Citation
Tailorale and deployable trans-corneal drainage devie fabricaed with nanaoporous liquid crystal elastomer

Material Information

Title:
Tailorale and deployable trans-corneal drainage devie fabricaed with nanaoporous liquid crystal elastomer
Creator:
Volpe, Ross ( author )
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Language:
English
Physical Description:
1 electronic file (73 pages) : ;

Subjects

Subjects / Keywords:
Glaucoma ( lcsh )
Ophthalmology ( lcsh )
Polymer liquid crystals ( lcsh )
Glaucoma ( fast )
Ophthalmology ( fast )
Polymer liquid crystals ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
by Ross Volpe.

Record Information

Source Institution:
University of Colorado Denver Collections
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
971500282 ( OCLC )
ocn971500282

Downloads

This item has the following downloads:


Full Text
TAILORABLE AND DEPLOYABLE TRANS-CORNEAL DRAINAGE DEVICE
FABRICATED WITH NANOPOROUS LIQUID CRYSTAL ELASTOMER
by
ROSS VOLPE
B.S. SUNY Environmental Science and Forestry, 2014
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment Of the requirements of the degree of Masters of Science Bioengineering
2016


2016
ROSS VOLPE ALL RIGHTS RESERVED
11


This thesis for the degree of Master of Science by Ross Volpe
Has been approved for the Bioengineering Program
By
Kendall S. Hunter, Chair Christopher M. Yakacki, Advisor David Ammar Amir Torbati
April 29, 2016
m


Volpe, Ross (M.S. Bioengineering)
Tailorable and Deployable Trans-Corneal Drainage Device Using Nanoporous Liquid Crystal Elastomer
Thesis directed by Assistant Professor Christopher M Yakacki
ABSTRACT
While the pathophysiology of glaucoma is widely unknown, current treatment relies on lowering intraocular pressure (IOP) in order to delay vision loss. The gold standard in the surgical treatment of glaucoma is the Ahmed valve. This device offers unpredictable intraocular pressure (IOP) lowering efficacy as well as a milieu of acute and chronic complications. On the forefront of glaucoma device interest is trans-comeal drainage, which offers a predictable IOP lowering performance while eliminating procedural outcomes that lead to many of the current complications seen (e.g. blebs). An investigation into a novel manufacturing process of a trans-comeal drainage device constructed from a porous liquid crystal elastomer (LCE) is proposed herein.
Using a sacrificial template of water soluble nanofibers, several hundred thousand submicron channels are created in a cylindrical LCE with diameter -200 microns. Tailorable sacrificial templates of nanofibers are formed via electrospinning of poly(vinylalcohol). LCE monomers are polymerized around these sacrificial templates, and once these fibers are dissolved from a bulk LCE, the remaining channels allow for constant and predictable flow through the material. Existence of pores is confirmed with SEM and fluorescent microscopy. Finally, the drainage efficacy is tested using a water column perfusion test.
Design of the device involves a two piece system: a silicone outer housing which is inserted into the cornea and a separate LCE filter. A collagenous exterior may be introduced to the outer housing to promote integration into the cornea. The LCE filter contains channels with controlled diameters to provide adequate drainage properties, while being small enough to block most comeal flora. The filter also may be chemically coated with a layer of copper to provide additional
IV


antimicrobial properties. The chemical and physical barriers to microbes yield an antimicrobial LCE, or A-LCE.
The shape switching property of LCEs allows this device to overcome several disadvantages of glaucoma devices which are currently available to patients. The first of which is a limited lifetime of devices caused by scarring over or migration of a device. The proposed device may be cooled down to be easily removed and replaced upon inactivation. A second shortcoming of current devices involves improper IOP control. Several styles of the proposed device could easily be made through manufacturing process which could provide varying levels of IOP relief. This would make the device appropriate for patients with mild to severe glaucoma. Finally, the use of heat activated expansion allows for simple, non-surgical insertion. This advantage may prove to be the most crucial in the development of the device. Non-surgical insertion and an attractive price point would make the device accessible in developing and third world countries where ophthalmic surgeons are rare or otherwise inaccessible. This unique attribute would open up new markets not currently realized with other glaucoma drainage devices.
The form and content of this abstract are approved. I recommend its publication.
Approved: Chris Yakacki
v


ACKNOWLEDGEMENTS
My parents Raymond and Andrea Volpe have always told me to not look to others to see what is possible, but to look within myself and achieve my dreams. Without advice like this and a constant backbone of support, I would not be here. My heartfelt thanks for making this possible.
I would like to thank my mentors for their patience in giving me advice and insight into the world of academia which few navigate with such grace and levelheadedness. Dr. Chris Yakacki and Dr. Amir Torbati.
I also thank Dr. Kendall Hunter for his will to see me succeed through unexpected and sometimes adverse situations.
Dr. Ammar gave invaluable insight into the world of ophthalmology at short notice, and always made time to give help and advice. Thank you.
Imaging was made possible by the much appreciated work of Dan Merkl (University of Wyoming) and Melissa Laughter (University of Colorado).
I would be at a loss without the acknowledgement and thanks of the University of Colorado Denver Calibration and Machine Laboratory, specifically Rich Wojzick, Jack and Tom and the many others who helped machine testing apparatuses and kept my smiling while I dug through 30 years of fittings and wiring.
Last but not least, the SMAB Lab has provided me a place to not only complete meaningful and life changing research, but has become a haven from the daily grind and a place I will never forget.
vi


DEDICATION
To my mother.


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION..........................................................1
Glaucoma..............................................................3
Existing Treatments..............................................3
Drugs.......................................................4
Surgery.....................................................5
Drainage Devices..........................................6
Liquid Crystal Elastomers............................................12
Background and Synthesis........................................12
Liquid Crystal Elastomers as Trans-Comeal Filter................17
Advantages of Liquid Crystal Elastomers....................18
Antimicrobial Properties.................................. 18
Electrospinning......................................................18
Background..................................................... 18
Tailoring Fiber Diameters.......................................20
II. MATERIALS AM) METHODS................................................21
Poly(vinylalcohol) Solution Preparation..............................21
viii


Electrospinner Fabrication...........................................21
Nanofiber Production.................................................22
Polymer Synthesis....................................................23
LCE Synthesis...................................................23
Test Samples....................................................25
Composite Formation................................................. 26
Nanochannel Formation................................................27
Imaging............................................................. 28
Perfusion Testing....................................................29
III. RESULTS............................................................. 34
Electrospinning..................................................... 32
Nanofiber/Polymer Composite......................................... 41
Porous LCE.......................................................... 41
Fluorescent Imaging................................................. 44
Perfusion Testing....................................................46
IV. DISCUSSION...................................................... 50
V. CONCLUSION.......................................................56
REFERENCES..............................................................59
IX


LIST OF FIGURES
Figure
1. Aqueous Humor Drainage Path...............................................1
2. Ocular Surface Irritation.................................................4
3. Trabeculectomy Procedure..................................................6
4. Acute Blebitis............................................................7
5. Traditional Drainage Devices..............................................7
6. MIGS Device...............................................................9
7. MicroOptx Trans-Comeal Device..............................................10
8. Liquid Crystal Elastomer Configurations................................... 12
9. TAMAP Reaction Visualization...............................................14
10. Liquid Crystal Elastomer Actuation........................................16
11. Trans-Comeal Filter.......................................................17
12. Electrospinning Unit Picture and Circuitry................................21
13. Electrospinning Collectors................................................22
14. Liquid Crystal Elastomer Monomers.........................................23
15. Cylindrical Composite Process Flow........................................26
16. Rectangular Composite Fabrication.........................................27
17. Pipette Perfusion Apparatus...............................................30
18. Water Column Perfusion Apparatus..........................................30
19. Disc Perfusion Fixture....................................................32
20. Photo of Perfusion Fixture............................................... 33
21. SEM Images of Randomly Oriented Fibers....................................34
x


22. Graph of Nanofiber Sizes vs. Solution Concentration......................36
23. SEM and Scanning Laser Topography Images of Aligned Fibers...............39
24. SEM Images of Fibers on a Wire...........................................40
25. Cylindrical Composite....................................................41
26. Cylindrical LCE Actuation................................................42
27. SEM Image of Nanochannels................................................43
28. Fluorescent Confocal Microscopy Image of Nanochannels....................44
29. Bar Graph Comparing Fluorescent Fibers and Sacrificial Template..........46
30. Pipette Perfusion Results................................................47
31. Large CylinderAVater Column Perfusion Results............................48
32. Disc Fixture/Water Column Perfusion Results..............................49
33. Poly(vinylalcohol) Synthesis.............................................51
xi


LIST OF TABLES
Table
1. Mowiol 8-88 Fiber Measurements.....................................37
2. Mowiol 10-98 Fiber Measurements....................................38
3. Fluorescent Fiber Measurements.....................................45
4. Perfusion Data and Conversions.....................................49
xii


LIST OF ABBREVIATIONS
AH Aqueous humor
IOP Intraocular pressure
GDD Glaucoma drainage device
NTG Normal tension glaucoma
LT Laser Trabeculoplasty
MIGS Micro invasive glaucoma surgery
LCE Liquid crystal elastomer
Ti Initiation temperature
TAMAP Two-stage thiol acrylate Micheal addition reaction
SEM Scanning electron microscopy
MMA Methyl methacrylate
DEGDMA Di(ethyleneglycol) dimethacrylate
PEGDMA Poly(ethyleneglycol) dimethacrylate
DMF Dimethylformamide
fBA tert(butyl)acrylate
PVA Poly(vinylalcohol)
AC Alternating current
DC Direct current
kV Kilovolt
X1U


CHAPTER I
INTRODUCTION
There is a dire need to develop an effective and consistent treatment to glaucoma, a disease that is predicted to effect 80 million people by 2020(5). Glaucoma is the second leading cause of blindness worldwide, with one in six cases leading to bilateral blindness(-/). The biological pathophysiology of glaucoma is widely unknown due to the difficulty of
Figure 1 A) Diagram showing aqueous humor (green arrow) in a healthy eye flow from the ciliary body around the iris, through the trabecular meshwork and out Schlemms Canal into an episcleral vein. B) Open angle glaucoma eye has limited egress through trabecular meshwork. (2)
1


studying ocular tissues in vitro. The only proven factor in decreasing the risk of blindness is by maintaining a healthy intraocular pressure (IOP) of 15 mmHG or lower(5-7). If pressure rises above this, ganglion cells rooted in the ocular nerve are damaged and lose function. Aqueous humor (AH) is the fluid that fills the eye and is constantly secreted from the ciliary processes posterior to the iris at a rate of 2 to 4 pl/minute(5).
In a healthy eye, the trabecular meshwork provides an egress for the AH in order to maintain a constant IOP at the interface of the iris and sclera (Figure 1, A). AH is subsequently shuttled into collection ducts in the anterior wall of Schlemm's canal and finally into aqueous veins. The pressure of the eye is determined by a balance between the production of aqueous humor and its outflow through the trabecular meshwork.
This study will focus on open angle glaucoma, the most prevalent form of the disease. In open angle glaucoma, cells in the trabecular meshwork morphologically or otherwise change, and AH cannot efficiently escape the eye (2). The most effective means of maintaining low IOP in patients has been shown to be glaucoma drainage devices (GDDs)(9), The design of these devices generally utse ab interno placement in order to facilitate drainage from inside of the eye to into another part of the eye. (10). This type of drainage device presents risk of various complications and efficacy issues that retract from patients willingness to undergo GDD procedures, especially due to the fact that the disease usually presents very late in life(5).
A transcomeal GDD would allow for the efficacy of a device based solution, while avoiding the many limitations of an ab interno device. The purpose of this study is to confirm the viability of manufacturing a transcorneal GDD from a nanoporous shape memory polymer created using a sacrificial nanofiber template.
2


Glaucoma
Glaucoma is characterized by damage to ganglion cells rooted in the ocular nerve. There are several types of glaucoma with the most common being open angle, also known as primary, glaucoma. This is the main type of glaucoma that may be treated with a glaucoma drainage device. In open angle glaucoma, the trabecular meshwork is still exposed despite ocular hypertension. It is known that the trabecular meshwork is unable to efficiently drain fluid, but a drug which corrects the cellular-level function is still unknown.
Another type of glaucoma is closed angle, which has a quick onset and is an acute disease resulting in fast deterioration of the optic nerve. The IOP spikes so dramatic because the iris presses against the trabecular meshwork, effectively closing the angle between them and eliminated the main egress of AH. This type of glaucoma requires quick reversal of angle closure in order to salvage vision, and a long term treatment such as a GDD is not usually considered.
A third, and perhaps most intriguing, type of glaucoma is called normal tension glaucoma (NTG). Patients with NTG experience loss of ganglion cell function despite a normal IOP. A treatment to NTG would give the community great insight into the disease (5).
Existing Treatments of Glaucoma
There is no known cure for the disease, and the cellular pathophysiology is largely unknown. The cells in the trabecular meshwork are believed to reorder their cytoskeleton, but studying this change presents significant difficulty as the cells do not maintain their character once taken from the living eye. There are, however, proven ways of managing glaucoma and
3


preventing further ocular nerve damage. By maintaining a normal IOP, which is around 10-20 mmHG, the nerves experience markedly decreased degeneration(-/). There are three strategies of managing glaucoma: drugs, surgery, and drainage devices.
Drugs
Topical drugs are usually the first line of offense used by doctors to combat high IOP. There are many drugs on the market, most notably are Bimatoprost, Travoprost, Latanoprost, and Timolol. All of these drugs are P-blockers, which cause vasodilation of the ciliary arteries leading to less blood flow to the eyes(77). Although this strategy does not target specific pathophysiology of glaucoma, it is nevertheless effective in reducing IOP -generally by around 30% (72).
Prostaglandins have been recently been used more commonly because they only require application once a day, compared to P-receptor antagonists which may be applied three or more times a day (13, 14). This type of drug increases the outflow of AH from the eye and have several side effects including eyelash growth and iris color change.
Figure 2 A patients eye exhibiting ocular surface irritation
4


Investigations on the efficacy of prostaglandins are limited by species-dependent reaction to these drugs (75).
There are several challenges in using topical drugs to lower IOP. First is the frequency of application, which may be one to two times a day (14). Patient adherence to daily topical drugs has been shown to be much less than they report to the clinician, as low as 59%, resulting in failure to lower IOP(76).There is an associated ocular surface irritation in some patients that makes this type of treatment unbearable (Figure 2). The chemical used in glaucoma drugs have a body-wide effect, sometimes causing respiratory and/or cardiovascular complications^ 7). Finally, the IOP lowering efficacy of drugs for patients with higher (over 21mmHg) IOP is not consistent (75).
Surgery
Each patient responds differently to medication, and if treatment with drugs alone is not sufficiently reducing IOP, doctors may suggest a surgical solution. Surgery is a more aggressive tactic in combatting high IOP, with effective results and many associated risks(7S).
Laser trabeculoplasty (LT) induces a thermal reconstruction of the trabecular meshwork, leading to higher drainage rate. Although LT is a common procedure due to its low risk of infection, it is ineffective in late stage glaucoma and subsequent treatments are generally minimally effective at further lowering IOP(79, 20). Furthermore, this technique requires a high degree of skill for the surgery.
5


Trabeculectomy is the gold standard invasive glaucoma procedure that aims to bypass the trabecular meshwork by dissection of the sclera, creating a scleral flap and opening a hole to the anterior chamber (Figure 3). Approximately 24,000 Medicare patients receive this treatment each year(27).This technique results in a bleb, or blister like pouch on the surface of the sclera which AH is allowed to drain into. Blebs have a tendency for long term complications including acute blebitis (shown in Figure 4) and leaking(70). The high
Cornea
Pupil
Lens
Conjunctiva

4*
Fluid exits through new opening

Meshwork
Sclera
Figure 3 Diagram of trabeculectomy procedure. An alternative egress for AH is introduced underneath the sclera.
Green arrow shows new path of AH.
morbidity of a trabeculecomy limits its use. The acute post-operative complications include choroidal effusion (13%), wound leaks (11%), shallow anterior chamber (10%), and anterior chamber bleeding (8%) (22). There are also many long term complications which include corneal edema (9%), dysesthesia (ocular discomfort) in 8%, fluid leaks (6%) and endophthalmitis (serious eye infection) in 5% of patients (22). These complications, along with scarring over of the anterior chamber shunt, lead to a 46.9% failure rate over five years (22).
6


Figure 4 Acute blebitis in a patient who has undergone a trabeculectomy
Drainage Devices
An alternative to traditional surgery is a glaucoma drainage device. These devices come in many varieties that act in several ways. The most common device is called an Ahmed Valve or tube shunt, several of which are shown in Figure 5. This device acts in a
Figure 5 (Left) A variety of shapes of tube shunts that may be placed in an eye to lower IOP. (Left) An animation of an Ahmed Valve after being sutured into place. In practice, the pouch would be covered by the sclera.
7


similar way to a trabeculectomy. A pouch is surgically attached under the sclera and a tube in inserted into the anterior chamber to provide an additional egress for AH. There are many variations of this type of valve shown in Figure 5 (Left). The top row shows Molteno implants. The middle row shows a Krupin slit valve (left) and an Ahmed valve (right) implant. The bottom row shows a variety of Baerveldt drainage devices.
Complications of this type of device include improper drainage (hypotony or failure to lower IOP), bleb infection and scleral tissue erosion. Early stage complications are experienced by approximately 21% of patients, and late stage complications occur in about 34% of patients (23). Furthermore, the relatively large size of these devices limits the number of additional devices that can be implanted upon failure. Doctors can generally only implant two devices into an eye before they deplete the accessible scleral tissue. If additional devices do not correct the problem, patients will undergo a cyclodestruction therapy, which generally consists of laser treatment around the ciliary body. This is aimed to decrease AH production but often results in significant collateral tissue damage resulting in a decline or loss of vision.
8


Micro-Invasive Glaucoma Surgery (MIGS) devices are an alternative to traditional GDDs in mild cases of glaucoma which aim to minimize surgical complications. These devices are inserted ab interno, meaning they are placed in the interior the eye and drain to another chamber inside of the eye. As seen in Figure 6, the iStent (Glaukos, USA) is a
titanium tube shaped device that is inserted into the trabecular meshwork to shunt AH directly into the Canal of Schlemm. Complications for iStent include fibrotic blockage, temporary IOP spike, corneal edema, stent obstruction by blood clot or iris, anterior chamber collapse, and vitreous incarceration (24).
A novel approach was achieved by Transcend Medical with their Cypass Microstent. This device shunts AH into the suprachoroidal space, a physiological drainage site which is secondary to the trabecular meshwork. It has been shown that the suprachoroidal space has a great capacity for drainage once an artificial fistula is introduced(25).
Although MIGS devices offer fewer and less severe post-operative complications, there efficacy is only shown for individuals with mild ocular hypertension (24). Additionally, due to their small size there is a high rate of scarring over of the devices inside the eye, where
Figure 6 Size of the iStent Drainage Device (Glaukos)
9


repair is fast and largely unpredictable. This leads to uncontrollable drainage rates resulting in either hypotony or no IOP lowering effect.
A transcomeal drainage method of reducing IOP offers several advantages compared to GDDs currently on the market. This strategy aims to shunt aqueous humor across the cornea directly onto the tear film, where it is naturally evacuated through the puncta. Microoptx (USA) is currently developing a transcorneal drainage device called the BG Implant (Figure 7, Top) and released some clinical data in the summer of 2015. This data
50
Flow rate: 2.5 pl/min (physiological)
o> 40 E
H 30
Q.
2
0 20 §
1 10T-t ..-,J' "if r n:
2 Q-
o)I|I|I|I|I|I|II
0 48 96 144 192 240 288 336
Time (hour)
Figure 7 (Top) The MicroOptx BG Implant next to an Ahmed Valve and a dime, for size reference. (Bottom) A showing in vitro data of the BG Implant maintaining a constant IOP at 12 mmHg over a 14 day period.
10


indicates that the BG Implant achieved a stable IOP at around 12 mmHg while maintaining a microbe free environment through the device (Figure 7, Bottom) (26). These results show a promising future in the use of transcorneal devices as they require a low level of surgical expertise, making them widely available in countries like India where technical expertise is not readily available, through monetary or geographic reasons.
Although Microoptx offers a viable device for the treatment of glaucoma, they are missing several key features. The device is coated in a collagen layer which helps the device adhere to the corneal tissue. This makes the device permanently attached in the eye and will not allow replacement if the device is clogged by proteins or debris in the AH. Also, the device is not tailorable to individual IOP reduction needs. This follows the one-size-fits-all approach similar to many devices, which ultimately is the reason for many post-operative complications.
Sizing of MIGS devises are generally on the millimeter scale. The exterior form factor usually differs from the actual drainage path, including anchors or features which allow the device to secure itself into the eye. Drainage paths vary enormously, depending on the downstream pressure barrier. For example, a device draining into the suprchoroidal space will provide much less internal resistance than a device draining to the tear film. This is because the suprachoroidal space contains intrinsic resistance to flow or AH accumulation whereas the tear film can accept as much AH as physiological flow rate. For comparison, 2-4 microliters per minute of physiological flow equates to about one drop every half hour.
11


Liquid Crystal Elastomers
Background and Synthesis
Liquid crystal elastomers (LCEs) are a class of smart material which are defined by thier combination of rubber elasticity of a lightly crosslinked polymer with self-organizing liquid crystals. This material system was first proposed by de Gennes etal in 1975(27) and then realized by Finkleman et al in 1981(25). Such a unique combination of subsystems gives LCEs the ability to reversibly and repeatedly change its shape and optical properties due to exposure to a stimulus, such as heat (29).
Figure 8 a) Grey mesogens in an end-on configuration, b) Mesogens in a side on configuration, c) A main chain LCE with mesogens integrated into the polymer chain.
The stimulus induced phase transition is made possible by the self-organization of the LCE monomers called mesogens. Mesogens are rigid molecules usually comprised of two or more aromatic rings connected in shapes like rods, discs, or bent banana shapes(20). These mesogens can be linked together with flexible spacer molecules to form a main chain LCE, or may be attached to a polymer as a side group to form an end-on or side-on LCE, pictured in Figure 8. Main chain LCEs have been of particular interest in the past few years because
12


of their high strain actuation compared to side-on or end-on LCEs as well as newly discovered synthesis pathway which allows for highly reproducible and tailorable samples to be made (31).
The unique properties of mesogens manifest in a phase transition at the isotropic temperature, or T,. Below Ti, mesogens will naturally align organize into monodomains which can form in two general configurations. When mesogens are aligned in only one axis, it is termed a nematic monodomain. When alignment occurs in two directions it is a smectic monodomain. The natural alignment of mesogens into a monodomain relies on their ability to move freely. However, when a crosslinked network is introduced to form an LCE, covalent bonds disrupt the free movement of the mesogens. This results in a phase of only localized order, called the polydomain. In all three cases, when the material is taken above T;, it will undergo a phase transition as the mesogens completely lose their order and become randomly oriented. This is now called the isotropic phase. An optical change can be observed while heating a sample as the opaque crystal structure turns unordered and transparent.
Although naturally forming in a polydomain, LCEs can be programmed into a monodomain. This is known as a liquid single crystal elastomer. In this state, the mesogens are aligned uniformly throughout the entire sample. Monodomain can be achieved through several means, depending on the chemistry used to synthesize the LCE. Traditionally programming of the monodomain was achieved at the same time as synthesis by three basic tactics: wiping a glass slide with polyimide in the desired direction of mesogens orientation, exposure to a strong (>1 Tesla) magnetic field, and mechanical stretching during the reaction (30). These methods of programming monodomain during synthesis can be unreliable, not easily reproducible and present strong limitations of sample geometry (only thin films are
13


able to be made with polyimide and magnetic orientation). Recently, Yakacki et al has proposed a two stage thiol acrylate Michal Addition photopolymerization (TAMAP) reaction to synthesize a reproducible, facile and tailorable nematic main chain LCE(29). Several recent studies show the effectiveness of the TAMAP methodology (32-34).
Photocrosslink (Stage 2)

(a) (b) (c)
Figure 9 Schematic of the second stage of a TAMAP reaction of an LCE. Sample starts after the first stage as a stable polydomain LCE (a). Sample is then stretched uniaxially to align mesogens within the polymer chains (b). Finally the polymer is photopolymerized to connect excess acrylates present in the sample (c). The polymer is now programed into a monodomain phase.
The TAMAP reaction represents a dramatic change in the way which researchers are able to synthesize and explore the mechanics of LCEs. Up to 2012, the only relatively simple method to create an LCE from a functionalized monomer was a free radical polymerization (30). This reaction was often hard to carry through in a homogeneous way, leading to uncontrollable liquid crystal domains.
Another reaction, proposed by Kupfer et al relies on the functionalization of a Si-H bond in the presence of a Pt catalyst (35). This reaction, deemed hydrosylilation, relies on the functionalization of a cross linking molecule with both a vinyl and an acrylate group, which exhibit two different reaction rates. The vinyl groups react quickly in the presence of a Pt catalyst while the acrylate groups react slower. This allows the researcher to stretch the
14


polymer into a monodomain once they feel the vinyl groups have reacted, but the acrylate groups have yet to react. This reaction allowed researchers to create a monodomain bulk polymer, albeit with quite unpredictable crosslinking densitities and degrees of alignment within the monodomain(3d).
To be successful, these reactions must be carried out under very strict temperature conditions. Furthermore, only side-chain LCEs were able to be synthesized from these reaction. This limited not only the application of the reaction, but the interest of researchers who were searching for truly applicable hands-free actuation. A Michael-Addition reaction offers a new perspective on a traditionally grueling and unpredictable synthesis.
15


Using the two stage Michael-Addition reaction, LCE can first be synthesized in the polydomain and subsequently programmed into monodomain during the second stage via photopolymerization-induced crosslinking during mechanical straining (Figure 9). The mechanical properties and glass transition temperature of the LCE have been shown to be tailorable through varying the crosslinking density during the first stage of the reaction. Figure 10 shows a strip of monodomain LCE at room temperature get heated up past its Ti and lift a small weight. It then cools back to room temperature and returns to its original shape.
Figure 10 A monodomain LCE has been programmed prior to this animation. The sample starts at room temperature with a small binder clip attached to the end. As the sample is heated up the polymer transitions to polydomain. Once heat is removed, the LCE sample cools back down and expands to its original position.
16


Liquid Crystal Elastomers as a Transcorneal Filter
Using this system, a cylindrical filter synthesized from LCE in the polydomain phase may be stretched along its axis into monodomain phase, locked in place via UV curing, and may switch between a longer thin cylinder at cool temperatures to a shorter thick cylinder at body temperatures. The filter will be designed in a way which expands and locks into an outer housing in the cornea once it reaches body temperature. Pictured in Figure 11 is the schematic of a device as the filter is placed into the outer housing, expands and locks in place and is subsequently removed by cooling of the device. This gives the transcorneal filter a unique advantage over any existing glaucoma drainage devices. Insertion and removal of the filter will be performed with an application device, not pictured.
Figure 11 A schematic describing the two-part device concept and the replacing of a filter, a) the silicone/collagen outer housing is placed in a pilot hole on the edge of the cornea, b) A chilled drainage filter is placed inside the outer housing. This schematic shows an Antimicrobial LCE (A-LCE) which has been chemically coated with a layer of copper, c) As the drainage filter reaches body temperature, the filter will expand causing it to lock in place and maintain a microbe-free barrier between the filter and the outer housing, d) Once the device needs replacing, the device is chilled and can be easily removed from the outer housing.
17


Advantages of Using LCE
One key feature of the LCE is the heat activated shape switching ability, which allows the device to be easily placed into, and taken out of, the outer housing. This gives surgeons the option to easily switch in different devices which have different drainage characteristics. This feature also affords patients reliable lifelong glaucoma relief by allowing replacement if the device is clogged by accumulation of protein in the aqueous humor.
Antimicrobial Properties
Introducing a transcomeal egress for aqueous humor has the potential to put patients at risk of infection. To combat this risk, the pores in the transcorneal filter are designed to be less than one micron, the average size of human corneal bacteria (6). This size based barrier, combined with the constant outflow of aqueous humor through the device, ensure that there will be no bacterial infection happening through the body of the drainage device. It is important to note that not all bacteria will be larger than one micron. The constant flow of AH will prevent these small bacteria from infiltrating the device.
Electrospinning
Background
In order to produce consistent, sub-micron channels in the LCE, a two polymer system is used that relies on a sacrificial template of nanofibers. A difference of solubility was taken advantage of in this study by using a water soluble polymer to form the sacrificial nanofibers. This allows the fibers to remain stable as the LCE monomers are infiltrating the fibers. Toluene was used as a solvent for the LCE and will not disturb the nanofiber system.
18


A similar study was conducted by Luo et al during work aimed at developing a self-healing polymer by embedding in it electrospun nanofibers to create a two-polymer composite(37). This system also relied on the solubility differences between the polymers, but differed from the system in this study by leaving the water soluble nanofibers intact in the polymer.
Another study was conducted with yet a more similar system of nanofiber composites, conducted by Bellan et al (38). In this study, a composite of water soluble nanofibers inside a bulk hydrophobic polymer (polydimethylsiloxane) was prepared and the electrospun fibers were subsequently dissolved out. This study aims to test if this technique is compatible with a shape memory polymer.
In both these studies, electrospinning was used to create the nanofibers. This has been shown to be a simple and fairly reliable way to produce continuous fibers on the nano- to micro-scale(3P-Â¥2). The process of electrospinning is based on a high voltage DC field dragging a conductive polymer solution across a charge gradient onto a grounded collector (39). The polymer is typically ejected from a syringe at a low flow rate while the positive charge is placed directly to the syringe needle.
The collector may have a variety of configurations, including a flat plate, rotating drum, wire, or pin electrode (pictured in Figure 345y). The type of collector plays a major part in the morphology of fibers collected as well as several other key parameters: voltage charge applied, tip-to-collector distance, polymer concentration, molecular weight of polymer, humidity of the room, and additions to the polymer solution (salts, pH control, etc.) (41). It is noted in literature that the parameters most influential on the fiber morphology are
19


polymer solution concentration and tip-to-collector distance (40-42). Nanofiber diameters are increased by increasing solution concentrations and decreasing tip-to-collector distance.
Tailoring Fiber Diameters
Tailoring the nanofiber diameters is an important step in developing a robust solution to transcomeal drainage because the size of the nanofibers produced will correlate to the diameter of pores left in the filter after dissolution of the fibers. The diameter of pores in the sample is perhaps the most impactful parameter of the filters drainage ability. If we consider only one pore which runs through the filter, its capacity to drain fluid at a given pressure is given by the Hagen-Poiseuille equation:
APnd4
Q = -------
v 128 fiL
Where Q is the flow rate, P is the pressure gradient, d is the pore diameter, p is the fluid viscosity and L is the length of the tube in question. It is clear that a small change in diameter will greatly affect flow rate through the tube. Therefore the diameter of the electrospun fibers is a key parameter is the overall filtration rate of a transcorneal GDD and will be investigated during this study.
In summary, a transcorneal glaucoma drainage device will be fabricated so as to maintain an IOP of ~15mmHG under physiologic AH production of 2-4 pl/min. The filter will be made using a sacrificial template of electrospun water soluble nanofibers embedded in a bulk LCE. The system will be qualified with SEM and fluorescent microscopy images as well as perfusion tests.
20


CHAPTER II
MATERIALS AND METHODS PVA Solution Preparation
Granualted polyvinyl alcohol (Mowiol 8-88, Sigma Aldrich) was dissolved in water at several concentrations (8%, 10%, 12%, 17%, 20% and 22%), by weight, by vigorously mixing at 85 oC overnight.
An additional granulated poly(vinylalcohol) (Mowiol 10-98, Sigma Aldrich, USA) was also prepared for electrospinning by dissolution in pure water. Concentrations of 8%, 10% and 13% were used in this study. Solutions were prepared by stirring at 95 C over 48 hours.
Electrospinner Fabrication
In order to avoid extreme and unnecessary prices of commercially available electrospinning machines, one was designed and fabricated in house. Two items are essential when making such a device an AC/DC converter (ESCC 0305, Astrodyne, Mansfield MA),
Figure 12 Image on the left shows the DC electric current output box. The diagram on the right shows the components and circuit configuration inside the box. a) AC current input, b) AC/DC power converter c) High voltage converter, d) LCD display e) Potentiometer f) Switch g) LCD display power input, h) High voltage ground, labeled GROUND on left, i) High voltage DC output, labeled HIGH VOLTAGE on left.
21


and a high voltage output (25A12 Series, Ultravolt, Ronkonkoma NY). An acrylic box was constructed to hold these two components, as well as an additional AC/DC converter to provide power to a rotating drum collector, an LCD readout screen displaying the current voltage (PM128, Jameco, Belmont CA), and adjustment knobs provided courtesy of the University of Colorado machine and calibration lab. In order to avoid the high voltage from deactivating other components, the high voltage output had to be removed from the main cabin and placed into a small acrylic box of its own and attached to the top of the main acrylic box, where it was accessible to wiring. Wiring schematic and photo of the electrospinning device is shown in Figure 12.
Nanofiber Production
Figure 13 (Left) The electrospinning setup (excluding the power supply and syringe pump). A polymer jet is dragged across a DC charge gradient from a needle tip onto a grounded wire in the form of continuous nanofibers. (Right) A pin-gap electrode is pictured with blue PVA nanofibers collecting across the gap. The two pins are separated by ~1 in. and are both grounded.
Nanofibers are produced by electrospinning a polymer jet of polyvinyl alcohol onto a wire collector. The polymer solution is placed in a 20ml syringe and is ejected from the tip of a 22 gauge blunted needle at a rate of 1 ml/h by a standard syringe pump. The positive side of the high voltage DC power converter 8is connected to the needle with a charge of 10 kV. The collector is a grounded 25 um copper wire with a length of around 7 inches. It is placed
22


perpendicular to the needle at a distance of 8 inches. The electospinning process is allowed to proceed for approximately 30 minutes, with attention to not allow web formation connecting the grounded wire to surrounding environment. All electrospinning is conducted at room temperature and a relative humidity of 14%. The experimental setup is pictured in Figure 13. Experiments were also done using a house made rotating drum collector and pin electrode collector in order to produce fibers arrays of aligned morphology.
The production of different shapes and sizes of fibers is able to be achieved by altering many parameters of the electrospinning process such as voltage charge, tip-to-collector distance, needle size, solution flow rate, collector geometry, and solution concentration. In a preliminary study, it was observed that solution concentration had the most drastic effect on fiber morphology and therefore was the one parameter that was tested throughout this study.
Polymer Synthesis
LCE Synthesis
LCE was synthesized according to Yakacki et al using a thiol-acrylate Michael Addition reaction. The mesogenic monomer used was RM 257 (Wilshire Technologies, Princeton NJ), and the flexible spacer was 2,2-(Ethylenedioxy) diethanethiol (EDDT) (Sigma
HS^o'
0 o
Figure 14 Chemical structures of the monomers involved in the TAMAP reaction, a) RM 257 b) EDDT and c) pentatetrakis (3-mercaptopropionate)
23


Aldrich, USA). The LCE was syntheized with 15% tetra-thiol crosslinkers, pentaerythritol tetrakis (3-mercaptopropionate). Figure 14 shows the monomers. Samples were synthesized with a stoichiometric ratio of functional acrylates to thiol groups, with 15% of the functional thiol groups belonging to the crosslinking molecule.
LCE was typically made in small batches according to the following: to a 20 ml scintillation vial, 1 gram of RM257 powder was added. This was dissolved in 0.30 g of toluene at 80 C. To the solution, 0.2942 g EDDT spacer and 0.021 g of pentatetrakis (3-mercaptopropionate) crosslinker was added. After thorough mixing, 0.245 g of a 50:1 toluene/DPA catalyst was added and thoroughly mixed, which initiated the reaction. Within 5 minutes of adding catalyst, LCE was poured into a mold and left to fully polymerize overnight.
The mold which LCE was polymerized in was generally a pipette, which was a cheap and readily available method of creating cylindrical samples. For preparing porous LCE, the tip of the pipette was used. First, a wire coated with electrospun nanofibers was inserted approximately one inch into the tip. The tip was then dipped in a mixture of LCE monomers soon after catalyst was added and the mixture was still quite fluid, after a vacuum treatment to remove air bubbles, the pipette was placed vertically to allow polymerization overnight.
In preparation of samples which showed the shape changing characteristics of the LCE, the wide end of the pipette was used. To synthesize these samples, the tip of the pipette was first plugged with wax. Soon after adding catalyst to a mixture of LCE monomers, the reaction was poured into the back end of the pipette until filled. Again, a vacuum treatment was applied and the LCE was left to polymerize overnight. After polymerization was
24


complete, the pipette was smashed and LCE was left overnight to evaporate all remaining solvent.
Test Samples
In order to quickly test composite systems, a shape memory polymer (SMP) which could be rapidly fabricated was used in place of LCE in some cases. A key attribute of the SMP used in place of LCE was its low glass transition temperature, which made the polymer rigid at room temperature. The SMPs were also amorphous which made them optically transparent, allowing imaging through samples possible.
Two different SMPs were used in this study. The first was a 5:1 tert(butyl aryl ate) tBA and di(ethyleneglycol) dimethacrylate (DEGDMA) mixture. The second was a 2:3 methyl methacrylate (MMA) and poly(ethyleneglycol) dimethacrylate (PEGDMA) mixture. Both of these were free radical driven polymerizations that were cured in 20 minutes under UV light. The MMA/PEGDMA mixture exhibited less shrinkage when polymerized which made possible synthesis of very thin, homogeneous samples.
25


Composite Formation
Figure 15 Picture schematic of a circular composite being formed. A 25 micron copper wire is a) electrospun onto and then b) inserted into a tube, either capillary tube, pipette or vial depending on the end use. The cross section of the tube with wire inside is seen on the bottom left, c) The nanofibers are then infiltrated by LCE polymer and vacuum treated to evacuate any air. In a later step, d) the the fibers are dissolved out leaving a nanoporous sample.
Composites were formed in a variety of ways, depending on their end use. These are diagramed in Figure 15. The first way was a mat composite, where the PVA nanofibers were sandwiched between two layers of LCE. To accomplish this, immediately after adding catalyst to the LCE monomers, the reaction was brushed on a glass slide coated with a hydrophobic rain repellent (used as a release agent). An electrospun fiber mat was laid on top of the first layer of LCE and then more LCE was spread on top of the nanofiber mat. Another glass slide coated in release agent was clamped on top of the composite overnight while the LCE completed polymerization.
To make prototypes suitable to test perfusion using basic laboratory equipment, composites are formed in pipettes. After a wire collector has been removed from the
26


elctrospinning apparatus by cutting both ends, it is strung through the pipette with the midpoint of the wire in the widest portion of the pipette. The wire is secured into place at the tip of the pipette with superglue, while also sealing off the tip of the pipette. A custom made, 3D printed cap is placed at the other end of the pipette which allows for centering of the wire, as well as supplying an ingress for LCE to be poured. The LCE monomers are prepared and quickly after adding catalyst, the reaction mixture is injected through the pipette cap opening with a syringe. The pipette is placed in a 15 mmHG vacuum for five minutes to degas the composite and is left to cure overnight.
1. Nanofibers 2. Engulf fibers with polymer 3. Dissolve out fibers
Figure 16 Picture schematic of creation of mat composites. Nanofibers are engulfed on top and bottom with LCE polymer. Fibers are dissolved out in a later step.
Composite samples using tBA and DEGDMA (5:1) were also synthesized in the front and back end of glass pipettes. These samples had a larger size (1mm and 5mm diameter) for the front and back end respectively, which made them easier to handle during perfusion tests. These samples were also formed in rectangular geometries by sandwiching tBA/DEGDMA and a flat sheet of nanofibers between two glass slides (Figure 16). The rigidity of the tBA/DEGDMA polymer at room temperature also aided in testing.
27


Nanochannel Formation
Once a composite has been removed from its mold (pipette, glass slides), it is dried overnight in an 80 C oven. The composite is then sliced into smaller pieces and placed in a vial of water. The composite is gently stirred in this solution for two days at 60 C.
Imaging
Nanopores are confirmed by first freezing samples and then slicing cross sections of the cylinders near the center. Carbon sputtered SEM imaging was used to obtain high resolution images of the individual nanopores. Imaging was kindly conducted by Dan Merkl from the University of Wyoming Mechanical Engineering department.
Fluorescent imaging was also performed in order to see continuous channels. Samples used in fluorescent imaging were prepared with a MMA and PEGDMA (3:2) mixture instead of LCE for several reasons. First of all, the MMA/PEGDMA is an amourphous glassy polymer at room temperature, which in lay terms is clear and hard. This makes handling and imaging possible at room temperature.
To prepare samples, the 3:2 mixture of MMA/PEGDMA was placed in a 1mm thick mold until it filled the mold half way. A section of an electrospun sheet, created on a flat plate collector, was placed on the MMA/PEGDMA and the mold was filled up all the way. The MMA/PEGDMA was subsequently crosslinked via UV irradiation. Once the composite was removed from the mold and the nanofibers were dissolved out in water for 2 days in 60 C water, the samples were dried in a 70 C vacuum oven. After the channels were evacuated, the samples were submerged in a 100 pM solution of fluorescein (Sigma Aldrich, USA). After soaking for 30 minutes, the submerged samples were introduced into a 15 inHg
28


(381 mmHg) vacuum 4 times in order to clear the air from the nanochannels and replace that space with fluorescein solution. After soaking for two days, the sample surfaces were dried with a wipe prior to imaging.
Confocal images were collected using a Nikon Eclipse Ti C2 LUN-A microscope (Nikon, Tokyo) equipped with two C2-DU3 high sensitivity PMT dtectors, 4 diode lasers (405/488/561/640 nm), and a motorized microscope stage with 3 axis navigation (X, Y and Z). A 488 nm laser setting was used to capture images.
All images were analyzed with ImageJ software.
Perfusion Testing
A liquid perfusion setup was assembled in house to confirm that the pores were continuous and would allow for steady and predicable outflow. This was done using a column-based water pressure system in conjunction with several test sample fixtures. In the first sample fixture setup (Figure 16), a cylindrical sample formed in the small end of a pipette was secured via super glue back into a pipette after dissolution of the nanofibers. The pipette was filled with a dye and the tip of the sample was dipped into a collection vial. The second perfusion setup achieved a higher back end pressure on the device. A test sample was secured this time in the large end of a pipette with superglue while other side of the pipette was superglued into a 64 inch long tube filled with water and dye. Perfusion was confirmed with visual confirmation of dye on the exposed end of the test sample.
29


Figure 17 A perfusion test setup including a pipette filled with red dye, a nanoporous filter secured with super glue in the tip of the pipette which is partially submerged in a collection bath where dye is collecting, indicating successful perfusion.
I
Figure 18 Water column perfusion assembly.
30


The third perfusion setup (shown in Figure 18, 19) achieved a greater number of pores exposed to the pressure of the water column. First, a flat sample was prepared by sandwiching a rectangular mesh between a 3:2 MMA/PEGDMA mixture. The nanofibers were dissolved out from the bulk polymer and the porous sample was imbedded in a strong fast-cast urethane (Master Fast-Cast Urethane, Dynacast) inside a cylindrical mold. Once set, a 1mm thick cross section of the epoxy cylinder containing the porous sample was cut on a lathe. This thin cylinder was placed between two flanged tube fittings and sealed with silicone gasket sealer on either side. This apparatus was inserted on the bottom of the water column and perfused with a 1:1 mixture of dimethylformamide (DMF) and water, which was collected in a vial. Visual confirmation of fluid in the collecting vial was used to qualify perfusion.
Figure 19 (Left) Schematic of the third type of perfusion setup. Perfusion fluid enters from a water column into a flanged aluminum fixture. A casted urethane disc containing a nanoporous MMA/PEGDMA sample is placed in the flow path and sealed with a silicone disc between each aluminum fixture. Fluid drains through the nanopores out of the other end and collects in a vial for further testing. (Right) Picture of the test disc used for perfusion qualification. Two rectangular pieces are seen imbedded in the disc, each of which contain nanopores. Using two pieces of porous sample yields a higher perfusion rate which will be easily detectable by visual confirmation.
31


Figure 20 Picture of the test fixture in line with a water column. Fluid will drain from the water column through the porous section of the test disc, and ultimately collect in a vial.
This test was performed for 90 hours with two different discs. One contained large pores fabricated from a sacrificial template made with 20% PVA electrospinning solution and one contained small pores with a 10% solution. Evaporation was accounted for by using a 10% glucose solution as perfusion fluid and back-calculating flow rate through dry weight of the fluid which perfused over the 90 hours.
The length of tubing necessary was determined by the amount of pressure needed to simulate physiological eye pressure for a given sample length. This value is reported to be between 15-30 mmHg for glaucomic patients. In order to simulate a 1mm long clinical device using a 6mm long test sample, a back pressure of 90-180 mmHg was needed. For example, a 6mm long test sample requires six time the eye pressure in order to simulate a 1mm clinical device. In this study, pressure of 20 mmHg was considered a disease-state IOP, requiring 64 inches of water pressure for a 6mm long test device. All perfusion setups were also tested with a control sample of identical geometry and no nanochannels to confirm efficacy of the setup.
32


CHAPTER III
RESULTS
Electrospinning
Using the house-made electrospinning apparatus, PVA nanofibers were fabricated in a variety of sizes topological morphologies. In general, as the concentration of PVA inside the syringe was increased, the fiber diameters increased and fiber formation was more stable.
Figure 21 SEM images of PVA nanofiber mats collected on a flat plate. All parameters were identical besides a change in PVA solution from 8%, 10%, 15% and 20%
33


At a low concentration, fibers formed in ribbon like morphologies (Figure 21.a). At a concentration of 10%, the fibers stabilize (Figure 21.b).At higher concentrations of 15% and 20%, fiber diameters increase (Figure 21.c, Figure 21.d). Fiber measurements for Mowiol 8-88 and Mowiol 10-98 are given in Table 1 and Table 2, respectively.
There were stark differences in the fiber formation between the two types of PVA used in this study Mowiol 8-88 and Mowiol 10-98. While Mowiol 10-98 yielded higher fiber diameters at a given concentration, its high level of hydrolysis hampered its ability to dissolve in water, therefore limiting the maximum fiber diameter achievable. The high hydrolysis of Mowiol 10-98 also made the fibers produced at lower concentrations more stable (Figure 22).
34


Average Fiber Diameters forTwo Types of PVA
D.6
05
04
04
Sb O.Z
5
0 1
0

L.
T / / F
r/ A / / /
> / ' H
r
0 S 10 IS 20 25 30
Law Hydrotyas
High Hydra lysis
Concentration [vu/w]
Figure 22 Graph showing results of average fiber diameters of various concentrations of two samples of PVA with different degrees of hydrolysis. The PVA with a high degree of hydrolysis produces stable fibers at lower concentrations, and produces higher fiber diameters compared to the low hydrolysis sample of the same concentration. The low hydrolysis PVA is able to be dissolved in water at higher concentrations and is able to produce an overall higher fiber diameter.
35


Table 1 Fiber diameter measurements of Mowiol 8-88 for concentrations of 8%, 10%, 15%, 20% and 25% (w/w). All measurements in nanometers.
Measurement 8% 10% 15% 20% 25%
1 152 217 205 407 263
2 210 106 142 468 436
3 77 132 173 520 712
4 105 106 226 414 431
5 143 94 116 229 323
6 175 126 94 213 421
7 118 67 15 288 190
8 139 177 157 576 612
9 196 137 106 379 487
10 66 54 195 325 237
11 180 70 95 223 229
12 261 152 94 402 351
13 193 164 91 523 276
14 141 142 229 269 446
15 114 76 124 380 569
Mean 151.3 121 147 374 399
SD 52.3 46 49 116 152
Min 66 54 91 213 190
Max 261 217 229 576 712
36


Table 2 Fiber diameter measurements of Mowiol 10-98 of concentrations 8%, 10% and 13% (w/w). All measurements in nanometers.
Measurement 8% 10% 13%
1 124 122 349
2 71 171 371
3 90 155 239
4 82 112 236
5 81 310 290
6 51 140 272
7 86 207 445
8 71 93 338
9 82 152 196
10 90 94 201
11 71 75 298
12 81 222 202
13 78 124 259
14 66 176 331
15 81 99 223
Mean 80 150 283
SD 16 61 72
Min 51 75 196
Max 124 310 445
37


Alternative fiber collectors yielded fiber morphologies which may be useful in the final design of a transcorneal drainage device. Using the pin gap electrode collector, arrays of aligned fibers were achieved (Figure 23). Collecting fibers on a wire was a useful way to form and easily manipulate a fiber array in a cylindrical shape. As shown in Figure 24, fibers are able to be collected on a wire with a stable and randomly oriented morphology.
Figure 23 (left) SEM image of aligned PVA nanofibers collected on a pin gap electrode. (Right) Laser scanning topology microscopy of aligned PVA nanofibers collected on a pin gap electrode. Color indicates depth with red being closest to the camera.
38


Figure 24 (Left) SEM image of PVA nanofibers collected on a 25 micron wire. The diameter of fibers collected around the fiber is approximately 200 microns. (Right) Magnified SEM image of the fibers pictured on the left. Fiber morphology indicates stable, branching fibers.
The investigation into electrospinning proved that tailorable fiber arrays of various morphologies and sizes are able to be made easily and reliably using several different collectors.
39


Nanofiber/Polymer Composite
Nanofibers produced via electrospinning PVA on a 25 micron wire were successfully incorporated into a cylindrically shaped memory polymer composite. These samples were made in a variety of shapes and with several shape memory polymers depending on the final use. Pictured in figure 25 is a tert-butyl acrylate (tBA) with di(ethylene glycol) dimethacrylate (DEGDMA) crosslinker shape memory polymer and a nanofiber-bearing 25 micron wire composite which was formed in a 100 micron ID capillary tube. This sample was prepared to show viability of creating a sample mimicking the size of a filter which would be incorporated into a glaucoma drainage device.
Porous LCE
Cylindrical samples of pure LCE were fabricated in pipettes to show the change in diameter with exposure to heat. The sample shown in Figure 26 starts at 0 C (Left) and contracts over 2 minutes of exposure to heat (Right). The diameter of the cylinder changes from 0.14 in to 0.20 in, a change of 43% the original diameter. The change in shape
Figure 25 A cylindrical composite (D = 100 micron) of nanofibers on a wire with tBA and DEGDMA crosslinker.
40


represents a molecular-level reorganization of the mesogens from a monodomain (Figure 26 Left) to polydomain (Figure 26 Right).
Figure 26 A cylindrical sample of LCE transitioning phases due to heating. (Left) Cylindrical sample starts at 0 oC and a diameter of 0.14 in. The sample is in the monodomain.(Right) After heating of the sample for 2 minutes, the sample transitions to polydomain and has a diameter of 0.2 in.
Porous cylindrical samples created with LCE were used to confirm the existence of channels within the sample. Cross sections of cylindrical nanoporous LCE samples, taken in the center of the sample, were analyzed under SEM imaging (Figure 27). The surface shown was within 100 microns of the wire, where hollow channels were expected to have taken the place of PVA nanofibers. The holes seen in the surface have both circular and elliptical topology, resulting from the random orientation of the fibers at this cross section.
41


Figure 27 SEM image of the surface of a cross sectioned cylindrical nanoporous LCE. The image was taken from a sample which used the fiber-on-a-wire approach to composite formation. The topology and size of the holes seen in this picture indicate that they were indeed formed from dissolved nanofibers. Some holes are elliptical in shape, indicating that a cylindrical fiber was passing through that cross section at an angle.
42


Fluorescent Imaging
Flattened composites for fluorescent imaging were also successfully created. These were made using several methods. The first was using a 1mm thick spacer between two glass slides. This produced an approximately 1mm thick rectangular sample. The second way in which flattened samples were created was without any spacer at all. These samples were approximately 80 microns in thickness. Using these two styles of sample thicknesses, images shown in Figure 28 were captured using fluorescent confocal microscopy. Both samples were created with nanofibers formed from electrospinning a 20% w/w solution of PVA. Fiber measurements from the 80 micron thick sample are shown in Table 3.
Figure 28 Fluorescent microscopy image showing nanochannels infiltrated with a 100 pM solution of fluorescein. Images from two samples are shown. On the left is a 1mm thick sample and the right shows an 80 micron thick sample. Both samples were made with nanofibers formed from electrospinning a 20% w/w concentration of PVA.
43


Table 3 Fiber diameter measurements taken from the 80 micron thick sample containing nanochannels formed from 20% w/w concentration PVA solution electrospun on a flat plate collector. Measurements are in nanometers.
Measurement Diameter
1 684
2 568
3 546
4 312
5 621
6 403
7 494
8 287
9 462
10 403
11 479
12 408
13 305
14 429
15 456
Mean 457
SD 114
Min 305
Max 684
The average fiber diameter of nanochannels are compared to the sacrificial nanofiber template in Figure 29. The average diameters were taken from a small area of the samples, and merely approximate the total average diameter with 15 measurements from each image. Using a two sample t-test, the means are significantly similar (P < .01).
44


Average Diameter of Sacrificial Fibers and Nanochannels
600 -| 500 -400 -300 -200 -100 -0 -
Electrospun Mat
Nanochannels
Figure 29 Bar graph comparing the average diameter of electrospun fibers and nanochannels (taken from fluorescent images). Both were formed from a 20% w/w PVA electrospinning solution.
Perfusion Testing
Perfusion was shown using three test fixtures. The first was a simple pipet apparatus. Figure 30 shows two pipette perfusion tests after 48 hours. The red dye was perfused through a 6mm porous cylinder, and is clearly seen to be collecting in the water bath. The green dye was used in a control setup where the 6mm cylinder did not contain any pores. The lack of green dye in the collecting bath shows that no dye is escaping the pipette around the edges of the cylinder.
45


Figure 30 A pipette based dye perfusion test. The red dye was perfused through a porous cylindrical sample, while the green dye was used in a control setup with a cylinder containing no pores. Perfusion is confirmed in the red test while the green dye did not penetrate the fixture.
The second perfusion setup used a water column to increase the pressure behind the filter in order to reach physiological conditions. The thicker cylinder diameter of this setup made handling and securing test samples in place with ease. Results were obtained using a 5x magnification microscope and confirmation of a blue dye reaching the other side of the filter. The results in Figure 31 show blue dye along the edges of a copper wire, with some blue dye surrounding the wire. It is apparent that the dye traveled through the filter and wetted the exposed face. Because flow was extremely slow, the water in the dye was able to evaporate, depositing spots of solid dye on the face of the sample.
46


Figure 31 Enhanced contrast image of the end of the large cylinder/water column test fixture after 48 hours. The blue seen next to the copper wire shows that dye has perfused through the cylinder. Blue dye seen around the wire is the result of surface wetting followed by evaporation of the dye.
The third perfusion test was performed using a disc apparatus which minimized the thickness of the sample the liquid perfused. Confirmation of perfusion was attained by visual confirmation of fluid reaching the back side of the fixture. A collection vial was placed beneath the test fixture, but after 48 hours fluid was still captured inside the aluminum test fixture, held in place through capillary action. Figure 32 shows the downstream end of the aluminum test fixture after 48 hours. Fluid is visible inside the test fixture, indicating it passed through the porous disc.
The dry weight of the evaporated perfusion fluid was used to provide a more comprehensive qualitative analysis. The table of measurements and conversion into flow rate
47


is listed in table 4. The results indicate that, as expected, less fluid passes through a sample of smaller pores compared to a sample containing larger pores.
Figure 32 Fluid collected in the downstream side of the disc perfusion apparatus. Image was taken after 48 hours of continuous perfusion
Table 4 Data obtained from 90 hour perfusion studies of two discs with different sized pores. Large and small pores were formed with a sacrificial template made from 20% and 10% PVA electrospinning solution, respectively.
Pore Size (Concentration) Dry Weight Volume Fluid Perfused Flow Rate
Large Pores (20%) 0.822 g 8,220 ul 1.52 ul/min
Small Pores (10%) 0.371 g 3,710 ul 0.68 ul/min
48


CHAPTER IV
DISCUSSION
Trans-comeal drainage is on the forefront of glaucoma device industry interest. It offers a predictable outflow of AH while avoiding bleb formation and complications associated with invasive surgery such as scarring and inflammation. Implanting a trans-corneal drainage device requires minimal surgical skill and is therefore well suited for treatment of glaucoma in developing countries. There are several additional considerations when designing a trans-comeal drainage device compared to traditional ab interno devices, which are not exposed directly to the environment. First of all, the trans-corneal device must provide a microbial barrier. Secondly, the device must rely solely on its intrinsic drainage properties to manage IOP there is no downstream pressure barrier. Lastly, the trans-corneal device must be secured in place without the use of barbs or sutures as many ab interno devices are. Fabrication of a trans-corneal drainage device as proposed in this investigation meets all three of these requirements, as well as offering additional advantages for users and surgeons.
The first aim of this study was creating tailorable nanofibers via electrospinning. It is widely reported in literature that electrospinning solution is the most sensitive parameter in changing the diameters of nanofibers produced(43-46). It was found that by raising the concentration of PVA (Mowiol 8-88) in the electrospinning solution from 10% to 25%, fibers from 54 nm to 712 nm could be fabricated, respectively.
Fundamental information about the nature of electrospinning was also gained in this process. By using two types of PVA, Mowiol 8-88 and Mowiol 10-98, relationships between
49


average molecular weight and fiber morphology were observed, as well as relationships between degree of hydrolysis and solubility of PVA in water. While using 8% w/w concentrations of both Mowiol 8-88 and Mowiol 10-98, drastically different fiber morphologies were observed. While solutions of Mowiol 10-98 produced fibers with a unique, branching and cylindrical morphology, solutions of Mowiol 8-88 produced fibers with a flattened, ribbon like morphology. As indicated by Tao, J, this was directly related to the entanglement concentration of the type of PVA used. Mowiol 10-98 would yield an entangled solution at lower concentrations due to the increased hydrophilic interactions. These would create stable nanofibers upon electrospinning while the same concentration of Mowiol 8-88 would produce flattened fibers. This is an important observation because only unique, cylindrical fibers will be suitable to achieve a drainage device with tailorable, controllable and predictable outflow.
The relationship between degree of hydrolysis of the PVA and its ability to dissolve in water was counter intuitive, but allowed fibers of higher diameters to be fabricated when taken advantage of. The manufacturing of PVA is achieved by hydrolyzing poly(vinylacetate), a polymer which is not water soluble (Figure 33). Contrary to common sense, increasing the degree of hydrolysis does not always increase the water solubility. At
cch2cha
v z | yn
0
1
c=o
ch3
polyvinyl acetate
NaOH
methanol
ACH2CHA
z | ' n
O
H
polyvinyl alcohol
Figure 33 Hydrolysis of poly(vinylacetate) into poly(vinylalcohol) (7)
50


very high (99%) levels of hydrolysis, solid PVA forms very stable crystal structures which are require heating above 100 C in water to dissolve. At a lower level of hydrolysis (85%), the remaining acetate groups act as steric hindrance to tight crystal structure formation and allow for much easier dissolution of PVA in water at temperature below 100 C. The majority of this study was performed with lower hydrolysis Mowiol 8-88 due to its ability to dissolve at high concentration and thus create a wide range of fiber diameters.
The difference in fiber diameters between the two types of PVA at an electrospinning solution concentration of 10% can be explained by the difference in surface tension between the solutions (47). The highly hydrolyzed PVA is more hydrophilic and thus has an increased surface tension compared to a less hydrophilic polymer solution. This increase in surface tension alters the Rayleigh instability relationships during splaying of the polymer jet in the electric field. This alteration causes earlier gelation of the fiber jet as it travels to the collector, thus increasing the size of fibers collected.
The use of multiple fiber collector geometries allowed manipulation of the fiber array patterns from randomly oriented to fully aligned. This is advantageous because when used in device fabrication, these two array patterns will yield very different drainage properties. As discussed earlier, the Hagan-Poiseuille equation dictates fluid flow through a single pipe. It states that drainage rate is proportional to the length of this tube. Considering with just one nanochannel in a 1mm long drainage device, path length fluid will travel if the nanochannel was formed with unaligned fibers will be much greater than 1mm. However, if an aligned fiber array is used to fabricate the drainage device, the path length of fluid flow through the device will be very close to, if not exactly, 1mm. In this study, unaligned fiber arrays were studied due to their ease of handling compared to unaligned fiber which generally are formed
51


between two electrodes (hanging in air) rather than directly on a collector. Future studies may investigate a manufacturing technique that allows for aligned fibers to be incorporated into the device.
The use of varying concentrations of PVA solution yielded tailorable and predictable nanofiber morphologies and topologies. This is the basis of obtaining a viable trans-corneal drainage device as the size and morphology of the fibers impacts not only the drainage rate, but the physical barrier to microbes on the surface of the eye. As the average size of corneal bound microbes is about 1 micron, any concentration of low hydrolysis PVA would be able to create a physical barrier to microbes if used to fabricate a sacrificial template for a GDD.
Synthesis of LCE was a simple one pot click reaction which required minimal chemical synthesis skill. Following Yakacki et al, the TAMAP reaction produced and elastomer which would set within 15-20 minutes of adding catalyst. This was helpful because the reaction mixture was able to be poured into a mold around PVA nanofibers when it still had a relatively low viscosity. Therefore the reaction mixture was able to fully penetrate the nanofiber web before setting.
The change of diameter of the cylinder shown in Figure 27 gives an example of the shape memory properties of this material. The chemistry used to synthesize these cylinders can be easily tailored to achieve a thinner initial diameter by decreasing the crosslink density during the first stage of the reaction. This, however, is at the expense of expansive strength of the polymer. Outside the scope of this study, but still of significant importance to the final product, would an investigation of the fixity and relaxation strength of the shape memory behavior of a cylindrical LCE.
52


In this investigation, only the first stage of the reaction was used. This was effective to observe whether the fiber mesh was being fully infiltrated by LCE monomers. In a second stage reaction, a cylindrical sample would be stretched lengthwise and cured with a UV light. Upon heating the sample up to body temperature, the device would expand both lengthwise and axially (see Figure 11). The strength at which the device locks in place is crucial for effective placement of the device as IOP swings of around 15 mmHG upon inversion of the head (48).
Visual confirmation of pores via several modes of imaging showed that once samples were treated to 48 hours of 60 oC water, or alternatively 30 minutes of sonication and 24 hours of 60 oC water, fibers were able to be dissolved out of the bulk LCE. In addition to LCE, MMA/DEGDMA was a useful bulk polymer to fabricate test samples. This polymer is glassy at room temperature, where LCE is quite rubbery. The hardness of MMA/DEGDMA made it possible to create rigid cylinders which could be reliably superglued into testing fixtures. Its amorphous nature differs from the LCE polydomain substructure and makes the MMA/DEGDMA optically clear while the LCE is opaque. This allowed the fluorescent images to be taken with MMA/DEGDMA (Figure 25) while the cross section SEM images clearly showed nanopores in the LCE sample (Figure 24). The mechanical and optical differences not only proved convenient, but the use of two different materials shows the robustness of the manufacturing process.
The fluorescent image seen on the right side of Figure 25 is a powerful visual which clearly shows fluid infiltrating the nanochannels. This indicates that fluid would pass through a drainage device manufactured in the same manner. This image differs from the left side of Figure 25 because the microscope used to take these images creates 15 micron deep Z-stacks
53


starting from the bottom of the sample. When imaging a 1mm thick sample, the camera only captures the bottom most fibers in the sample. The 80 micron thick sample shows many more fibers within the 15 micron deep window that was available to image. These images show that fluid may be easily infiltrated into the nanochannels created after dissolution of nanofibers embedded in a shape memory polymer.
To prove without a doubt that fluid may pass through a nanoporous drainage device fabricated with an electrospun sacrificial template, water column perfusion tests were performed with several indicators and a variety of test fixtures.
The first perfusion test involved supergluing a porous MMA/DEGDMA cylindrical sample into the tip of a pipette. This was a delicate process that required glue around the entire edge of the pipette tip while not allowing any glue to touch the flat top or bottom surfaces of the sample as the leave the channels clear. Once the sample was secured and sealed into the tip of the pipette and the pipette was filled with a colored dye, the tip of the filter was dipped into a collection bath to wait for colored dye to perfuse the sample. Encouraging results were obtained from these tests, with an indicator dye appearing to perfuse through the nanochannels into a collection bath. While using a control sample containing no pores, dye was not able to perfuse into the collection vial, proving a robust seal around the cylinder.
The two other perfusion set ups also showed qualitative perfusion, with visual confirmation being the indication of a positive result. In the future, a more elaborate perfusion setup would be needed to quantify the flow rate at a given pressure. A programmable syringe pump would be needed to obtain these results, such as Pump 11 Plus by Harvard Apparatus.
54


CHAPTER V
CONCLUSION
The purpose of this investigation was to create a manufacturing technique for a tailorable trans-corneal glaucoma drainage device using a sacrificial nanofiber template within a shape switchable liquid crystal elastomer. After fabrication of an electrospinning power unit, several concentrations of poly(vinylalcohol) were used to create sacrificial nanofiber templates on flat plate and wire collectors. SEM images of these arrays showed various morphologies and topologies of the fibers, including randomly oriented and highly aligned. The diameters of the fibers produced varied from around 50 nm to 750 nm, which once dissolved out of a bulk polymer yielded nanochannels of the same diameter. It was shown in imaging and perfusion tests that not only could these nanochannels be infiltrated with a liquid, a fluid may pass through them in a controlled fashion.
There is one instance in the literature published by Bellan el al which uses sacrificial electrospun nanofibers as a template for nanochannels in a bulk polymer (38). The current study expands on this in several ways. While Bellan et al focus on creating these channels in a poly(dimethylsiloxane), or PDMS, substrate, the current study uses a functional shape memory elastomer. This implies greater potential for end usage of such a device including, but not limited to, a trans-corneal glaucoma drainage device. Additionally, the study performed by Bellan et al did not deeply explore the relationships between electrospun fiber morphology and the various electrospinning parameters such as solution concentration. These relationships play a key role in the tunable nature of such a device, especially considering the Hagen-Pouseuille equation which dictates flow as a function of channel diameter to the fourth power.
55


There were limitations throughout this study that, if alleviated, would allow the proper qualification of this transcorneal drainage device. Of particular importance is the perfusion testing. These tests were designed to indicate a binary result did the fluid pass through or not. There was little means of measuring exact flow through test fixtures using the low-tech solutions that were obtained for little to no cost. As a result, there was no way to quantify the drainage rate of a trans-corneal filter containing convoluted and tortuous nanochannels paths. As indicated previously, the Hagen-Pouseuille equation gives the relationship between flow, pressure, and channel diameter. Another equation would better suit this scenario if such quantifiable drainage information was available. The Darcy equation relates flow and pressure inside a porous medium. The tortuous nature of the nanochannels create a scenario that mimics a porous medium closer than a group of pipes. The Darcy equation contains a K factor, which is derived from perfusion data. If an advanced perfusion set-up (Pump n Plus, Harvard Apparatus) was available, a quantifiable relationship between PVA solution concentration used in electrospinning and drainage rate of a trans-corneal device could be derived.
Another limitation of the study was time. Aligned fiber arrays may prove necessary in future studies if the drainage rates of filter devices using randomly oriented fiber arrays proved too slow. Because of the different nature of aligned fibers (which collect between to grounded units opposed to directly on a collector), separate manufacturing technique as well as manufacturing fixtures must still be designed.
Lowering IOP in patients with glaucoma remains the cornerstone of limiting risk of vision loss, and while there are many approaches to this, no current strategy is without complications. This study proposed a unique solution to replace the gold standard both in surgical and drainage device treatments of Glaucoma. A trans-corneal device which is
56


replaceable, tunable and easily deployed into patients has the potential to greatly reduce the number of those who lose their vision from the disease. It will especially make an impact in areas of the world where ophthalmic surgeons and surgical arenas are not readily available.
57


REFERENCES
1. F. Medlege, H. Rouault, N. Belgacem, A. Blayo. (Google Patents, 2010).
2. J. S. Distelhorst, G. M. Hughes, Open-angle glaucoma. American family physician 67, 1937 (May 1, 2003).
3. H. A. Quigley, A. T. Broman, The number of people with glaucoma worldwide in 2010 and 2020. British journal of ophthalmology 90, 262 (2006).
4. D. Peters, Visual Impairment and Vision-Related Quality of Life in Glaucoma. (2015).
5. C. N.-T. G. S. Group, Comparison of glaucomatous progression between untreated patients with normal-tension glaucoma and patients with therapeutically reduced intraocular pressures. American journal of ophthalmology 126, 487 (1998).
6. I. Agis, The Advanced Glaucoma Intervention Study (AGIS): 7. The relationship between control of intraocular pressure and visual field deterioration. American journal of ophthalmology 130, 429 (2000).
7. R. N. Weinreb, T. Aung, F. A. Medeiros, The pathophysiology and treatment of glaucoma: a review. Jama 311, 1901 (2014).
8. R. W. Nickells, G. R. Howell, I. Soto, S. W. John, Under pressure: cellular and molecular responses during glaucoma, a common neurodegeneration with axonopathy. Annual review of neuroscience 35, 153 (2012).
9. P. R. Lichter et al., Interim clinical outcomes in the Collaborative Initial Glaucoma Treatment Study comparing initial treatment randomized to medications or surgery. Ophthalmology 108, 1943 (2001).
10. A. E. Laing, L. K. Seibold, J. R. SooHoo, M. Y. Kahook, Evaluation of bleb characteristics after implantation of the EX-PRESS glaucoma filtration device. Molecular vision 18, 10 (2012).
11. E. M. Van Buskirk, D. R. Bacon, W. H. Fahrenbach, Ciliary Vasoconstriction After Topical Adrenergic Drugs. Am. J. Ophthalmol. 109, 511 (1990).
12. R. van der Valk et al., Intraocular Pressure-Lowering Effects of All Commonly Used Glaucoma Drugs: A Meta-analysis of Randomized Clinical Trials. Ophthalmology 112, 1177 (2005).
13. R. K. Parrish, P. Palmberg, W. P. Sheu, X. L. T. S. Group, A comparison of latanoprost, bimatoprost, and travoprost in patients with elevated intraocular pressure: a 12-week, randomized, masked-evaluator multicenter study. American journal of ophthalmology 135, 688 (May, 2003).
14. P. Watson, J. Stjernschantz, A Six-month, Randomized, Double-masked Study Comparing Latanoprost with Timolol in Open-angle Glaucoma and Ocular Hypertension. Ophthalmology 103, 126.
15. J. Stjernschantz, G. Selen, M. Astin, B. Resul, Microvascular effects of selective prostaglandin analogues in the eye with special reference to latanoprost and glaucoma treatment.
Progress in retinal and eye research 19, 459 (Jul, 2000).
16. C. O. Okeke et al., Adherence with topical glaucoma medication monitored electronically the Travatan Dosing Aid study. Ophthalmology 116, 191 (Feb, 2009).
17. J. A. Han, W. H. Frishman, S. Wu Sun, P.-M. Palmiero, R. Petrillo, Cardiovascular and Respiratory Considerations With Pharmacotherapy of Glaucoma and Ocular Hypertension. Cardiology in Review 16, (2008).
18. D. Chiselita, Non-penetrating deep sclerectomy versus trabeculectomy in primary open-angle glaucoma surgery. Eye 15, 197 (2001).
58


19.
J. S. M. Lai, J. K. H. Chua, C. C. Y. Tham, D. S. C. Lam, Five-year follow up of selective laser trabeculoplasty in Chinese eyes. Clinical & Experimental Ophthalmology 32, 368 (2004).
20. M. O. M. Wong, J. W. Y. Lee, B. N. K. Choy, J. C. H. Chan, J. S. M. Lai, Systematic review and meta-analysis on the efficacy of selective laser trabeculoplasty in open-angle glaucoma. Survey of Ophthalmology.
21. P. Y. Ramulu, K. J. Corcoran, S. L. Corcoran, A. L. Robin, Utilization of Various Glaucoma Surgeries and Procedures in Medicare Beneficiaries from 1995 to 2004. Ophthalmology 114, 2265 (2007).
22. S. J. Gedde et al., Postoperative complications in the Tube Versus Trabeculectomy (TVT) study during five years of follow-up. Am J Ophthalmol 153, 804 (May, 2012).
23. S. J. Gedde et al., Treatment outcomes in the Tube Versus Trabeculectomy (TVT) study after five years of follow-up. Am J Ophthalmol 153, 789 (May, 2012).
24. L. M. Brandao, M. C. Grieshaber, Update on Minimally Invasive Glaucoma Surgery (MIGS) and New Implants. Journal of ophthalmology 2013, (2013).
25. J. F. Jordan et al., A Novel Approach to Suprachoroidal Drainage for the Surgical Treatment of Intractable Glaucoma. Journal of glaucoma 15, (2006).
26. C. Pulling. (OIS, 2015), vol. 2015.
27. P. De Gennes, Reflexions sur un type de polymeres nematiques. CR Acad. Sci. Ser. B 281, 2 (1975).
28. H. Finkelmann, H. J. Kock, G. Rehage, Investigations on liquid crystalline polysiloxanes 3. Liquid crystalline elastomersa new type of liquid crystalline material. Die Makromolekulare Chemie, Rapid Communications 2, 317 (1981).
29. C. M. Yakacki et al., Tailorable and programmable liquid-crystalline elastomers using a two-stage thiol-acrylate reaction. RSC Advances 5, 18997 (2015).
30. F. Brommel, D. Kramer, H. Finkelmann, in Liquid Crystal Elastomers: Materials and Applications. (Springer, 2012), pp. 1-48.
31. A. R. Garcia-Marquez, B. Heinrich, N. Beyer, D. Guillon, B. Donnio, Mesomorphism and Shape-Memory Behavior of Main-Chain Liquid-Crystalline Co-Elastomers: Modulation by the Chemical Composition. Macromolecules 47, 5198 (2014/08/12, 2014).
32. T. Ware, T. White, Programmed Liquid Crystal Elastomers with Tunable Actuation Strain. Polymer Chemistry, (2015).
33. C. Ahn, X. Liang, S. Cai, Inhomogeneous stretch induced patterning of molecular orientation in liquid crystal elastomers. Extreme Mechanics Letters, (2015).
34. M. O. Saed, A. H. Torbati, D. P. Nair, C. M. Yakacki, Synthesis of Programmable Main-chain Liquid-crystalline Elastomers Using a Two-stage Thiol-acrylate Reaction. Journal of visualized experiments: JoVE, (2016).
35. J. Kiipfer, H. Finkelmann, Nematic liquid single crystal elastomers. Die Makromolekulare Chemie, Rapid Communications 12, 717 (1991).
36. C. Yakacki et al., Tailorable and programmable liquid-crystalline elastomers using a two-stage thiol-acrylate reaction. RSC Advances 5, 18997 (2015).
37. X. Luo, P. T. Mather, Shape Memory Assisted Self-Healing Coating. ACS Macro Letters 2, 152 (2013/02/19, 2013).
38. L. M. Bellan, E. A. Strychalski, H. G. Craighead, Nanochannels fabricated in polydimethylsiloxane using sacrificial electrospun polyethylene oxide nanofibers. Journal of Vacuum Science & Technology B 26, 1728 (2008).
39. A. Greiner, J. H. Wendorff, Electrospinning: A fascinating method for the preparation of ultrathin fibres. Angew Chem Int Edit 46, 5670 (2007).
59


40. D. H. Reneker, A. L. Yarin, E. Zussman, H. Xu, Electrospinning of nanofibers from polymer solutions and melts. AdvAppi Mech 41, 43 (2007).
41. W. E. Teo, S. Ramakrishna, A review on electrospinning design and nanofibre assemblies. Nanotechnology 17, R89 (2006).
42. S. Chuangchote, A. Sirivat, P. Supaphol, Mechanical and electro-rheological properties of electrospun poly (vinyl alcohol) nanofibre mats filled with carbon black nanoparticles. Nanotechnology 18, 145705 (2007).
43. J. Tao, Effects of Molecular weight and Solution Concentration on Electrospinning of PVA. (2003).
44. B. Ding, Preparation and Characterization of Nanoscaled PVA Fibers via Electrospinning. Fibers and Polymers 3, 73 (2002).
45. K. H. Hong, Preparation and properties of electrospun poly(vinyl alcohol)/silver fiber web as wound dressings. Polymer Engineering & Science 47, 43 (2007).
46. Q. Li, Preperation and Properties of Electrospun PVA Nanofibers. International Conference on Sollid Dielectrics, (2007).
47. J.-C. Park et al., Electrospun poly(vinyl alcohol) nanofibers: effects of degree of hydrolysis and enhanced water stability. Polym J 42, 273 (2010).
48. R. N. Weinreb, J. Cook, T. R. Friberg, Effect of Inverted Body Position on Intraocular Pressure. American journal of ophthalmology 98, 784.
60


Full Text

PAGE 1

TAILORABLE AND DEPLOYABLE TRANS CORNEAL DRAINAGE DEVICE FABRICATED WITH NANOPOROUS LIQUID CRYSTAL ELASTOMER by ROSS VOLPE B.S. SUNY Environmental Science and Forestry 2014 A thesis submitted to the Faculty of the Graduate School of the University o f Colorado in partial fulfillment Of the requirements of the degree of Masters of Science Bioengineering 2016

PAGE 2

ii 2016 ROSS VOLPE ALL RIGHTS RESERVED

PAGE 3

iii This thesis for the degree of Master of Science by Ross Volpe Has been approved for the B ioengineering Program By Kendall S. Hunter, Chair Christopher M. Yakacki, Advisor David Ammar Amir Torbati April 29, 2016

PAGE 4

iv Volpe, Ross (M.S. Bioengineering) Tailorable and Deployable Trans Corneal Drainage Device Using Nanoporous Liquid Crystal Elastomer Thesis directed by Assistant Professor Chris topher M Yakacki ABSTRACT While the pathophysiology of glaucoma is widely unknown, current treatment relies on lowering intraocular pressure (IOP) in order to delay vision loss. The gold stand ard in the surgical treatment of glaucoma is the Ahmed valve. This device offers unpredictable intraocular pressure (IOP) lowering efficacy as well as a milieu of acute and chronic complications. On the forefront of glaucoma device interest is trans cornea l drainage, which offers a predictable IOP lowering performance while eliminating procedural outcomes that lead to many of the current complications seen (e.g. blebs). An investigation into a novel manufacturing process of a trans corneal drainage device c onstructed from a porous liquid crystal elastomer (LCE) is proposed herein. Using a sacrificial template of water soluble nanofibers, several hundred thousand sub micron channels are created in a cylindrical LCE with diameter ~200 microns. Tailorable sacr ificial templates of nanofibers are formed via electrospinning of poly(vinylalcohol). LCE monomers are polymerized around these sacrificial templates, and o nce these fibers are dissolved from a bulk LCE, the remaining channels allow for constant and predic table flow through the material. E xistence of pores is confirmed with SEM and fluorescent microscopy. Finally, the drainage efficacy is tested using a water column perfusion test. Design of the device involves a two piece system: a silicone outer housing which is inserted into the cornea and a separate LCE filter. A collagenous exterior may be introduced to the outer housing to promote integration into the cornea. The LCE filter contains channels with controlled diameters to provide adequate drainage prope rties, while being small enough to block most corneal flora. The filter also may be chemically coated with a layer of copper to provide additional

PAGE 5

v antimicrobial properties. The chemical and physical barriers to microbes yield an antimicrobial LCE, or A LCE The shape switching property of LCEs allows this device to overcome several disadvantages of glaucoma devices which are currently available to patients. The first of which is a limited lifetime of devices caused by scarring over or migration of a device. The proposed device may be cooled down to be easily removed and replaced upon inactivation. A second shortcoming of current devices involves improper IOP control. Several styles of the proposed device could easily be made through manufacturing process whi ch could provide varying levels of IOP relief. This would make the device appropriate for patients with mild to severe glaucoma. Finally, the use of heat activated expansion allows for simple, non surgical insertion. This advantage may prove to be the most crucial in the development of the device. Non surgical insertion and an attractive price point would make the device accessible in developing and third world countries where ophthalmic surgeons are rare or otherwise inaccessible. This unique attribute wou ld open up new markets not currently realized with other glaucoma drainage devices. The form and content of this abstract are approved. I recommend its publication. Approved: Chris Yakacki

PAGE 6

vi ACKNOWLEDGEMENTS My parents Raymond and Andrea Volpe have always told me to not look to others to see what is possible, but to look within myself and achieve my dreams. Without advice like this and a constant backbone of support, I would not be here. My heartfelt thanks for making this possible. I would like to t hank my mentors for their patience in giving me advice and insight into the world of academia which few navigate with such grace and levelheadedness Dr. Chris Yakacki and Dr. Amir Torbati. I also thank Dr. Kendall Hunter for his will to see me succeed th rough unexpected and sometimes adverse situations. Dr. Ammar gave invaluable insight into the world of ophthalmology at short notice and always made time to give help and advice Thank you. Imaging was made possible by the much appreciated work of Dan M erkl (University of Wyoming) and Mel issa Laughter (University of Colorado) I would be at a loss without the acknowledgement and thanks of the University of Colorado Denver Calibration and Machine Laboratory, specifically Rich Wojzick, Jack and Tom and th e many others who helped machine testing apparatuses and kept my smiling while I dug through 30 years of fittings and wiring. Last but not least the SMAB Lab has provided me a place to not only complete meaningful and life changing research but has beco me a haven from the daily grind and a place I will never forget.

PAGE 7

vii DEDICATION T o my mother

PAGE 8

viii TABLE OF CONTENTS CHAPTER I. INTRODUCTION 1 Glaucoma 3 Existing Treatments 3 Drugs 4 Surgery 5 Drainage Devices 6 Liquid Crystal Elastomers 12 Background and Synthesis 12 Liquid Crystal Elastomers as Trans Corneal Fi lter 1 7 Advantages of Liquid Crystal Elastomers 18 Antimicrobial Properties 18 Electrospinning 1 8 Background 18 Tailoring Fiber Diameters 2 0 II. MATERIALS AND METHODS 21 Poly(vinylalcohol) Solution Preparation 21

PAGE 9

ix Electrospinner Fabrication 21 Nanofiber Production 22 Polymer Synthesis .. 23 23 25 Composite Formation 26 Nanochannel Formation 2 7 Imaging 28 Perfusion Te sting 29 III. RESULTS 34 Electrospinning 32 41 Porous LCE 41 Fluorescent Im 44 Perfusion Testing 46 IV. DISCUSSION 50 V. CONCLUSION 56 REFERENCES 59

PAGE 10

x LIST OF FIGURES Figure 1. Aqueous Humor Drainage Path 1 2. Ocular Surface Irritation 4 3. Trabeculectomy Procedure 6 4. Acute Blebitis 7 5. Traditional Drainage Devices 7 6. MIGS Device 9 7. MicroOptx Trans Corneal D evice 10 8. Liquid Crystal Elastomer Configurations 12 9. TAMAP Reaction Visualization 14 10. Liquid Crystal Elastomer Actuation 1 6 11. Trans Corneal Filter 17 12. Electr ospinning Unit Picture and Circuitry 21 13. Electrospinning Collectors 22 14. Liquid Crystal Elastomer Monomers 23 15. Cylindrical Composite Process Flow 26 16. Rectangular Composite Fabrication 27 17. Pipette Perfusion A p paratus 30 18. 30 19. 32 20. Photo of Perfusion 33 21. SEM Images of Randomly O riented Fibers 34

PAGE 11

xi 22. Graph of Nanofiber Sizes vs. Solution Concentration 36 23. SEM and Scanning Laser Topography Images of Aligned Fibers 39 24. SEM Images of Fibers on a Wire 40 25. Cylindrical Composite 41 26. Cylindrical LCE Ac 42 27. SEM Image of Nanochannels 43 28. Fluorescent Confocal Microscopy Image of Nanochannels 44 29. 46 30. P ipette Perfusion 47 31. Large Cylinder/Water Column Pe 48 32. Disc Fixture/Water Column Perf 49 33. Poly(v inylalcohol) Synthesis 51

PAGE 12

xii LIST OF TABLES Table 1. Mow iol 8 88 Fiber Measurements 37 2. Mowiol 10 98 Fiber Measurements 38 3. Fluorescent Fiber Measurements 45 4. 49

PAGE 13

xiii LIST OF ABBREVIATIONS AH Aqueous humor IOP Intraocular pressure GDD Glaucoma drainage device NTG Normal tension glaucoma LT Laser Trabeculoplasty MIGS Micro invasive glaucoma surgery LCE Liquid crystal elastomer Ti Initiation temperature TAMAP Two stage thiol acrylate Mic heal addition reaction SEM Scanning electron microscopy MMA Methyl methacrylate DEGDMA Di(ethyleneglycol) dimethacrylate PEGDMA Poly(ethyleneglycol) dimethacrylate DMF Dimethylformamide fBA tert(butyl)acrylate PVA Poly(vinylalcohol) AC Alte rnating current DC Direct current kV K ilovolt

PAGE 14

1 CHAPTER I INTRODUCTION There is a dire need to develop an effective and consistent treatme nt to glaucoma, a disease that is predicted to effect 80 million people by 2020 ( 3 ) Glaucoma is the second leading cause of blindness worldwide, with one in six cases leading to bilateral b lindness ( 4 ) The biological pathophysiology of glaucoma is widely unknown due to the difficulty of Figure 1 A) Diagram showing aqueous humor (green arrow) in a healthy eye flow from the ciliary body around the iris, through the trabecular ep iscleral vein. B) Open angle glaucoma eye has limited egress through trabecular meshwork ( 2 )

PAGE 15

2 studying ocular tissues in vitro T he only proven factor in decreasing the risk of blindness is by maintaining a healthy intraocular pressure (IOP) of 15 mmHG or lower ( 5 7 ) If pressure rises above this, ganglion cells rooted in the ocular nerve are damaged and lose function. Aqueous humor (AH) is the fluid that fills the eye and is constantly secreted from the ciliary processes posterior t o the iris ( 8 ) In a healthy eye, the trabecular meshwork provides an egress for the AH in order to maintain a constant IOP at the interface of the iris and sclera (Figure 1 A ) AH is subsequently shuttled into collection ducts in the anterior wall of Schlemm's canal and finally into aqueous veins. The pressure of the eye is determined by a balance between the production of aqueous humor and its outflow through the trabecular meshwork. This study will focus on open angle glaucoma, the most prevalent form of the disease. I n open angle glaucoma, cells in the trabecular meshwork morphologically or otherwise change and AH cannot effici ently escape the eye ( 2 ) The most effective means of maintaining low IOP in patients has been shown to be glaucoma drainage devices (GDD s ) ( 9 ) The design of these devices generally utse ab interno placement in order to facilitate drainage from inside of the eye to into another part of the e ye ( 10 ) This type of drainage device presents risk of various complications and efficacy issues that retract from patie usually presents very late in life ( 5 ) A transcorneal GDD would allow for the efficacy of a device based solution, while avoiding the many limitations of an ab interno device. The purpose of this study is to confirm the viability of manufacturing a transcorneal GDD from a nanoporous shape memory polymer created using a sacrificial nanofiber template.

PAGE 16

3 Glaucoma Glaucoma is characterized by damage to ganglion cells rooted in the ocular nerve. Ther e are several types of glaucoma with the most common being open angle, also known as primary glaucoma. This is the main type of glaucoma that may be treated with a glaucoma drainage device. In open angle glaucoma, th e trabecular meshwork is still exposed despite ocular hypertension It is known that the trabecular meshwork is unable to efficiently drain fluid, but a drug which corrects the cellular level function is still unknown. Another type of glaucoma is close d a ngle, which has a quick onset and is an acute disease resulting in fast deterioration of the optic nerve. The IOP spikes so dramatic because the iris presses against the trabecular meshwork, effectively closing the angle between them and eliminated the mai n egress of AH. This type of glaucoma requires quick reversal of angle closure in order to salvage vision, and a long term treatment such as a GDD is not usually considered. A third, and perhaps most intriguing, type of glaucoma is called normal tension g laucoma (NTG). Patients with NTG experience loss of ganglion cell function despite a normal IOP. A treatment to NTG would give the community great insight into the disease ( 8 ) Existing Treatments of Glaucoma There is no known cure for the disease, and the cellular pathophysiology is largely unknown. The cells in the trabecular meshwork are believed to reorder their cytoskeleton, but studying this change present s significant difficulty as the cells do not maintain their character once taken from the living eye. There are, however, proven ways of managing glaucoma and

PAGE 17

4 preventing further ocular nerve damage. By maintaining a normal IOP, which is around 10 20 mmHG, t he nerves experience markedly decreased degeneration ( 4 ) There are three strategies of managing glaucoma: drugs, surgery, and drainage devices. Drugs Topical drugs are usually the first line of offense used by doctors to combat high IOP. There are many drugs o n the market, most notably are B imatoprost, Travoprost, Latanopros t, and T blockers, which cause vasodilation of the ciliary arteries leading to less blood flow to the eyes ( 11 ) Although this strategy does not target specific pathophysiology of glaucoma, it is nevertheless effective in reducing IOP generally by around 30% ( 12 ) Prost a glandins have been recently been u sed more commonly because they only receptor antagonists which m ay be applied three or more times a day ( 13 14 ) This type of drug increases the outflow of AH from the eye and have several side effects including eyelash growth and iris color change. Figure 2 A patients eye exhibiting ocular surface irritation

PAGE 18

5 Investigations on the efficacy of prostaglandins are limited by specie s dependent reaction to these drugs ( 15 ) There are several challenges in using topical drugs to lower IOP. First is the frequency of application, which may be one to two times a day ( 14 ) Patient adherence to daily topical drugs has been shown to be much less than they report to the clinician, as low as 59%, resulting in failure to lower IOP ( 16 ) There is an associated ocular surface irritation in some patients that makes this type of treatm ent unbearable (Figure 2 ) The chemical used in glaucoma drugs have a body wide effect, sometimes causing respiratory and/or cardiovascular complications ( 17 ) Finally, the IOP lowering efficacy of drugs for patients with higher (ove r 21mmHg) IOP is not consistent ( 15 ) Surgery Each patient responds differently to medication, and if treatment with drugs alone is not sufficiently redu cing IOP doctors may suggest a surgical solution. Surgery is a more aggressive tactic in combatting high IOP, with effective results and many associated risks ( 18 ) Laser trabeculoplasty (LT) induces a thermal reconstruction of the trabecular meshwork, leading to highe r drainage rate. Although LT is a common procedure due to its low risk of infection, it is ineffective in late stage glaucoma and subsequent treatments are generally minimally effective at further lowering IOP ( 19 20 ) Furthermore, this technique requires a high degree of skill for the surgery.

PAGE 19

6 Trabeculectomy is the gold standard invasive g laucoma procedure that aims to byp ass the trabecular meshwork by dissection of the sclera, creating a scleral flap and opening a hole to the anterior chamber (Figure 3 ) Approximately 24,000 Medicare patients receive this treatment each year ( 21 ) .This technique results in a bleb, or blister like pouch on the surface of the sclera which AH is allowed to drain into. Ble bs have a tendency for long term complications including acute blebitis (shown in Figu re 4 ) and leaking ( 10 ) The high morbidity of a trabeculecomy limits its use. T he acute post operative complications include choroidal effusion (13%), wound leaks (11%), shallow anterior chamber (10%), and anterior chamber bleeding ( 8%) ( 22 ) There are also many long term complications which include corneal edema (9%), dysesthesia (ocular discomfort) in 8%, fluid leaks (6%) and endophthalmitis ( serious eye infection) in 5% of patients ( 22 ) These complications, along with scarring over of the anterior chamber shunt, lead to a 46.9% failure rate over five ye ars ( 23 ) Figure 3 Diagram of trabeculectomy procedure. An alternative egress for AH is introduced underneath the sclera. Green arrow s hows new path of AH.

PAGE 20

7 Drainage Devices An alternative to traditional surgery is a glaucoma drainage device. These devices come in many varieties that act in several ways. The most common device is called an Ahmed Valve or tube shunt severa l of which are shown in Figure 5 This device acts in a Figure 4 Acute blebitis in a patient who has undergone a trabeculectomy Figure 5 ( Left ) A variety of shapes of tube shunts that may be placed in an eye to lower IOP. (Left) An animation of an Ahmed Valve after being sutured into place. In practice, the pouch would be covered by the sclera.

PAGE 21

8 similar way to a trabeculectomy. A pouch is surgically attached under the sclera and a tube in inserted into the anterior chamber t o provide an additional egress for AH. There are many variations of this type of valve shown in Figure 5 (Left). The top row shows Molteno implants. The middle row shows a Krupin slit valve (left) and an Ahmed valve (right) implant. The bottom row shows a variety of Baerveldt drainage devices. Complications of this type of device include improper drainage (hypotony or failure to lower IOP), bleb infection and scleral tissue erosion. Early stage complications are experienced by approximately 21% of patients, and late stage complications occur in about 34% of patients ( 23 ) Furthermore, t he relatively large size of these devices limits the number of additional devices t hat can be implanted upon failure. Doctors can generally only implant two devices into an eye before they deplete the accessible scleral tissue. If additional devices do not correct the problem, patients will undergo a cyclodestruction therapy, which gener ally consists of laser treatment around the ciliary body. This is aimed to decrease AH production but often results in significant collateral tissue damage resulting in a decline or loss of vision.

PAGE 22

9 Micro Invasive Glaucoma Surgery (MIGS) devices are an alt ernative to traditional GDDs in mild cases of glaucoma which aim to minimize surgical complications. These devices are inserted ab interno meaning they are placed in the interi or the eye and drain to another chamber inside of the eye. As seen in Figure 6 t he iStent (Glauk os, USA) is a titanium tube shaped device that is inserted into the trabecular meshwork to shunt AH directly into the Canal of Schlemm. Complications for iStent include fibrotic blockage, temporary IOP spike, corneal edema, stent obstruct ion by blood clot or iris, anterior chamber collapse, and vitreous incarceration ( 24 ) A novel approach was achieved by Transcend Medical with their Cypass Microstent. This device sh unts AH into the suprachoroidal space, a physiological drainage site which is secondary to the trabecular meshwork. It has been shown that the suprachoroidal space has a great capacity for drainage once an artificial fistula is introduced ( 25 ) Although MIGS devices offer fewer and less severe post operative complications, there effic acy is only shown for individuals with mild ocular hypertension ( 24 ) Additionally, due to their small size there is a high rate of scarring over of the devices inside the eye, where Figure 6 Size of the iStent Drainage Device (Glaukos)

PAGE 23

10 repair is fast and largely unpredictable. This leads to uncontrol lable drainage rates resulting in either hypotony or no IOP lowering effect. A transcorneal drainage method of reducing IOP offers several advantages compared to GDDs currently on the market. This strategy aims to shunt aqueous humor across the cornea d ir ectly onto the tear film, where it is naturall y evacuated through the puncta. Microoptx (USA) is currently developing a transcorneal drainage device called the BG Implant (Figure 7, Top) and released some clinical data in the summer of 2015. This data Figure 7 (Top) The MicroOptx BG Implant next to an Ahmed Valve and a dime, for size reference. (Bottom) A showing i n vitro data of the BG Implant mainta ining a constant IOP at 12 mmHg over a 14 day period.

PAGE 24

11 indi cates that the BG Implant achieved a stable IOP at around 1 2 mmHg while maintaining a microbe free environment through the device (Figure 7 Bottom ) ( 26 ) These results show a promising future in the use of transcorneal devices as they require a low level of surgi cal expertise, making them widely available in countries like India where technical expertise is not readily available, through monetary or geographic reasons. Although Microoptx offers a viable device for the treatment of glaucoma, they are missing severa l key features. The device is coated in a collagen layer which helps the device adhere to the cornea l tissue. This makes the device permanently attached in the eye and will not allow replacement if the device is clogged by proteins or debris in the AH. Als o, the size fits approach similar to many devices, which ultimately is the reason for many post operative complications. Sizing of MIGS devises are generally on the mill imeter scale. The exterior form factor usually differs from the actual drainage path, including anch ors or features which allow the device to secure itself into the eye. Drainage paths vary enormously, depending on the downstream pressure barrier. For exam ple, a device draining into the suprchoroidal space will provide much less internal resistance than a device draining to the tear film. This is because the suprachoroidal space contains intrinsic resistance to flow or AH accumulation whereas the tear film can accept as much AH as physiological flow rate. For comparison, 2 4 microliters per minute of physiological flow equates to about one drop every half hour.

PAGE 25

12 Liquid Crystal Elastomers Background and Synthesis Liquid crystal elastomer s (LCE s ) are a class of smart material which are defined by thier combination of rubber elasticity of a lightly crosslinked polymer with self organizing liquid crystals. This material system was first proposed by de Gennes et al in 1975 ( 27 ) a nd then realized by Finkleman et al in 1981 ( 28 ) Such a unique combination of subsystems gives LCEs the ability to reversibly and repeatedly change its shape and optical properties due to exposur e to a sti mulus such as heat ( 29 ) The stimulus induced phas e transition is made possible by the self organization of the LCE monomers called mesogens. Mesogens are rigid molecules usually comprised of two or more aromatic rings ( 30 ) These mesogens can be linked together with flexible spacer molecules to form a main chain LCE, or may be attached to a polymer as a side group to form an end on or side on LCE pictured in F igure 8 Main chain LCEs have been of particular interest in the past few years because Figure 8 a) Grey mesogens in an end on configuration. b) Mesogens in a side on configuration. c) A main chain LCE with mesogens integrated into the polymer chai n.

PAGE 26

13 of their high strain actuation compared to side on or end on LCEs as well as newly discovered synthesis pathway which allows for highly reproducible and tailorable samples to be made ( 31 ) The unique properties of mesogens manifest in a phase transition at the i sotropic temperature, or T i Below T i mesogens will naturally align organize into monodomains which can form in two general configurations. When mesogens are aligned in only one axi s, it is termed a nematic monodomain When alignment occurs in two directions it is a smectic monodomain. The natural alignment of mesogens into a monodomain relies on their ability to move freely. However, when a crosslinked network is introduced to form an LCE, covalent bonds disrupt the free movement of the mesogens. This results in a phase of only localized order, called the polydom ain In all three cases, when the material is taken a bove T i it will undergo a phase transition as the mesogens completely lose their order and become randomly oriented. This is now called the isotropic phase. An optical change can be observed while heating a sample as the opaque crystal structure turns unordered and transparent Although naturally forming in a polydomain, LCEs can be programmed into a monodomain. This is known as a li quid single crystal elastomer. In this state, the mesogens are aligned uniformly throughout the entire sample. Monodomain can be achieved through several means, depending on the chemistry used to synthesize the LCE. Traditional ly programming of the monodomain was achieved at the same time as synthesis by three basic tactics: wiping a glass slide with polyimide in the desired direction of mesogens orientation, exposure to a strong (>1Tesla) magne tic field, and mechanical stretching during the reaction ( 30 ) These methods of programming monodomain during synthesis can be unreliable, not easily reproducible and present strong limitations of sa mple geometry (only thin films are

PAGE 27

14 able to be made with polyimide and magnetic orientation). Recently, Yakacki et al has proposed a two stage thiol acrylate Michal Addition photopolymerization (TAMAP) reaction to synthesize a reproducible, facile and tailo rable nematic main chain LCE ( 29 ) Several recent studies show the effectiveness of the TAMAP methodology ( 32 34 ) The TAMAP reaction represents a dramatic change in the way which researchers are able to synthesize and exp lore the mechanics of LCEs. Up to 2012, the only relatively simple method to create an LCE from a functionalized monomer was a free radical polymerization ( 30 ) This reaction was often hard to carry through in a homogeneous way, leading to uncontrollable liquid crystal domains. Another reaction, proposed by Kupfer et al relies on the functionalization of a Si H bond in the presence of a Pt catalyst ( 35 ) This reaction, deemed hydrosylilation, relies on the functionalization of a cross linking molecule with both a vinyl and an acrylate g roup, which exhibit two different reaction rates. The vinyl groups react qu ickly in the presence of a Pt catalyst while the acrylate groups react slower. This allows the researcher to stretch the (a) (b) (c) Figure 9 Schematic of the second stage of a TAMAP reaction of an LCE. Sample starts after the first stage as a stable pol ydomain LCE (a). Sample is then stretched uniaxially to align mesogens within the polymer chains (b). Finally the polymer is photopolymerized to connect excess acrylates present in the sample (c). The polymer is now programed into a mon o domain phase.

PAGE 28

15 polymer into a monodomain once they feel the vinyl groups have reacted, but the acrylate groups have yet to react. This reaction allowe d researchers to create a monodomain bulk polymer, albeit with quite unpredictable crosslinking densitities and degrees of alignment within the monodomain ( 36 ) To be successful, t he se reaction s must be carried out under very strict temperature conditions Furthermore, only side chain LCEs were able to be synthesi zed from the s e reaction. This limited not only the application of the reaction but the interest of researchers A Michael Addition reaction offers a new perspective on a traditionally grueling and unpredictable synthesis.

PAGE 29

16 Using th e two stage Michae l Addition reaction, LCE can first be synthesized in the polydomain and subsequently programmed into monodomain during the second stage via photopolymerization induced crosslinking during mechanical straining (Figure 9 ) The mechanical properties and glass transition temperature of the LCE have been shown to be tailorable through varying the crosslinking density during the first stage of the reaction. Figure 10 shows a strip of monodomain LCE at room temperature get heated up past its Ti and lift a small weight. It then cools back to room temperature and returns to its original shape. Fi gure 10 A monodomain LCE has been programmed prior to this animation. The sample starts at room temperature with a small binder clip attached to the end. As the sample is heated up the polymer transitions to polydomain. Once heat is removed, the LCE samp le cools back down and expands to its original position.

PAGE 30

17 L iquid C rystal E lastomers as a Transcorneal Filter Using this system, a cylindrical filter synthesized from LCE in the polydomain phase may be stretched along its axis into monodomain phase, locked in place via UV curing and may switch between a longer thin cylinder at cool temperatures to a shorter thick cylinder at body temperatures. The filter will be designed in a way which expands and locks into an outer housing in the cornea once it reaches body temperature. Pictured in Figure 1 1 is the schematic of a device as the filter is placed into the outer housing, expands and locks in place and is subsequently removed by coo ling of the device. This gives the transcorneal filter a unique advantage over any existing glaucoma drainage devices. Insertion and removal of the filter will be performed with an application device, not pictured. Figure 11 A schematic describing the two part device concept and the replacing of a filter. a) the silicone/collagen outer housing is placed in a pilot hole on the edge of the cornea. b) A chilled drainage filter is placed inside the outer housing. This schematic shows an Antimicrobial LCE (A LCE) which has been chemically coated with a layer of copper. c) As the drainage filter reaches body temperature, the filter will expand causing it t o lock in place and maintain a microbe free barrier between the filter and the outer housing. d) Once the device needs replacing, the device is chilled and can be easily removed from the outer housing.

PAGE 31

18 Advantages of Using LCE One key featu re of the LCE is the heat activated shape switching ability, which allows the device to be easily placed into, and taken out of, the outer housing. This gives surgeons the option to easily switch in different devices which have different drainage character istics This feature also affords patients reliable lifelong glaucoma relief by allowing replacement if the device is clogged by accumulation of protein in the aqueous humor. Antimicrobial Properties Introducing a transcorneal egress for aqueous humor has the potential to put patients at risk of infection. To combat this risk, the pores in the transcorneal filter are designed to be less than one micron, the average size of human corneal bacteria ( 6 ) This size based barrier, combined with the constant outflow of aqueous humor through the device ensure that there will be no b acterial infection happening through the body of the drainage device. It is important to note that not all bacteria will be larger than one micron. The constant flow of AH will prevent these small bacteria from infiltrating the device. Electrospinning Back ground In order to produce consistent, sub micron channels in the LCE, a two polymer system is used that relies on a sacrificial template of nanofibers. A difference of solubility was taken advantage of in this study by using a water soluble polymer to for m the sacrificial nanofibers. This allows the fibers to remain stable as the LCE monomers are infiltrating the fibers. Toluene was used as a solvent for the LCE and will not disturb the nanofiber system.

PAGE 32

19 A similar study was conducted by Luo et al during w ork aimed at developing a self healing polymer by embedding in it electrospun nanofibers to create a two polymer composite ( 37 ) This system also relied on the solubility differences between the polymers, but differed from the system in this study by leaving the water soluble nanofibers intact in the polymer. Another study was conducted with yet a more similar system of nanofiber composites, conducted by Bellan et al ( 38 ) In this study a composite of water soluble nanofibers inside a bulk hydrophobic polymer (polydimethylsiloxane) was prepared and the electro spun fibers were subsequently dissolved out. This study aims to test if this technique is compatible with a shape memory polymer. In both these studies, electrospinning was used to create the nanofibers. This has been shown to be a simple and fairly reliab le way to produce continuous fibers on the nano to micro scale ( 39 42 ) The process of electrospinning is based on a high voltage DC field dragging a conductive p olymer solution across a charge gradient onto a grounded collector ( 39 ) The polymer is typically ejected from a syringe at a low flow rate while the positive cha rge is placed directly to the syringe needle. The collector may have a variety of configurations, including a flat plate, rotating drum, wire, or pin electrode (pictured in Figure 345y). The type of collector plays a major part in the morphology of fibers collected as well as several other key parameters: voltage charge applied, tip to collector distance, polymer concentration, molecular weight of polymer, humidity of the room, and additions to the polymer solution (salts, pH control, etc.) ( 41 ) It is noted in literature that the parameters most influential on the fiber morphology are

PAGE 33

20 polymer solution concentr ation and tip to collector distance ( 40 42 ) Nanofiber diameters are increased by increasing solution concentrations and decreasing tip to collector distance. Tai loring Fiber Diameters Tailoring the nanofiber diameters is an important step in developing a robust solution to transcorneal drainage because the size of the nanofibers produced will correlate to the diameter of pores left in the filter after dissolution of the fibers. The diameter of pores in the only one pore which runs through the filter, i t s capacity to drain fluid at a given pr essure is given by the Hagen Poi seuille equation: fluid viscosity and L is the length of the tube in question. It is clear that a small change in diameter will greatly af fect flow rate through the tube. Therefore the diameter of the electrospun fibers is a key parameter is the overall filtration rate of a transcorneal GDD and will be investigated during this study. In summary, a transcorneal glaucoma drainage device will be fabricated so as to maintain an IOP of ~15mmHG unde r physiologic AH production of 2 4 will be made using a sacrificial template of electrospun water soluble nanofibers embedded in a bulk LCE. The system will be qualified with SEM and fluorescent microscopy images as well as perfusion tests.

PAGE 34

21 CHAPTER II MATERIALS AND METHO DS PVA Solution Preparation Granualted polyvinyl alcohol (Mowiol 8 88 Sigma Aldrich) was dissolved in water at several concentrations (8%, 10%, 12%, 17%, 20% and 22%), by weight, by vigorously mixing at 85 oC o vernight. An additional granulated poly(vinyl alcohol) ( Mowiol 10 98 Sigma Aldrich, USA) was also prepared for electrospinning by dissolution in pure water. Concentrations of 8%, 10% and 13% were used in this study. Solutions were prepared by stirring at 95 o C over 48 hours. Electro spinner Fabricati on In order to avoid extreme and unnecessary prices of commercially available electrospinning machines, one was designed and fabricated in house. Two items are essential when making such a device an AC/ DC converter (ESCC 0305, Astrodyne, Mansfield MA) Figure 12 Image on the left shows the DC elec tric current output box. The diagram on the right shows the components and circuit configuration inside the box. a) AC current input. b) AC/DC power converter c) High voltage converter. d) LCD display e) Potentiometer f) Switch g) LCD display power input. h) High voltage ground, labeled

PAGE 35

22 a nd a high voltage o utput (25A12 Series, Ultravolt, Ronkonkoma NY). An acrylic box was constructed to hold these two components, as well as an additional AC/DC converter to provide power to a rotating drum collector, an LCD readout screen displaying the cur rent voltage (PM128, Jameco, Belmont CA), and adjustment knobs provided courtesy of the University of Colorado machine and calibration lab. In order to avoid the high voltage from deactivating other components, the high voltage output had to be removed fro m the main cabin and placed into a small acrylic box of its own and attached to the top of the main acrylic box, where it was accessible to wiring. Wiring schematic and photo of the electrospinning device is shown in F igure 12 Nanofiber Production Nanofi bers are produced by electrospinning a polymer jet of polyvinyl alcohol onto a wire collector The p olymer solution is placed in a 20ml syringe and is ejected from the tip of a 22 gauge blunted ne e dle at a rate of 1 ml/h by a standard syringe pump. The po sitive side of the high voltage DC power converter 8 is connected to the needle with a charge of 10 kV. The collector is a grounded 25 um copper wire with a length of around 7 inches. It is placed Figure 13 (Left) The electrospinning setup (excluding the power supply and syringe pump). A polymer jet is dragged across a DC cha rge gradient from a needle tip onto a grounded wire in the form of continuous nanofibers. (Right) A pin gap electrode is pictured with blue PVA nanofibers collecting across the gap. The two pins are separated by ~1 in. and are both grounded.

PAGE 36

23 perpendicular to the needle at a distance of 8 inches. The e lectospinning process is allowed to proceed for approximately 30 minutes, with attention to not allow web formation c onnecting the grounded wire to surrounding environment. All electrospinning is conducted at room temperature and a relative humidity of 14 % The experimenta l setup is pictured in Figure 13 Experiments were also done using a house ma de rotating drum collector and pin electrode collector in order to produce fibers arrays of aligned morphology. The production of different shapes and sizes of fibers is able to be achieved by altering many parameters of the electrospinning process such as voltage charge, tip to collector distance, needle size, solution flow rate, collector geometry, and solution concentration. In a preliminary study, it was obs erved that solution concentration had the most drastic effect on fiber morphology and therefore was the one parameter that was tested throughout this study. Polymer Synthesis LCE S ynthesis LCE was synthesized according to Yakacki et al using a thiol acryl ate Michael Addition reaction. The mesogenic monomer used was RM 25 7 (Wilshire Technologies, Princeton NJ), and the flexible spacer was 2,2 (Ethylenedioxy) diethanethiol ( EDDT ) (Sigma a b c Figure 14 Chemical structures of the monomers involved in the TAMAP r eaction. a) RM 257 b) EDDT and c) pentatetrakis (3 mercaptopropionate)

PAGE 37

24 Aldrich, USA) The LCE was syntheized with 15% tetra thiol crosslinkers pentaerythritol tetrakis (3 mercaptopropionate) Figure 1 4 shows the monomers. Samples were synthesized with a stoichiometric ratio of functional acrylates to thiol groups, with 15% of the functional thiol groups belonging to the crosslinking molecule. L CE was typically made in small batches according to the following: to a 20 ml scintillation vial, 1 gram of RM257 powder was added. This was dissolved in 0.30 g of toluene at 80 o C. To the solution, 0.2942 g EDDT spacer and 0.021 g of pentatetrakis (3 merc aptopropionate) crosslinker was added. After thorough mixing, 0.245 g of a 50:1 toluene/ DPA catalyst was added and thoroughly mixed, which initiated the reaction. Within 5 minutes of adding catalyst, LCE was poured into a mold and left to fully polymerize overnight. The mold which LCE was polymerized in was generally a pipette, which was a cheap and readily available method of creating cylindrical samples. For preparing porous LCE, the tip of the pipette was used. First, a wire coated with electrospun nanof ibers was inserted approximately one inch into the tip. The tip was then dipped in a mixture of LCE monomers soon after catalyst was added and the mixture was still quite fluid. after a vacuum treatment to remove air bubbles, the pipette was placed vertica lly to allow polymerization overnight. In preparation of samples which showed the shape changing characteristics of the LCE, the wide end of the pipette was used. To synthesize these samples, the tip of the pipette was first plugged with wax. Soon after a dding catalyst to a mixture of LCE monomers, the reaction was poured into the back end of the pipette until filled. Again, a vacuum treatment was applied and the LCE was left to polymerize overnight. After polymerization was

PAGE 38

25 complete, the pipette was smash ed and LCE was left overnight to evaporate all remaining solvent. Test Samples In order to quickly test composite systems, a shape memory polyme r (SMP) which could be rapidly fabricated was used in place of LCE in some cases A key attribute of the SMP us ed in place of LCE was its low glass transition temperature, which made the polymer rigid at room temperature. The SMPs were also amorphous which made them optically transparent, allowing imaging through samples possible. Two different SMPs were used in t his study. The first was a 5:1 tert(butylarylate) t BA and di(ethyleneglycol) dimethacrylate ( DEGDMA ) mixture. The second was a 2:3 methyl methacrylate ( MMA ) and poly(ethyleneglycol) dimethacrylate ( PEGDMA ) mixture. Both of these were free radical driven po lymerizations that were cured in 20 minutes under UV light. The MMA/PEGDMA mixture exhibited less shrinkage when polymerized which made possible synthesis of very thin, homogeneous samples.

PAGE 39

26 C omposite Formation Composites were formed in a variety o f ways, depending on their end use. These are diagramed in F igure 1 5 The first way was a mat composite, where the PVA nanofibers were sandwiched between two layers of LCE. To accomplish this, immediately after adding catalyst to the LCE monomers, the reac tion was brushed on a glass slide coated with a hydrophobic rain repellent (used as a release agent). An electrospun fiber mat was laid on top of the first layer of LCE and then more LCE was spread on top of the nanofiber mat. Another glass slide coated in release agent was clamped on top of the composite overnight while the LCE completed polymerization. To make prototypes suitable to test perfusion using basic laboratory equipment, composites are formed in pipettes. After a wire collector has been removed from the Figure 15 Picture schematic of a circular composite being formed. A 25 micron copper wire is a) electrospun onto and then b) inserted into a tube, either capillary tube, pipette or vial depending on the end use. The cross section of the tube with wire inside is seen on the bottom left. c) The nanofibers are then infiltrated by LCE polymer and vacuum treated to evacuate any air. In a later step, d) the the fibers are dissolved out le aving a nanoporous sample.

PAGE 40

27 elctrospinning apparatus by cutting both ends, it is strung through the pipette with the midpoint of the wire in the widest portion of the pipette. The wire is secured into place at the tip of the pipette with superglue, while also sealing off the tip of the pipette. A custom made, 3D printed cap is placed at the other end of the pipette which allows for centering of the wire, as well as supplying an ingress for LCE to be poured. The LCE monomers are prepared and quickly after adding catalyst, the reaction mixture is injected through the pipette cap opening with a syringe. The pipette is placed in a 15 mmHG vacuum for five minutes to degas the composite and is left to cure overnight. Composite samples using tBA and DEGDMA (5:1) were also synthesize d in the front and back end of glass pipettes. These samples had a larger size (1mm and 5mm diameter) for the front and back end respectively, which made them easier to handle during perfusion tests. These samples were also formed in rectangular geometries by sandwiching tBA/DEGDMA and a flat sheet of nanofibers bet ween two glass slides (Figure 16 ). The rigidity of the tBA/DEGDMA polymer at room temperature also aided in testing. Figure 16 Picture schematic of creation of mat composites. Nanofibers are engulfed on top and bottom with LCE polymer. Fibers are dissolved out in a later step.

PAGE 41

28 Nanochannel Formation Once a composite has been removed from its mold (pipe tte, glass slides), it is dried overnight in an 80 o C oven. Th e composite is then sliced into smaller pieces and placed in a vial of water The composite is gently stirred in this solution for two days at 60 o C. Imaging Nanopores are confirmed by first fr eezing samples and then slicing cross sect ions of the cylinders near the center Carbon sputtered SEM imaging was used to obtain high resolution ima ges of the individual nanopores Imaging was kindly conducted by Dan Merkl from the University of Wyoming Me chanical Engineering department. Fluorescent imaging was also performed in order to see continuous channels. Samples used in fluorescent imaging were prepared with a MMA and PEGDMA (3:2) mixture instead of LCE for several reasons. First of all, the MMA/PE GDMA is an amourphous glassy polymer at room temperature, which in lay terms is clear and hard. This makes handling and imaging possible at room temperature. To prepare samples, the 3:2 mixture of MMA/PEGDMA was placed in a 1mm thick mold until it filled the mold half way. A section of an electrospun sheet, created on a flat plate collector, was placed on the MMA/PEGDMA and the mold was filled up all the way. The MMA/PEGDMA was subsequently crosslinked via UV irradiation. Once the composite was removed fro m the mold and the nanofibers were dissolved out in water for 2 days in 60 o C water, the samples were dried in a 70 o C vacuum oven. After the channels were evacuated, the samples we USA). After soaking for 30 minutes, the submerged samples were introduced into a 15 inHg

PAGE 42

29 (381 mmHg) vacuum 4 times in order to clea r the air from the nanochannels and replace that space with fluorescein solution. After soaking for two days, the sample surfaces were dried with a wipe prior to imaging. Confocal images were collected using a Nikon Eclipse Ti C2 LUN A microscope (Nikon, Tokyo) equipped with two C2 DU3 high sensitivity PMT dtector s, 4 diode lasers (405/488/561/640 nm), and a motorized microscope stage with 3 axis navigation (X, Y and Z). A 488 nm laser setting was used to capture images. All images were analyzed with ImageJ software. Perfusion T esting A liquid perfusion setup was assembled in house to confirm that the pores were continuous and would allow for steady and predicable outflow. This was done using a column based water pressure system in conjunction with several test sample fixtures. In the first sample fixture setup (Fi gure 16 ) a cylindrical sample formed in the small end of a pipette was secured via super glue back into a pipette after dissolution of the nanofibers. The p ipette was filled with a dye and the tip of the sample was dipped into a collection vial. The secon d perfusion setup achieved a higher back end pressure on the device. A test sample was secured this time in the large end of a pipette with superglue while other side of the pipette was superglued into a 64 inch long tube filled with water and dye. Perfusi on was confirmed w ith visual confirmation of dye on the exposed end of the test sample.

PAGE 43

30 Figure 17 A perfusion test setup including a pipette fill ed with red dye, a nanoporous filter secured with super glue in the tip of the pipette which is partially submerged in a collection bath where dye is collecting, indicating successful perfusion. Figure 18 Water column perfusion assembly

PAGE 44

31 The third perfusion setup (shown in F igure 1 8 19 ) achieved a greater number of pores exposed to the pressure of the water column. Fi rst, a flat sample was prepared by sandwiching a rectangular mesh between a 3:2 MMA/PEG DMA mixture. The nanofibers were dissolved out from the bulk polymer and the porous sample was imbedded in a strong fast cast urethane (Master Fast Cast Urethane, Dynaca st) inside a cylindrical mold. Once set, a 1mm thick cross section of the epoxy cylinder containing the porous sample was cut on a lathe. This thin cylinder was placed between two flanged tube fittings and sealed with silicone gasket sealer on either side. This apparatus was inserted on the bottom of the water column and perfused with a 1:1 mixture of dimethylformamide (DMF) and water, which was collected in a vial. Visual confirmation of fluid in the collecting vial was used to qu a lify perfusion. Figure 19 (Left) Schematic of the third type of perfusion setup. Perfusion fluid enters from a water column into a flanged aluminum fixture. A casted urethane disc containing a nanoporous MMA/PEGDMA sample is placed in the flow path and sealed with a silicone disc between each aluminum fixture. Fluid drains through the nanopores out of the other end and collects in a vial for further testing. (Right) Picture of the test disc used for perfusion qualification. Two rectangular pieces are seen imbedded in the disc, each of which contain nanopores. Using two pieces of porous sample yields a higher perfusion rate which will be easily detectable by visual confirmation.

PAGE 45

32 This t est was performed for 90 hours with two different discs. One contained large pores fabricated from a sacrificial template made with 20% PVA electrospinning solution and one contained small pores with a 10% solution. Evaporation was accounted for by using a 10% glucose solution as perfusion fluid and back calculating flow rate through dry weight of the fluid which perfused over the 90 hours. The length of tubing necessary was determined by the amount of pressure needed to simulate physiological eye pressure for a given sample length. This value is reported to be between 15 30 mmHg for glaucomic patients. In order to simulate a 1mm long clinical device using a 6mm long test sample, a back pressure of 90 180 mmHg was needed. For example, a 6mm long test sample requires six time the eye pressure in order to s imulate a 1mm clinical device. In this study, pressure of 20 mmHg was considered a disease state IOP requiring 64 inches of water pressure for a 6mm long test device. All perfusion setups were also tested wi th a control sample of identical geometry and no nanochannels to confirm efficacy of the setup. Figure 20 Picture of the test fixture in line with a water column. Fluid will drain from the water column through the porous section of the test disc, and ultimately collect in a vial.

PAGE 46

33 CHAPTER III RESULTS Electrospinning Using the house made electrospinning apparatus, PVA nanofibers were f abricated in a variety of sizes topological morpholo gie s. In general, as the concentration of PVA inside the syringe was increased, the fiber diameters increased and fiber formation was more stable. b Figure 21 SEM images of PVA nanofiber mats collected on a flat plate. All parameters were identical besides a ch ange in PVA solution from 8%, 1 0%, 15 % and 20% corresponding to a c respectively.

PAGE 47

34 At a low concentration, fibers form ed in ribbon like morphologies (Figure 2 1 .a). At a conc entration of 10%, t he fibers stabilize (Figure 2 1 .b).At higher concentrations of 15% and 20%, fiber dia meters increase (Figu re 21.c, Figure 21 .d). Fiber measurements for Mowiol 8 88 and Mowiol 10 98 are given in Table 1 and Table 2, respectively. There were stark differences in the fiber formation between the two types of PVA used in this study Mowiol 8 88 and Mowiol 10 98 While Mowiol 10 98 yielded higher fiber diameters at a given concentration, its high level of hydrolysis hampered its ability to dissolve in water, there fore limiting the maximum fiber diameter achievable. The high hydrolysis of Mowiol 10 98 also made the fibers produced at lower concentrations more stable (Figure 2 2 ).

PAGE 48

35 Figure 22 Graph showing results of average fiber diameters of various concentrations of two samples of PVA with different degrees of hydrolysis. The PVA with a high degree of hydrolysis produces stable fibers at lower concentrations, and produces higher fiber diameters compared to the low hydrolysis sample of the same concentration. The low hydrolysis PVA is able to be dissolved in water at higher concentrations and is able to produce an overall h igher fiber diameter.

PAGE 49

36 Table 1 Fiber diameter measurements of Mowiol 8 88 for concentrations of 8% 10%, 15%, 20% and 25% (w/w). All measurements in nanometers. Measurement 8% 10% 15% 20% 25% 1 152 217 205 407 263 2 210 106 142 468 436 3 77 13 2 173 520 712 4 105 106 226 414 431 5 143 94 116 229 323 6 175 126 94 213 421 7 118 67 15 288 190 8 139 177 157 576 612 9 196 137 106 379 487 10 66 54 195 325 237 11 180 70 95 223 229 12 261 152 94 402 351 13 193 164 91 523 276 14 141 142 229 26 9 446 15 114 76 124 380 569 Mean 151 .3 121 147 374 399 SD 52.3 46 49 116 152 Min 66 54 91 213 190 Max 261 217 229 576 712

PAGE 50

37 Table 2 Fiber diameter measurements of Mowiol 10 98 of concentrations 8%, 10% and 13% (w/w). All measurements in nanometers. Measurement 8% 10% 13% 1 124 122 349 2 71 171 371 3 90 155 239 4 82 112 236 5 81 310 290 6 51 140 272 7 86 207 445 8 71 93 338 9 82 152 196 10 90 94 201 11 71 75 298 12 81 222 202 13 78 124 259 14 66 176 331 15 81 99 223 Mean 80 150 283 S D 16 61 72 Min 51 75 196 Max 124 310 445

PAGE 51

38 Alternative fiber collectors yielded fiber morphologies which may be useful in the final design of a transcorneal drainage device. Using the pin gap electrode collector, arrays of aligne d fibers were achieved (Figure 23 ). Collecting fibers on a wire was a useful way to form and easily manipulate a fiber array in a cylindr i cal shape. As shown in Figure 2 4 fibers are able to be collected on a wire with a stable and randomly oriented morphology. Figure 23 (left) SEM image of aligned PVA nanofibers collected on a pin gap electrode. (Right) Laser scanning topology microscopy of aligned PVA nanofibers collected on a pin gap electrode. Color indicate s depth with red being closest to the camera.

PAGE 52

39 The investigation into electrospinning proved that tailorable fiber arrays of various morphologies and sizes are able to be made easily and reliably using several different collectors. Figure 24 (Left) SEM image of PVA nanofibers collected on a 25 micron wire. The diameter of fibers collected around the fiber is approximately 200 microns. (Right) Magnified SEM image of the fibers pict ured on the left. Fiber morphology indicates stable, branching fibers.

PAGE 53

40 Nanofiber/Polymer Composite N anofibers produced via electrospinning PVA on a 25 micron wire were successfully incorporated into a cylindrical ly shape d memory polymer composite. These sam ples were made in a variety of shapes and with several shape memory polymers depending on the f inal us e. Pictured in figure 25 is a tert butyl acrylate (tBA) with di(ethylene glycol) dimethacrylate ( DEGDMA ) crosslinker shape memory polymer and a nanofiber bearing 25 micron wire composite which was formed in a 100 micron ID capillary tube. This sample was prepared to show viability of creating a sample mimicking the size of a filter which would be incorporated into a glaucoma drainage device. Porous LCE Cyl indrical samples of pure LCE were fabricated in pipettes to show the change in diameter with exposure to heat. The sample shown in Figure 2 6 starts at 0 o C (Left) and contracts over 2 minutes of exposure to heat (Right). The diameter of the cylinder change s from 0.14 in to 0.20 in, a change of 43% the original diameter. The change in shape Figure 25 A cylindrical composite (D = 100 micron) of nanofibers on a wire with tBA and DEGDMA crosslinker.

PAGE 54

41 represents a molecular level reorganization of the mesogens from a monodomain (Figure 2 6 Left) to polydomain (Figure 2 6 Right). Porous c ylindrical samples created with L CE were used to confirm the existence of channels within the sample. Cross sections of cylindrical nanoporous LCE samples taken in the center of the sample were analyze d under SEM imaging (F igure 27 ). The surface shown was within 100 microns of the wire, where hollow channels were expected to have taken the place of PVA nanofibers The holes seen in the surface have both circular and elliptical topology, resulting from the random orientation of the fibers at this cross section. Figure 26 A cylindrical sample of LCE transitioning phases due t o heating. (Left) Cylindrical sample starts at 0 oC and a diameter of 0.14 in. The sample is in the monodomain.(Right) After heating of the sample for 2 minutes, the sample transitions to polydomain and has a diameter of 0.2 in.

PAGE 55

42 Figure 27 SEM image of the surface of a cross sectioned cylindrical nanoporous LCE. The image was taken from a sample which used the fiber on a wire approach to composite formation. The topology and size of the holes seen in this picture indicate that they were indeed formed fr om dissolved nanofibers. Some holes are elliptical in shape, indicating that a cylindrical fiber was passing through that cross section at an angle.

PAGE 56

43 Fluorescent Imaging Flattened composites for fluorescent imaging were also successfully created These were made using several methods. The first was using a 1mm thick spacer between two glass slides. This produced an approximately 1mm thick rectangular sample. The second way in which flattened sample s were created was without any spacer at all. These samples were approximately 80 microns in thickness. Using these two styles of sample thicknesses, images shown in Fig ure 28 were captured using fluorescent confocal microscopy. B oth samples were created with nanofibers formed from electrospinning a 20% w/w solution of PVA. Fiber measurements from the 80 micron t hick sample are shown in Table 3 Figure 2 8 Fluorescent microscopy image showing nanochannels infiltrated with a 100 fluorescein. Images from two samples are shown. On the left is a 1mm thick sample and the right shows an 80 micron thick sample. Both samples were made with nanofibers formed from electrospinning a 20% w/w concentration of PVA.

PAGE 57

44 The average fiber diameter of nanochannels are compared to the sacrifici a l nanofiber template in Figure 29 The average diameters were taken from a small area of the samples, and merely approximate the total average diameter with 15 measurements fr om each image. Using a two sample t test, the means are significantly similar (P < .01). Table 3 Fiber diameter me asurements taken from the 80 micron thick sample containing nanochannels formed from 20% w/w concentration PVA solution electrospun on a flat plate collector Measurements are in nanometers. Measurement Diam eter 1 684 2 568 3 546 4 312 5 621 6 403 7 494 8 287 9 462 10 403 11 479 12 408 13 305 14 429 15 456 Mean 4 57 SD 114 Min 305 Max 684

PAGE 58

45 Perfusion Testing Perfusion was shown using three test fixtures. The first was a simple pipet apparatus. Figure 30 shows two pipette perfusion tests after 48 hours. The red dye was perfused through a 6mm porous cylinder, and is clearly seen to be collecting in the water bath. The green dye was used in a control setup where the 6mm cylinder did not contain any pores. The lack of green dye in the collecting bath shows that no dye is escaping the pipette around the edges of the cylinder. Figure 29 Bar graph comparing the average diameter of electrospun fibers and nanochannels (taken from fluorescent images). Both were formed from a 20% w/w PVA electrospinning solution 0 100 200 300 400 500 600 Average Diameter of Sacrificial Fibers and Nanochannels Electrospun Mat Nanochannels

PAGE 59

46 The s econd perfusion setup used a water column to increase the pressure behind the filter in order to reach physiological conditions. The thicker cylinder diameter of this setup made handling and securing test samples in place with ease. Results were obtained u sing a 5x magnification microscope and confirmation of a blue dye reaching the other side of the filter The results in Figure 31 show blue dye along the edges of a copper wire, with some blue dye surrounding the wire. It is apparent that the dye traveled through the filter and wetted the exposed face. Because flow was extremely slow, the water in the dye was able to evaporate, depositing spots of solid dye on the face of the sample. Figure 30 A pipette based dye perfusion test. The red dye was perfused through a porous cylindrical sample, while the green dye was used in a control setup with a cylinder containing no pores. Perfusion is conf irmed in the red test while the green dye did not penetrate the fixture.

PAGE 60

47 The third perfusion test was performed using a disc apparatus which mini mized the thickness of the sample the liquid perfused Confirmation of perfusion was attained by visual confirmation of fluid reaching the back side of the fixture. A collection vial was placed beneath the test fixture, but after 48 hours fluid was still c aptured inside the aluminum test fixture, held in place through capillar y action. Figure 3 2 shows t he downstream end of the aluminum test fixture after 48 hours. Fluid is visible inside the test fixture, indicating it passed through the porous disc. The d ry weight of the evaporated perfusion fluid was used to provide a more comprehensive qualitative analysis. The table of measurements and conversion into flow rate Figure 31 Enhanced contrast image of the end of the large cylinder /water column test fixture after 48 hours. The blue seen next to the copper wire shows that dye has perfused thr ough the cylinder. Blue dye seen around the wire is the result of surface wetting followed by evaporation of the dye.

PAGE 61

48 is listed in table 4. The results indicate that, as expected, less fluid passes through a sam ple of smaller pores compared to a sample containing larger pores. Figure 32 Fluid collected in the downstream side of the disc perfusion apparatus. Image was taken after 48 hours of continuous perfus ion Table 4 Data obtained from 90 hour perfusion studies of two discs with different sized pores. Large and small pores were formed with a sacrificial template made from 20% and 10% PVA electrospinning solution, respectively. Pore Size (Concentration) Dry Weight Volume Fluid Perfused Flow Rate Large Pores (20%) 0.822 g 8,220 ul 1.52 ul/min Small Pores (10%) 0.371 g 3,710 ul 0.68 ul/min

PAGE 62

49 CHAPTER IV DISCUSSION Trans corneal drainage is on the forefront of glaucoma device indust ry interest It offers a predictable outflow of AH while avoiding bleb formation and complicat ions associated with invasive surgery such as scarring and inflammation Implanting a trans corneal drainage device requires minimal surgical skill and is therefore well suited for treatment of glaucoma in developing countries. There are several additional considerations when designing a trans corneal drainage device compared to traditional ab interno devices, which are not exposed directly to the environment. First of all, the trans corneal device must provide a microbial barrier. Secondly, the device must rely solely on its intrinsic drainage properties to manage IOP there is no downstream pressure barrier. Lastly, the trans corneal device must be secured in place with out the use of barbs or sutures as many ab interno devices are. Fabrication of a trans corneal drainage device as proposed in this investigation meets all three of these requirements, as well as offering additional advantages for users and surgeons. The first aim of this study was creating tailorable nanofibers via electrospinning. It is wi dely reported in literature that electrospinning solution is the most sensitive parameter in changing the diameters of nanofibers produced ( 43 46 ) It was found that by raising the concentration of PVA (Mowiol 8 88) in the electrospinning solution from 10% to 25% fibers from 54 nm to 712 nm could be fabricated respectively Fundamental information about the nature of electrospinning was also gained in this process. By using two types of PVA, Mowiol 8 88 and Mowiol 10 98, relationships between

PAGE 63

50 average molecular weight and fiber morphology were observed, as well as relationships between degree of hydrolysis and solubility of PVA in water While using 8% w/w concentrat ions of both Mowiol 8 88 and Mowiol 10 98, drastically different fiber morphologies were observed. While solutions of Mowiol 10 98 produced fibers with a unique, branching and cylindrical morphology, solutions of Mowiol 8 88 produced fibers with a flattene d, ribbon like morphology. As indicated by Tao, J, this was directly related to the entanglement concentration of the type of PVA used. Mowiol 10 98 would yield an entangled solution at lower concentrations due to the increased hydrophilic interactions. Th ese would create stable nanofibers upon electrospinning while the same concentration of Mowiol 8 88 would produce flattened fibers. This is an important observation because only unique, cylindrical fibers will be suitable to achieve a drainage device with tailorable, controllable and predictable outflow. The relationship between degree of hydrolysis of the PVA and its ability to dissolve in water was counter intuitive, but allowed fibers of higher diameters to be fabricated when taken advantage of. The man ufacturing of PVA is achieved by hydrolyzing poly(vinylacetate), a polymer which is not water so luble (Figure 33 ) Contrary to common sense, increasing the degree of hydrolysis does not always increase the water solubility. At Figure 33 Hydrolysis of poly(vinylacetate) into poly(vinylalcohol) ( 1 )

PAGE 64

51 very high (99%) levels of hyd rolysis, solid PVA forms very stable crystal structures which are require heating above 100 o C in water to dissolve. At a lower level of hydrolysis (85%), the remaining acetate groups act as steric hindrance to tight crystal structure formation and allow f or much easier dissolution of PVA in water at temperature below 100 o C. The majority of this study was performed with lower hydrolysis Mowiol 8 88 due to its ability to dissolve at high concentration and thus create a wide range of fiber diameters. The dif ference in fiber diameters between the two types of PVA at an electrospinning solution concentration of 10% can be explained by the difference in surface tension between the solutions ( 47 ) The highly hydrolyzed PVA is more hydrophilic and thus has an increased surface tension compared to a less hydrophilic polymer solution. This incre ase in surface tension alters the Rayleigh instability relationships during splaying of the polymer jet in the electric field. This alteration causes earlier gelation of the fiber jet as it travels to the collector, thus increasing the size of fibers colle cted. The use of multiple fiber collector geometries allowed manipulation of the fiber array patterns from randomly oriented to fully aligned. This is advantageous because when used in device fabrication, these two array patterns will yield very different drainage properties. As discussed earlier, the Hagan Pois euille equation dictates fluid flow through a single pipe. It states that drainage rate is proportional to the length of this tube. Considering with just one nanochannel in a 1mm long drainage device path length fluid will travel if the nanochannel was formed with unaligned fibers will be much greater than 1mm. However, if an aligned fiber array is used to fabricate the drainage device, the path length of fluid flow through the device will be very cl ose to, if not exactly, 1mm. In this study, unaligned fiber arrays were studied due to their ease of handling compared to unaligned fiber which generally are formed

PAGE 65

52 between two electrodes (hanging in air) rather than directly on a collector. Future studies may investigate a manufacturing technique that allows for aligned fibers to be incorporated into the device. The use of varying concentrations of PVA solution yielded tailorable and predictable nano fiber morphologies and topologies. This is the basis of obtaining a viable trans corneal drainage device as the size and morphology of the fibers impacts not only the drainage rate, but the physical barrier to microbes on the surface of the eye. As the average size of corneal bound microbes is about 1 micron, a ny concentration of low hydrolysi s PVA would be able to create a physical barrier to microbes if used to fabricate a sacrificial template for a GDD. Synthesis of LCE was a simple one pot click reaction which required minimal chemical synthesis skill. Foll owing Yakacki et al the TAMAP reaction produced and elastomer which would set within 15 20 minutes of adding catalyst. This was helpful because the reaction mixture was able to be poured into a mold around PVA nanofibers when it still had a relatively low viscosity. Therefore the reaction mixture was able to fully penetrate the nanofiber web before setting. The change of diameter of the cylinder shown in Figure 27 gives an example of the shape memory properties of this material. The chemistry used to synt hesize these cylinders can be easily tailored to achieve a thinner initial diameter by decreasing the crosslink density during the first stage of the reaction. This, however, is at the expense of expansive strength of the polymer. Outside the scope of this study, but still of significant importance to the final product, would an investigation of the fixity and relaxation strength of the shape memory behavior of a cylindrical LCE.

PAGE 66

53 In this investigation, only the first stage of the reaction was used. This wa s effective to observe whether the fiber mesh was being fully infiltrated by LCE monomers. In a second stage reaction a cylindrical sample would be stretched lengthwise and cured with a UV light. Upon heating the sample up to body temperature, the device would expand both lengthwise and axially (see Figure 11 ). The strength at which the device locks in place is crucial for effective placement of the device as IOP swings of around 15 mmHG upon inversion of the head ( 48 ) Visual confirmation of pores via several modes of imaging showed that once samples were treated to 48 hours of 60 oC water, or alternatively 30 minutes of sonication and 24 hours of 60 oC water, fibers were able to be dissolved out of the bulk LCE. In addition to LCE, MMA/DEGDMA was a useful bulk polymer to fabricate test samples. This polymer is glassy at room temperature, where LCE is quite rubbery. The hardness of MMA/DEGDMA ma d e it possible to create rigid cylinders which could be reliably superglued into testing fixtures. Its amorphous nature differs from the LCE polydomain substructure and makes the MMA/DEGDMA optically clear while the LCE is opaque. This allowed the fluoresce nt images to be taken with MMA/DEGDMA (Figure 25 ) while the cross section SEM images clearly showed nanopo res in the LCE sample (Figure 24 ). The mechanical and optical differences not only proved convenient, but the use of two different materials shows the robustness of the manufacturing process. The fluorescent image see n on the right side of Figure 25 is a powerful visual which clearly shows fluid infiltrat ing the nanochannels. This indicate s that fluid would pass through a drainage device manufactured in the same manner. This image differs from the left side of Figure 25 because the microscope used to take these images creates 15 micron deep Z stacks

PAGE 67

54 starting from the bottom of the sample. When imaging a 1mm thick sample, the camera only captures the bott om most fibers in the sample. The 80 micron thick sample shows many more fibers within the 15 micron deep window that was available to image. These images show that fluid may be easily infiltrated into the nanochannels created after dissolution of nanofibe rs embedded in a shape memory polymer. To prove without a doubt that fluid may pass through a nanoporous drainage device fabricated with an electrospun sacrificial template, water column perfusion tests were performed with several indicators and a variety of test fixtures The first perfusion test involved supergluing a porous MMA/DEGDMA cylindrical sample into the tip of a pipette. This was a delicate process that required glue around the entire edge of the pipette tip while not allowing any glue to touch the flat top or bottom surfaces of the sample as the leave the channels clear. Once the sample was secured and sealed into the tip of the pipette and the pipette was filled with a colored dye, the tip of the filter was dipped into a collection bath to wai t for colored dye to perfuse the sample. Encouraging results were obtained from these tests, with an indicator dye appearing to perfuse through the nanochannels int o a collection bath While using a control sample containing no pores, dye was not able to perfuse into the collection vial, proving a robust seal around the cylinder. The two other perfusion set ups also showed qualitative perfusion, with visual confirmation being the indication of a positive result. In the future, a more elaborate perfusion setup would be needed to quantify the flow rate at a given pressure. A programmable syringe pump would be needed to obtain these results, such as Pump 11 Plus by Harvard Apparatus.

PAGE 68

55 CHAPTER V CONCLUSION The purpose of this investigation was to create a ma nufacturing technique for a tailorable trans corneal glaucoma drainage device using a sacrificial nanofiber template within a shape switchable liquid crystal elastomer. After fabrication of an electrospinning power unit, several concentrations of poly(viny lalcohol) were used to create sacrificial nanofiber templates on flat plate and wire collectors. SEM images of these arrays showed various morphologies and topologies of the fibers, including randomly oriented and highly aligned. The diameters of the fiber s produced varied from around 50 nm to 750 nm, which once dissolved out of a bulk polymer yielded nanochannels of the same diameter. It was shown in imaging and perfusion tests that not only could these nanochannels be infiltrated with a liquid, a fluid ma y pass through them in a controlled fashion. There is one instance in the literature published by Bellan et al which uses sacrificial electrospun nanofibers as a template for nanochannels in a bulk polymer ( 38 ) The current study expands on this in several ways. While Bellan et al focus on creating these channels in a poly(dimethylsiloxane), or PDMS, substrate, the current study uses a functional shape memory elastomer. Th is implies greater potential for end usage of such a device including, but not limited to, a trans corneal glaucoma drainage device. Additionally the study performed by Bellan et al did not deeply explore the relationships between electrospun fiber morpho logy and the various electrospinning parameters such as solution concentration. These relationships play a key role in the tunable nature of such a device, especially considering the Hagen Pouseuille equation which dictates flow as a function of channel di ameter to the fourth power.

PAGE 69

56 There were limitations throughout this study that, if alleviated, would allow the proper qualification of this transcorneal drainage device. Of particular importance is the perfusion testing. These tests were designed to indicat e a binary result did the fluid pass through or not. There was little means of measuring exact flow through test fixtures using the low tech solutions that were obtained for little to no cost. As a result, there was no way to quantify the drainage rate o f a trans corneal filter containing convoluted and tortuous nanochannels paths. As indicated previously, the Hagen Pouseuille equation gives the relationship between flow, pressure, and channel diameter. Another equation would better suit this scenario if such quantifiable drainage information was available. The Darcy equation relates flow and pressure inside a porous medium. The tortuous nature of the nanochannels create a scenario that mimics a porous medium closer than a group of pipes. The Darcy equatio n contains a K factor, which is derived from perfusion data. If an advanced perfusion set up ( Pump 11 Plus, Harvard Apparatus ) was available, a quantifiable relationship between PVA solution concentration used in electrospinning and drainage rate of a tran s corneal device could be derived. Another limitation of the study was time. Aligned fiber arrays may prove necessary in future studies if the drainage rates of filter devices using randomly oriented fiber arrays proved too slow. Because of the different n ature of aligned fibers (which collect between to grounded units opposed to directly on a collector), separate manufacturing technique as well as manufacturing fixtures must still be designed. Lowering IOP in patients with glaucoma remains the cornerstone of limiting risk of vision loss, and while there are many approaches to this, no current strategy is without complications. This study proposed a unique solution to replace the gold standard both in surgical and drainage device treatments of Glaucoma. A t rans corneal device which is

PAGE 70

57 replaceable tunable and easily deployed into patients has the potential to greatly reduce the number of those who lose their vision from the disease. It will especially make an impact in areas of the world where ophthalmic sur geons and surgical arenas are not readily available.

PAGE 71

58 REFERENCES 1. F. Medlege, H. Rouault, N. Belgacem, A. Blayo. (Google Patents, 2010). 2. J. S. Distelhorst, G. M. Hughes, Open angle glaucoma. American family physicia n 67 1937 (May 1, 2003). 3. H. A. Quigley, A. T. Broman, The number of people with glaucoma worldwide in 2010 and 2020. British journal of ophthalmology 90 262 (2006). 4. D. Peters, Visual Impairment and Vision Related Quality of Life in Glaucoma. (2015 ). 5. C. N. T. G. S. Group, Comparison of glaucomatous progression between untreated patients with normal tension glaucoma and patients with therapeutically reduced intraocular pressures. American journal of ophthalmology 126 487 (1998). 6. I. Agis, The A dvanced Glaucoma Intervention Study (AGIS): 7. The relationship between control of intraocular pressure and visual field deterioration. American journal of ophthalmology 130 429 (2000). 7. R. N. Weinreb, T. Aung, F. A. Medeiros, The pathophysiology and tr eatment of glaucoma: a review. Jama 311 1901 (2014). 8. R. W. Nickells, G. R. Howell, I. Soto, S. W. John, Under pressure: cellular and molecular responses during glaucoma, a common neurodegeneration with axonopathy. Annual review of neuroscience 35 153 (2012). 9. P. R. Lichter et al. Interim clinical outcomes in the Collaborative Initial Glaucoma Treatment Study comparing initial treatment randomized to medications or surgery. Ophthalmology 108 1943 (2001). 10. A. E. Laing, L. K. Seibold, J. R. SooHoo, M. Y. Kahook, Evaluation of bleb characteristics after implantation of the EX PRESS glaucoma filtration device. Molecular vision 18 10 (2012). 11. E. M. Van Buskirk, D. R. Bacon, W. H. Fahrenbach, Ciliary Vasoconstriction After Topical Adrenergic Drugs. Am. J. Ophthalmol. 109 511 (1990). 12. R. van der Valk et al. Intraocular Pressure Lowering Effects of All Commonly Used Glaucoma Drugs: A Meta analysis of Randomized Clinical Trials. Ophthalmology 112 1177 (2005). 13. R. K. Parrish, P. Palmberg, W. P. Sheu, X. L. T. S. Group, A comparison of latanoprost, bimatoprost, and travoprost in patients with elevated intraocular pressure: a 12 week, randomized, masked evaluator multicenter study. American journal of ophthalmology 135 688 (May, 2003). 14. P. Wats on, J. Stjernschantz, A Six month, Randomized, Double masked Study Comparing Latanoprost with Timolol in Open angle Glaucoma and Ocular Hypertension. Ophthalmology 103 126. 15. J. Stjernschantz, G. Selen, M. Astin, B. Resul, Microvascular effects of selec tive prostaglandin analogues in the eye with special reference to latanoprost and glaucoma treatment. Progress in retinal and eye research 19 459 (Jul, 2000). 16. C. O. Okeke et al. Adherence with topical glaucoma medication monitored electronically the Travatan Dosing Aid study. Ophthalmology 116 191 (Feb, 2009). 17. J. A. Han, W. H. Frishman, S. Wu Sun, P. M. Palmiero, R. Petrillo, Cardiovascular and Respiratory Considerations With Pharmacotherapy of Glaucoma and Ocular Hypertension. Cardiology in Revi ew 16 (2008). 18. D. Chiselita, Non penetrating deep sclerectomy versus trabeculectomy in primary open angle glaucoma surgery. Eye 15 197 (2001).

PAGE 72

59 19. J. S. M. Lai, J. K. H. Chua, C. C. Y. Tham, D. S. C. Lam, Five year follow up of selective laser trabec uloplasty in Chinese eyes. Clinical & Experimental Ophthalmology 32 368 (2004). 20. M. O. M. Wong, J. W. Y. Lee, B. N. K. Choy, J. C. H. Chan, J. S. M. Lai, Systematic review and meta analysis on the efficacy of selective laser trabeculoplasty in open ang le glaucoma. Survey of Ophthalmology 21. P. Y. Ramulu, K. J. Corcoran, S. L. Corcoran, A. L. Robin, Utilization of Various Glaucoma Surgeries and Procedures in Medicare Beneficiaries from 1995 to 2004. Ophthalmology 114 2265 (2007). 22. S. J. Gedde et al Postoperative complications in the Tube Versus Trabeculectomy (TVT) study during five years of follow up. Am J Ophthalmol 153 804 (May, 2012). 23. S. J. Gedde et al. Treatment outcomes in the Tube Versus Trabeculectomy (TVT) study after five years of follow up. Am J Ophthalmol 153 789 (May, 2012). 24. L. M. Brando, M. C. Grieshaber, Update on Minimally Invasive Glaucoma Surgery (MIGS) and New Implants. Journal of ophthalmology 2013 (2013). 25. J. F. Jordan et al. A Novel Approach to Suprachoroidal Drainage for the Surgical Treatment of Intractable Glaucoma. Journal of glaucoma 15 (2006). 26. C. Pulling. (OIS, 2015), vol. 2015. 27. P. De Gennes, Rflexions sur un type de polymres nmatiques. CR Acad. Sci. Ser. B 281 2 (1975). 28. H. Finkelmann, H. J. Kock, G. Rehage, Investigations on liquid crystalline polysiloxanes 3. Liquid crystalline elastomers a new type of liquid crystalline material. Die Makromolekulare Chemie, Rapid Communications 2 317 (1981). 29. C. M. Yakacki et al. Tailorable and p rogrammable liquid crystalline elastomers using a two stage thiol acrylate reaction. RSC Advances 5 18997 (2015). 30. F. Brmmel, D. Kramer, H. Finkelmann, in Liquid Crystal Elastomers: Materials and Applications (Springer, 2012), pp. 1 48. 31. A. R. Ga rca Mrquez, B. Heinrich, N. Beyer, D. Guillon, B. Donnio, Mesomorphism and Shape Memory Behavior of Main Chain Liquid Crystalline Co Elastomers: Modulation by the Chemical Composition. Macromolecules 47 5198 (2014/08/12, 2014). 32. T. Ware, T. White, Pr ogrammed Liquid Crystal Elastomers with Tunable Actuation Strain. Polymer Chemistry (2015). 33. C. Ahn, X. Liang, S. Cai, Inhomogeneous stretch induced patterning of molecular orientation in liquid crystal elastomers. Extreme Mechanics Letters (2015). 34. M. O. Saed, A. H. Torbati, D. P. Nair, C. M. Yakacki, Synthesis of Programmable Main chain Liquid crystalline Elastomers Using a Two stage Thiol acrylate Reaction. Journal of visualized experiments : JoVE (2016). 35. J. Kpfer, H. Finkelmann, Nematic liquid single crystal elastomers. Die Makromolekulare Chemie, Rapid Communications 12 717 (1991). 36. C. Yakacki et al. Tailorable and programmable liquid crystalline elastomers using a two stage thiol acrylate reaction. RSC Advances 5 18997 (2015). 37 X. Luo, P. T. Mather, Shape Memory Assisted Self Healing Coating. ACS Macro Letters 2 152 (2013/02/19, 2013). 38. L. M. Bellan, E. A. Strychalski, H. G. Craighead, Nanochannels fabricated in polydimethylsiloxane using sacrificial electrospun polyethylen e oxide nanofibers. Journal of Vacuum Science & Technology B 26 1728 (2008). 39. A. Greiner, J. H. Wendorff, Electrospinning: A fascinating method for the preparation of ultrathin fibres. Angew Chem Int Edit 46 5670 (2007).

PAGE 73

60 40. D. H. Reneker, A. L. Yarin E. Zussman, H. Xu, Electrospinning of nanofibers from polymer solutions and melts. Adv Appl Mech 41 43 (2007). 41. W. E. Teo, S. Ramakrishna, A review on electrospinning design and nanofibre assemblies. Nanotechnology 17 R89 (2006). 42. S. Chuangchote, A. Sirivat, P. Supaphol, Mechanical and electro rheological properties of electrospun poly (vinyl alcohol) nanofibre mats filled with carbon black nanoparticles. Nanotechnology 18 145705 (2007). 43. J. Tao, Effects of Molecular weight and Solution Concen tration on Electrospinning of PVA. (2003). 44. B. Ding, Preparation and Characterization of Nanoscaled PVA Fibers via Electrospinning. Fibers and Polymers 3 73 (2002). 45. K. H. Hong, Preparation and properties of electrospun poly(vinyl alcohol)/silver f iber web as wound dressings. Polymer Engineering & Science 47 43 (2007). 46. Q. Li, Preperation and Properties of Electrospun PVA Nanofibers. International Conference on Sollid Dielectrics (2007). 47. J. C. Park et al. Electrospun poly(vinyl alcohol) n anofibers: effects of degree of hydrolysis and enhanced water stability. Polym J 42 273 (2010). 48. R. N. Weinreb, J. Cook, T. R. Friberg, Effect of Inverted Body Position on Intraocular Pressure. American journal of ophthalmology 98 784.