USF Libraries
USF Digital Collections

Microsphere-aided characterization of stimuli-responsive polymer networks

MISSING IMAGE

Material Information

Title:
Microsphere-aided characterization of stimuli-responsive polymer networks
Physical Description:
Book
Language:
English
Creator:
Bello, Carlos A
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Soft lithography
Confocal microscopy
Microstructures
Poly(NIPAAm)
Delaunay triangulation
Dissertations, Academic -- Chemical & Biomedical Engineering -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: The fabrication and characterization of surface-anchored hydrogel microstructures are described. The hydrogel structures are constructed from poly(N-isopropylacrylamide), or poly(NIPAAm), which is a well-known thermoresponsive polymer that swells and contracts with changes in temperature. When patterned on a surface, these structures can experience a variety of shape changes induced by nonuniform swelling. Depending on the aspect ratio, patterns can, for instance buckle upon swelling and form wave-like patterns. Such structural changes replicate oscillatory motion of the smooth muscle cells and can be used to transport objects in microfluidics. The work, herein, investigates methods of pattern production and introduces a new technique for characterizing local swelling in the patterns. In order to achieve the latter, fluorescent microspheres were embedded in hydrogel patterns and their positions were mapped in three-dimensions using confocal microscopy. The measurements permit, for the first time, swelling maps of the structures based on relative movements of the microspheres. This information will ultimately aid in understanding how swollen macroscopic structures are related to gradients in localized swelling.
Thesis:
Thesis (M.S.Ch.)--University of South Florida, 2008.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Carlos A. Bello.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 89 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002007183
oclc - 401725965
usfldc doi - E14-SFE0002788
usfldc handle - e14.2788
System ID:
SFS0027105:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Microsphere-Aided Characterization of Stimuli-Responsive Polymer Networks by Carlos A. Bello A thesis submitted in partial fulfillment of the requirement s for the degree of Master of Science in Chemical Engineering Department of Chemical & Biomedical Engineering College of Engineering University of South Florida Major Professor: Ryan G. Toomey, Ph.D. John T. Wolan, Ph.D. Martin Muschol, Ph.D. Date of Approval: November 5, 2008 Keywords: soft lithography, confocal micro scopy, microstructures, poly(NIPAAm), Delaunay triangulation Copyright 2008, Carlos A. Bello

PAGE 2

DEDICATION To my mom

PAGE 3

ACKNOWLEDGEMENTS First, I wish to sincerely thank my Pr ofessor, Dr. Toomey for guiding me throughout my research and providing valuable insight over the past three years. It is because of Dr. Toomeys advice and support that my graduate career has been rewarding. I will always be grateful for the time he t ook to enhance my knowledge and understanding on a wide variety of topics, which have all been vital to the success of my graduate studies. I am also grateful to Samuel DuPont for his continued support, as well as Dr Wangs group, who manufactured the SU-8 molds. Thanks to my committ ee members, Dr. Wol an and Dr. Muschol, who have played an integral part in my gr aduate education. A special thanks to my lab partners who have played an integr al part in my graduate education: Mr. Ajay Vidyasagar, Ms. Leena Patra, Ms. Ophir Ortiz, Ms. C hamila Siyambalapitiya Mr. Ryan Cates, Ms. Gulnur Efe and Mr. Abiola Shitta.

PAGE 4

i TABLE OF CONTENTS LIST OF TABLES.................................................................................................iii LIST OF FI GURES...............................................................................................iv ABSTRACT .........................................................................................................vii CHAPTER ONE: INTRODUCT ION......................................................................1 CHAPTER TWO: FABRICATION OF POLY(NIPAAM) MICROSTRUCT URES......................................................................6 2.1. Photolithography Process.............................................................10 2.1.1. Piranha Cl ean...............................................................10 2.1.2. Dehydratio n Bake .........................................................11 2.1.3. Coat ..............................................................................12 2.1.4. Soft Ba ke......................................................................13 2.1.5. Exposur e.......................................................................15 2.1.6. Post Exposure Bake ( PEB)........................................... 17 2.1.7. Developm ent.................................................................17 2.2. Silicon-PDMS Molding..................................................................18 2.3. Silanization of Gl ass Substrates....................................................21 2.4. Photopolymerization of Hydrogel with Embedded Microspher es.................................................................................22 2.5. Fabrication of Fl uidic Chamber .....................................................25 2.6. Remarks........................................................................................26 CHAPTER THREE: CHARACTERIZATION OF SWELLING OF HYDROGELS BY SPINNING DISK CONFOCAL MICROSCOP Y................................................................................36 3.1. Introducti on...................................................................................36 3.2. Confocal Microsc opy Introduction.................................................37 3.2.1. Advantages of Confocal Micro scopy.............................37 3.2.2. Disadvantages of Confocal Microscopy........................38 3.3. Spinning Disk Confocal Mi croscopy In troduction..........................39 3.3.1. Mono Scanning Confocal Microscopy Introduction.......41 3.3.1.1. Operati on of the Dual-Spinning Disk Confocal Microscope ...........................................................42

PAGE 5

ii 3.3.1.2. Advantages of the Dual-Spinning Disk Confocal Microscope ...........................................................44 3.4. Image Processing and 3D R endering of Stacks............................45 3.4.1. Three-dimensional Image Proce ssing...........................45 3.4.2. Surface and Volume Renderin g....................................46 3.4.3. 3D and 4D Animation of Confocal Images....................47 3.4.4. Fluorescence Images of Fluorescent Microspheres......48 3.5. Image Processing Routines to Locate Particle Positions in Three Dimens ions .........................................................................49 CHAPTER FOUR: RESULTS, DISC USSION, AND CO NCLUSIONS................62 4.1. Soft Lithography Delimitati ons and Approaches...........................62 4.2. 3D Reconstr uction .........................................................................64 4.2.1. Fluidic Chamber Considerat ions...................................64 4.3. Image Regist ration ........................................................................64 4.4. Pixel Reso lution............................................................................65 4.5. Image Processing of the St acks....................................................66 4.6. Delaunay Triangulatio ns Resu lts...................................................66 4.7. Measuring Local Swelling.............................................................67 4.8. Conclusi ons..................................................................................68 REFERENC ES...................................................................................................84

PAGE 6

iii LIST OF TABLES Table 2-1. Recommended soft bake param eters...........................................14

PAGE 7

iv LIST OF FIGURES Figure 1-1. Swelling behavior: unconstrained gels vs. surface attached gels.................................................................................................4 Figure 1-2. Microstructures of poly(NIPAAm) below and above LCST..............5 Figure 2-1. SU-8 microstruc ture......................................................................27 Figure 2-2. Development of SU-8 master molds by photolithography.............28 Figure 2-3. PDMS mold with bubbles..............................................................29 Figure 2-4. Fabrication of PDMS molds..........................................................30 Figure 2-5. Processing fo r poly(NIPAAm) pa tterning.......................................31 Figure 2-6. Silanization mechanism of glass surfaces by using a monolayer of 3-(trichlorosilyl) propyl methacrylate (TPM)..............32 Figure 2-7. Fluidi c chamber in 3D ...................................................................33 Figure 2-8. Front vi ew fluidic chamber............................................................34 Figure 2-9. Side view of fluidic chamber ..........................................................35 Figure 3-1. Schematic di agram of the confocal mi croscope principle..............51 Figure 3-2. Schematic diagram dem onstrating the principle of optical sectioning......................................................................................52 Figure 3-3. Diagram sh owing generation of optic al sections in the confocal mi croscope.....................................................................53 Figure 3-4. Comparison of specim en illumination in confocal (A) and conventional (B) microscopes.......................................................54 Figure 3-5. Comparison of confoc al and conventional light microscope when the object lens is focused well below the fluorescent sample, depicted by the small rectangul ar box.............................55

PAGE 8

v Figure 3-6. Nipkow di sk...................................................................................56 Figure 3-7. Spinning-disk confocal mi croscope...............................................57 Figure 3-8. The raster scan pattern that (A) confocal laser scanning microscopes and (B) spinning-disk confocal microscopy commonly use to generate im ages...............................................58 Figure 3-9. Schematic diagram of the optical path in a monoscanning confocal mi croscope.....................................................................59 Figure 3-10. Epif luorescence micro scope princi ple...........................................60 Figure 3-11. Backgroun d subtraction using t he wavelength of the fluorescence mi crospher es...........................................................61 Figure 4-1. Cross-sectional area of trenc hes...................................................70 Figure 4-2. Un-buckling of tr enches below and above the LCST....................71 Figure 4-3. SU-8 patte rned silicon wafer.........................................................72 Figure 4-4. Loss of contrast bet ween lower and upper focal pl anes................73 Figure 4-5. Three-dimens ional reconstruction of a spherical fluorescent microspher e..................................................................................74 Figure 4-6. Cross-sectional view of microspheres where too few optical sections were taken resulting in a squashed, threedimensional rec onstructi on...........................................................75 Figure 4-7. Bright field and 3D images of isolated trench below and above the LCST ............................................................................76 Figure 4-8. Delaunay triangulation of a poly(NIPAAm) trench (A) below (B) above t he LCST ......................................................................77 Figure 4-9. Mask with different trenches wid ths and gap s...............................78 Figure 4-10. 3D of micr ostructure patterns........................................................79 Figure 4-11. Mean distanc e of microspheres vs. different size trenches widths and gaps (25 um thick mold)..............................................80 Figure 4-12. Mean distanc e of microspheres vs. different size trenches widths and gaps (5 um thick mo ld)................................................81

PAGE 9

vi Figure 4-13. Steps to m easure the local swelling in different sections of the microstr ucture.........................................................................82 Figure 4-14. Calculating mean distance of microspheres in different zsections.........................................................................................83

PAGE 10

vii Microsphere-Aided Characterization of Stimuli-Responsive Polymer Networks Carlos A. Bello ABSTRACT The fabrication and characterizati on of surface-anchored hydrogel microstructures are described. The hydr ogel structures are constructed from poly(N-isopropylacrylamide), or po ly(NIPAAm), which is a well-known thermoresponsive polymer that sw ells and contracts with changes in temperature. When patterned on a surface, these structures can experience a variety of shape changes induced by nonuniform swelling. Depending on the aspect ratio, patterns can, for inst ance buckle upon swelling and form wave-like patterns. Such structural changes replic ate oscillatory motion of the smooth muscle cells and can be used to transport objects in microfluidics. The work, herein, investigates methods of pattern production and introduces a new technique for characterizing local swelling in the patterns. In or der to achieve the latter, fluorescent microspheres were embedded in hydrogel patterns and their positions were mapped in three-dimensi ons using confocal microscopy. The

PAGE 11

viii measurements permit, for the first time, sw elling maps of the structures based on relative movements of the microspheres. This information will ultimately aid in understanding how swollen macroscopic stru ctures are relat ed to gradients in localized swelling.

PAGE 12

1 CHAPTER ONE: INTRODUCTION The set of techniques used for th e fabrication of patterned poly(Nisopropylacrylamide), poly(NIPAAm), hy drogel microstructures on glass substrate, and the methods for measuring the local swelling inside a hydrogel structure are describe here. Poly(NIPAAm), is a well-know thremoresponsive polymer that exhibits a lower critical solution temperature (LCST) when mixed with water, 1 wherein the polymer phase separates above the LCST. The transition temperature can also be affect ed by the addition of salt, alcohol, pH and surfactants. 2 The unusual phase behavior is a consequence of the entropy gain accompanying dehydration of the amide moieties above the LCST. For cross-linked gels, the LCST results in shrinkage in size. These materials have been of great interest in applications such as chemical/biological sensors, drug delivery, biomaterials, and chem ical/biological separations. 3, 4 There is a growing interest in using patterned poly(NIPAAm) microgels in microfluidics for their act uation properties. However, t here are no techniques to characterize local swelling of surfaceattached networks while the swelling of unconfined networks is well understood (see Figure 1-1 ). Surface-attachment of network results in non-uniform swelling that leads to a variety of swollen structures (see Figure 1-1 and Figure 1-2 ).

PAGE 13

2 The polymer trenches were fabric ated by a technique employing soft lithography. Soft lithographic methods us e an elastomeric stamp or mold prepared by casting the liquid prepolymer of an elastomer against a master that has a patterned relief structure. 5 Glass surfaces were first acrylated with a trichlorosilane surface attach monolayer (S AM) to ensure covalent attachment of the poly(NIPAAm) arrays and thereby im prove hydrogel adhesion to the substrate. A poly(NIPAAm) hydrogel containing 0.005 wt% of fluorescent microspheres was polymerized onto the tr eated substrate. As a result, threedimensional, cross-linked poly(NIPAAm) hydrogel microstructures were attached to the surface. Widths and gaps between the structures were varied from 5 20 m by the use of different patterns in a photo mask, while height varied from 5 25 m, depending on the SU-8 photoresis t used (SU-8 2025 or SU-8 2005). The hydrogel patterns were hydrated for as long as 2 weeks without any notice of delamination from the substrate. When the hydrogel structures are below the LCST, they form either a wave-like pattern known as buckling or a mushroom-like shape. 6 The contraction/swelling behavior of the tr enches was characterized with spinningdisk confocal microscopy, SDCM. This thesis presents an approach to soft lithography making micrometerscale patterns of poly(NIPPAm) hydrogel monoliths with embedded fluorescent microspheres on solid surfaces in order to characterize the swelling behavior of the gel. Whereas mechanically pinn ing a polymer hydrogel to a surface

PAGE 14

3 generally confines swelling in the usual direction, adequately high aspect ratios allow lateral swelling to the surface. This type of swelling makes possible the creation of intelligent surf aces, described previously by Castellanos (2007). In addition, a set of imaging processing steps is described here in order to directly quantify the local spatial structure of the surface-attached hydrogels by measuring the distribution of the microspheres inside the polymer.

PAGE 15

Unconstrained Gel L0 L0 L0L L L L0 L0 L0 L L0 Surface attached Gel Unconstrained Gel L0 L0 L0L L L Unconstrained Gel L0 L0 L0 L0 L0 L0L L L L0 L0 L0 L L0 Surface attached Gel L0 L0 L0 L L0 Surface attached Gel Figure 1-1. Swelling behavior: unconstraine d gels vs. surface attached gels 4

PAGE 16

Above LCST Below LCST Mold: Heigh Mold: gap Mold: width umumum 251510 Mold: Heigh Mold: gap Mold: width umumum 51 51 0 Above LCST Below LCST Mold: Heigh Mold: gap Mold: width umumum 251510 Mold: Heigh Mold: gap Mold: width umumum 51 51 0 Figure 1-2. Microstructures of poly( NIPAAm) below and above LCST 5

PAGE 17

6 CHAPTER TWO: FABRICATION OF POLY(NIPAAM) MICROSTRUCTURES Thick (>100 m), high aspect-ratio microstructures have traditionally been developed by using LIGA microstereo lithography 7 and deep reactive ion etching (DRIE) processes. 8, 9 However, the problem with t hese techniques is that they often involve equipment that is not readily available due to its complexity, and as a result, is expensive. On the other hand, the techniques used for soft lithography are relatively inexpensive, and the proc edures used are relatively easy since it can be done in the laborator y. Soft lithography creates by contact printing micropatterns of self-assembled monolayers, SAMs, 10 which occurs when molecules or objects spontaneously organize into steady, well-defined structures by non covalent forces. 11 Finally, soft lithography does not have all of the limitations that optical diffr action and optical transparency have. 12 The technique used to make microfluid ic channels during this research was based on the patterning of SU-8 photor esist. The properties of these SU-8 photoresist patterning have made possibl e to create many new microfluidic devices, whose capabilities have been well demonstrated. 13 On the other hand, PDMS was used as the elastomer for the majority of the research associated with soft lithography. PDMS has several proper ties that make it well suited for soft-lithography, such as its biocompatibility, permeability to gases, as well as its

PAGE 18

7 ability to be use with cell cultures. 14 In addition, PDMS is see-through above 300 nm, and since it is elastomeric, it c an reach non planar surfaces conformably. 15 Once the photoresist patterns are layered on silicon substrates using soft lithography, they can be uses as masters to make the necessary PDMS stamps. Glass surface was treated with 3-(trichlorosilyl)propyl methacrylate to create a SAM with pendant acrylate groups. Glass su rfaces or silicon were first acrylated with a trichlorosilane SAM to ensure a covalent connection of the PEG arrays and thus strengthen hydrogel adhesion to the substrate. 16 Micropatterned substrates were set up by the photopolymerization of NIPAAm onto surfaces functionalized with 3-(trichlorosilyl)-p ropyl methacrylate monolayer. These surfaces ensured the attachment of the hydrogel to the substrate as well as allowing for cell adhesion in areas lacking hydrogel as a result of its hydrophobicity. 17 The production of high-aspect ratio microfluidic channels is still under extensive research. SU-8 is an epoxybased line of near-ultraviolet (UV) radiation-sensitive ph otoresists. It has high contras t, as well as suitable chemical and mechanical properties, which are capable of developing thick photoresist structures in a single photolithographic step. 18 SU-8 photoresists have yielded aspect ratios of over 25 19, 20 and microstructures up to 1.5 mm high using a single coating step. 21

PAGE 19

8 Soft lithography techniques have yielded results which include very high aspect-ratio micro-structures with ve rtical sidewalls and high resolution. 22 Various groups have found that a micro-structur e 1.2 mm in height can be formed by double coating the SU-8 photoresist layers 13 Multicoating, on the other hand, is a very time-consuming process. The us e of NANOTM SU-8 (MicroChem Corp., Newton, MA) photoresist to increase thi ck microstructures for MEMS has drawn significant attention. 18, 21, 23 Another process that is used to generate complex SU-8 features is the creation of multi-level microstructures by performing various photolithographic steps of exposure and development. 24 For example, the spinning and exposing of multiple layers of SU-8 sequentially followed by a simultaneous (single) developing step hav e achieved both, two and three-level microstructures. 25, 26 One drawback associated with the use of SU-8 photoresist is related to stress, adhesion selectivity, and resist stripping. During the development process, the SU-8 photoresist can simply be removed from the surfaces of such materials. The thermal expansion c oefficient mismatch is large on many surfaces, such as silicon or glass, which are suitable for spinning the SU-8 photoresist. This results in a large amount of stress at the material interface, due to the size reduction of t he SU-8 photoresist while cros s-linking during curing. The stress effect is pronounced in large SU -8 photoresist structures and, if poor adhesion is obtained during processing, the photoresist delaminates easily. Coating, baking, the devel oping efficiency, and the tremendous residual stress

PAGE 20

9 formed after curing are just some of t he challenges involved in making an ultrathick micro-structure with SU-8 photoresist. The flatness of the surface is also a major issue. SU-8 photoresists present many challenges despite its popularity. For instance, it hinders reproducti on and limits potential applications. 21, 27, 28 Furthermore, differences in exposure throughout the photoresist layer can result in partial curing of the resist, residual st ress build-up, as well as poor adhesion to the underlying silicon wafer. This can resu lt in distortions, such as lifting off the entire SU-8 pattern. 21, 25-27, 29-31 The consistency of exposure provided by an appropriate contact between the photoresis t and mask is particularly important for thicker SU-8 layers and is highly reliant on other processing steps. 21 First, the variations of thickness in t he SU-8 layer across the wafer is a result of flowing of the uncross linked photoresis t during the soft-bake step. 21, 29 Secondly, the creation of edge beads during the spin coating procedure likewise produces variations in thickness across the wafer. Third, multiple SU-8 layers can enhance residual stresses, as well as subsequent bowing of the silicon wafer. 18, 21, 29, 32 Components of the developing step also play a crucial role in the processing of SU-8 microstructures (see Figure 2-1 ). Whereas low-aspect ratio features are typically insensitive to developing parameters, high-aspect ratio microstructures many times require optimiz ation of the development time as well as the use of agitation. 32 On the other hand, forceful agitation of the SU-8 Developer could lead to the destructi on of high aspect ratio structures. 33

PAGE 21

10 2.1. Photolithography Process Two types of photoresists were used to fabricate different thicknesses of SU-8 structures. SU-8 2005 produces 3 7.5 m structures, and SU-8 2025 produces 20 75 m structures. 2.1.1. Piranha Clean In order to start the phot olithography process, it is best to clean and dry substrates before using SU-8 resist. Doing so will provide maximum process reliability. For the most accurate result s, substrates should be cleaned with a piranha wet etch, using H 2 O 2 and H 2 SO 4, along with a de-ionize d water rinse. The n-type <100> Single-Side-Polished 2 silicon wafers were cleaned with a piranha solution, which removes organic material from substrates. Piranha cleaning solution is a warm mixture of sulfuric acid (H 2 SO 4 ) and hydrogen peroxide (H 2 O 2 ) typically at 100 o C 150 o C, this mixture reacts violently with organic materials and must be handled with extreme care. Therefore, it is performed in a glass container with extra protective equipment like rubber gloves, heavy apron with sleeves, and full-face shie ld in addition to safety glasses. Moreover, this process is performed in a well-ventilated wet bench. The H 2 SO 4 to H 2 O 2 ratio is 3:1, and once the H 2 SO 4 is added to H 2 O 2 the reaction must be allowed to occur within a time range of 1 to 2 minutes. Next, the silicone wafers are slowly added to the solution and allow ed to develop for 10 minutes. Once

PAGE 22

11 the silicone wafers have fully developed, they must be removed. Excess rinsing with de-ionized water is required in order to remove excess amount of acidic residue; consequently, achieving a surface t hat is free of excess acidic remains. Once all excess acidic remains have been removed, the silicone wafers must be allowed to dry with a cont inuous flow of nitrogen. 2.1.2. Dehydration Bake For the soft bake process, a level hotpl ate should be used. It is important to use one with good thermal control, as we ll as uniformity. The vendor does not recommend the use of convectional ovens. The dehydration bake was performed for 10 mins at 150 o C. The wafer should be removed from the hotplate after the allotted time. This will help optimize the baking conditions. Once it has cooled to room temperature, it should be re turned to the hotplate. In the event that the film 'wrinkles', the wafer should be put back on the hotplate for a couple of more minutes. The same cool down and heat up cycle should be repeated until the wrinkles are no longer visible in the film.

PAGE 23

12 2.1.3. Coat The purpose of spin coating is to appl y a uniform layer of SU-8 over the entire surface of the silicone wafer. A 2 ml pipette was used in the first attempt to make a smooth layer of the SU-8. A sm all amount of acetone was applied to the edge of the wafer to eliminate the exce ss build-up of photoresist accumulated throughout the various spinning cycles. The elimination of excess edge bead allows the mask to be put in close cont act with the silicone wafer, thus enhancing the resolution. SU-8 2000 is best when used for imaging ne ar vertical sidewalls in thick films. This is due to its high optical transmission, which is above 360 nm. Some of the many features of the SU-8 2000 include high aspect ratio imaging, vertical sidewalls, improved coating properties, and 0.5 to > 200 m film thickness in one coat. A spin coater Model p6700 is us ed for the coating of the photoresist. The MicroChem recommended coating conditions are as follows: STATIC Dispense: Approximately 1ml of SU-8 2000 per inch of substrate diameter. Spread Cycle: Increase to 500 rpm at 100 rpm/second acceleration. This will take 5 seconds. Spin Cycle: Increase to final spin speed at an acceleration of 300 rpm/second and maintain at said rate for a total of 30 seconds.

PAGE 24

13 There were some challenges in the coat ing of the SU-8 which affected the consistency of the results from exper iment to experiment. It was found that pulling photoresist directly from bottle r educed the formation of bubbles on top of the surface. In addition, a high accele ration (>10,000 rpm/sec) seemed to be more favorable for a better spreading of the SU-8. Also, there were some samples done with an expired SU-8 whic h generated ticker structures. Finally, the photoresist should be kept inside the fr idge. However, it was observed that when it was spin at cold temperatures it gave unpredicted result s. Therefore, the SU-8 needs to be at room temperature when it is spin coat. 2.1.4. Soft Bake A level hotplate should be used during this process. It is also important to use one that has good thermal control, as we ll as uniformity. Since it is possible for a skin to form on the resist, the vendor strongly advises against using convection ovens. The formation of this skin could in turn st op the evolution of solvent, which would either prevent the film from drying or lead to an extended bake time. Table 2-1 lists what is recommended for the soft bake times and temperatures for the used SU-8 2000 products based on the selected film thicknesses.

PAGE 25

Table 2-1. Recommended soft bake parameters Product NameThicknessPre-bakeSoft bake (um) @ 65 oC @ 95 oCSU-8 2005 3 1 2 512 612 7.5 1 2 SU-8 2025 20 2 5 75 3 9 In order to prevent any complications with step coverage on the silicon, the sample was given an hour to relax once the spin coating was complete. A pre-exposure bake was done by extending and increasing the te mperature as a means to reduce stress and to permit solvent evaporation. The sample was baked on a hotplate at 65 C for 5 minutes. The tem perature was immediately increased, and baking continued at 95 C for 40 minutes. Soft bake process: 65 C for 1 min on the hot plate, 95 C for 5 min in the oven, 10 min of cool down period (relaxa tion) at room temperature (~25 C). It is worth mentioning so me of the challenges and conclusions of the soft bake step. It is better to use soft bake ti mes for high aspect ratio to reduce the retained solvent level since doing so can lessen the risk of exposed resist loss, adhesion failure, and swelling. It is possi ble that a build-up of photoresist could form around the edge of the substrat e during the spin coat process. 14

PAGE 26

15 2.1.5. Exposure The mask was design on AutoCAD by another group member. The color of the mask body is black for the most part and will not absorb the UV light. Even still, only the see-through areas of the mask will absorb the UV light and, as a result, become a permanent fixture. A Karl Suss Mask Aligner UV light source was used for exposure. A mask, which defined the locations of t he SU-8 photoresist on the wafer, was used, while exposure continued for 35 sec onds. The SU-8 photoresist structures on the cantilevers formed the trenches. In order to eliminate t he UV radiation below 350 nm a Hoya UV-34 filter has been used. Excessive dose below 350nm may has been proven to result in over exposure of the top portion of the resist film resulting in exaggerated negative sidewall profil es or T-topping. 34 In addition, Chuang (2005) showed that smaller and uniform gap between photoma sk and resist reduced the diffraction error. 28 He proved that the edges of mask patte rns usually pose diffraction effect during contact printing, which is produced by Fresnel diffraction. When the thickness of photoresist and the gap bet ween photomask and resist increases, the diffraction error gets larger. Consequently it results in non-st raight sidewall of resist structures and constrains the application of microstructures in high-aspectratio range. 28 Revzin (2001) used contact printing to reduced the diffraction caused by the air gap. 16 According to Revzin (2001), the square root of the gap

PAGE 27

16 between a photomask and polymer layer is directly proportional to the minimum resolution. 16 Consequently, the gap has to be mini mized in order to get images at resolutions critical to our experiments. The exposure time (dose) has been varied to obtained different thicknesses. Besides, it is well known fr om literature that ticker films require higher dosage. The exposure process was in itiated by having a vacuum contact between the mask and the silicon wafer. Th en the UV filter was located between the UV source and the mask. Finally, 100 mJ (10.0 mW cm 2 for 10.0 s or 15.0 mW cm 2 ) were applied to the substrate. The thickness non-uniformity caused by standard resist spinner and hot plate soft-baker can easily pose 30 m air gap when contacting with photomasks for resist thick ness ranging from 0.1 to 1 mm. Printing mask pattern through the aforementioned air gap causes the resulting patterns deform. The key phenomenon that affects the size of pat terns and sidewall profile in the UV lithography printing is the Fresnel diffraction which gives more energy on the pattern edges. 28 Chuan (2002) described a novel way to compensate the air gap between photomask using glycerol between the ma sk and the sample. This technique highly reduced the UV light diffraction on pho toresist. Also, this method increases the side wall straightness. 28

PAGE 28

17 2.1.6. Post Exposure Bake (PEB) Post exposure bake (PEB) was carried out at 65 o C for 3 minutes and 95 o C for 12 minutes. After this, the samples we re left at room temperature to avoid excess stress due to thermal gradients. It was concluded that extended PEB ti mes at lower temperatures give better results than the samples treated at higher temperatures. In addition, an image of the mask should be seen in the photoresist coating after one minute of PEB, which indicated that there was insufficient exposure, heating or both. 2.1.7. Development The development process consists of gentle agitation for 8 minutes. The substrates were developed using a MicroC hem SU-8 photoresist developer. This step clears out deep trenches. Various techniques were tested before finding a suitable technique for the patterning r equirements. For exam ple, one technique in particular consisted of placing the wa fer on a stir rod and t hen immersing it in developer. The problem with this, however, is that a significant amount of heat is absorbed when drying the acet one that the temperature of the substrate drops significantly. This, in turn, could either cause extreme thermal expansion or contraction. In either ca se, this will cause stress cracks to the film. Finally, the samples were heated on the hotplate at 15 o C for 1 min after development.

PAGE 29

18 If, during the IPA rinse, a white film is formed, it means that the unexposed photoresist was underdeveloped. A quick fix would be to add more SU-8 developer to the substrate. Doing so will remove the white film and finish the development process. Rinse steps should be repeated. In order to dry the sample it is important to keep the patterned face of the wafer away from strong nitrogen flow cu rrents since these blows of nitrogen can delaminate the SU-8 in this st ep of the process. Therefore, it is important to blow dry the back side of the wafer in order to dr y it, and then it is necessary to put the wafer in a spin coater and spin it dry at 2000-3000 rpm for about 15 seconds. Figure 2-2 shows each step of the entir e photolithography process. 2.2. Silicon-PDMS Molding Soft lithographic methods use an elasto meric stamp or mold, prepared by casting the liquid prepolymer of an elastomer against a master that has a patterned relief structure. 5 Most of the research based on soft lithography has used poly(dimethylsiloxane) (PDMS) as the preferred elastomer. PDMS has become the primary material for the lowvolume manufacturing of microfluidic devices owing to its ease of processi ng in standard laboratory conditions, and low cost. In addition, PDMS has several pr operties that make it well suited for patterning proteins and cells. It is biocompatible, permeable to gases, and can be

PAGE 30

19 used for cell culture. Furthermore, it is optically transparent down to about 300 nm, and it can contact non-planar surfaces conformably due to its elasticity. 5 The elastic characteristic of PDMS allows it to be released easily, even from complex and fragile structures. Also, PD MS provides a surface that is low in interfacial free energy (ca. 21.6x10 -3 Jm -2 ) and chemically inert. 12 Finally, it has been shown that threedimensional structures can be built up layer by layer, e.g. fluidic channels and mixers, 35, 36 valves, 37 and pumps connected to an active microfluidic matrix to perform programmable tasks 38 with the use of the PDMS elastomer. T he fabrication of the SU-8 patterned photoresist structures was discussed in pr evious sections. These masters were used for PDMS embossing. A mix of PDMS, (Dow Corning), and curing agent was prepared in a 10:1 ratio. The prepolymer mixture was degasse d with a mechanical vacuum pump to remove any air bubbles in the mixture and to insure complete mixing between the two parts 5 (see Figure 2-3 which shows a PDMS molds that was not degassed with the mechanical pump). The PDMS mix was placed on top of the patterned SU-8 photoresist and placed on a hot plate at 75 O C for 60 minutes in order to obtain the solidified PDMS mold. Afterwar ds, the sample was allowed to cool down for 2 hours in order to obtain the desired final patterns. Finally, excess PDMS was cut off at the edges in order to expose the openings of the channels;

PAGE 31

20 this is done in order to apply capillar y force to the pre-polymer solution (see Figure 2-4 and Figure 2-5 for a detail illustration of the process). The elastomeric character of PDMS is al so the origin of some of the most serious technical problems that must be solved befor e soft lithography can be used in forming complex patterned struct ures. For instance, PDMS molds were not release easily from the silicon wafers even though of its elasticity and low surface energy. After curing the PMDS, this was peeled out from the mold revealing the microfluidic channels. Obse rvations in a conventional microscope before and after the peeling of the PDMS (not shown in this thesis) points up the destruction of the PDMS mold. Therefore, special attent ion was taken during this process in order to minimize this destruction. There are several other disadvantages that may limit the performance of PDMS for certain types of applications. Fo r example, PDMS shrinks by a factor of about 1% upon curing. 12 Also, gravity, adhesion an d capillary forces exert stress on the elastomeric features and cause them to collapse and generate defects in the pattern that is formed. 39 This could occur, for example, if the aspect ratio of the relief features is too lar ge. Then the PDMS microstructures could collapse or fall under their own weight; Delamarche (1997) showed that the aspect ratios of the relief structures on PDMS surfaces had to be between about 0.2 and 2 in order to obtain defect-free stamps. 39 On the other hand, when the aspect ratio is too low, insufficient relie f exists on the surfac e of the stamp to

PAGE 32

21 withstand the compressive forces (due to adhesion between stamp and substrate) typical of printing. 2.3. Silanization of Glass Substrates In early experiments poly(NIPAAm) microstructures were generated on glass surfaces cleaned with Acetone/Me thanol/IPA but not modified with and adhesion promoting monolayer. Even though patterns could be develop, surface attachment was weak and microstructure s easily delaminated upon hydration due to swelling of the crosslinked polymer. A slightly better attachment of the network was obtained by treating the substrate with 3-aminopropyl triethoxysilane. However, by using a self assembled monolayer of 3(trichlorosilyl)propyl methacrylate (T PM, Aldrich) a reactive surface was covalently affixed to the gel duri ng photopolymerization, which prevented delamination. The treatment of glass surfaces with chlorosilanes or alkoxysilanes is a very common and effective way to fo rm dense, self-assembled monolayers. 40, 41 In order to ensure covalent attachment of the poly(NIPAAm) arrays and thereby improve hydrogel adhesion, an adhesion layer was created using a monolayer of TPM. The modification of the substrates was perfo rmed using standard protocols for silane surface modifications. 42 In brief, cover slides were cleaned with the standard solvent cleaning process using Acetone/Methanol/IPA and DI water.

PAGE 33

22 Subsequently, in order to remove organic contaminant s and to promote hydroxylation (OH groups) on the surface, the cover sli des were plasma-cleaned (Harrick Plasma Cleaner/Sterilizer PDC32G) for 10 minutes at 1 Torr with a medium RF power. During this process organic contaminants are removed by chemical reaction with highly reactive ox ygen radicals and ablation by energetic oxygen ions. Substrates were then treated fo r 5 min, at room te mperature, in a 1 mM solution of TPM in a 4:1 ratio of heptane-carbon tetrachloride in an N 2 atmosphere, followed by wa shing with hexane and water. Figure 2-6 depicts the silanization mechanism of glass surfaces. 2.4. Photopolymerization of Hydr ogel with Embedded Microspheres N -isopropylacrylamide (NIPAAm) (97% Al drich) was recrystallized from nhexane solution. The crosslinking agent was N N methylenebisacrylamide (BIS, Chemzymes Ultra Pure). The initia tor for the polymerization reaction was 2dimethoxy-2-phenylacetophenone (DMPA, Aldrich). 500 nm polystyrene yellowgreen fluorescence microspheres (Molecula r Probes, Invitrogen, CA) were used as molecular probes. Once the surface of the coverslip is treated with the silane binder, PDMS is placed on the surface of the substrate. Structured PDMS sheets can easily be bonded to a glass plate or another sheet of polymer owing to its excellent adhesion properties. For a non-permanent bond for low-pressure applications,

PAGE 34

23 the PDMS sheet can simply be pressed against the closing sheet and makes conformal contact with that surface. Consequently, the relief stru cture in the mold forms a network of empty channels. A lowviscosity liquid pre-polymer is placed at the open ends of the network of channels; then by capillary action the liquid spontaneously fills the channels due to t he pressure difference between the two hydraulically connected regions of the liquid mass. After the pre-polymer solution accept ed the shape of the PDMS patterns, it was photopolymerized, In-situ. 16.5 wt% NIPAAm, 0.4 wt% BIS, and 0.005 wt% of microspheres were dissolved in isopropanol. The solution was then bubbled for 20 minutes in order to remove the remaining oxygen. Photopolymerization was then initiat ed by adding 0.16 wt% of DMPA and using an uncollimated, 365 nm, 300 mW/cm 2 light source (EFOS Ultracure 100ss Plus, UV spot lamp, Mississauga, Ontario) under an inert nitrogen environment for 8 minutes. Finally, after filling the channels and photo polymerizing the prepolymer into a solid, the PDMS mold is removed, and a network of poly(NIPAAm) remains on the surface of t he substrate. The formation of hydrogel microstructures was based upon the UV initiated free-radical polymerization of the acrylate end groups in NIPAAm. The radical polymerization begins by the di ssociation of DMPA upon exposure to the UV radiation. This creates highly reactive methyl radicals, which then attack unsaturated carbon-carbon double bonds (C=C) of acrylate functionalities on the macromer. Since two reactive center s per macromer ar e created, propagation

PAGE 35

24 results in the formation of a highly cross-linked polyacrylate network. This network represents a three-dimensiona l structure capable of entrapping the fluorescent microspheres. It should be stated that the C=C conversion cannot be completely correlated to polymerization in the gel s because of the presence of oxygen during illumination. A fraction of acrylate groups will be converted to oxygenated species rather than participate in the polym erization process due to the reaction of oxygen with radicals formed in t he gel which effectively terminates propagation. In addition, the surface of the covers lip will also parti cipate in the freeradical polymerization reaction due to t he unsaturated C=C f unctionalities from the TPM monolayer. Methacrylate groups on the surface of the substrate react with radical species present near the surfac e to effectively anchor the gel to the surface of the substrate. 16 Thus, delamination or detachment of poly(NIPAAm) networks from the substrate did not o ccur, even though ext ensive swelling did occur in the gels upon expos ure to water. Filling up of the channels and subsequent polymerization of the pre polymer solution does have several limitatio ns. First, it requires a hydraulically connected network of capillar ies; it cannot, therefore, form isolated structures or patterns on contoured surfaces. Also, although capillary filling over a short distance (ca. 1 cm) can be achieved quickly and efficiently, the rate of filling over

PAGE 36

25 a large distance decreases significantly due to the viscous drag of the fluid in the capillary. Moreover, the forward ends of capillaries may fill incompletely if the hydraulic drag is sufficiently high. 43 Interestingly, capillaries with closed ends may fill completely if they are short; the gas in them appears to escape by diffusing into the PDMS. 12 In addition, even though there is large amount of solvent in the polymer solution, the solvent can be easily evaporated at room temperature since both sides of the channels are open. Therefore, the so lution should be polymerized immediately after the channels are completely filled, which occurs relatively fast 44 in the cross-sectional sections used for this report. Lastly, a PDMS structure theoretically could be reused after being peeled off the coverslip; in practice, however, the surfaces ar e likely to be contaminated and a second sealing does not hold tight. 2.5. Fabrication of Fluidic Chamber The fluidic chamber was constructed by sealing a 2 cm (diameter) rubber ring between a no. 1 cover slide and a 3" x 1" plexi glass sheet. In addition, several 5 mm (diameter) rings were sealed in the corner s of the cover slide with nail varnish in order to give more rigidity to the structure (see Figure 2-7 Figure 2-8 Figure 2-9 ).

PAGE 37

26 2.6. Remarks Due to mechanical forces associated with volume changes in the gel during water induced swelling, the pol y(NIPAAm) microstructures easily detached from surfaces not treated with TPM. However, hydrogel structures fabricated on TPM treated glass substrates did not show an increase in diameter at the base of the st ructures because they were cova lently fixed to the substrate surface. Nonetheless, poly(NIPAAm) array elements became swollen and increased in height by nearly 30% 50%. Height changes of 32% and 51% were observed for gel microstructures made from 5 um and 25 um PDMS molds, respectively. Such behavior is expected because surface anchoring of the gel prevents the hydrogel array elements from swelling laterally at their base but does not inhibit swelling upward.

PAGE 38

Figure 2-1. SU-8 microstructure 27

PAGE 39

28 Figure 2-2. Development of SU-8 master molds by photolithography

PAGE 40

Figure 2-3. PDMS mold with bubbles 29

PAGE 41

Figure 2-4. Fabrication of PDMS molds 30

PAGE 42

Figure 2-5. Processing for poly(NIPAAm) patterning 31

PAGE 43

Figure 2-6. Silanization mechanism of glass su rfaces by using a monolayer of 3(trichlorosilyl)propyl methacrylate (TPM) 32

PAGE 44

Figure 2-7. Fluidic chamber in 3D 33

PAGE 45

Figure 2-8. Front view fluidic chamber 34

PAGE 46

Figure 2-9. Side view of fluidic chamber 35

PAGE 47

36 CHAPTER THREE: CHARACTERIZATION OF SWELLING OF HYDROGELS BY SPINNING DISK CONFOCAL MICROSCOPY 3.1. Introduction Optical microscopy is a technique that is widely used in many systems when investigating the sub micrometer to micrometer range. When light enters dense samples, such as biological ti ssue, it goes through many scattering events, which can leave the image blurry, thus making it difficult for deep visualization within the sample. Marvin Minsky faced this very significant problem in the 1950s. He was looking fo r a way to visualize the nervous system, and the nerve cells connected to it. He developed a two-step solution. The first step was reducing aberrant ra ys of scattered light from around the target point of interest. This was accomplished by using point by point illumination of the sample. The second step was to ta ke a pinhole-sized opening in the image plane on the other side of the objective, wh ich was used to eliminate out-of-focus light. This prototype for the first conf ocal microscope used a carbon-arc lamp as the light source and a translating stage to visualize each point of the sample. Other subsequent improvements use a method of scanning the laser beam

PAGE 48

37 instead of translating the sample. These tw o processes helped establish confocal microscopy as a powerful tool in various scientific fields. 45 3.2. Confocal Microscopy Introduction As conventional optical microscopes use a halogen lamp and other surface light sources, diffus ed light from places other than the point of interest are mixed with the lig ht from the point of intere st. Moreover, when planar photo detectors like cameras or eyes are used, blurred, out-of-fo cus light from the points shifted toward the optical axis are observed, overlap with the observed light. For these reasons, the conventiona l optical microscope has a limited spatial resolution. 46 On the contrary, confocal microscope systems use laser beams to illuminate samples with pin-point accura cy, to eliminate di ffused light from sources other than the point of intere st. In addition, pin-hole windows are provided in front of the light detector to cut off light from sources other than the point of interest for high resolution (see Figure 3-1 ). 3.2.1. Advantages of Confocal Microscopy There are several advantages of using confocal microscopy rather than electron microscopy and conventional ligh t. First, the confocal microscope visually divides the images into sections (see Figure 3-2 and Figure 3-3 ). In

PAGE 49

38 doing so, it reduces any problematic ar eas that would otherwise be observed using electron microscopic techniques and conventional light. Also, both living and fixed samples can be viewed with greater precision since optical sectioning is generally noninvasive. In addition, information can be collected from a welldefined plane instead of from the entire sample thi ckness due to the shallow depth of field of the optical sections (see Figure 3-4 ). As a result, the image is black since it is no longer collecting light that is out of focus when the focal plane is above or below the specimen. By reducin g the light that is out of focus, this improves the clarity, contra st, and detection sensitivity. Figure 3-5 shows the comparison between confocal and conventio nal light microscope when the object lens is focused well below the fluorescent sample. Another advantage of confocal microsc opy is vertical sectioning. By taking optical sections in different pl anes, cells are scanned both laterally as well as in depth. The optical sections can also be obtained digitally, which is beneficial because the prints can be done quickly, as opposed to using darkroom materials, which take longer to develop. 3.2.2. Disadvantages of Confocal Microscopy Unlike the conventional fluoresce nce light microscope, confocal microscopes are limited by the available wavelengths of light that are produced

PAGE 50

39 by laser lines. If the laser light has not been properly attenuated, the intensity of the beam could be harmful to living cells. Price is another issue to consider when using a confocal microscope. Depending on the one used, they can cost over ten time s more than a typical conventional fluorescence light microscope. However, this can be remedied if shared among researchers in a centralized fa cility. This is especially beneficial when taking into account the size of the microscope. A confocal microscope takes up more space than a conventional fluorescence light microscope since it needs more space for computer hardwar e, a laser, and scan head. 3.3. Spinning Disk Confocal Microscopy Introduction One type of confocal microscope uses a spinning Nipkow disk, a stationary light source, and a stage. The Nip kow disk, which is used to raster scan the sample, is a black, metallic-coated, glass disk that has various openings 20-60 um in diameter. They are arranged in an interlacing pattern. Figure 3-6 depicts the pattern arrangement of the Nipko w disk. One of the special features of this microscope is that specimens can be viewed directly in their true colors as a confocal image in the reflected as we ll as the fluorescence mode. Since the disk can spin at speeds of up to 3600 rp m, the images can be viewed in real time, which allows direct confocal focusing through the sample (see Figure 3-7 ).

PAGE 51

40 Spinning disk confocal microscopes scan the specimen with thousands of spots of light concurrently, as opposed to CLSMs, which use one spot of light to scan the specimen (see Figure 3-8 which illustrates this difference). Light from each spot traces a single scan over succe ssive parts of the specimens. Spinning disk confocal microscopes are also diffe rent from CLSMs: instead of using coherent laser light to sc an the specimen, a broad-spectrum, white light is the illumination source as in conventional light and fluorescence microscopy. The white light source can eit her be mercury or xenon, allowing for the selection of any filter combination employed for conventional epifluorescence. Dyes excited by UV can be imaged confocally along with fluorescein, rhodamine, and infrared dyes without the added expense of UV and infrared lasers. Cells that are labeled with multiple fluorochromes can theref ore be imaged confocally in their true colors. Some spinning disk systems use laser light for fluorochrome excitation. The design of spinning disk confocal mi croscopes is also different from the CLSMs in that only x y sectioning is po ssible. Another noteworthy difference from CLSMs is that a com puter is not needed for most spinning disk confocal microscopes to work. In older spinning disk systems, the Nip kow disk typically transmitted only 1% of the illumination, re sulting in dim fluorescence images. Unless fluorescent samples were bright, an intensifying vi deo camera or cooled CCD camera was needed to generate sufficient signals and improv e the signal-to-noise ratio. The relatively large size (60 m) of the confocal apertures in the Nipkow disk, as

PAGE 52

41 compared to CLSMs compromised optical sectioning so that optical section thickness was approximately 1 m for sp inning disk confocal microscopes as compared to 0.5 m for CLSMs with t he same objective. Newer models of confocal microscopes have addressed the lack of brightness using microlenses. There are two types of spinning disk c onfocal microscopes: the tandem scanning (not included in this thesis) and monoscanning microscopes. 3.3.1. Mono Scanning Confocal Microscopy Introduction The monoscanning spinning disk confocal microscope (see Figure 3-9 ) also uses a spinning Ni plow disk with spirally arranged openings. It uses noncoherent white light for illumination, and produces real time images. This microscope detects light through the same aperture from which a given area of the sample was illuminated. Therefore, illumination and imaging are performed at the same time using the same aperture which allows for easier alignment of the Nipkow disk. To reduce unwanted light reflected by the Nipkow disk, one version of the confocal microscope, the K2-S Bio, tilts the Nipkow disk and uses polarized light. Newer models of the monoscanning confocal microscope use a second disk which has 20,000 microlenses that are aligned and spin together with the apertures in the Nipkow disk, unlike older models, which suffered from loss of light. This amplifies the light trans mitted to and from the Nipkow disk.

PAGE 53

42 Furthermore, background light in the mi croscope that was developed by the Yokogawa Electric Company is reduced si nce the dichroic mirror is positioned between the Nipkow disk and the disk containi ng the microlens array. Confocal microscopes using the microlens e array have been used to examine microcirculatory events and calcium ion dynamics. Finally, the data shown in this report is obtain from the Perkin Elmer Ultra View confocal microscope system, which is an updated version of the Yokogawa unit that employs dual spinning disks wit h microlens technology together with either argon ion or kryp ton-argon ion laser, and digital cooled CCD camera detection to produce images. Both fast and slow scan rates are available. Fast temporal data acquisition with frame transfer or interline cameras at 12 or 14 bits per pixel lets images be captured at video rate s or faster (50 frames/s and as fast as 120 frames/s). This is especially usef ul for imaging ionic fluctuations in living cells and tissues. 3.3.1.1. Operation of the Dual-S pinning Disk Confocal Microscope Please refer to Figure 3-7 for an illustration of t he system. The expanded and collimated laser beam illuminates the upper Nipkow disk containing roughly about 20,000 microlenses. Each mi crolens focuses the beam onto the corresponding pinhole. In doing so, it significantly raises the fraction of the illuminating beam that is transmitted by the main Nipkow disk containing the

PAGE 54

43 pinhole array. The beams then go down to fill the opening of the objective lens. The objective lens produces a reduced im age of the pinholes into the sample focal plane. Fluorescence is captured by the objective lens exit pinhole and eliminates fluorescence from out-of-focus regions, and thus transmits fluorescence that started fr om the sample region illumi nated by that particular pinhole. On the other hand, for samples that have fluorescence distributed over large depths, some out-of-focus fl uorescence can escape through nearby pinholes in multiple pinhole systems. The rays sent out by the exit pinholes are averted by the dichromatic beam splitter, which is between the two Nipkow disks, and go on to the image plane. The principle of operation of the dual-spinning-disk confocal is described in detail elsewhere. 47 Briefly, in the Ultraview unit, the light input from the fiber goes through some beam-shaping optics and is delivered onto a disk spinning at 1800 rpm. The disk contains a helical array of 20,000 microlenses spaced at 250m center-to-center distance which collect the light and each creates a scanning laser point source as the disk spins. The point sources pass through a reverse multibandpass dichromatic mi rror (Chroma, Brattleboro VT) (see Figure 3-10 for an explanation of the dichromatic mirror) and through a Nipkow disk spinning at the same rate and containing 50m pinholes positioned precisely conjugate to the microlenses. The scanni ng point sources leave the confocal head and enter the inverted microscope via a c-mount-coupled side camera port and are focused onto the specimen through the objective lens. Fluorescence

PAGE 55

44 emission from the specimen is then colle cted by the objective, and out-of-focus emission is removed in the Ultraview un it by the spinning Nipkow confocal pinhole array. The focused emission that passes through the pinholes is reflected by the reverse dichroma tic mirror away from t he microlenses, through a multibandpass emission filt er, and to the camera. 48 3.3.1.2. Advantages of the Dual-S pinning Disk Confocal Microscope The dual-spinning-disk system has im portant advantages over both the laser scanning confocal microscopes, as well as traditional single Nipkow disk confocals. First, the pinhole array on the Nipkow disk of the Ultraview is made to scan twelve image frames per rotation, which are 360 frames per second. The laser scanning confocal, on the other hand, only does ~1 frame per second. Also, multiple point sources scanning t he specimen gives real-time confocal imaging, which allows one to see a full c onfocal image through both the ocular as well as on a camera. On the other hand, in a single scanning beam in a laserscanning confocal, the entire sample mu st be registered serially, which makes low-noise CCD detection impossible, and le aves it to be limited to noisy photomultiplier tube detectors, which are unsuitable for FSM. It is also important to note that in traditional Nipkow disk conf ocals, only about 5% of the incident illumination makes it thr ough the pinholes to scan the specimen. However, the microlenses in the Ultraview amplifies t he illumination efficiency to theoretically

PAGE 56

40%. 46 The positioning of the reverse dichromatic mirror between the two spinning disks stops the emission from the specimen from being defocused by the microlenses in the upper disk. 48 For multispectral imaging by the confocal, a dual-bandpass reverse dichromatic mirror is used, as well as an emission filter set fo r the 488 nm ex/525 em and 568 ex/605 em (Chroma). Various other filter sets are now available from Chroma, including dual sets for excitation at 442/514 and 457/514, and a triple set for 488/568/647. It is import ant to note that reverse dichroics are complicated and not easy to manufacture. Also, the ones made for this application are on a very specialized substrate so as to not misalign light coming from the microlenses as it goes through the pi nholes, making them very expensive. 48 3.4. Image Processing and 3D Rendering of Stacks 3.4.1. Three-dimensional Image Processing Since the scanning-disk confocal micr oscope is used directly for the noninvasive serial optical sect ioning, yielding high resoluti on images free from out-offocus blur, it permits the di rect acquisition and study of complete in-focus 3D data sets. 45

PAGE 57

46 Once a series of images has been collected, the 3D data set, which these comprise, may be manipulated in m any ways. A great advantage of this technique is that optical sections are already in spatial register, and thus do not require computational alignment before subsequent processing. The availability of high resolution non-overlapping optical section series makes it possible to undertake the accurate quantization of stru ctural features in the sample, using a variety of unbiased es timators for 3D stereological analysis, ImageJ. In addition, th ere exists a wide variety of image processing and analysis procedures that can be applied to fu rther enhance the images, or to extract features for subsequent m easurements or display. Confocal images, for all their adv antages, are not perfect, and it is possible to enhance them further by subs equent computational image restoration processing to remove residual blurring due to the 3D point spre ad function of the confocal system. 3.4.2. Surface and Volume Rendering By making use of surface rendering procedures that are standard on graphics work stations, compact fluorescent objects can be displayed as surface shadowed or surface luminant. 45 In other words, 3D re ndering methods convert 3D data sets into 2D pictures, or a sequenc e of 2D pictures, in order to provide

PAGE 58

47 interpretation of the 3D structure. Among a multitude of possible projection methods are the summed-intensity, ma ximum-value, voxel-gradient shading and alpha blending. The maximum-intensity proj ection is demonstrated in this thesis. For more complex procedures, the primary problem relies on the defining of the object surface required for t hese procedures from the original unsegmented 3D voxel image, using some kind of threshold algorithm; ImageJ standard threshold algorithm was used for the processed images shown in this work. Once the voxels are defined by ident ifying the surface of the object to be displayed, e.g. microspheres, the other voxels of the image are discarded, to produce a very much smaller data set which can be handled easily by the workstation. More sophisticated volume r endering algorithms allow altering of the constants that model transparency or refl ectivity, so that the rendered objects can be displayed as opaque or semi-transparent. For instance, it will be possible to see structures within other struct ures, such as the microspheres within poly(NIPAAm) trenches. 3.4.3. 3D and 4D Animation of Confocal Images The ultimate development of confocal images display is to compute not just a single stack of images forming a 3D structure, but time incremental series of such stacks that may subsequently be displayed as an animation sequence. This 4D animation displayed may be achiev ed by storing the complete animation

PAGE 59

48 sequence of two time points at any one time the first being overwritten from the disc by that for the time point-after-nex t while the second is being displayed. Such a four-dimensional (x, y, z, t) disp lay of a time-lapse series of confocal images collected from a 3D structur e enables one to study phenomena such as hydrogel dynamics. However, the amount of memory required for the storage and displayed of pre-computed animation se quences is vast. Therefore, special graphic workstations should be use. A fe llow group member is investigating the expediency of such methods to analyze the buckling of patterned poly(NIPAAm) hydrogels (unpublished data). 3.4.4. Fluorescence Images of Fluorescent Microspheres To investigate the local swelling of poly(NIPAAm) microstructures attach to pendant surfaces, fluorescent microspheres were embedded into the polymer solution. The fluorescence microspheres are optimally excited at 468 nm, with an emission maximum at 508 nm. Figure 3-11 (A) shows a fluorescent micrograph of the patterned poly(NIPAAm) above the LCST. Figure 3-11 (A) is the original image and Figure 3-11 (B) is after wavelet application. Figure 3-11 (A) and Figure 3-11 (B) demonstrate the capability of the wavelet analysis to select features of interest in the bead data and to reduce background. In both figures 15% threshold is set relative to the ma ximum in each image. Therefore, the lower 15% intensity pixels are blank ed out. As you can see, in Figure 3-11 (A), the

PAGE 60

49 broad background fluorescence (with lo w spatial frequencies) has been subtracted by the wavelet application. This is a convenient way to automatically select objects with certain morphologie s in an image for further analysis. For instance, the fluorescence informati on can then be fitted only from the bead locations selected by the wavelet applicati on. In further sections, the practicality of such an application to analyze automatically large sets of data is highly useful. 3.5. Image Processing Routines to Lo cate Particle Positions in Three Dimensions Real-space images are collected with a spinning-disk confocal microscope. Images sizes are 6.45 m in the xand y-directions. Image stacks typically contain approximately 150 400 images, each offset by 0.05 0.15 m in the z-direction. Fluorescence is excited using with the 488 nm line of an Ar+ ion gas laser; the fluorescence is filt ered with a 500 nm long-pass filter, and is detected with a photomultiplier tube. Images are analyzed using image processing routines to filter the images and locate particle positions in three dimensions. Images with the computerdetermined particle positions overlaid ar e routinely checked to ensure accuracy. All the image processing data acquire for this thesis was obtained using ImageJ which is a public domain, Java-based im age processing pr ogram developed at the National Institutes of Health, ImageJ.

PAGE 61

50 The procedure in the location of the particles positions involved background subtraction and Gaussian filtering of the images and location of the particle centers in 3D based on regiona l intensity maxima. Subsequently, the locations of the centers were refined to subpixel accuracy based on moments of the local intensity distribution. Once the location of the centers was obtained a Matlab routine was used in the reconstruction of 3D geometric figures using Delaunay triangulation procedur es. Delaunay triangulation is a triangulation such that no point in a set of P points is insi de the circum-hypersphere of any simplex in a triangulation DT(P). In other words, it is a set of lines connecting each point to its natural neighbors. The Delaunay tr iangulation is related to the Voronoi diagram which is a special kind of decom position of a metric space determined by distances to a specified discrete set of objects in the space. The use of Delaunay triangulations is particularly su ited when we do not want to force any constraints on the set of points to be connected. By means of Delaunay triangulations, we directly quantify the local spatial structure of the surface-attached hydrogel s by measuring the distribution of the microspheres inside the polymer.

PAGE 62

51 photodetector objective lens focal plane In sample confocal aperture light source illumination aperture dichoic mirror x z Figure 3-1. Schematic diagram of the confocal microscope principle

PAGE 63

Figure 3-2. Schematic diagram demonstrating the principle of optical sectioning 52

PAGE 64

53 Focal plane Figure 3-3. Diagram showing generati on of optical sections in the confocal microscope

PAGE 65

54 Objective lens AB Coverslip Plexiglass Focal plane in sample Figure 3-4. Comparison of specimen illumination in confocal (A) and conventional (B) microscopes

PAGE 66

55 Objective lens Figure 3-5. Comparison of confocal and conv entional light microscope when the object lens is focused well below the fl uorescent sample, depicted by the small rectangular box

PAGE 67

Figure 3-6. Nipkow disk 56

PAGE 68

Figure 3-7. Spinning-disk confocal microscope 57

PAGE 69

58 x y x A B Figure 3-8. The raster scan pattern that (A) confocal laser scanning microscopes and (B) spinning-disk conf ocal microscopy commonly use to generate images

PAGE 70

59 objective lens focal plane In sample dichoic mirror arc lamp motor Nipkow disk /4 wave plate polarizer eye piece analyzer x z Focal plane Figure 3-9. Schematic diagram of the optical path in a monoscanning confocal microscope

PAGE 71

60 Figure 3-10. Epifluorescence microscope principle

PAGE 72

A B A B Figure 3-11. Background subtraction using the wavelength of the fluorescence microspheres 61

PAGE 73

62 CHAPTER FOUR: RESULTS, DISCUSSION, AND CONCLUSIONS As stated in Chapter One, the study reported here examined in detail the techniques used for the fabrication of poly(NIPAAm) microstructures on silane coated surfaces. These techniques required the overcoming of several challenges that will be address in this se ction. In addition, the successful attachment of the polymer networks to the glass surface will be illustrated with spinning-disk confocal microscopy im ages. Furthermore, the swelling and desswelling ( Figure 4-1 shows a cross-sectional section of the trenches in the collapse state); the buckli ng and un-buckling effect of the polymer trenches (see Figure 4-2 ) will be analyzed by tracking the microspheres displacement within the sample. Finally, the techniques used to measure the local swelling of the structure will be correlated with an m-f ile routine that uses Delaunay triangulations to find the distance of the neighbor ing microspheres. 4.1. Soft Lithography Delim itations and Approaches The fabrication of SU-8 microstructures (see Figure 4-3 ) required the overcoming of several challenges that have been mentioned along this thesis. These process obstacles and complications resulted in SU-8 delaminating,

PAGE 74

63 feature alteration, underdevelopment as well as overdevelopment. All of these problems can be associated with the spinning, ramping velocities exposure time, and development time. As you can see, each part of the process presented a certain degree of challenge. Feature roughness was highly sensitive to variations in exposure uniformity, which arose from SU-8 thickness, non-uniformity as well as non-uniform contact between photo mask and the SU-8 layer. Therefore, special care was taken at the initial of the soft bake period especially for thicknesses higher than 25 m. In addition, the bead formation during the spinning step was highly characterized in order to minimize the bead removal process which can be pretty invasive to the SU-8. Moreover, the contact between the photo mask and SU-8 was decreased at maximum by upgrading from an initial process which used to use hard contact to a process in which vacuum was used in order to maximize contact and minimize air diffraction. On the other hand, the total thickne ss of the SU-8 microstructure was another important variable in exposure uniformity. In addition, optimizing exposure dosage (power and time) also improved the exposure uniformity. Crack-like distortions (wrinkles or cracks) that manifest during the post-exposure bake can result from underexposure of the SU-8. Finally, bake temperatures, developing time, and agitation played a significant role at the time of characterizing the process of fabric ating the SU-8 microstructures.

PAGE 75

64 4.2. 3D Reconstruction There were several parameters that had to be considered before acquiring a z-series suitable for 3D reconstruction in order to obtain a stack with minimal distortions. 4.2.1. Fluidic Chamber Considerations It was necessary to employ the greatest care when preparing a sample for 3-D reconstruction. Improper sample preparation will produce distorted 3-D images. It is important to notice that the further one focuses in the z-axis of the sample, the more prominent loss of c ontrast and shadowing become so that upper bright planes appear dim, while lower ones are bright. As you can see from Figure 4-4 (A) shows the z-axis of the micr ospheres closest to the lens; these microspheres appear larger than the ones fr om a higher z position (B). One could increase the illumination intensity to overcome some of the problem, yet photobleaching may become severe. 4.3. Image Registration Another requirement for the effe ctive 3-D reconstruction is image registration. Each optical section of t he z-series must be in register with its

PAGE 76

65 neighbors during image acquisi tion; otherwise distortion is introduced into the 3D reconstruction. Hence, care must be taken to prevent the specimen from moving during the acquisition of the z-series. Since the heating of the chamber is done by flowing hot air through the sample the chamber had to be sealed on the top. On the other hand, it wa s not possible to take measurements on the confocal microscopy when the microstructures where fabricated wit hout the TPM binder since there was not a relative structure where to focus the microscope at. 4.4. Pixel Resolution A third important consideration before obtaining a z-series is the pixel resolution for the conditions used (objecti ve, magnification, etc) so that the proportional dimensions of the spec imen are maintained during the 3D reconstruction. For example, in Figure 4-5 (A), too few sections were taken resulting in a squashed, three-dimensional reconstruction (also, refer to Figure 46 which shows an actual exampl e from our data). In (B) t he pixel resolution in the xy plane was used to generate the correct di stance to obtain optical sections in the z axis. As expected, a near-perfect sphe re is expected. (C) Shows that in this example too many optical section were taken, and the spher e appears stretched after three-dimensional reconstruction.

PAGE 77

66 4.5. Image Processing of the Stacks In order to obtain a 3D reconstruction of the microstructure there are some steps that should be considered if one is working with ImageJ. First, the stack should be downloaded as an image seque nce. Then, a minimum and a maximum threshold should be applied in order to reduce background noise. In the data shown in this thesis a 15% thres hold is set relative to the maximum in each image. After the background noise has been clear out with the threshold tool on ImageJ, we can proceed to find the x, y, and z positions of the microspheres. This was done by choosing a region of interest (ROI) through out the stack (the ROI will be a single trench) (see Figure 4-7 ). Once the ROI is selected one can proceed to crop this region, so only the microspheres inside the trench will be accounted in the calculations. This will r educe the amount of time ImageJ takes to calculate the coordinat es. Finally, using the ImageJ algorithm called D Object Counter the volume, the surface area, the intensity and the coordinates of the spher es is calculated. 4.6. Delaunay Triangulations Results As mentioned in the previous chapter, a Delaunay triangulation algorithm obtained from Matlab was used to calc ulate the mean dist ribution of the microspheres in the network. Figure 4-8 (A) and (B) shows the Matlab output from the Delaunay triangulat ion algorithm of a poly(NI PAAm) trench that has 15

PAGE 78

67 m gap between trenches and 10 m width. (C) and (D) shows the histogram of the distances between the microspheres. The calculated mean distance for (A) was 4.97 m and 2.85 m for (B); this is a signifi cant 57% difference between the swollen and collapse state. Figure 4-11 and Figure 4-12 show the mean distance of microspheres in different trenches widths (for additi onal information please refer to Figure 4-9 and Figure 4-10 ). Also, the gaps between the tr enches are changed in order to characterize the closing and opening of these trenches. 4.7. Measuring Local Swelling Another useful application of t he method presented here entails the calculation of the local swelling of the microstructure by sectioning along the zaxis (refer to Figure 4-13 ). By dividing the z-axis into several sections one can calculate the mean distance of the micros pheres in those sections. This method is especially useful in trying to understand how different patterns of surfaceattached microstructures of poly(NI PAAm) swelled and collapse along their zaxis. Figure 4-14 shows the Matlab plots of a 52 m microstructure below the LCST that was divided into four sections. The mean di stance of the microspheres and the standard error of t he mean were calculated in order to understand the swelling behavior of the trench in each sect ion. The trench in this example came from a PDMS mold with a gap distance of 15 m, a width of 10 m, and a

PAGE 79

68 thickness of 25 m. The mean distance of the micr ospheres in the four sections indicates that the trench reaches a maximum swelling approximately three fourths of its thickness. This behavior resembles a mushroom like shape like the ones reported by Castellanos (2007). 4.8. Conclusions The successful attachment of surf ace-attached poly(NIPAAm) networks to the glass surface was illus trated with spinning-disk conf ocal microscopy images. In addition, the Delaunay triangulati on algorithm obtained from Matlab was proven to successfully calculate the mean distribution of the microspheres in the network. Mean distance of microspheres in different trenches widths and thicknesses was calculated in order to understand the trenches separation influence in the swelling of the patterned hydrogels. These measurements create, for the first time, swelling maps of the structures based on relative movements of fluorescent microspheres. Th is information will ultimately aid in understanding how swollen macroscopic stru ctures are related to gradients in localized swelling Furthermore, the quantitative data from this technique may be used to verify the validity of existing hydrogel expansion and contraction models or lead to development of refined theorie s regarding hydrogel response.

PAGE 80

69 Finally, although this set of tec hniques has been demonstrated for a thermo-responsive hydrogel, its application is not limited only to this class of materials. It can also be applied to hydrogels engineered to respond to other external stimuli, such as ph, light, etc. In addition, this technique could possibly be used on other translucent materials utilized in the construction of more intelligent device designs.

PAGE 81

Figure 4-1. Cross-sectional area of trenches 70

PAGE 82

Figure 4-2. Un-buckling of trenches bel ow and above the LCST 71

PAGE 83

Figure 4-3. SU-8 patterned silicon wafer 72

PAGE 84

B A Figure 4-4. Loss of contrast between lo wer and upper focal planes 73

PAGE 85

AB C AB C Figure 4-5. Three-dimensional reconstructi on of a spherical fluorescent microsphere 74

PAGE 86

75 Figure 4-6. Cross-sectional view of mi crospheres where too few optical sections were taken resulting in a s quashed, three-dimens ional reconstruction

PAGE 87

A. B. C. A. B. C. Figure 4-7. Bright field and 3D images of isolated trench below and above the LCST. Cropped trench from a bright fiel d image (A). 3D image from below (B) and above (C) the LCST 76

PAGE 88

A B C D Figure 4-8. Delaunay triangulation of a poly(NIPAAm) trench (A) below (B) above the LCST 77

PAGE 89

1 2 3 4 5 6 7 8 9 10 W X Y Z 1 2 3 4 5 6 7 8 9 10 W X Y Z Rowg w ( in m ) ( in m ) 12015 21510 31010 4510 555 635 744 8104 925100 10103 Figure 4-9. Mask with different trenches widths and gaps 78

PAGE 90

79 h = thickness of microstructure w = trenches width g = gap between trenches g w h Figure 4-10. 3D of microstructure patterns

PAGE 91

Figure 4-11. Mean distance of microspheres vs. different size trenches widths and gaps (25 um thick mold) 80

PAGE 92

Figure 4-12. Mean distance of microspheres vs. different size trenches widths and gaps (5 um thick mold) 81

PAGE 93

A. B. C. Divide the z-axis into sections 1 2 3 4 Cropped ROI A. B. C. Divide the z-axis into sections 1 2 3 4 A. B. C. Divide the z-axis into sections 1 2 3 4 Divide the z-axis into sections 1 2 3 4 Divide the z-axis into sections 1 2 3 4 Cropped ROI 82 Figure 4-13. Steps to measure the local swelli ng in different sections of the microstructure. Crop ROI (A ). Arrange stack in 3D ( B ). Divide z-axis into sections and calculate microsphere distances in these sections ( C )

PAGE 94

Figure 4-14. Calculating mean distance of micr ospheres in different z-sections 83

PAGE 95

84 REFERENCES 1. J. Hoffmann, M. Plotner, D. Kuckli ng and W. J. Fischer, Photopatterning of thermally sensitive hydrogels useful for microactuators. Sens. Actuator APhys. 1999, 77, (2), 139-144. 2. H. Tanaka, H. Tomita, A. Takasu, T. Hayashi and T. Nishi, Morphological And Kinetic Evolution Of Surface Patte rns In Gels During The Swelling Process Evidence Of Dynamic Pattern Ordering. Phys. Rev. Lett. 1992, 68, (18), 2794-2797. 3. A. Suzuki, M. Yamazaki and Y. Kobi ki, Direct observation of polymer gel surfaces by atomic force microscopy. Journal of Chemical Physics 1996, 104, (4), 1751-1757. 4. A. Suzuki, S. Yoshikawa and G. Bai, Shrinking pattern and phase transition velocity of poly(N-isopropylacrylamide) gel. Journal of Chemical Physics 1999, 111, (1), 360-367. 5. B. H. Jo, L. M. Van Lerberghe, K. M. Motsegood and D. J. Beebe, Threedimensional micro-channel fabrication in polydim ethylsiloxane (PDMS) elastomer. Journal of Microelectromechanical Systems 2000, 9, (1), 7681. 6. A. Castellanos, S. J. DuPont, A. J. Heim, G. Matthews, P. G. Stroot, W. Moreno and R. G. Toomey, Size-e xclusion "capture and release" separations using surface-patte rned poly(N-isopropylacrylamide) hydrogels. Langmuir 2007, 23, (11), 6391-6395. 7. A. Bertsch, S. Jiguet and P. R enaud, Microfabrication of ceramic components by microstereolithography. Journal of Micromechanics and Microengineering 2004, 14, (2), 197-203.

PAGE 96

85 8. P. A. Clerc, L. Dellmann, F. Gretillat, M. A. Gretilla t, P. F. Indermuhle, S. Jeanneret, P. Luginbuhl, C. Marxer, T. L. Pfeffer, G. A. Ra cine, S. Roth, U. Staufer, C. Stebler, P. Thiebaud and N. F. de Rooij, Advanced deep reactive ion etching: a versatile t ool for microelectromechanical systems. J. Micromech. Microeng., 1998, 8, (4), 272-278. 9. Y. Hanein, C. G. J. Schabmueller, G. Holman, P. Lucke, D. D. Denton and K. F. Bohringer, High-aspect ratio submicrometer needles for intracellular applications. J. Micromech. Microeng. 2003, 13, (4), S91-S95. 10. J. L. Wilbur, R. J. Ja ckman, G. M. Whitesides, E. L. Cheung, L. K. Lee and M. G. Prentiss, Elastomeric optics. Chem. Mat. 1996, 8, (7), 1380-1385. 11. J. A. Rogers, R. J. Jackman, O. J. A. Schueller and G. M. Whitesides, Elastomeric diffraction gratings as photothermal detectors. Appl. Optics 1996, 35, (34), 6641-6647. 12. Y. N. Xia and G. M. Whitesides, Soft lithography. Annual Review of Materials Science, 1998, 28, 153-184. 13. A. Mata, A. J. Fleischman and S. Ro y, Fabrication of multi-layer SU-8 microstructures. J. Micromech. Microeng. 2006, 16, (2), 276-284. 14. R. S. Kane, S. Takayama, E. Ost uni, D. E. Ingber and G. M. Whitesides, Patterning proteins and cells using soft lithography. Biomaterials 1999, 20, (23-24), 2363-2376. 15. K. L. Chavez and D. W. Hess, Removal of resist materials using acetic acid. Journal of the Electrochemical Society, 2003, 150, (4), G284-G291. 16. A. Revzin, R. J. Russell, V. K. Yadav alli, W. G. Koh, C. Deister, D. D. Hile, M. B. Mellott and M. V. Pishko, Fabr ication of poly(ethylene glycol) hydrogel microstructures using photolithography. Langmuir 2001, 17, (18), 5440-5447.

PAGE 97

86 17. W. G. Koh, A. Revzin, A. Simonian T. Reeves and M. Pishko, Control of mammalian cell and bacteria adhesion on substrates micropatterned with poly(ethylene glycol) hydrogels. Biomedical Microdevices 2003, 5, (1), 1119. 18. H. Lorenz, M. Despont, N. Fahrni, J. Brugger, P. Vettiger and P. Renaud, High-aspect-ratio, ultrathick, negative-tone near-UV photoresist and its applications for MEMS. Sens. Actuator A-Phys. 1998, 64, (1), 33-39. 19. K. D. Vora, B. Y. Shew, E. C. Ha rvey, J. P. Hayes and A. G. Peele, Specification of mechanical support stru ctures to prevent SU-8 stiction in high aspect ratio structures. J. Micromech. Microeng. 2005, 15, (5), 978983. 20. J. Zhang, M. B. Chan-Park and S. R. Conner, Effect of exposure dose on the replication fidelity and profile of very high aspect ratio microchannels in SU-8. Lab on a Chip 2004, 4, (6), 646-653. 21. C. H. Lin, G. B. Lee, B. W. Chang and G. L. Chang, A new fabrication process for ultra-thick microfluid ic microstructures utilizing SU-8 photoresist. Journal of Micromechanics and Microengineering, 2002, 12, (5), 590-597. 22. E. H. Conradie and D. F. Moore, SU-8 thick phot oresist processing as a functional material for MEMS applications. J. Micromech. Microeng. 2002, 12, (4), 368-374. 23. C. L. Cromer, T. B. Lucatorto, T. R. Obrian and M. Walhout, Improved dose metrology in optical lithography. Solid State Technology 1996, 39, (4), 75-78. 24. J. Tien, C. M. Nelson and C. S. Chen, Fabricat ion of aligned microstructures with a single elastomeric stamp. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, (4), 1758-1762. 25. K. Kim, D. S. Park, H. M. Lu, W. Che, J. B. Lee and C. H. Ahn, A tapered hollow metallic microneedle array using backside exposure of SU-8. J. Micromech. Microeng. 2004, 14, (4), 597-603.

PAGE 98

87 26. J. Carlier, S. Arscott, V. Thomy, J. C. Fourrier, F. Caron, J. C. Camart, C. Druon and P. Tabourier, Integrated mi crofluidics based on multi-layered SU-8 for mass spectrometry analysis. J. Micromech. Microeng. 2004, 14, (4), 619-624. 27. M. B. Chan-Park, J. Zhang, Y. H. Y an and C. Y. Yue, F abrication of large SU-8 mold with high aspect ratio mi crochannels by UV exposure dose reduction. Sensors and Actuators B-Chemical 2004, 101, (1-2), 175-182. 28. Y. J. Chuang, F. G. Tseng and W. K. Lin, Reduction of diffraction effect of UV exposure on SU-8 negative thick phot oresist by air gap elimination. Microsystem Technologies 2002, 8, (4-5), 308-313. 29. M. C. Peterman, P. Huie, D. M. Bloom and H. A. Fishman, Building thick photoresist structures from the bottom up. J. Micromech. Microeng. 2003, 13, (3), 380-382. 30. X.-Z. Zhang, G.-M. Sun and C. C. Chu, Temperature sensitive dendriteshaped PNIPAAm/Dex-AI hybrid hydrogel particles: formulation and properties. European Polymer Journal 2004, 40, (9), 2251-2257. 31. B. Bohl, R. Steger, R. Zengerle and P. Koltay, Multi-layer SU-8 lift-off technology for microfluidic devices. J. Micromech. Microeng. 2005, 15, (6), 1125-1130. 32. H. K. Chang and Y. K. Kim, UV -LIGA process for high aspect ratio structure using stress barrier and C-shaped etch hole. Sens. Actuator APhys. 2000, 84, (3), 342-350. 33. J. D. Williams and W. Wang, Microfabrication of an electromagnetic power relay using SU-8 based UV-LIGA technology. Microsyst. Technol. 2004, 10, (10), 699-705. 34. J. Wang and Z. Chen, Self-Act uated, Thermo-Responsive Hydrogel Valves for Lab on a Chip. Biomedical Microdevices, 2005, 7, (4), 313-322.

PAGE 99

88 35. J. M. K. Ng, I. Gitlin, A. D. St roock and G. M. Whitesides, Components for integrated poly(dimethylsilox ane) microfluidic systems. Electrophoresis 2002, 23, (20), 3461-3473. 36. A. D. Stroock, S. K. W. Dertinger, A. Aj dari, I. Mezic, H. A. Stone and G. M. Whitesides, Chaotic mixer for microchannels. Science 2002, 295, (5555), 647-651. 37. W. H. Grover, A. M. Skelley, C. N. Liu, E. T. Lagally and R. A. Mathies, Monolithic membrane valves and diaphr agm pumps for practical largescale integration into glass microfluidic devices. Sens. Actuator B-Chem. 2003, 89, (3), 315-323. 38. H. Takao, K. Miyamura, H. Ebi, M. Ashiki, K. Sawada and M. Ishida, A MEMS microvalve with PDMS diaphr agm and two-chamber configuration of thermo-pneumatic actuator for in tegrated blood test system on silicon. Sens. Actuator A-Phys. 2005, 119, (2), 468-475. 39. E. Delamarche, H. Schmid, B. Mic hel and H. Biebuyck, Stability of molded polydimethylsiloxane microstructures. Adv. Mater. 1997, 9, (9), 741-746. 40. P. Silberzan, L. Leger, D. Ausserr e and J. J. Benattar, SILANATION OF SILICA SURFACES A NEW METHOD OF CONSTRUCTING PURE OR MIXED MONOLAYERS. Langmuir, 1991, 7, (8), 1647-1651. 41. R. Banga, J. Yarwood, A. M. Mo rgan, B. Evans and J. Kells, FTIR AND AFM STUDIES OF THE KINETICS AND SELF-ASSEMBLY OF ALKYLTRICHLOROSILANES AND (PERFLUOROALKYL)TRICHLOR OSILANES ONTO GLASS AND SILICON. Langmuir 1995, 11, (11), 4393-4399. 42. J. B. Brzoska, I. Benazouz and F. Rondelez, Silanization of Solid Substrates a Step to ward Reproducibility. Langmuir 1994, 10, (11), 4367-4373. 43. D. Kim and D. J. Beebe, Hydrogel -based reconfigurable components for microfluidic devices. Lab Chip 2007, 7, (2), 193-198.

PAGE 100

89 44. M. J. Park, W. M. Choi and O. O. Park, Patterning pol ymer light-emitting diodes by micromoldi ng in capillary. Current Applied Physics, 2006, 6, (4), 627-631. 45. N. F. Sheppard, M. J. Lesho and P. McNally, MICROFABRICATED CONDUCTIMETRIC PH SENSOR. Sens. Actuator B-Chem. 1995, 28, (2), 95-102. 46. L. E. Helseth and T. M. Fisc her, Fundamental limits of optical microrheology. Journal of Colloid and Interface Science 2004, 275, (1), 322-327. 47. T. Tanaami, S. Otsuki, N. Tomos ada, Y. Kosugi, M. Shimizu and H. Ishida, High-speed 1-frame/ms scanni ng confocal microscope with a microlens and Nipkow disks. Appl. Optics 2002, 41, (22), 4704-4708. 48. M. C. Adams, W. C. Salmon, S. L. Gupton, C. S. Cohan, T. Wittmann, N. Prigozhina and C. M. Waterman-St orer, A high-speed multispectral spinning-disk confocal microscope system for fluorescent speckle microscopy of living cells. Methods 2003, 29, (1), 29-41.


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 002007183
003 fts
005 20091009142540.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 090617s2008 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002788
035
(OCoLC)401725965
040
FHM
c FHM
049
FHMM
090
TP145 (Online)
1 100
Bello, Carlos A.
0 245
Microsphere-aided characterization of stimuli-responsive polymer networks
h [electronic resource] /
by Carlos A. Bello.
260
[Tampa, Fla] :
b University of South Florida,
2008.
500
Title from PDF of title page.
Document formatted into pages; contains 89 pages.
502
Thesis (M.S.Ch.)--University of South Florida, 2008.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
3 520
ABSTRACT: The fabrication and characterization of surface-anchored hydrogel microstructures are described. The hydrogel structures are constructed from poly(N-isopropylacrylamide), or poly(NIPAAm), which is a well-known thermoresponsive polymer that swells and contracts with changes in temperature. When patterned on a surface, these structures can experience a variety of shape changes induced by nonuniform swelling. Depending on the aspect ratio, patterns can, for instance buckle upon swelling and form wave-like patterns. Such structural changes replicate oscillatory motion of the smooth muscle cells and can be used to transport objects in microfluidics. The work, herein, investigates methods of pattern production and introduces a new technique for characterizing local swelling in the patterns. In order to achieve the latter, fluorescent microspheres were embedded in hydrogel patterns and their positions were mapped in three-dimensions using confocal microscopy. The measurements permit, for the first time, swelling maps of the structures based on relative movements of the microspheres. This information will ultimately aid in understanding how swollen macroscopic structures are related to gradients in localized swelling.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor: Ryan G. Toomey, Ph.D.
653
Soft lithography
Confocal microscopy
Microstructures
Poly(NIPAAm)
Delaunay triangulation
690
Dissertations, Academic
z USF
x Chemical & Biomedical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.2788