carri ed out with the objective of reducing the device power consumption and thermal efficiency. The microfluidics design and fabrication was studied and analyzed to define the proper processing integration procedure. The SU-8 photores ist patterned layer between the two device substrates demonstrated an optimal performance that included a flow range of hundreds of microliters per minute, which make it appropriate for the final applications. Results confirmed the existence of a laminar flow region, which was expected for these types of microfluidics ch annels. The existence of a laminar flow region simplifies the selection process for a sample deflecting mechanism in order to obtain particle-polymer layer interaction.
78 The experiments performed to test the E coli attachment-release mechanism to the polymer structures, due to the flow veloci ty, unveiled the possibi lity of developing a shear controlled sorting stra tegy as a function of particle adherence. Future work in this area is proposed to define new polymer layer topographies to expl oit this principle. The dielectrophoresis force effect in the device was successfully tested and proved to be adequate, as expected. The particle deflection increased the particles attachment density within the polymer tren ches, which improved the sorting capacity of the device. The combination of DEP and a thermally responsive polymer in microfluidics for sorting or particle manipula tion has not been reporte d in the literature. The polymers biocompatibility, when applied for cell manipulation, overcomes the DEP major drawback of cell damage due to exposure to a large electric field. Future work is proposed for the design and analysis of new DEP microelectrode array geometries in order to enhance the deflection of the particle to the polymer structures. Current BioMEMS-based assays suffer fr om low detection resolution when the target is either in low concentration or is embedded in a complex or dirty matrix. A potential application for the strategy deve loped is a preconcentration process that enhances sample purity in microfluidic plat forms with low power requirements. Another possible application is a multi-zone filteri ng device when the separation and manipulation of mixed multiple bio-species is required for diagnostics, sensing and detection. The integration of different micro and nano fabrication processes supports the development of the strategy presented in th is dissertation, which represents a major contribution to the field. Furt her research work is proposed to optimize the microfluidics subsystem, the tuning of poly(NIPAAM) pr operties and the different poly(NIPAAm)
79 patterning processes, their effects on the BioMEMS/NEMS device. In addition, the fabrication and integration of the DEP microelectrode arra y and the thermal actuation should be studied with respect to optimization.
80 REFERENCES  Bashir, R., "BioMEMS: State-of-the-A rt in Detection, Opportunities and Prospects", Advanced Drug Deliver y Reviews, vol. 56(11), pp 1565-1586, 2004.  Mastrangelo, C. H. and M. A. Burns, et al., "Microfabricated Devices for Genetic Diagnostics", Proceedings of the IEEE, vol. 86(8), pp 1769-1787, 1998.  Dong, L.C. and A. S. Hoffman, "Synthe sis and Application of Thermally Reversible Heterogels for Drug Delivery", Journal of Controlled Release, vol. 13(1), pp 21-31, 1990.  Hoffman, A. S. and A. Afrassiabi, "R eversible Swelling and Deswelling of Hydrogels", Abstracts of Papers of th e American Chemical Society, vol. 191, pp 204-POLY, 1986.  Falconnet, D. and G. Csucs, et al., "Surface Engineering Approaches to Micropattern Surfaces for Cell-Based Assays", Biomaterials, vol. 27(16), pp 3044-3063, 2006.  Lee, G. B. and S. H. Chen, et al., "Microfabricated Plastic Chips by Hot Embossing Methods and their Applicatio ns for DNA Separation and Detection", Sensors and Actuators B-Chem ical, vol. 75(1-2), pp 142-148, 2001.  Gambari, R. and M. Borgatti, et al., "App lications to Cancer Research of Devices based on Delectrophoresis, (DEP) ", Technology in Cancer Research & Treatment. Vol. 2(1), pp 31-39, 2003.  Krebs, M. D. and A. M. Zapata, et al ., "Detection of Biological and Chemical Agents using Differential Mobility Spectrometry, (DMS), Technology", IEEE Sensors Journal, vol. 5(4), pp 696-703, 2005.  Liao, C. S. and G. B. Lee, et al., "M iniature RT-PCR System for Diagnosis of RNA-Based Viruses", Nucleic Acids Research, vol. 33(18), 2005.  Lin, C. H. and G. B. Lee, et al., "Vertical Focusing Ddevice Utilizing a Dielectrophoretic Force and its Application on Microflo w Cytometry", Journal of Microelectromechanical Syst ems, vol. 13(6), pp 923-932, 2004.
81  Yang, J. and D. Y. Kwok, "Analytical Tr eatment of Electrokinetic Microfluidics in Hydrophobic Microchannels", Analytica Chimica Acta, vol. 507(1), pp 39-53, 2004.  Yang, S. Y. and S. K. Hsiung, et al ., "A Cell Ccounting/Ssorting System Incorporated with a Micr ofabricated Flow Cytometer Chip", Measurement Science & Technology, vol. 17(7), 2006.  Ashkin, A., "Forces of a Single-Beam Gradie nt Laser Trap on a Dielectric Sphere in the Ray Optics Regime", Biophysi cal Journal, vol. 61(2), pp 569-582, 1992.  Yin, X. and A. S. Hoffman, et al., "Poly(N-isopropylacrylamid e-co-propylacrylic acid) Copolymers that Respond Sharply to Temperature and pH", Biomacromolecules, vol. 7(5), pp 1381-1385, 2006.  Ziaie, B., A. and Baldi, et al., "Har d and Soft Micromachining for BioMEMS: Review of Techniques and Examples of Applications in Microfluidics and Drug Delivery", Advanced Drug Delivery Reviews, vol. 56(2), pp 145-172, 2004.  Huang, W. H. and W. Cheng, et al., "T ransport, Location and Quantal Release Monitoring of Single Cells on a Microfluid ic Device", Analytical Chemistry, vol. 76(2), pp 483-488, 2004.  Choi, S. and J. K. Park, "Microfluidic System for Dielectrophoretic Separation based on a Trapezoidal Electrode Arra y", Lab on a Chip, vol. 5(10), pp 11611167, 2005.  Papagiakoumou, E., D. and Pi etreanu, et al., "Evaluati on of Trapping Efficiency of Optical Tweezers by Dielectrophoresis", Journal of Biomedical Optics, vol. 11(1), 2006.  Arai, F. and H. Maruyama, et al,. "Pinpoi nt Injection of Microtools for Minimally Invasive Micromanipulation of Mi crobes by a Laser Trap", Ieee-Asme Transactions on Mechatronics, Vol, 8(1), pp 3-9, 2003.  Numata, T. and A. Takayanagi, et al., "Manipulation of Metal Nanoparticles using Fiber-Optic Laser Tweezers with a Microspherical Fo cusing Lens", Japanese Journal of Applied Physics, Part 1Regular Papers Brie f Communications & Review Papers, vol, 45(1A), pp 359-363, 2006.  Grayson, A.C.R. and R. S. Shawgo, et al., "A BioMEMS Review: MEMS Technology for Physiologically Integrated Devices", Proceedings of the IEEE, vol. 92(1), pp 6-2, 2004.
82  Mariella, R. (2002). "MEMS for bio-assays." Biomedical Microdevices 4(2): 7787.  Wang, Z. and J. El-Ali, et al., "Measurem ents of Scattered Light on a Microchip Fow Cytometer with Integrated Polymer based Optical Elements", Lab on a Chip, Vol. 4(4), pp 372-377, 2004.  Byun, I. and W. Kim, et al.,"A Micr o Flow Cell Cytometry based on MEMS Technologies using Silicon and Optical Fibers ", Journal of Materials Science, vol. 38(22), pp 4603-4605, 2003.  W. K. Wu, C. K. Liang and J. Z. Huang, MEMS-Based Flow Cytometry: Microfluidics-Based Cell Identificati on System by Fluorescent Imaging, Proceedings of the 26th Annual Interna tional Conference of the IEEE-EMBS San Francisco, CA, USA, September, 2004.  W. H. Grover, A. M. Skelley, C. N. Liu, E. T. Lagally and A. Mathies, Practical Valves and Pumps for Large-Scale Inte gration into Microfluidic Analysis Devices, Micro Total Analysis Systems and Proceedings of the TAS Symposium, 2002.  Yao, B. and Luo, et al., "A Microfluidic Device based on Gravity and Electric Forces Driving for Flow Cytometry and Fuorescence Activated Cell Sorting", Lab on a Chip, vol. 4(6), pp 603-607, 2004.  Koch, M., Evans, A. G. R. and Brunnschwe iler, A., Design a nd Fabrication of a Micromachined Coulter Ccounter, Journal of Micromechanics and Microengineering, vol. 9, pp. 159-161, 1999.  Tang, H. and Y. F. Gao, "An Impedance Microsensor with Coplanar Electrodes and Vertical Sensing Apertures", Ieee Sensors Journal, vol. 5(6), pp 1346-1352, 2005.  Svoboda, K. and S. M. Block, "Biological Applications of Optical Forces." Annual Review of Biophysics and Biom olecular Structure, vol. 23, pp 247-285, 1994.  Block, S. M. and L. S. B. Goldstein, et al., "Bead Moveme nt by Single Kinesin Molecules Studied with Optical Tw eezers", Nature, vol. 348(6299), pp 348-352, 1990.  Wang, M. D. and H. Yin, et al., "S tretching DNA with Optical Tweezers", Biophysical Journal, vo l. 72(3), pp 1335-1346, 1997.
83  Konig, K., Y. and Tadir, et al., "Effects of Ultraviolet Exposure and Near Infrared Laser Tweezers on Human Spermatozoa", Human Reproduction, vol. 11(10), pp 2162-2164, 1996.  Loken, M. R. and L. A. Herzenberg, "Analysis of Cell-Populations with a Fluorescence-Activated Cell Sorter", A nnals of the New York Academy of Sciences, vol. 254, pp 163-171, 1975.  Fu, A. Y. and C. Spence, et al., "A Mi crofabricated Fluorescence-Activated Cell Sorter", Nature Biotech nology, vol. 17(11), pp 1109-1111, 1999.  Ho, C. T. and R. Z. Lin, et al., "Micro machined Electrochemical T-Switches for Cell Sorting Applications", La b on a Chip, vol. 5(11), pp 1248-1258, 2005.  Suzuki, H., C. and M. Ho, et al., "A Chaotic Mixer for Magnetic Bead-Based Micro Cell Sorter", Journal of Microele ctromechanical Systems, vol. 13(5), pp 779-790, 2004.  Wu, X. S. and A. S. Hoffman, et al ., "Synthesis and Characterization of Thermally Reversible Macroporous Poly (N-Isopropylacrylamide) Hydrogels", Journal of Polymer Science Part aPolymer Chemistry, vol. 30(10), pp 2121-2129, 1992.  Yoshida, R., K. Uchida, Y. Kaneko, K. Sakai, A. Kikuchi, Y. Sakurai and T. Okano, Comb-Type Grafted Hydrogels with Rapid De-Swelling Response to Temperature-Changes, Na ture, vol. 374(6519), pp 240-242, 1995.  Yamada N., T. Okano, H. Sakai, F. Ka rikusa, Y. Sawasaki, and Y. Sakurai, Thermoresponsive Polymeric Surfaces Control of Attachment and Detachment of Cultured-Cells, Makromolekulare Chemie-Rapid Communications, vol. 11(11), pp 571-576, 1990.  Peppas, Hydrogels in Medicine and Pharmacy, CRC Press, Boca Raton FL., 1986.  Chilkoti, A., M.R. Dreher, D.E. Meyer and D. Raucher, Targeted Drug Delivery by Thermally Responsive Polymers, Advanced Drug Delivery Reviews, vol. 54(5), pp 613-630, 2002.  Schild, H.G., Poly(N-Isopropylacrylamide) Experiment; Theory and Application, Progress in Polymer Science, vol. 17(2), pp 163-249, 1992.
84  Erbil, C., E. Kazancioglu, et al. (2004). "Synthesis, Characterization and Thermoreversible Behaviours of Poly(dimethyl siloxane)/P oly(N-isopropyl acrylamide) Semi-Interpenetrating Networ ks." European Polymer Journal 40(6): 1145-1154.  Suzuki, A. and T. Tanaka (1990). "Phase-Transition in Polymer Gels Induced by Visible-Light." Nature 346(6282): 345-347.  Suzuki, A., Phase Transition in Gels of Submillimeter Size Induced by Iteraction with Stimuli, Advances in Polymer Science, vol. 110, pp 199-240, 1993.  Ding, Z. L. and G. H. Chen, et al., "Unusual Properties of Thermally Sensitive Oligomer-Enzyme Conjugates of Poly(N-i sopropylacrylamide)-T rypsin", Journal of Biomedical Materials Res earch, vol. 39(3), pp 498-505, 1998.  Sershen, S. R. and S. L. Westcott, et al., "Temperature-Sensitive PolymerNanoshell Composites for Photothermally Modulated Drug Delivery", Journal of Biomedical Materials Research, vol. 51(3), pp 293-298, 2000.  Ramkissoon-Ganorkar and C., F. Liu, et al ., "Modulating Insulin -Release Profile from pH Thermosensitive Polymeric Bead s through Polymer Molecular Weight", Journal of Controlled Release, vol. 59(3), pp 287-298, 1999.  Toomey, R.., D. Freindank and J. Ruhe, Swelling Behavior of Thin, SurfaceAttached Polymer Networks, Macr omolecules, vol. 37(3), pp 882-887, 2004.  Ratner. B. and Bryant J., Biomaterials: Where We have Been and Where We are Going, Annu. Rev. Biomed Eng., vol. 6, pp 41-75, 2004.  Qiu, Y. and K. Park, "Environment-Se nsitive Hydrogels for Drug Delivery", Advanced Drug Delivery Reviews, vol. 53(3), pp 321-339, 2001.  Hu, Z. B. and Y. Y. Chen, et al., (). "Polymer Gels with Engineered Environmentally Responsive Surface Pa tterns", Nature, vol. 393(6681), pp 149152, 1998.  Kim, S. Y. and H. S. Shin, et al., "Pr operties of Electrores ponsive Poly(vinyl alcohol)/Poly(acrylic acid) IPN Hydrogels under an Electric Stimulus", Journal of Applied Polymer Science, vol. 73(9), pp 1675-1683, 1999.  Zrinyi, M., "Intelligent Polymer Gels Controlled by Magnetic Fields", Colloid and Polymer Science, vol. 278(2), pp 98-103, 2000.  Tanaka, T., Phase-Transition in Gels and a Single Polymer, Polymer, vol. 20(11), pp 1404, 1979.
85  Sun, Y. M. and T. L. Huang, "Pervaporation of Ethanol-Eater Mixtures through Temperature-Sensitive Poly(vinyl alc ohol-g-N-isopropyacrylamide) Membranes", Journal of Membrane Scie nce, vol. 110(2), pp 211-218, 1996.  Gutowska, A. and Y. H. Bae, et al., "Heparin Release from Thermosensitive Polymer-Coatings in-Vivo Studies", Jour nal of Biomedical Materials Research, vol. 29(7), pp 811-821, 1995.  Chen, G. H. and A. S. Hoffman (1995). "A New Temperature-Responsive and PhResponsive Copolymer for Possibl e Use in Protein Conjugation." Macromolecular Chemistry and Physics 196(4): 1251-1259.  Hawker, C. J. and E. Elce, et al., "Well-Defined Ra ndom Copolymers by a ''Living'' Free-Radical Polymerization Process", Macromolecules, vol. 29(7), pp 2686-2688, 1996.  Prior-Cabanillas and A., I. Quijada-Garrido, et al., "Influence of the Swelling History on the Swelling Kinetics of Stimuli-Responsive Poly [(Nisopropylacrylamide)-Co-(methacrylic acid) ] Hydrogels", Polymer, vol. 46(3), pp 685-693, 2005.  Kuckling, D., J. and Hoffmann, et al., "Photo Cross-Linkable Poly(Nisopropylacrylamide) Copolymers III: Micr o-Fabricated Temperature Responsive Hydrogels", Polymer, vol. 44(16), pp 4455-4462, 2003.  Brazel, C. S. and N. A. Peppas, "Synthesis and Characterization of Thermomechanically and Chemomechanically Responsive Poly(NIsopropylacrylamide-Co-Methacrylic Acid ) Hydrogels", Macromolecules, vol. 28(24), pp 8016-8020, 1995.  Gutowska, A. and Y. H. Bae, et al., "Thermosensitive Interpenetrating Polymer Networks Synthesis, Characterization, and Macromolecular Release", Macromolecules, vol. 27(15), 4167-4175, 1994.  Onuki, A., Theory of Phase Transition in Polymer Gels, Advances in Polymer Science, vol. 109. pp 63-121, 1993.  Hirotsu, S. and A. Onuki, Volume Phase Transition of Gels under Uniaxial Tension, Journal of the Physical Society of Japan, vol. 58(5), pp 1508-1511, 1989.  Sigma Aldrich Aldrichimica Acta vol. 38, No 3, 2005.  Vozzi, G., C. Flaim, et al. (2003). "Fab rication of PLGA Scaffolds Using Soft Lithography and Microsyringe Deposition." Biomaterials 24(14): 2533-2540.
86  Zhang, K. and X. Y. Wu (2004). "Tem perature and pH-Responsive Polymeric Composite Membranes for Controlled Delivery of Proteins and Peptides." Biomaterials 25(22): 5281-5291.  Kane, R. S., S. Takayama, et al. (1999). "Patterning Prot eins and Cells using Soft Lithography." Biomaterials 20(23-24): 2363-2376.  Chavez, K. L. and D. W. Hess, "Removal of Resist Materials using Acetic Acid", Journal of the Electrochemical Society, vol. 150(4), pp G284-G291, 2003.  Ehrfeld, W. and H. Lehr (1995). "Deep X-Ray-Lithography for the Production of 3-Dimensional Microstructures from Meta ls, Polymers and Ceramics." Radiation Physics and Chemistry 45(3): 349-365.  Carlier, J., S. and Arscott, et al., "Integrated Microfluidics based on MultiLayered SU-8 for Mass Spectrometry Anal ysis", Journal of Micromechanics and Microengineering, vol. 14(4), pp 619-624, 2004.  Bachmann, D., B. and Schoberle, et al., "Fabrication and Characterization of Folded SU-8 Suspensions for MEMS Applications", Sensors and Actuators APhysical, vol. 130, pp 379-386, 2006.  Kohlmeier, T., V. and Seidemann, et al ., "An Investigation on Technologies to Fabricate Microcoils for Miniaturiz ed Actuator Systems", Microsystem Technologies, Micro and Nanosystems Information Storage and Processing Systems, vol. 10(3), pp 175-181, 2004.  Lorenz, H., M. and Despont, et al., "H igh-Aspect-Ratio, Ultra-thick, NegativeTone Near-UV Photoresist and its A pplications for MEMS", Sensors and Actuators A-Physical, vol. 64(1), pp 33-39, 1998.  Conradie, E. H. and D. F. Moore, "S U-8 Thick Photoresist Processing as a Functional Material for MEMS Applica tions", Journal of Micromechanics and Microengineering, vol. 12(4), pp 368-374, 2002.  Loechel, B., "Thick-Layer Resists for Surface Micromachining", Journal of Micromechanics and Microengineering, vol. 10(2), pp 108-115, 2000.  L'Hostis, E., P. and E. Michel, et al., "M icroreactor and Electrochemical Detectors Fabricated using Si and EPON SU-8", Se nsors and Actuators, B-Chemical, vol. 64(1-3), pp 156-162, 2000.  Anhoj, T. A. and A. M. Jorgensen, et al., "The Effect of Soft Bake Temperature on the Polymerization of SU-8 Photores ist", Journal of Micromechanics and Microengineering, vol. 16(9), pp 1819-1824, 2006.
87  Liu, J., B. and Cai, et al., "Process Research of High Aspect Ratio Microstructures using SU-8 Resist", Microsystem Technologies, Micro and Nanosystems Information Storage and Processi ng Systems, vol. 10(4), pp 265-268, 2004.  Lee, K. Y. and N. LaBianca, et al., "Micromachining Applications of a High Resolution Ultra-Thick Photoresist", J ournal of Vacuum Science & Technology, vol. B 13(6), pp 3012-3016, 1995.  Mata, A., A. and J. Fleischman, et al. "Fabrication of Multi-Layer SU-8 Microstructures", Journal of Micromech anics and Microengineering, vol. 16(2), pp 276-284, 2006.  MicroChem SU8 products, http://www.mic rochem.com/products/su_eight.htm.  Senol Mutlu1, Cong Yu2, Frantisek Svec2, Carlos H. Mastrangelo3, Jean M.J. Frchet2 and Yogesh B. Gianchandani 1, A Thermally Responsive Polymer Microvalve without Mechanical Parts Photo-Patterned in a Parylene Channel, Transducers Conference Boston, pp 802-805, 2003.  J.Z. Hilt, A. Gupta, R. Bashir and N.A. Peppas, Ultra-Sensitive Biomems Sensors based on Microcantilevers Patte rned with Environmentally Responsive Hydrogels, Biomed. Microdevices, pp 177, September 2003.  M. G. Rosenfield, M. G. R. Thomson, P. J. Coane, K. T. Kwietniak, J. Keller, D. P. Klaus, R. P. Volant, C. R. Blair, K. S. Tremaine, T. H. Newman, and F. J. Hohn, "Electron-Beam Lithography for A dvanced Device Prototyping: Process Tool Metrology", J. Vac. Sci. Technol., vol. B11(6), pp 2615-2620, 1993.  J. C. Nabity and M. N. Wybourne, "A Versatile Pattern Generator for HighResolution Electron-Beam Lithography", Rev. Sci. Instrum., vol. 60(1), 1989.  M. Hatzakis, "PMMA Copolymers as High Se nsitivity Electron Resists", J. Vac. Sci. Technol., vol. 16(6), pp 1984, 1979.  McCord, M. A. and R. F. W. Peas e (1986). "Lithography with the Scanning Tunneling Microscope." Journal of V acuum Science & Technology B 4(1): 8688.  MicroCHEM Nano PMMA Product Manual.  J. HI Daniel, D. F. Moore and J. F. Walker, Focused Ion Beams for Micro Fabrication, Engineering Science and Education Journal, April 1998.
88  Brugger, J., Beljakovic, G., Despont, M., De Rooij, N. E, and Vettiger, Silicon Micro/Nanomechanical Device Fabricati on based on Focused Ion Beam Surface Modification and KOH Etching, Microelectron Eng., 1997.  Fiedler, S., S. G. Shirley, et al., "Diele ctrophoretic Sorting of Particles and Cells in a Microsystem", Analytical Chem istry, vol. 70(9), pp 1909-1915, 2002, P.  Wang, X. B. and J. Vykoukal, et al., "Sep aration of Polystyr ene Microbeads using Dielectrophoretic/Gravitati onal Field-Flow-Fractionati on." Biophysical Journal, vol. 74(5), pp 2689-2701, 1998.  Brown, A. P. and W. B. Betts, et al., "E valuation of a Dielect rophoretic Bacterial Counting Technique", Biosensors & Bi oelectronics, vol. 14(3), pp 341-351, 1999.  Yang, J. and Y. Huang, et al., "Cell Sepa ration on Microfabricated Electrodes using Dielectrophoretic/Gravitational Fiel d Flow Fractionation", Analytical Chemistry, vol. 71(5), pp 911-918, 1999.  ANSOFT: http://www.ansoft.com/products/hf/hfss/.  Yanwei Zhang, Yongxia Zhang and R. B. Marcus, Thermally Actuated Microprobes for a New Wafer Pr obe Card, IEEE Journal of Microelectromechanical systems, vol. 8, No. 1, MARCH 1999.  Ji-Hyuk Lim, Yong-Soo Lee, Hyung-Taek Lim, Seog-Soon Baek, Keon Kuk and Yong-Soo Oh, Visualization of Bubbles Generated by Micro Heaters with Various Current Density Distributions, IEEE Journal of Microelectromechanical Systems, 2003.  Kovacs, G.T.A., Micromachined Transducers Sourcebook, Mc. Graw Hill, pp 795, 1998.  M.A. Unger, H. Chou, T. Thorsen, A. Scherer and S.R. Quake, Monolithic Microfabricated Valves and Pumps by Mu ltilayer Soft Lithography, Science, vol. 288288, pp 113-116, 2000.  D.C. Duffy, J.C. McDonald, O.J.A. Sc hueller and G.M. Whitesides, Rapid Prototyping of Microfluidic Systems in Poly(dimethylsiloxane), Anal. Chem, vol. 70, pp.4974-4984, 1998.  Jo, B.-H., Van Lerberghe, L.M., Motsegood, K.M. and Beebe, D.J., Three Dimensional Micro-Channel Fabrication in PDMS Elastomer, J. MEMS, Vol. 9, pp 76-81, 2000.
89  D. J. Campbell, K. J. Beckman, C. E. Cald eron, P. W. Doolan, R. H. Moore, A. B. Ellis and G. C. Lisensky, Replicatio n and Compression of Bulk Surface Structures with Polydimethylsiloxane Elastomer, J. Chem. Educ. vol. 76(537), 1999.  F. Laermer and D. Schilp, Method of anis otropically Etching Silicon, U.S. Pat. 5 501 893.  J. K. Bhardwaj and H. Ashraf, Advan ced Silicon Etching using High Density Plasmas, Proc. SPIE-Micromachining/ Microfabrication Process Technology, vol. 2639, Austin, TX, pp. 224, 1995.  M. Khoo and C. Liu, Micro-Magnetic Sili cone Elastomer Membrane Actuator, Sensors and Actuators, vol. 89(3), 2001.  Jackman, R. J. and T. M. Floyd, et al., "Microfluidic Systems with on-line UV Detection Fabricated in Photodefinable Epoxy", Journal of Micromechanics and Microengineering, vol. 11(3), pp 263-269, 2001.
ABOUT THE AUTHOR Alexandro Castellanos Mier received a Bachelors Degree in Mechanic-Electric Engineering from Universidad Veracruzana, Mexico in 1991. He received a M.S. in Control Engineering from Instituto Tecnologico y de Estudios Superiores de Monterrey Campus Monterrey in 1997. Alexandro started teaching electrical engineering before he entered the Masters program and continued on as a full time professor. Alexandro rose to the position of Chairman for the Elect ronics and Telecommunications program at Universidad Veracruzana Facultad de Ingenier ia where he presided until he entered the Ph.D. program at the Univers ity of South Florida in 2003. While in the Ph.D. program at the Un iversity of South Florida, Alexandro functioned for three years as the integrated circuit pr ocessing laboratory teaching assistant in the electrical engineering department. Alex andro coauthored publications related to this dissertation. In addition, he submitted papers and participated at international conferences of the Ibero-Am erican science and technology consortium, (ISTEC), and the IEEE. Alexandro also particip ated as an invited l ecturer at different Universities in Mexico.