University of South Florida

Scholar Commons
Graduate Theses and Dissertations Graduate School

2010

Influence of crosslink density on swelling and

conformation of surface-constrained Poly(N-
isopropylacrylamide) hydrogels
Ryan S. Cates
University of South Florida

Follow this and additional works at: http://scholarcommons.usf.edu/etd

Part of the American Studies Commons

Scholar Commons Citation

Cates, Ryan S., "Influence of crosslink density on swelling and conformation of surface-constrained Poly(N-isopropylacrylamide)
hydrogels" (2010). Graduate Theses and Dissertations.
http://scholarcommons.usf.edu/etd/1592

This Thesis is brought to you for free and open access by the Graduate School at Scholar Commons. It has been accepted for inclusion in Graduate
Theses and Dissertations by an authorized administrator of Scholar Commons. For more information, please contact [email protected].
 Influence of Crosslink Density on Swelling and Conformation of Surface-Constrained
Poly(N-Isopropylacrylamide) Hydrogels

by

Ryan S. Cates

A Thesis submitted in partial fulfillment

of the requirements for the degree of
Master of Science in Chemical Engineering
Department of Chemical and Biomedical Engineering
College of Engineering
University of South Florida

Major Professor: Ryan G. Toomey, Ph.D.

Nathan D. Gallant, Ph.D.
Babu Joseph, Ph.D.

Date of Approval:
March 31, 2010

Keywords: Crosslinked polymer, thermoresponsive polymers, Flory-Rehner theory, soft

lithography, confocal microscopy.

©Copyright 2010, Ryan S. Cates

 TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. iv

LIST OF FIGURES ............................................................................................................ v

ABSTRACT ...................................................................................................................... ix

CHAPTER ONE: INTRODUCTION, MOTIVATION AND BACKGROUND ....................... 1

1.1. Introduction to Poly(N-Isopropylacrylamide) ....................................................... 1

CHAPTER TWO: FABRICATION OF POLY(N-ISOPROPYLACRYLAMIDE)

MICROSTRUCTURES ...................................................................................................... 8

2.1 Creation of Photolithographic Mask Using AutoCAD............................................... 9

2.1.1. Monolithe I ....................................................................................................... 9

2.1.2. Monolithe II .................................................................................................... 14

2.2. Photolithography Process ..................................................................................... 16

2.2.1. Piranha Clean ................................................................................................ 16

2.2.3. Photoresist Application .................................................................................. 17

2.2.4. Soft Bake ....................................................................................................... 19

2.2.5. Exposure ........................................................................................................ 19

2.2.6. Hard Bake ...................................................................................................... 21

2.2.7. Development .................................................................................................. 21

2.2.8. Characterization ............................................................................................. 22

2.2.9. Extended Hard Bake ...................................................................................... 22

i
 2.3. Soft Lithography ................................................................................................... 23

2.3.1. Micro Molding in Capillary (MIMIC) ................................................................ 28

2.4. Application of Silane Binder to Substrate ............................................................. 28

2.5. Photopolymerization of Microgels ......................................................................... 33

2.6. Fabrication of Fluid Cells ...................................................................................... 36

2.7. Remarks ............................................................................................................... 38

CHAPTER THREE: CHARACTERIZATION OF EXTENT OF SWELLING AND

CROSSLINK DENSITY ................................................................................................... 39

3.1. Introduction ........................................................................................................... 39

3.2. Characterization of Extent of Swelling in Bulk Gels .............................................. 39

3.2.1. Experimental Procedure ................................................................................ 40

3.2.2. Remarks and Results ..................................................................................... 46

3.3. Crosslink Density Determination .......................................................................... 63

3.4 Remarks and Results ............................................................................................ 65

CHAPTER FOUR: CHARACTERIZATION OF CONFORMATION OF PNIPAAm

MICROSTRUCTURES BY LASER SCANNING CONFOCAL MICROSCOPY............... 69

4.1. Laser Scanning Confocal Microscopy Introduction............................................... 69

4.2. Characterization of Structural Morphology by Confocal Microscopy .................... 72

4.3. Remarks and Results ........................................................................................... 72

CHAPTER FIVE: SUMMARY, CONCLUSION AND FUTURE WORK ........................... 77

5.1. Preparing Stock Solutions .................................................................................... 77

5.2. Sample Preparation .............................................................................................. 78

ii
 5.3. Experimental Setup and Data Collection .............................................................. 78

5.4. Confocal Microscopy ............................................................................................ 78

5.5. Future Works ........................................................................................................ 79

5.5.1. Soft Lithography ............................................................................................. 79

5.5.2. Preparing Stock Solutions .............................................................................. 79

5.5.3. Confocal Microscopy Characterization .......................................................... 80

REFERENCES ................................................................................................................ 81

iii
 LIST OF TABLES

Table 2-1: Dimensions key for Monolithe I schematic ..................................................... 14

Table 2-2: SU-8 spin recipes used and the results achieved .......................................... 18

Table 2-3: Recommended soft bake times for different SU-8 thicknesses ..................... 18

Table 2-4: 125% of the recommended UV dosages for different SU-8s and

desired thicknesses ...................................................................................... 18

Table 2-5: Recommended post bake time for different SU-8 samples and

thicknesses .................................................................................................. 218

Table 2-6: Developing times for different SU-8 samples and thicknesses ...................... 18

Table 3-1: Initial prepolymer solutions recipe .................................................................. 40

Table 3-2: BIS dilution chart ............................................................................................ 41

Table 3-3: Revised prepolymer recipe for equal volume contribution ........................... 433

Table 3-4: Dilution chart for BIS ...................................................................................... 43

Table 3-5: Revised dilution chart for BIS ......................................................................... 43

Table 3-6: Crosslink density values (mol/cm3) calculated from the swelling data

using the Flory-Rehner equation .................................................................. 66

iv
 LIST OF FIGURES

Figure 1-1: A representation of the swelling and deswelling of PNIPAAm……………......2

Figure 1-2: Hydrogel in three stages: fully swollen (left), after polymerization

with ambient moisture content (center), fully collapsed (right). *Eraser

added for size reference* ................................................................................ 3

Figure 1-3: Cross-sectional profile of a surface-confined hydrogel in its swollen

(a) and unswollen (b) states…………………………………. ......................... .. 4

Figure 1-4: Surface confined microgel in the swollen and unswollen state. The

swollen state displays mechanical deformation of the microgel. ..................... 5

Figure 1-5: PNIPAAm samples: fully swollen at highest crosslink density (20 wt %

BIS, left), at the lowest crosslink density (1.43 wt%, right) and the

initial, unswollen state (top). ......…………………………………………………..6

Figure 1-6: Surface-confined PNIPAAm samples swelling as a function of

crosslink density...............................................................................................7

Figure 2-1: Photolithography process………………… ………………………….……….. ....9

Figure 2-2: Underdeveloped photoresist pattern and the trapping of the molding

polymer......................................................... ................ .................................11

Figure 2-3: AutoCAD schematic of the “Monolithe I” photolithographic mask.. ………….13

Figure 2-4: AutoCAD schematic of “Monolithe II” photolithographic mask. …….………..15

Figure 2-5: Close-up of the 40x20x40 µm structure on the AutoCAD schematic of

the “Monolithe II” photolithographic mask……………………… . ……………..16

Figure 2-6: Air bubbles in a PDMS stamp still on an SU-8 master mold………… …......26

Figure 2-7: PDMS relief in petri dish…………………………………………… …………...27

v
Figure 2-8: Micro injection molding in capillaries (MIMIC) process………………..……...28

Figure 2-9: Silanization mechanism………………………………………………… ..... ……30

Figure 2-10: Dirty slide leading to heavy silane deposition………………………… . …….31

Figure 2-11: Polymer microstructure with surrounding polymer scum layer…………. ….32

Figure 2-12: Deformed polymer pad (top-right corner) with surrounding scum

layer exhibiting surface wrinkling……………………………………... ……....32

Figure 2-13: Slide with crystallized monomers (prepolymer solution) on it......... .. ......…34

Figure 2-14: Confocal microscope image of crystallized monomer on the slide

surface………………………………………………………………………… . ..34

Figure 2-15: Top and side view of assembled fluid cell................................ ...................36

Figure 2-16: Top and bottom view of fluid cell with sample slide adhered.......... .... ....... 37

Figure 2-17: Chronicle evolution of fluid cell design. First design (left) was an

enclosed design that used a plexiglass cover with ports. The latter

two (to the right) were open top designs with open tops…. ...... ..................37

Figure 3-1: Solubility plot of BIS concentration verses water to acetone volume

ratios.. .......................................................................................................... ..42

Figure 3-2: Swelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations... ................................................................................................

47

Figure 3-3: Swelling of 1ml PNIPAAm samples with only four select weight ratios

of BIS/NIPAAm shown................................................................................... 48

Figure 3-4: Swelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations... ............................................................................................ 49

Figure 3-5: Swelling of 1ml PNIPAAm samples with only four select weight ratios

of BIS/NIPAAm shown................................................................................... 50

vi
Figure 3-6: Swelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations... ............................................................................................ 51

Figure 3-7: Swelling of 1ml PNIPAAm samples with only four select weight ratios

of BIS/NIPAAm shown................................................................................... 52

Figure 3-8: Swelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations... ............................................................................................ 53

Figure 3-9: Swelling of 1ml PNIPAAm samples with only four select weight ratios

of BIS/NIPAAm shown................................................................................... 54

Figure 3-10: Deswelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.... ........................................................................................ 55

Figure 3-11: Deswelling of 1ml PNIPAAm samples with only four select weight

ratios of BIS/NIPAAm shown..... ................................................................. 56

Figure 3-12: Deswelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.... ........................................................................................ 57

Figure 3-13: Deswelling of 1ml PNIPAAm samples with only four select weight

ratios of BIS/NIPAAm shown..... ................................................................. 58

Figure 3-14: Deswelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.... ........................................................................................ 59

Figure 3-15: Deswelling of 1ml PNIPAAm samples with only four select weight

ratios of BIS/NIPAAm shown..... ................................................................. 60

Figure 3-16: Deswelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.... ........................................................................................ 61

Figure 3-17: Deswelling of 1ml PNIPAAm samples with only four select weight

ratios of BIS/NIPAAm shown..... ................................................................. 62

Figure 3-18: Collapsing gel that forms an outer skin layer.. ............................................ 67

Figure 3-19: A PNIPAAm that fractured in the process of swelling.. ............................... 68

vii
Figure 4-1: A schematic of the confocal microscope concept.... ..................................... 70

Figure 4-2: Difference in illumination volume between wide-field and confocal

microscopy.. .................................................................................................. 71

Figure 4-3: Increased number of stacks to compensate for uneven adhesive layer.. ..... 73

Figure 4-4: PNIPAAm monolith in the collapsed state.... ................................................ 74

Figure 4-5: PNIPAAm monolith in the collapsed state..... ............................................... 74

Figure 4-6: PNIPAAm monolith exhibiting surface deformations in the swollen

state.. ............................................................................................................. 75

Figure 4-7: PNIPAAm monolith exhibiting bulk mechanical deformation in the

swollen state .................................................................................................. 75

viii
 INFLUENCE OF CROSSLINK DENSITY ON SWELLING AND CONFORMATION OF

POLY(N-ISOROPYLACRYLAMIDE) HYDROGELS

Ryan Scott Cates

ABSTRACT

A stimuli-responsive microgel is a three-dimensional polymer network that is able

to absorb and expel a solvent (commonly water). These materials are unique in the fact

that their sponge-like behavior can be actuated by environmental cues, like temperature,

ion concentration, pH, and light. Because of the dynamic properties of these materials

they have found applications in drug-delivery systems, micro-assays, selective filtration,

artificial muscle, and non-fouling surfaces. The most well-known stimuli-responsive

polymer is Poly(N-isopropylacrylamide) or PNIPAAm and it experiences a switchable

swelling or deswelling over a critical temperature ( T c ≅ 3 2 °C). Below the critical

temperature, the gel begins mixing with the surrounding solvent and swells; above this

temperature, the opposite is true. The unconstrained hydrogel will continue to swell in all

directions until equilibrium is established between its propensity for mixing with the

surrounding solvent and the elastic restoring forces of the gel matrix. The strength of the

elastic restoring forces is dependent on the interconnectedness of the polymer network

and is therefore a function of crosslink density. An increase in crosslink density results in

ix
a decreased swelling and vice versa. If the hydrogel is mechanically constrained to a

surface, it can experience various wrinkling and buckling conformations upon swelling,

as the stresses associated with its confinement are relieved. These conformation

characteristics are a strong function of geometry (aspect ratio) and extent of swelling

(i.e. crosslink density). In order to capitalize on the utility of this material, it is imperative

that its volume transition is well characterized and understood.

Toward this end, pNIPAAm gels have been created with 1x10-7 to 2x10-3 mol/cm3

crosslink density and characterized. This was done by first examining its bulk,

unattached swelling ability and then by evaluating its microscale properties as a surface-

confined monolithe. The latter was achieved through the use of confocal microscopy and

copolymerization with a fluorescent monomer. This method allows for a detail analysis of

the deformations experienced (bulk-structural bending and surface undulating) and will

ultimately lend itself to the correlation between crosslink density and the onset of

mechanical phenomena.

x
 CHAPTER ONE: INTRODUCTION, MOTIVATION AND BACKGROUND

1.1. Introduction to Poly(N-Isopropylacrylamide)

Stimuli responsive polymers are unique materials with applications in drug-

delivery agents[1], microfluidics[2-5], molecular capture and release[6], and non-fouling

surfaces[7, 8]. Poly(N-isopropylacrylamide) or PNIPAAm, is the most well-known

thermo-responsive hydrogel or water containing gel, which experiences a phase change,

causing swelling or deswelling, over some critical temperature ( Tc ≅ 32 °C) [9-11]. Below

the lower critical solution temperature (LCST), the gel is soluble in the surrounding

solvent and the interconnected chains begin to extend and further mix with the solvent.

Here the free energy of mixing (i.e. the energy that can be converted to do work) is

greater than the elastic energy of the polymer chains (i.e. spring-like energy) which had

previously kept the polymer matrix coiled or collapsed. Visuals of these coiled and

collapsed states are provided in figure 1-1. Effectively, the driving force to mix with the

solvent outweighs the forces keeping the polymer matrix collapsed. The exact reverse

relationship is observed above the LCST. The swelling or deswelling of the gel matrix

comes to a final equilibrium state when free energy is minimized (i.e. the two energies

balance one another) [12, 13].

1
Figure 1-1: A representation of the swelling and deswelling of PNIPAAm.

PNIPAAm can take on different physical and chemical properties through

copolymerization and functionalization which gives its gels application specific tunability

[15]. Through copolymerization with other monomers, PNIPAAm is able to be responsive

to environmental cues other than temperature. Additional, copolymerization can be used

to control the threshold value of stimulus needed to induce a transition and can even

alter the rate of the volume phase transition (e.g. from a highly non-linear to a linear

rate).

If these PNIPAAm copolymers are to be used onboard devices and are

mechanically adhered to a surface, it is found that, upon swelling, these gels assume

different structural conformations as they change in size[14]. When this gel is confined to

a rigid substrate and swollen, swelling primarily occurs in the direction normal to the

2
surface with less swelling in parallel directions[3, 15]. This is the result of limitations in

solvent penetration due to surface confinement, and leads to anisotropic distribution of

osmotic pressure throughout the gel and results in biaxial compressive stresses. The

compressive stresses are not experienced in untethered gels and are the direct result of

the immobilization of one plane of the gel[3, 15, 16]. The swelling nature of both

unconfined and confined PNIPAAm gels are displayed in figures 1-2 and 1-3[17].

Figure 1-2: Hydrogel in three stages: fully swollen (left), after polymerization with

ambient moisture content (center), fully collapsed (right). *Eraser added for size

reference*

3
Figure 1-3: Cross-sectional profile of a surface-confined hydrogel in its swollen (a) and

unswollen (b) states.

To relieve these compressive stresses, mechanical distortions in the form of

surface wrinkling or edge undulating results[15, 18-20]. Additionally, it has been found

that the geometry of the surface confined gel plays a significant role in determining the

mechanical deformation, figure 1-4[21, 22]. In addition to geometry, the threshold and

magnitude of mechanical deformations is contingent on the methods of preparation and

the chemical composition of the media; including copolymers, initiators, and cross-

linkers[23-25]. This dependence on crosslink density for both unconstrained and

constrained PNIPAAm gels is show in figures 1-5 and 1-6.

4
Figure 1-4: Surface confined microgel in the swollen and unswollen state. The swollen

state displays mechanical deformation of the microgel.

5
Figure 1-5: PNIPAAm samples: Fully swollen at highest crosslink density (20 wt % BIS,

left), at the lowest crosslink density (1.43 wt%, right) and the initial, unswollen state (top).

6
Figure 1-6: Surface-confined PNIPAAm samples swelling as a function of crosslink

density.

Herein, the dependence of swelling on cross-link density is examined; first

through macrogel swelling test and then in surface patterned microgels. Both gels are

synthesized using a fluorescent comonomer, which will enable the characterization of

the patterned microgels by confocal microscopy.

7
 CHAPTER TWO: FABRICATION OF POLY(N-ISOPROPYLACRYLAMIDE)

MICROSTRUCTURES

Traditionally high aspect ratio microstructures or those greater in height than

width, have been produced by a variety of techniques including photolithography and ion

etching, such as deep reactive ion etching (DRIE) or focused ion beam (FIB)

techniques[26, 27]. Because the equipment to run these processes is expensive and

their operation is complicated to learn (often requiring a full time technician) the

widespread use of them is limited[28]. By contrast, soft lithography techniques require

comparatively inexpensive equipment and are straightforward to learn and apply[28].

Additionally, soft lithography can circumvent diffraction limitations associate with

projection photolithography, produce quasi three dimensional structures, create

structures or patterns on non-planer surfaces, and can be used with a wide variety of

surface chemistries[28]

In this study techniques from both approaches are used. Photolithography is first

used to generate a silicon patterned mold that can be used to create an elastomer (soft)

relief mold. SU-8 photoresist epoxy is spun onto 4” silicon wafer and patterned using a

sodalime quartz / chrome photolithography mask and UV irradiation. Once these

“master” molds have been fabricated the siloxane elastomer is poured over the mold to

generate a relief pattern. The photolithographic process is outline in the flow diagram in

figure 2-1. This relief pattern is then used in a micro injection molding in capillary

(MIMIC) technique to fabricate patterned, PNIPAAm microstructures. Additional surface

treatments are also used to chemically adhere the structures to the substrate surface.

8
Using both techniques, allows for easy, repeatable fabrication of “soft” molds and a

patterning process that may be carried out in an ordinary laboratory.

Figure 2-1: Photolithography process.

2.1 Creation of Photolithographic Mask Using AutoCAD

Over the course of this investigation two masks were designed (Monolithe I and

II) in AutoCAD and manufactured by Advanced Reproductions, Inc. These mask were

created with the known physical properties and limitation of SU-8 photoresist, PDMS and

PNIPAAm in mind and were tailored to investigate unknown or poorly understood

phenomena of PNIPAAm microstructures.

2.1.1. Monolithe I

Monolthe I was designed based on the results from the previously used

photolithography mask, which yielded microgel monolith structures that displayed buck

out-of-plane bending and edge undulations and cylindrical microgels that showed in-

plane twisting upon swelling. More information was desired from these two shapes

9
regarding all of these phenomena and they were included in greater geometric variety in

Monolithe I. Additionally these structures were spaced more liberally from one another to

avoid problems with underdevelopment and PDMS fouling in the trenches between

adjacent microstructures, shown in figure 2-2.

10
Figure 2-2: Underdeveloped photoresist pattern and the trapping of the molding

polymer.

11
 This mask was designed to investigate the affect that width, length and curvature

had on the surface-confined swelling of monolithe structures and how diameter and arc

length altered the swelling of surface confined cylindrical structures. To this end, the

following structures were designed: single beams varying in length (100 µm to 5 mm)

and width (5 µm – 100 µm), single waves varying lengths (1 mm & 3 mm), arc angles

(10° - 90°) and periodicities (1 – 3), single saw waves varying in length (200 & 400 µm),

width (5 µm – 100 µm) and periodicity (1 – 4), and circles varying in diameter (0.5 mm &

1.5 mm), arc angle (5° - 180°), and periodicity (1-4). A full schematic of this design is

displayed in figure 2-3 and its key in table 2-1.

12
Figure 2-3: AutoCAD schematic of the “Monolithe I” photolithographic mask.

13
Table 2-1: Dimensions key for Monolithe I schematic.

Structure Dimensions (µm) Structure Dimensions (µm)

Axy Dxyz
x=1-4 (length) 100.500.1000.5000 x=1 (length) 3000
y=a-e (Arc
y=a-e (width) 5.10.20.50.100 Angle) 25.20.15.10.5
z=1-2
Bxyz (Periodicity) 1.2
x=1-2 (length) 200.400. Exyz
y=a-e (width) 5.10.20.50.100 x=1-2 (length) 1000.3000.
z=1-4 y=a-d (Arc
(Periodicity) 1.2.3.4 Angle) 90.50.25.10
z=1-3
Cxyz (Periodicity) 1.2.3
x=1-2
(Diameter) 500.1500.
y=a-e (Arc
Angle) 180.90.50.25.5
z=1-3
(Periodicity) 1.2.3

2.1.2. Monolithe II

Monolithe II was design to further investigate the unique properties of single,

linear, surface-confined structures and to speed up the production of these samples,

through the creation of battery groups. These groups included single, constant width

monolithes ranging from 5-100 µm, monolithes descending in width from 100-5, 100-10,

60-4, 40-20 and monolithes that descend and then ascend to their original width. With

the remaining mask space, patterns for creating low aspect ratio bar structures with

different geometric pits, and free standing doughnut structures were designed to look at

potential encapsulation applications. Lastly, single lines of text were also worked into the

design. An AutoCAD drawn schematic of Monolithe II is shown in figure 2-4 and a close-

up of a select structure in displayed figure 2-5.

14
Figure 2-4: AutoCAD schematic of “Monolithe II” photolithographic mask.

15
Figure 2-5: Close-up of the 40x20x40 µm structure on the AutoCAD schematic of the

“Monolithe II” photolithographic mask.

2.2. Photolithography Process

To achieve different aspect ratios, the thickness of the SU-8 micromolds (height

of the photoresist) was varied by different SPIN speeds and photoresist viscosity.

Photoresists SU-8 3025, SU-8 2035 or SU-8 100 were used to achieve photoresist

thicknesses of 25-50 µm, 40-70 µm and 80-100 µm samples respectively.

2.2.1. Piranha Clean

Before beginning the photolithography process, it is important that the

silicon substrate is cleared of all impurities to insure proper adhesion of the photoresist.

16
One way by which this is accomplished is through wet etching, in this case a Piranha

etch. This was accomplished by using a solution of H2SO4 and H2O2 (3:1 volume ratio)

which produces a very exothermic oxidation reaction[29-31]. This solution is capable of

removing organic residues and hydroxylating the substrate’s surface, all without altering

surface topography[29-31]. Because of the differences in density and the tendency of

acid to flash boil when mixed incorrectly, it is advised that the sulfuric acid is added to

the hydrogen peroxide and that all of this is performed with extensive PPE (e.g. a face

mask, chemical resistant gloves and apron) [29]. In most microprocessing clean rooms,

this process is performed on a well ventilated hood because of the possibility of noxious

products and because of the exothermic nature of this reaction.

Once the Piranha solution is mixed, the n-type, single-side polished 4” silicon

wafers were added and allowed to etch for 15 minutes. At the end of 15 minutes, the

reaction was slowed to a stop by dilution with a constant stream of de-ionized water (DI

water) for approximately 5 minutes. From here the wafers were removed, further rinsed

with DI and dried by a continuous flow of nitrogen. To insure all residual water was

removed and would not interfere with the future photoresist film, the wafers were further

dried by a dehydration bake for 30 minutes at 250°C.

2.2.3. Photoresist Application

In order to achieve a uniform coating of photoresist across the surface of the

wafer, a spin coating technique was used. Different film thicknesses were achieved by

varying the viscosity of the photoresist, SU-8, and by using different spin recipes

recommended by the manufacturer, MicroChem. SU-8 was chosen because of its high

optical transmission, its ability to reliably reproduce high aspect ratio, high resolution

structure (+/- 5 µm) and for its ease of coating. For smaller features (those <5 µm), other

17
techniques such as, Deep Reactive Ion Etching or Focus Ion Beam, could be used. A

Model P6700 spin coater was used for photoresist application.

Spin coating regime suggested by MicroChem[32]:

• Dispense: 1 ml of photoresist per inch of wafer diameter on top of wafer-

spin coater setup.

• Start-up: Start spin process at 500 RPM for 5-10 seconds (5 seconds

used) with an acceleration of 100 rpm/second. This cycle is responsible

for the initial spreading of the photoresist.

• Spread cycle: Begin the spread cycle at the recommended speed and

duration at 300 rpm/second. Speeds and durations are provided by

MicroChem.

Table 2-2 list some of the recipes used for SU-8 application and the resulting structure

heights.

Table 2-2: SU-8 Spin recipes used and the results achieved

SU-8 Thickness Goal Spread Speed Spin Speed Achieved

(µm) (RPM) (sec @ RPM) Height (µm)
2035 35 10 @ 500 35 @ 4000 32
2035 45 10 @ 500 35 @ 3200 45
100 100 10 @ 500 35 @ 3000 80

It was found that dispensing the photoresist from the vendor’s (MicroChem)

container (i.e. as opposed to the use of eyedroppers pipettes or other containers) kept at

room temperature and un-stirred prevented most common application defect. These

defects included bubbles, spider webs (uneven, often stringing spreading) and severe

edge beads. The age of the SU-8 was also a major factor in the consistency of large

18
batches of wafers. All this was accounted for in the end with profile characterization. In

order to correct for edge beading, a cotton swab coated with photoresist developer was

used to remove any excess photoresist. The soft bake directly followed to facilitate

solvent evaporation and reflowing of the SU-8.

2.2.4. Soft Bake

After the photoresist application, the sample was heated on a hot plate in

accordance with the vendor’s specification in Table 2-3.

Table 2-3: Recommended soft bake times for different SU-8 thicknesses.

SU-8 Thickness Recommended Prebake Recommended Soft Bake

Product (µm) Time @65°C (Min) Time @65°C (Min)

2025 35 3 6
2035 45 3 6
100 100 20 50

Heating the sample before exposing it allows for the photoresist to reflow and

compensate for any non-uniformities in the applied film. it is imperative that the hot plate

is level and is placed on a level surface to prevent variation in structure height[32]. The

vendor also advises against using a convection oven as this could lead to the formation

of a skin on the photoresist, which could further blemish results[32].

2.2.5. Exposure

The photomasks discussed above were used to selectively block UV light from

reaching the sample’s surface, thereby preferentially crosslinking portions of the

19
photoresist film. To achieve this, a Carl Suss Mask Aligner was used and the samples

were illuminated for periods of time that varied as a function of light intensity and were

dictated by photoresist type and thickness. The samples were actually dosed with 125%

of the prescribed values of the vendor to insure complete developing, as problems with

underdevelopment were experienced following their prescription verbatim. In Table 2- 4

a list of exposure amounts can be found for all pertinent samples.

Table 2-4: 125% of the recommended UV dosages for different SU-8s and desired

thicknesses.

SU-8 Thickness Recommended UV Dosage Exposure Time

Product (µm) (mJ/cm2) (Sec.)
Lamp Output
[39 mW/cm2]
2025 30 155 10
2035 45 160 10
100 100 240 15

To screen out UV radiation below 350 nm of wavelength, a Hoya UV-34 filter was

used. It is important to limit the amount of high intensity UV radiation because it has the

propensity to cause overdevelopment of the superficial layers of the photoresist,

resulting in uneven sidewalls or “T-topping” upon developing[5]. To further increase

pattern resolution, the photoresist/photomask separation distance was minimized to

reduce errors resulting from diffraction[33]. This technique was also cited by Revzin

(2001) who found that the minimum resolution is directly proportional to the square root

of the photoresist/photomask separation distance[34].

Due to separation distance imposed by an unlevel photoresist film, variation in

height and structure resolution were seen in all samples, In the future this could be

20
further avoided by using a technique developed by Chuang (2002), whereby glycerol is

placed in the photoresist/photomask gaps to reduce Fresnel diffraction[33].

2.2.6. Hard Bake

After exposing the sample, a post exposure hard bake was performed in

accordance with the vendor’s recommendation. During this time, the appearance of the

photomask image was used as confirmation that the sample was given an ample UV

dosage to develop the photoresist. In Table 2-5, a list bake times and temperatures can

be found for all pertinent samples.

Table 2-5: Post exposure bake times recommended by Microchem Corp. for different

SU-8 samples and thicknesses.

SU-8 Product Thickness (µm) Bake Time (Min) Bake Time (Min)
@ (65°C) @ (65°C)
2025 30 1 6
2035 45 1 6
100 100 1 12
*These samples were allowed to return to room temperature before being developed*

2.2.7. Development

All samples were developed in a shallow Pyrex dish under gentle agitation for the

amount of time recommended by the vendor[32]. The developer used was a proprietary

SU-8 photoresist developer provided by MicroChem. After the prescribed time, the

sample was removed from the developer and rinsed with lab grade isopropyl alcohol,

IPA, to quench the developing reaction and to test for completion. If a white film resulted,

this indicated incomplete developing and additional time was given. Once the sample

21
finished developing, it was removed and rinsed with IPA and DI water and then dried

with a stream of nitrogen. In table 2-6 a list developing times can be found for all

pertinent samples.

Table 2-6: Developing times for different SU-8 samples and thicknesses.

SU-8 Product Thickness (µm) Developing time (min.)

2025 30 6.5
2035 45 7
100 100 15

When working with large structures or those with high aspect ratios, it is

important to use caution when using nitrogen streams to dry the sample, as the high

pressure could delaminate the SU-8 structures.

2.2.8. Characterization

After developing the microstructures, all samples were examined under a light

microscope to check for broken patterns or other irregularities. If a sample was too badly

marred or was not to specification, the SU-8 was removed by Reactive ion etching at

100 mTorr and 10°C with 200 W with 80 sccm O2 and 8 sccm CF4. To verify the

microstructure heights, a Tencor Alphastep 200 Profilometer was used.

2.2.9. Extended Hard Bake

Per recommendation of the vendor, the final samples were heated on a hot plate

to 200 °C for a period of approximately thirty minutes. In past attempts to use SU-8

microstructures as master molds for soft lithography, delamination of the SU-8 has been

an issue. This final annealing step increases the strength of the mold thereby decreasing

22
the likelihood of delamination. After heating, the samples were allowed to return to room

temperature on the hotplate to avoid thermal shocking.

2.3. Soft Lithography

Soft lithography is a non-photolithographic field that is founded on self-assembly

and mold casting over patterned relief structures for the fabrication of structures ranging

from 30 nm to 100 µm[28, 35]. Microcontact printing (µCP), replica molding (REM),

solvent-assisted micromolding (SAMIM), microtransfer molding (µTM) and micromolding

in capillaries (MIMIC) are the most prevalent staple techniques of soft lithography[28,

35]. With one or multiple of these techniques three-dimensional microstructures can be

built, often even stacked layer-by-layer, to give rise to such micro devices as: mixers[36],

fluidic channels[37], valves[38], pumps[39], and tweezers[40].

At the center of all of these methods, the patterned elastomeric block material

must be selected according to the materials that are to be patterned. These materials

include: unsensitized polymers (e.g. polyurethane, polyethylene, epoxy, etc.), colloidal

materials and biological macromolecules[28]. Poly(dimethylsiloxane), PDMS, is a

prevalent patterned block material because of its ease of processing, relative

inexpensive and favorable chemical characteristics[28]. In addition to being

biocompatible, which makes PDMS ideal for patterning proteins and cells, it is

permeable to gases, optically transparent to about 300 nm, and can coat small surface

features with relatively high fidelity (~+/- 5 µM) [40, 41]. The elasticity of cured PDMS, in

conjunction with its low interfacial surface energy (21.6x10-3 Jm-2) and chemically inert

nature, allow for an easy removal of the patterned block from complicated and fragile

positive reliefs. Lastly, PDMS is not hydroscopic and will not swell in ambient humidity

23
and is very durable, which allows for repeated use (50+ stampings) over a several month

span without significant loss in structural fidelity[28].

While PDMS has many favorable characteristics for its application as a patterned

relief mold, it is not without its drawbacks and limitations. While the elasticity and

interfacial surface energy are favorable, PDMS still imposes mechanical stress on

photoresist microstructures, especially those of a high aspect ratio, and can delaminate

or otherwise destroy these microstructures. We have observed that in small features of

the photoresist microstructures (<~10µm), the PDMS will flow and polymerize but may

sometimes tear away from the bulk relief mold upon liftoff. This lowers the fidelity of the

relief mold and potentially fills in small features on the photoresist pattern. Also, upon

curing there is a loss in volume of the PDMS, as it will shrink about 1%[28]. This affect is

noteworthy but not significant to the microstructure fidelity. Furthermore, adhesive,

capillary and gravitational forces impose stresses on the PDMS stamp, causing the

patterns to collapse or bow and thus produce defective prints[39]. Such defects are a

strong function of geometry. Delamarche (1997) notes that the aspect ratio in these

PDMS relief stamps must be within 0.2 and 2 in order to achieve defect free stamps[39].

The photolithography process discussed above, was used to generate SU-8

patterned structures that serve as the positive mold for the PDMS stamp. PDMS

prepolymer solutions were mixed from two parts, an elastomeric base and curing agent,

in a 10:1 ratio, respectively. This prepolymer solution was then placed under vacuum

(101.6 mm Hg) until all air bubbles were removed. This ensures complete mixing of the

two prepolymer constituents and prevents air bubbles from coming in contact with the

positive relief surface, thus causing pockets and other imperfections in the PDMS mold,

as show in figure 2-6[28]. This prepolymer mixture was then poured over the SU-8

patterned wafer and heated at 75°C for 60 minutes to cure the PDMS. Directly

afterwards, the PDMS/wafer sample was removed, briefly allowed to return to room

24
temperature and then the PDMS was removed from the wafer. It is important that these

PDMS molds are removed from the wafer in the direction of the length of the SU-8

structures to prevent delamination of these structures. Leaving the PDMS on the wafer

for extended periods of time allows for the PDMS to further cure and lose more of its

elasticity. We have found that this increases the likelihood of damaging SU-8 patters,

especially those of high aspect ratio, and therefore such practice is avoided. At this

point, the PDMS molds are placed face down on a clean, dust-free, cutting surface and

cut with an Exacto knife. A Petri dish was used as both a stencil for cutting and a

container for storing. This is done so that the samples may be labeled and stored in a

place where they are least likely to pick up dust or other particulate matter that may

compromise the molds integrity. Later these samples will be revisited and structures of

interest can be cut directly out of the PDMS without removing it from the Petri dish, as

shown in figure 2-7.

25
Figure 2-6: Air bubbles in a PDMS stamp still on an SU-8 master mold.

26
Figure 2-7: PDMS relief in petri dish.

27
2.3.1. Micro Molding in Capillary (MIMIC)

In MIMIC, the PDMS relief molds are placed, patterned side down, onto a

substrate which forms a series of empty corridors[42]. At the open end of these

corridors, a prepolymer solution is pooled and capillary forces pull the fluid through the

corridors. Here the polymer is cured and the PDMS relief is removed, leaving behind the

patterned polymer microstructures. It is important to note that the PDMS should be

removed in the direction that runs with the length of the microstructures to prevent

damaging the tops of thin structures. The entire MIMIC process is shown in figure 2-8.

Figure 2-8:Micro injection molding in capillaries (MIMIC) process.

2.4. Application of Silane Binder to Substrate

In early attempts to deposit PNIPAAm structures onto glass surfaces, only

rudimentary cleaning methods were used (acetone/methanol/isopropanol) but no

fixatives were applied to the surfaces. The patterns could be cured on these surfaces

and the small forces present were sufficient to adhere them to the surface, however,

28
when these structures swelled those weak forces were easily overcome and

delamination often occurred. In response to this, all future surfaces were chemical

enhanced with 3-(trichlorosilyl)propyl methacrylate (TPM, Aldrich). This self-assembled

monolayer covalently bound the pNIPAAm microstructures to the glass slide during

photopolymerization.

In the field of self-assembled monolayers (SAMs), chlorosilanes or alkoxysilanes

are employed on silica or glass substrates to achieve dense, thick (~22 Հ) monolayers

with less surface roughness[43, 44]. Using a similar approach, glass cover slips

(~0.170μm thick) were treated with a monolayer of TPM to give it adhesive

properties[45]. In summary, the cover slides were first cleaned using a standard solvent

rinse of methanol/acetone/isopropanol/DI water and then dried with high purity dry

nitrogen. Subsequently, the slides were further cleaned with O2 plasma in a Harrick

Plasma Cleaner/Sterilizer PDC-32G for 15 minutes at approximately 500 mTorr of

pressure and at 6.8 Watts of RF power. Plasma cleaning removes any residual organic

deposits by chemical reaction with highly reactive oxygen radicals and removal by

oxygen ions and promotes hydroxylation (formation of OH groups) of the surface which

will help with monolayer application. From here the glass slides were treated in 3

chemical baths in glove bag, which provided an oxygen free environment. These slides

were first treated for 5 minutes in a 1 mM solution of TPM in a 4:1 ratio of heptane to

carbon tetrachloride at room temperature and atmospheric pressure. This was followed

by subsequent soaking in hexane and then in DI water; each for 5 minutes. The

mechanism for silanization is shown in figure 2-9.

29
Figure 2-9: Silanization mechanism.

30
 It was found that if too much time passed between plasma cleaning and TPM

application or if precautions were not taken to ensure that samples were kept in closed,

clean containers and free of particulates, that a snowing effect could be expected when

the TPM was applied. This effect, shown in figure, is speculated to be indicative of

thicker deposition of TPM, as shown in figure 2-10. Use of these slides with MIMIC,

yields poor adhesion between the patterned PDMS mold and the substrate, which leads

to the creation of scum layers (polymer that seeps under the mold and polymerizes).

This is shown in figures 2-11 and 2-12. Also, if silanized slides were not used

immediately (usually within a few day) the likelihood of structure delamination seemed to

increase.

Figure 2-10: Dirty slide leading to heavy silane deposition.

31
 100 µm

Figure 2-11: Polymer microstructure with surrounding polymer scum layer.

Figure 2-12: Deformed polymer pad (top-right corner) with surrounding scum layer
exhibiting surface wrinkling.

32
2.5. Photopolymerization of Microgels

Microgels were created by MIMIC and used a prepolymer acetone/water (4:1)

solution containing the following: 200 mg n-isopropylacrylamide monomer (97% Aldrich),

2 mg of 2-dimethoxy-2-phenylacetophenone (DMPA, Aldrich) as the photoinitiator, 1 mg

polyfluor 570 fluorescent monomer (Methacryloxyethyl thiocarbonyl Rhodamine B)

(Polysciences, Inc), and the crosslinker N,N'-Methylenebisacrylamide (varied from 0-40

mg, BIS, Chemzymes Ultra Pure).

Patterns of interest were cut out of the PDMS relief molds and placed pattern-

down on the silanized surface. Because of PDMS’s excellent adhesive properties, no

further measure is required to bind them to the surface. As discussed in the section on

MIMIC, applying these relief patterns to the surface creates a network of corridors. At

the opening of these corridors, the pre-polymer solution is pooled and pressure created

by capillary forces fills the corridors with the solution. Immediately after the pre-polymer

solution has had time to fill the pattern, the assembly is photopolymerized using an

uncollimated, 365 nm, 300 mW/cm2 light source (EFOS Ultracure 100ss Plus, UV spot

lamp, Mississauga, Ontario) for 5 minutes. Here it is imperative to start the

polymerization as soon as possible, else solvent evaporation and subsequent

crystallization (figure 2-13 and 2-14) are eminent[46]. The mixing of the pre-polymer

solution from the prefabricated stocks as well as the MIMIC process described above,

were all performed in a nitrogen environment.

33
Figure 2-13: Slide with crystallized monomers (prepolymer solution) on it.

Figure 2-14: Confocal microscope image of crystallized monomer on the slide surface.

34
 The formation of the hydrogel microstructures is performed through the photo-

initiate free-radical polymerization of the acrylate end group in the NIPAAm monomers.

In this reaction, DMPA is dissociated by the UV radiation, creating methyl radicals that

then react with the carbon-carbon double bonds of the acrylate function group on the

NIPAAm monomer. This reaction propagates between until it is terminated by the

carbon-carbon double bond becoming oxygenated.

The silanated cover slide will also participate in this free radical reaction because

of the availability of the vinyl groups (C=C) offered by the monolayer. This happens

when the methacrylate radicals present in the NIPAAm monomer react with the vinyl

groups and it forms covalent bonds which anchor the polymer to the monolayer[34].

While this process has been very useful in creating these microstructures, it is

very inflexible in the variety of structures that can be made. This is due to the fact that

the structures must be hydraulically connected. Thus isolated structures are an

impossibility or must be created as hydraulically connected patterns and modified after

the fact; a very tedious and altogether impractical endeavor. Also, this method, which

relies on capillary forces, is only valid so long as the viscous forces at the corridor

boundaries do not dominate; that is long patterns will experience too much viscous drag

and will not fully fill with the pre-polymer mixture. However, interestingly enough,

patterns that are short enough will fill even if the corridors have closed ends. It is thought

that this occurs because small amounts of gas are able to diffuse into the PDMS

elastomer[28]. Lastly, mechanically these PDMS relief patterns could be reused multiple

times but due to the absorption of acetone and contamination, subsequent uses don’t

yield desired results.

35
2.6. Fabrication of Fluid Cells

The fluid chamber was created by adhering individual glass cover slides

(Corning, No. 1½, 25 mm2), each containing a single microstructure, to a plexiglass slide

having dimensions of standard microscope slides. The plexiglass overlay was

prefabricated with holes, such that each microstructure would be encased in the

plexiglass (forming a fluid well). This design is shown in figure 2-15 and 2-16. The

design presented in figures 2-15 and 2-16 is the most recent design but the evolution of

the fluid cell design can be seen in figure 2-17. Nail polish was used to adhere the

plexiglass to the glass substrate. From here DI water can easily be added or removed

from this minimalistic design that easily accommodates the inverted confocal

microscope.

Figure 2-15: Top and side view of assembled fluid cell.

36
Figure 2-16: Top and bottom view of fluid cell with sample slide adhered.

Figure 2-17: Chronicle evolution of fluid cell design. First design (left) was an enclosed

design that used a plexiglass cover with ports. The latter two (to the right) were open top

designs with open tops.

37
2.7. Remarks

All pattern PNIPAAm microstructures that are created with this method seem to

display a decrease in structure height at the midway point between hydraulic pads. It is

not certain whether this is caused by bowing of the PDMS relief mold or if the tops of the

structures in that part of the pattern are being torn off. The latter is not supported by

closer examination of the PDMS stamp after polymerization. Further investigation is

needed.

38
 CHAPTER THREE: CHARACTERIZATION OF EXTENT OF SWELLING AND

CROSSLINK DENSITY

3.1. Introduction

The most important characteristic of thermo-responsive gels is their ability to

readily swell and deswell. This makes absorption capacity the most crucial property of

these gels. The primary variables that control this property are the ion content of the

polymer matrix, the crosslink density and the solvent interaction parameter ( χ1 ) [47].

Because PNIPAAm is a nonionic polymer and the solvent used is constant, the crosslink

density is the remaining variable which must be determined experimentally. Of the

available experiments for determining crosslink density (stretching, compressing,

indenting, etc…), equilibrium swelling was chosen to characterize this parameter.

Knowledge of the swelling limits of the hydro gel can then be combined with equilibrium

swelling theory and Flory-Huggins mixing theory to determine the crosslink concentration

[47].

3.2. Characterization of Extent of Swelling in Bulk Gels

In order to get a general understanding of the swelling dependence on crosslink

density, marcoscale swelling test were conducted for 15 different samples over a spread

of 0-20 wt% BIS (weight ratio BIS/NIPAAm ~= wt% BIS). This was accomplished by

fabricated 15 (1ml) microgel samples at different crosslink concentration and placing

them in water baths below and then above the LCST and measuring the changes in

39
mass. With the given density of water (0.997 g/cm3 and 0.992 g/cm3) at these operating

temperatures (21.5°C and 40°C, on average), the volume change and hence the degree

of swelling can be calculated.

3.2.1. Experimental Procedure

PNIPAAm gels were fabrication in 12 x 75 mm disposable culture tubes in a nitrogen

environment at room temperature and pressure. The procedure is as follows:

3.2.1.1. Preparing the Stock Solutions

The desired prepolymer solution contains 200mg of monomer N-isopropylacrylamide, 2

mg of photoinitiator 2,2-dimethoxy-2-phenylacetophenone (DMPA), 0.5 mg of

fluorescents monomer polyfluor 570 (Methacryloxyethyl thiocarbonyl Rhodamine B), and

5 mg of crosslinker N,N’-methylenebisacrylammide (MBS) all in a 1 ml portion of solvent.

In order to make multiple prepolymer solutions with accuracy, weighing out small

quantities was avoided and instead stock solutions of all of the constituents were

created. To make for easy variability of the MBS, the mass of each of the solute needed

in the final solution was added to 1 ml of acetone and then the concentrations of all four

were quadrupled to yield the following in table 3-1.

Table 3-1: Initial prepolymer solutions recipe.

Conc. of Stocks
Solute mg. of solute (mg/ml) 4 x Conc. wt% solutes
NIPAAm 200 200 800 96.3855%
DMPA 2 2 8 0.9639%
Rhodamine 0.5 0.5 2 0.2410%
MBS 5 5 20 2.4096%

40
 For this base recipe, which has been used in previous work by DuPont et al[17],

the current weight ratio of BIS to NIPAAm is 2.5%. It is desire to have a maximum weight

ratio of 20% and a minimum of 0%. At 20 wt%, the concentration of the BIS solution

climbs to 160 mg/ml. To achieve this, a battery of 15 samples were created to cover this

span by altering the concentration of the BIS stock with acetone to maintain and equal ¼

volume contribution to the prepolymer mix, as shown by table 3-2.

Table 3-2: BIS dilution chart.

Solution MBS (mg) wt% MBS/NIPAAm Actetone (µl) BIS stock (µl)
1 0 0.00% 250.00 0.00
2 11.429 1.43% 246.43 3.57
3 22.857 2.86% 239.39 10.61
4 34.286 4.29% 229.13 20.87
5 45.714 5.71% 216.04 33.96
6 57.143 7.14% 200.60 49.40
7 68.571 8.57% 183.41 66.59
8 80.000 10.00% 165.07 84.93
9 91.429 11.43% 146.20 103.80
10 102.857 12.86% 127.41 122.59
11 114.286 14.29% 109.21 140.79
12 125.714 15.71% 92.04 157.96
13 137.143 17.14% 76.27 173.73
14 148.571 18.57% 62.10 187.90
15 160.000 20.00% 0.00 250.00

Unfortunately, BIS has a solubility of ~8 mg/ml in pure acetone. Trying to stay as close to

the original solvent as possible, so that data collect on microgel from previous work

would still be prudent, a solubility test with acetone/water mixtures conducted with BIS,

as shown in figure 3-1.

41
 Solubility of BIS in a Water/Acetone Solution
140

Concentration of BIS (mg/ml) 120

100

80

60 Series1
Poly. (Series1)
40

20
y = 810.67x3 - 1563.4x2 + 784.76x + 8.1714
R² = 0.9998
0
0% 20% 40% 60% 80% 100% 120%
Water (Volume %)

Figure 3-1: Solubility plot of BIS concentration verses water to acetone volume ratios

Based on this, a stock of BIS was made with a 25/75% water/acetone mixture

and it was found to form a suspension with a low settling rate. In lieu of switching to a

different solvent this was used. Also, to maintain a 25/75% solvent mixture, the DMPA

stock was quadrupled again (now 8x the original) and added as 1 part DMPA stock to 3

parts DI water for the DMPA’s contribution to the prepolymer solution. The final recipe

and dilution table now read as follows in tables 3-3 & 3-4.

42
Table 3-3: Revised prepolymer recipe for equal volume contribution.

Solute mg of solute Conc. of Stocks (mg/ml) 4 x Conc. wt% solutes

NIPAAm 200 200 800 96.3855%
DMPA 2 2 16 0.9639%
Rhodamine 0.5 0.5 2 0.2410%
MBS 40 40 160 19.2771%

Table 3-4: Dilution chart for BIS.

Solution MBS (mg) wt% MBS/NiPAAm Actetone (µl) Stock (µl)

1 0 0.00% 250.00 0.00
2 11.429 1.43% 246.43 3.57
3 22.857 2.86% 239.39 10.61
4 34.286 4.29% 229.13 20.87
5 45.714 5.71% 216.04 33.96
6 57.143 7.14% 200.60 49.40
7 68.571 8.57% 183.41 66.59
8 80.000 10.00% 165.07 84.93
9 91.429 11.43% 146.20 103.80
10 102.857 12.86% 127.41 122.59
11 114.286 14.29% 109.21 140.79
12 125.714 15.71% 92.04 157.96
13 137.143 17.14% 76.27 173.73
14 148.571 18.57% 62.10 187.90
15 160.000 20.00% 0.00 250.00

The stock solutions for NIPAAm were made with 5 ml of acetone for the stock, while 20

ml was used for the remaining solutes.

The first two batches of samples and swell test were conducted using this recipe

and method until it was discovered that an error had been made; the solvent added to

the BIS stock was pure acetone instead of a 25/75 volume ratio of water and acetone.

This would have thrown off the whole ratio of the solvent which was to be maintained at

43
25/75 to preserve optimal solvating conditions. The dilution table was reworked, as

shown in table 3-5, and the next two batches of samples and swell test were conducted

with this correction.

Table 3-5: Revised dilution chart for BIS.

Solution mg MBS wt% MBS/NiPAAm 25/75% Water/Actetone ul Stock

1 0 0.00% 250.00 0.00
2 11.429 1.43% 246.43 3.57
3 22.857 2.86% 239.39 10.61
4 34.286 4.29% 229.13 20.87
5 45.714 5.71% 216.04 33.96
6 57.143 7.14% 200.60 49.40
7 68.571 8.57% 183.41 66.59
8 80.000 10.00% 165.07 84.93
9 91.429 11.43% 146.20 103.80
10 102.857 12.86% 127.41 122.59
11 114.286 14.29% 109.21 140.79
12 125.714 15.71% 92.04 157.96
13 137.143 17.14% 76.27 173.73
14 148.571 18.57% 62.10 187.90
15 160.000 20.00% 0.00 250.00

3.2.1.2. Sample Preparation

The 15 samples were prepared five at a time with the same procedure for each

of the 4 total batches. Each batch was placed in a test tube rack outside of the nitrogen

tent and was dispensed its prescribed BIS concentration and the 3 parts DI water that

was to be added to the DMPA’s aliquot (187.5 µl). In order to maintain a consistent

suspension of BIS, the BIS stock solution was first agitated on a Vortex Genie 2 (Model

G-560) for 10 minutes and then place on a stir plate to maintain agitation for the duration

of the sample preparation. The remaining ingredients were added in the glove bag under

a positive pressure of nitrogen. After all of the ingredients were added the culture tubes

44
were all stirred well to insure proper mixing of the prepolymer solution. At this point the

samples were irradiated with a UV spot lamp for 15 minutes under constant agitation.

Once polymerized, all samples were removed from the glove bad and either shaken

from their culture tubes or the tubes were gently fractured and the sample were

removed. All samples were then placed in 60 ml graduated glass bottles and labeled

with their respective BIS concentration.

3.2.1.3. Experimental Setup and Data Collection

Once all samples were finished, 20 ml of DI water at 21.5 °C (on average) was

added to each specimen container and then they were sealed. Mass measurements

were recorded by removing the samples and drying off any superficial water before

recording a data point. Data of all of the samples were taken every 30 minutes during

the initial 3 hours of swelling due to the steep rate of change. After each recorded mass,

20 ml of fresh DI water was added to replace the used water. This was done do limit the

affects of excess solvent or unreacted monomer. All subsequent readings occurred only

as often as needed.

After 120 hours of data collection had past, all sample’s DI water was replaced

by heated DI water at approximately 40 °C and then they were placed in 1000 ml

beakers containing 200 ml of DI water also at approximately 40 °C. The increments at

which data was nearly identical to regiment followed for the swelling portion of the

experiment. Deswelling data was also taken for the duration of 120 hours.

After deswelling, all samples were placed in labeled Pyrex Petri dishes and

heated in an Isotemp® vacuum oven (Model 280A) furnace at 75 °C under a constant

2.92 mmHg of vacuum for 24 hours. Samples were then removed and their dry mass

was recorded. All samples were stored in their dry state.

45
3.2.2. Remarks and Results

To calculate the crosslink density from extent of swelling data, one only needs

the volume of the original sample, the most swollen state and the dry state but additional

data was collected over the duration of the swelling so that a qualitative assessment of

the data may be made. The swelling and deswelling data for all 4 batches is listed in

figures 3-2 through 3-17. It is important to note that samples 1 and 2 were created with

the acetone rich recipe. Sample 3 and 4 were made with the recipe that keeps a

constant 25/75 volume % of water and acetone, respectively.

46
 Swelling Sample I: Weight vs. Time at Different wt% of BIS
10
1.43%
9
2.86%
8
4.29%
Normalized Weight (gr.)

7 5.71%
6 7.14%

5 8.57%
10.00%
4
11.43%
3 12.86%
2 14.29%

1 15.71%
17.14%
0
18.57%
0 20 40 60 80 100 120 140
20.00%
Time (Hrs.)

Figure 3-2: Swelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.

47
 Swelling Sample I: Weight vs. Time at Different wt% of BIS
10

8
Normalized Weight (gr.)

5 1.43%
2.86%
4
10.00%
3 20.00%
2

0
0 20 40 60 80 100 120 140
Time (Hrs.)

Figure 3-3: Swelling of 1ml PNIPAAm samples with only four select weight ratios of

BIS/NIPAAm shown.

48
 Swelling Sample II: Weight vs. Time at Different wt% of BIS
7
1.43%
6 2.86%
4.29%
Normalized Weight (gr.)

5
5.71%
7.14%
4
8.57%
3 10.00%
11.43%
2 12.86%
14.29%
1 15.71%
17.14%
0
18.57%
0 20 40 60 80 100 120 140
20.00%
Time (Hrs.)

Figure 3-4: Swelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.

49
 Swelling Sample II: Weight vs. Time at Different wt% of BIS
7

6
Normalized Weight (gr.)

4
1.43%
3 2.86%
10.00%
2 20.00%

0
0 20 40 60 80 100 120 140
Time (Hrs.)

Figure 3-5: Swelling of 1ml PNIPAAm samples with only four select weight ratios of

BIS/NIPAAm shown.

50
 Swelling Sample III:Weight vs. Time at Different wt% of BIS
4.5
1.43%
4
2.86%
3.5 4.29%
Normalized Weight (gr.)

5.71%
3
7.14%
2.5 8.57%
10.00%
2
11.43%
1.5 12.86%
14.29%
1
15.71%

0.5 17.14%
18.57%
0 20.00%
0 20 40 60 80 100 120 140
Time (Hrs.)

Figure 3-6: Swelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.

51
 Swelling Sample III:Weight vs. Time at Different wt% of BIS
4.5

3.5
Normalized Weight (gr.)

2.5 1.43%
2.86%
2
10.00%
1.5 20.00%

0.5

0
0 20 40 60 80 100 120 140
Time (Hrs.)

Figure 3-7: Swelling of 1ml PNIPAAm samples with only four select weight ratios of

BIS/NIPAAm shown.

52
 Swelling Sample IV: Weight vs. Time at Different wt% of BIS
3.5
1.43%
3 2.86%
4.29%
Normalized Weight (gr.)

2.5
5.71%
7.14%
2
8.57%
1.5 10.00%
11.43%
1 12.86%
14.29%
0.5 15.71%
17.14%
0
18.57%
0 20 40 60 80 100 120 140
20.00%
Time (Hrs.)

Figure 3-8: Swelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.

53
 Swelling Sample IV: Weight vs. Time at Different wt% of BIS
3.5

3
Normalized Weight (gr.)

2.5

2
1.43%
1.5 2.86%
10.00%
1 20.00%

0.5

0
0 20 40 60 80 100 120 140
Time (Hrs.)

Figure 3-9: Swelling of 1ml PNIPAAm samples with only four select weight ratios of

BIS/NIPAAm shown.

54
 Collapsing Sample I: Weight vs. Time at Different wt% of BIS
1.2
1.43%
1 2.86%
4.29%
Normalized Weight (gr.)

0.8 5.71%
7.14%

0.6 8.57%
10.00%
11.43%
0.4
12.86%
14.29%
0.2
15.71%
17.14%
0
18.57%
0 20 40 60 80 100 120 140
20.00%
Time (Hrs.)

Figure 3-10: Deswelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.

55
 Collapsing Sample I: Weight vs. Time at Different wt% of BIS
1.2

1
Normalized Weight (gr.)

0.8

0.6 1.43%
2.86%
10.00%
0.4
20.00%

0.2

0
0 20 40 60 80 100 120 140
Time (Hrs.)

Figure 3-1: Deswelling of 1ml PNIPAAm samples with only four select weight ratios of

BIS/NIPAAm shown.

56
 Collapsing Sample II: Weight vs. Time at Different wt% of BIS
1.2
1.43%
1 2.86%
4.29%
Normalized Weight (gr.)

0.8 5.71%
7.14%

0.6 8.57%
10.00%
11.43%
0.4
12.86%
14.29%
0.2
15.71%
17.14%
0
18.57%
0 20 40 60 80 100 120
20.00%
Time (Hrs.)

Figure 3-2: Deswelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.

57
 Collapsing Sample II: Weight vs. Time at Different wt% of BIS
1.2

1
Normalized Weight (gr.)

0.8

0.6 1.43%
2.86%
10.00%
0.4
20.00%

0.2

0
0 20 40 60 80 100 120
Time (Hrs.)

Figure 3-3: Deswelling of 1ml PNIPAAm samples with only four select weight ratios of

BIS/NIPAAm shown.

58
 Collapsing Sample III: Weight vs. Time at Different wt% of BIS
1.2
1.43%
1 2.86%
4.29%
Normalized Weight (gr.)

0.8 5.71%
7.14%

0.6 8.57%
10.00%
11.43%
0.4
12.86%
14.29%
0.2
15.71%
17.14%
0
18.57%
0 20 40 60 80 100 120 140
20.00%
Time (Hrs.)

Figure 3-4: Deswelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.

59
 Collapsing Sample III: Weight vs. Time at Different wt% of BIS
1.2

1
Normalized Weight (gr.)

0.8

0.6 1.43%
2.86%
10.00%
0.4
20.00%

0.2

0
0 20 40 60 80 100 120 140
Time (Hrs.)

Figure 3-5: Deswelling of 1ml PNIPAAm samples with only four select weight ratios of

BIS/NIPAAm shown.

60
 Swelling Sample IV: Weight vs. Time at Different wt% of BIS
1.2
1.43%
1 2.86%
4.29%
Normalized Weight (gr.)

0.8 5.71%
7.14%

0.6 8.57%
10.00%
11.43%
0.4
12.86%
14.29%
0.2
15.71%
17.14%
0
18.57%
0 20 40 60 80 100 120 140
20.00%
Time (Hrs.)

Figure 3-66: Deswelling of 1ml PNIPAAm samples with variable BIS (crosslinker)

concentrations.

61
 Swelling Sample IV: Weight vs. Time at Different wt% of BIS
1.2

1
Normalized Weight (gr.)

0.8

0.6 1.43%
2.86%
10.00%
0.4
20.00%

0.2

0
0 20 40 60 80 100 120 140
Time (Hrs.)

Figure 3-7: Deswelling of 1ml PNIPAAm samples with only four select weight ratios of

BIS/NIPAAm shown.

62
3.3. Crosslink Density Determination

The theory that allows for the determination of the crosslink density based on the

swelling limits is the equilibrium swelling theory, which state that a polymer will absorb its

neighboring solvent until the solvent chemical potentials inside and outside of the

polymer are equal. In terms of osmotic swelling pressure, this can be written as,

Πmix +Πelastic +Πion +Πelec = 0 (3.1)

It is important to note that this equation makes the assumption that all contribution to the

swelling pressure are independent[47]. Here the Πmix is the tendency of the polymer of

dissolve into the solvent, Πelastic is the elastic response of the network due to

crosslinking, Πion is the contribution to osmotic pressure due to differences in ion

concentration between the gel and the water and Πelec is the electrostatic interactions of

charges on the polymer chains[47]. Typically the Πelec is very small in comparison to the

to Πion and, for sake of brevity, these will not be discussed at any further length as

PNIPAAm is a nonionic polymer and these terms will not be used in its calculations[47].

While theoretical a priori calculations can’t accurately determine the equilibrium

swelling of hydrogels, some theoretical predictions of different parameters of swelling do

agree with empirical results. The Flory-Huggins theory defines the mixing contribution by

introducing the polymer-solvent interaction parameter ( χ1 ), as follows:

Πmix = [RT / V1][ln(1−υ2 ) +υ2 + χυ 2

1 2]
(3.2)

63
where R is the gas constant, T is absolute temperature, V1 is the solvent molar volume

and υ2 is the polymer volume fraction[47]. The polymer-solvent interaction parameter

takes into account the free energy changes that are caused by mixing and is a function

of both temperature and concentration[47]. Its value is typically between 0 and 1, where

low values of χ1 are indicative of good solvents (those that favor minimizing and cause

the polymer to swell) and higher values indicate poor solvents (elasticity of the polymer

matrix dominates causing the polymer to collapse)[47]. While χ1 has theoretical

backing, it is most often determined empirically and there exist many sources for these

values[47, 48]. Heuristics are available for determining χ1 but since they are inaccurate

in systems where hydrogen bonding is significant, it is unlikely they would be a

responsible approach for hydrogels[47, 48].

The Πelastic term is primarily governed by the elastic restraining forces of the

crosslinked polymer chains and is a limiting parameter on extent of swelling[47]. These

elastic restraining forces are entropic in nature because stretching of the polymer matrix

reduces the number of available chain conformations[47]. For an ideal polymer matrix

that is crosslinked in the bulk state, this can be shown as:

Πelastic = −RT ρx[υ21 3 − 0.5υ2 ]. (3.3)

If the polymer is prepared in solution, where the chains are in their “relaxed”

conformation, equation (3.3) becomes:

Πelastic = −RTυ2,r ρx [(υ2 υ2,r ) − 0.5(υ2 υ2,r )]

13
(3.4)

64
where υ2,r is the polymer volume fraction at the time of polymerization[47]. Equation

(3.4) reduces back to equation (3.3) in the case that a polymer is not polymerized in the

present of the solvent, i.e. υ2,r =1[47].

As stated before, for a nonionic gel, the Flory-Rehner theory represents the total gel

pressure as a sum of the osmotic pressure contributions of mixing and elasticity.

Combining equations (3.2) and (3.4) we arrive at an equation for the total swelling

pressure:

[ln(1 −υ2,s ) + υ2,s + χ1υ2,s 2 ] / V1 +υ2,r ρx [(υ2,s υ2,r ) − 0.5(υ2,s υ2,r )]

13
(3.5)

This relationship can easily be solved for the crosslink density from here and is

reported to be a good approximation for organic solvents. Because the solvent-

interaction parameter cannot account for the hydrogen bonding that will take place in this

hydrogel, this calculation of crosslink density stands only as an approximation.

Additionally, this calculation assumes that the polymerization goes to completion, which

is not always the case.

3.4 Remarks and Results

As expected, the sample without any crosslinker polymerized and exhibited a

noticeable change in viscosity but did not congeal and therefore was not included in the

swelling test. As for the remaining samples, they calculated crosslink densities are listed

below in tables 3-6.

65
Table 3-6: Crosslink density values (mol/cm3) calculated from the swelling data using

the Flory-Rehner equation.

Batch

Sample 1 2 3 4

1 N/A N/A N/A N/A

2 8.11E-09 1.15E-08 3.78E-07 2.33E-07

3 3.63E-06 5.13E-06 3.93E-06 1.27E-06

4 4.26E-05 4.39E-05 4.38E-05 2.73E-05

5 1.33E-04 8.58E-05 1.24E-04 5.99E-05

6 4.21E-04 4.16E-04 3.73E-04 2.28E-04

7 7.56E-04 4.54E-04 1.53E-03 5.55E-04

8 2.76E-04 2.49E-04 1.00E-03 6.58E-04

9 9.10E-04 1.21E-03 1.27E-03 4.45E-04

10 1.56E-03 1.36E-03 2.20E-03 1.38E-03

11 3.72E-03 2.50E-03 2.37E-03 7.12E-04

12 2.06E-03 2.79E-03 4.09E-03 2.93E-03

13 2.05E-03 2.80E-03 1.43E-03 1.83E-03

14 6.89E-03 3.82E-03 1.42E-03 7.12E-04

15 3.89E-03 3.47E-03 1.45E-03 1.15E-03

In order to makes these calculations, an interaction parameter of 0.5 and a dry

density of 1.37 g/cm3 were used to solve the Flory-Rehner equation[49-51].

There is a significant swelling difference (~60%) between samples created with

the acetone rich recipe and the intended 25/75 mix. Plots of swelling and deswelling

kinetics are displayed in figures 3-2 through 3-17. A possible explanation for this may be

that the additional water present is enough to have a swelling effect on the PNIPAAm

66
samples as they are polymerizing. This swelling would limit chain conformations and

subsequently hinder network formation. Hence, the water rich system would be less

developed and not capable of swelling to the same magnitude as the acetone rich

sample.

Additional variability in the data is likely the result of structural defects that

occurred during deswelling. Upon deswelling some samples, typically those with the

lowest crosslink density formed an opaque skin inhibited deswelling, while other

fractured across the top or along multiple planes, expediting deswelling. These two

effects are shown in figures 3-18 and 3-19.

Figure 3-8: Collapsing gel that forms an outer skin layer.

67
Figure 3-9: A PNIPAAm that fractured in the process of swelling.

Also, the swelling data collected has confirmed that there are only small changes

in the extent of swelling or in swelling kinetics when the crosslinker concentration is

beyond 5-7 wt% BIS/NIPAAm. Therefore, investigating the swelling properties of these

densities is of little worth to the nature of this work, as these gels aren’t likely to swell

enough to enter a deformation regime.

68
 CHAPTER FOUR: CHARACTERIZATION OF CONFORMATION OF PNIPAAm

MICROSTRUCTURES BY LASER SCANNING CONFOCAL MICROSCOPY

4.1. Laser Scanning Confocal Microscopy Introduction

Wide-field fluorescence microscopy is commonly used for the investigation of

organic or inorganic substances on a micron or submicron range by using florescence or

phosphorescence phenomena, as opposed to adsorption and reflection[52]. It operates

by evenly irradiating the sample with light and exciting those portions of the sample

simultaneously causing it to fluoresce[52]. The emitted fluorescence is captured by

photodetectors or cameras and then translated into a coherent image[52]. This method

is limited by the microscopes ability to filter out background information, concisely control

the depth of field and collect optical sections of thicker specimens[53]. These are the

exact limitation that Marvin Minsky sought to overcome in the 1950’s when he invented

confocal microscopy.

In contrast to fluorescence microscopy, confocal microscopy uses pin-point

illumination and a pinhole aperture that is positioned in the conjugate plane (confocal)

with the illumination point on the specimen and a second pinhole aperture in front of the

detector (a photomultiplier tube)[53]. A schematic of the confocal microscope setup and

a comparison of wide field vs confocal illumination volumes can be seen in figures 4-1

and 4-2. Once the laser is emitted it is reflected by the chromatic mirror and rastered

across a specific focal plane of the specimen where it excites secondary fluorescence

and sends light back through the dichromatic mirror to the detector pinhole aperture,

filtering out-of-focus or out-of-plane light and propagating all other light to the

69
photomultiplier detector[53]. With the out-of-focus light excluded from the emissions

data, a more accurate account of the specimen can be rendered.

Figure 4-1: A schematic of the confocal microscope concept.

70
Figure 4-2: Difference in illumination volume between wide-field and confocal

microscopy.

On the downside, confocal microscopy is limited by the number of excitation

wavelengths that are available by today’s lasers, which emit relatively narrow bands and

are costly (especially in the UV spectrum)[53]. Concurrently, the high price of purchasing

and operating a confocal microscope limits the number of facilities that can afford

one[53]. Also, the high intensity lasers that are used in laser scanning microscopy can

be harmful to living cells or organism[53]. This problem was circumvented by Nipkow

disk confocal microscopy by using a multiphoton source at lower intensities but since

this is neither a limitation to the current investigation nor an technique employed, this will

not be discussed further[53].

71
4.2. Characterization of Structural Morphology by Confocal Microscopy

The images take with the scanning laser confocal microscope are individual

slices or well-defined planes within a sample[53]. These may be taken from the x-y, x-z,

or y-z planes over some desired thickness. Additionally, if an area of interest (AOI) is

greater that the scan area, the boundaries of this AOI may be entered into the software

interface and the microscope will split the AOI into separate image stacks that can then

be fused together to form the larger AOI.

The micrographs shown have been taken with a Leica TCS SP5 confocal laser

scanning microscope (CLSM) through a 20X/0.7NA (Leica Microsystems, Germany). An

Argon laser line with an emission at 543 nm was used to excited the fluorescent

microstructures and an AOBS was used to filter the images. The images were recorded

with photomultiplier detectors using LAS AF software (version 2.1.0, Leica

Microsystems, Germany). Image stack were taken at z-steps (thickness spacing)

between 0.25 and 1 µm.

The 3-D images of the PNIPAAm microstructures were rendered using Imaris

5.5.0 (Bitplane, Inc., St. Paul, Minnesota) software package. The RAW data from the

scanning confocal microscope was imported into Imaris and then refined using the

software’s thresholding ability to remove background fluorescence. From here the

individual stacks were assembled to create and iso-surface of the microgel.

4.3. Remarks and Results

Because of the inherent inaccuracies that come with adhering glass slides to

Plexiglass covers with nail polish, samples have often been displacement between

opposite ends that measure as much as 40 µm. Unfortunately, an AOI that runs

72
diagonally through a sample requires that the number of scan slice be increased to

capture the data and the user must deal with the excess data. An illustration of this

drawback is displayed in figure 4-3.

Figure 4-3: Increased number of stacks to compensate for uneven adhesive layer.

To this point, all PNIPAAm monolithe structures have been made using the first

recipe (i.e. the one that uses pure acetone for dilution) for varying BIS concentrations

and all resulting structures have exhibited sharply angled side-walls accompanied by

deep concavities across the width. This concavity yield structural edges that vary along

the monolithe between 15 µm and 18 µm on average and concave valleys that run the

length of the structure and are an average of ~10 µm in height, as displayed in figure 4-

4. A thinner portion of the stack, about 20 µm in width, looks closer to what has been

seen in the past but is <1/5th the height of the mold. This is show in figure 4-5. The

swollen states of both of these also yield unusual findings. While they both exhibit the

expected deformations, there are small nuances that do not follow with previous

observation. The low aspect ratio structure displays edge undulation but the effect is less

pronounced than usual and higher aspect ratio end buckles but the structural bulge that

is often present on the outside of the wave is now pronounce on the inside. Both of

these structures are show in figures 4-6 and 4-7.

73
Figure 4-4: PNIPAAm monolith in the collapsed state.

Figure 4-5: PNIPAAm monolith in the collapsed state.

74
 Edge Undulations

100 µm

Figure 4-6: PNIPAAm monolith exhibiting surface deformations in the swollen state.

Figure 4-7: PNIPAAm monolith exhibiting bulk mechanical deformation in the swollen

state.

75
 Structures of this nature have only been observed using the variable BIS recipes

described herein and have not been observed using original solution recipes. It is

currently believed that the observed effect is the result of increased surface tension of

the prepolymer solution that results in a meniscus or some other phenomena when

applied to mimic. To this point, all other potential causes (PMDS or SU-8 height and

PNIPAAm being torn off) have been disproven.

76
 CHAPTER FIVE: SUMMARY, CONCLUSION AND FUTURE WORK

This study has examined the techniques used to reproducible construct surface-

confined Poly(N-isoropylacrylamide) microgels, measure crosslink density of these gels

and how to characterize their buckling behavior in real time with the use of confocal

microscopy. Many challenges were faced in these endeavors and these trials and

triumphs will be covered here. Lastly, current and future approaches for establishing

correlations between crosslink density and mechanical deformations will be discussed.

5.1. Preparing Stock Solutions

Ultimately, using a suspended solution and a stir plate is not the most ideal

approach but is suitable for the nature of this swelling test. A more analytical approach

should be sought out to eliminate sources of potential error that are eminent with this

approach. These source might include, changing BIS concentration over time due to the

high volatility of the solvent, recrystallization of BIS along the edges of the container due

to the volatility of this mixture and its supersaturated nature, and the error in samples

taken when the stock is intermittently remove from the stir plate and allow to stop mixing

before withdraw. Moreover, this method was used out of convenience and necessity of

time.

77
5.2. Sample Preparation

There may be a small loss in accuracy when samples were removed by

fracturing the culture tubes if the samples are marred in any way. When this approach

was used, it was done with the utmost care to avoid such incidents.

5.3. Experimental Setup and Data Collection

It is important to handle these samples with care as they are dried and weighed

each time. It was found in the “beta test” (not included in the 4 batches listed herein),

that dropping samples or picking this up with one’s fingers, were likely to result in a

fractured sample. This made the remaining data collection very toilsome and the data

less accurate.

Additionally, the beta test revealed that samples should be thoroughly collapsed

before they are sent to dry in the vacuum furnace or they run the risk of exploding. Extra

care was taken to avoid this.

5.4. Confocal Microscopy

These image stacks can also be take repeatedly over time. The 4-D data can

allow for the in-situ view of PNIPAAm structures, as they transition from collapsed to

swollen or vice-versa. However, due to the relative speed/quality of the resulting image

stacks, changes may out-pace the raster ability. To overcome this, the nature of the

solvent (water) can be changed by added salts. Do so allows for a step by step

transformation of the PNIPAAm structures and allows for high-resolution image stack to

be acquired at each stage. This technique is currently being employed to study the

relationship between crosslink density and buckling conformations.

78
5.5. Future Works

5.5.1. Soft Lithography

A mentioned drawback to the MIMIC technique is that it requires the molding

stamp patterns to have attached corridors through which the prepolymer solution may

flow. With this technique, patterns without these fluid ducts can be achieved however

this must be done through manual cleaving of the ducts after the polymer is formed.

Since this is both in accurate and potentially damaging to the PNIPAAm structure, other

soft lithography methods need to be investigate. Additionally, there are several patters

on “Monolithe II” that would support such a technique and the methods for its application

are well established. This could potentially give rise to free-standing, actuatable surfaces

depressions for capturing or storing, free-standing monolithes and surfaces mimicking

biological smooth muscle[54, 55].

5.5.2. Preparing Stock Solutions

In light of some of the drawbacks mentioned above, future stock solutions will be

prepared in a more analytical fashion. Instead of pipetting from a suspension of the

crosslinker, excess solvent will be used to fully solvate the mass of BIS needed in the

respective recipe and then evaporated off. The solvent required for this aliquot can be

added in after the fact and will allow for the entire process to take place inside of the

glove bag. Adding this extra solvent to the NIPAAm stock solution would also help with

lower its propensity to crystallize on the edges of the stock container.

79
5.5.3. Confocal Microscopy Characterization

Additional characterization still needs to be done to determine the effect that

crosslink density has on surface confined structures and buckling morphologies. These

structural conformations will be further studied and used to valid or disprove current

elastic theory of surface confined hydrogels.

80
 REFERENCES

[1] J.P. Rolland, B.W. Maynor, L.E. Euliss, A.E. Exner, G.M. Denison, J.M.
DeSimone, Direct Fabrication and Harvesting of Monodisperse, Shape-Specific
Nanobiomaterials, Journal of the American Chemical Society 127 (2005) 10096-10100.

[2] A. Dongshin Kim, J.B.A. David, Lab on a Chip Hydrogel-based reconfigurable

components for microfluidic devices, 2006.

[3] M.E. Harmon, M. Tang, C.W. Frank, A microfluidic actuator based on

thermoresponsive hydrogels, Polymer 44 (2003) 4547-4556.

[4] Q. Luo, S. Mutlu, Y.B. Gianchandani, F. Svec, J.M.J. Fréchet, Monolithic valves
for microfluidic chips based on thermoresponsive polymer gels, Electrophoresis 24
(2003) 3694-3702.

[5] J. Wang, Z. Chen, M. Mauk, K.-S. Hong, M. Li, S. Yang, H. Bau, Self-Actuated,
Thermo-Responsive Hydrogel Valves for Lab on a Chip, Biomedical Microdevices 7
(2005) 313-322.

[6] A. Castellanos, S.J. DuPont, A.J. Heim, G. Matthews, P.G. Stroot, W. Moreno,
R.G. Toomey, Size-Exclusion "Capture and Release" Separations Using Surface-
Patterned Poly(N-isopropylacrylamide) Hydrogels, Langmuir 23 (2007) 6391-6395.

[7] L.K. Ista, V.H. Perez-Luna, G.P. Lopez, Surface-Grafted, Environmentally

Sensitive Polymers for Biofilm Release, Appl. Environ. Microbiol. 65 (1999) 1603-1609.

[8] M. Ebara, J.M. Hoffman, P.S. Stayton, A.S. Hoffman, Surface modification of
microfluidic channels by UV-mediated graft polymerization of non-fouling and 'smart'
polymers, Radiation Physics and Chemistry 76 (2007) 1409-1413.

[9] Y. Hirokawa, T. Tanaka, Volume phase transition in a nonionic gel, The Journal
of Chemical Physics 81 (1984) 6379-6380.
[10] M. Shibayama, T. Tanaka, Volume phase transition and related phenomena of
polymer gels, Responsive Gels: Volume Transitions I, 1993.

[11] H.G. Schild, Poly(N-isopropylacrylamide): experiment, theory and application,

Progress in Polymer Science 17 (1992) 163-249.

[12] P.J. Flory, J.J. Rehner, Statistical Mechanics of Cross-Linked Polymer Networks
II. Swelling, The Journal of Chemical Physics 11 (1943) 521-526.

[13] P.J. Flory, J.J. Rehner, Statistical Mechanics of Cross-Linked Polymer Networks
I. Rubberlike Elasticity, The Journal of Chemical Physics 11 (1943) 512-520.

81
[14] A. Vidyasagar, J. Majewski, R. Toomey, Temperature Induced Volume-Phase
Transitions in Surface-Tethered Poly(N-isopropylacrylamide) Networks, Macromolecules
41 (2008) 919-924.

[15] R. Toomey, D. Freidank, J. Ruhe, Swelling Behavior of Thin, Surface-Attached

Polymer Networks, Macromolecules 37 (2004) 882-887.

[16] S. Seelenmeyer, I. Deike, S. Rosenfeldt, C. Norhausen, N. Dingenouts, M.

Ballauff, T. Narayanan, P. Lindner, Small-angle x-ray and neutron scattering studies of
the volume phase transition in thermosensitive core--shell colloids, The Journal of
Chemical Physics 114 (2001) 10471-10478.

[17] S.J. DuPont, Swelling-Induced Instabilities in Surface-Confined Poly(N-

isopropylacrylamide) Microgels, Soft Matter (In Review).

[18] T. Mora, A. Boudaoud, Buckling of swelling gels, Eur. Phys. J. E 20 (2006) 119-
124.

[19] R.C. Hayward, B.F. Chmelka, E.J. Kramer, Template Cross-Linking Effects on
Morphologies of Swellable Block Copolymer and Mesostructured Silica Thin Films,
Macromolecules 38 (2005) 7768-7783.

[20] T. Tanaka, S.-T. Sun, Y. Hirokawa, S. Katayama, J. Kucera, Y. Hirose, T. Amiya,

Mechanical instability of gels at the phase transition, Nature 325 (1987) 796-798.

[21] W. Hong, Z. Liu, Z. Suo, Inhomogeneous swelling of a gel in equilibrium with a

solvent and mechanical load, International Journal of Solids and Structures 46 (2009)
3282-3289.

[22] M.A. Biot, Exact Theory of Buckling of a Thick Slab, Appl. sci. Res. 20 (1963) 5.

[23] H. Inomata, N. Wada, Y. Yagi, S. Goto, S. Saito, Swelling behaviours of N-

alkylacrylamide gels in water: effects of copolymerization and crosslinking density,
Polymer 36 (1995) 875-877.

[24] K.F. Freed, A.I. Pesci, Computation of the crosslink dependence of the effective
Flory interaction parameter .chi. for polymer networks, Macromolecules 22 (1989) 4048-
4050.

[25] S.H. Gehrke, M. Palasis, M.K. Akhtar, Effect of synthesis conditions on

properties of poly(N-isopropylacrylamide) gels, Polymer International 29 (1992) 29-36.

[26] P.A. Clerc, et al., Advanced deep reactive ion etching: a versatile tool for
microelectromechanical systems, Journal of Micromechanics and Microengineering 8
(1998) 272.

[27] Y. Hanein, C.G.J. Schabmueller, G. Holman, uuml, P. cke, D.D. Denton, ouml,
K.F. hringer, High-aspect ratio submicrometer needles for intracellular applications,
Journal of Micromechanics and Microengineering 13 (2003) S91-S95.

82
[28] Y. Xia, G. Whitesides, SOFT LITHOGRAPHY, Annual Review of Materials
Science 28 (1998) 153-184.

[29] Piranha etch | University of British Columbia's AMPEL Nanofabrication and

Microfabrication Facility, University of British Columbia's AMPEL Nanofabrication and
Microfabrication, 2010.

[30] M.C. Hull, L.R. Cambrea, J.S. Hovis, Infrared Spectroscopy of Fluid Lipid
Bilayers, Analytical Chemistry 77 (2005) 6096-6099.

[31] K.J. Seu, A.P. Pandey, F. Haque, E.A. Proctor, A.E. Ribbe, J.S. Hovis, Effect of
Surface Treatment on Diffusion and Domain Formation in Supported Lipid Bilayers,
Biophysical Journal 92 (2007) 2445-2450.

[32] SU-8 2000 Permanenet Epoxy Negative Photoresist, MicroChem, 2009.

[33] Y.J. Chuang, F.G. Tseng, W.K. Lin, Reduction of diffraction effect of UV
exposure on SU-8 negative thick photoresist by air gap elimination, Microsystem
Technologies 8 (2002) 308-313.

[34] A. Revzin, R.J. Russell, V.K. Yadavalli, W.-G. Koh, C. Deister, D.D. Hile, M.B.
Mellott, M.V. Pishko, Fabrication of Poly(ethylene glycol) Hydrogel Microstructures Using
Photolithography, Langmuir 17 (2001) 5440-5447.

[35] X.-M. Zhao, Soft lithographic methods for nano-fabrication, J. Mater. Chem. 7
(1997) 6.

[36] A.D. Stroock, S.K.W. Dertinger, A. Ajdari, I. Mezic, H.A. Stone, G.M. Whitesides,
Chaotic Mixer for Microchannels, Science 295 (2002) 647-651.

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

[38] H. Takao, K. Miyamura, H. Ebi, M. Ashiki, K. Sawada, M. Ishida, A MEMS

microvalve with PDMS diaphragm and two-chamber configuration of thermo-pneumatic
actuator for integrated blood test system on silicon, Sensors and Actuators A: Physical
119 (2005) 468-475.

[39] E. Delamarche, H. Schmid, B. Michel, H. Biebuyck, Stability of molded

polydimethylsiloxane microstructures, Advanced Materials 9 (1997) 741-746.

[40] A.W. Feinberg, A. Feigel, S.S. Shevkoplyas, S. Sheehy, G.M. Whitesides, K.K.
Parker, Muscular Thin Films for Building Actuators and Powering Devices, Science 317
(2007) 1366-1370.

[41] B.H. Jo, L.M. Van Lerberghe, K.M. Motsegood, D.J. Beebe, Three-dimensional
micro-channel fabrication in polydimethylsiloxane (PDMS) elastomer, Journal of
Microelectromechanical Systems 9 (2000) 76-81.

83
[42] D. Qin, Y. Xia, J. Rogers, R. Jackman, X.-M. Zhao, G. Whitesides,
Microfabrication, Microstructures and Microsystems, Microsystem Technology in
Chemistry and Life Science, 1998, pp. 1-20.

[43] R. Banga, J. Yarwood, A.M. Morgan, B. Evans, J. Kells, FTIR and AFM Studies
of the Kinetics and Self-Assembly of Alkyltrichlorosilanes and
(Perfluoroalkyl)trichlorosilanes onto Glass and Silicon, Langmuir 11 (1995) 4393-4399.

[44] P. Silberzan, L. Leger, D. Ausserre, J.J. Benattar, Silanation of silica surfaces. A

new method of constructing pure or mixed monolayers, Langmuir 7 (1991) 1647-1651.

[45] J.B. Brzoska, I.B. Azouz, F. Rondelez, Silanization of Solid Substrates: A Step
Toward Reproducibility, Langmuir 10 (1994) 4367-4373.

[46] M.J. Park, W.M. Choi, O.O. Park, Patterning polymer light-emitting diodes by
micromolding in capillary, Current Applied Physics 6 (2006) 627-631.

[47] L. Brannon-Peppas, Absorbent Polymer Technology, Elsevier Publishing

Company, Amsterdam, 1990.

[48] A.F.M. Barton, Handbook of Solubility Parameters and Other Cohesion

Parameters, CRC Press, Boca Raton, 1983.

[49] H. Feil, Y.H. Bae, J. Feijen, S.W. Kim, Effect of comonomer hydrophilicity and
ionization on the lower critical solution temperature of N-isopropylacrylamide
copolymers, Macromolecules 26 (1993) 2496-2500.

[50] K. Tauer, D. Gau, S. Schulze, A. Völkel, R. Dimova, Thermal property changes of

poly(N-isopropylacrylamide) microgel particles and block copolymers, Colloid & Polymer
Science 287 (2009) 299-312.

[51] A.K. Lele, M.M. Hirve, M.V. Badiger, R.A. Mashelkar, Predictions of Bound Water
Content in Poly(N-isopropylacrylamide) Gel†Macromolecules 30 (1997) 157-159.

[52] B. Herman., Fluorescence microscopy, 2nd ed. ed., Oxford, UK : Bios Scientific
Publishers ; New York : Springer in Association with the Royal Microscopical Society,
1998.

[53] F.T.J. Claxton N S, Davidson M W., Laser scanning confocal microscopy, 2008.

[54] D.E. Ingber, Mechanical control of tissue growth: Function follows form, PNAS
102 (2005) 11571-11572.

[55] E. Armstrong, A. Schleicher, H. Omran, M. Curtis, K. Zilles, The Ontogeny of

Human Gyrification, Cereb. Cortex 5 (1995) 56-63.

84