.. .. ha . .
. .99....
Aug

.. . . .

. {ha} a

i 5.5!. . .u.

. Van“... (1..
I

.. ¢ . W
L. 1! ill. :9. .
UV 3.. in. ”it!!! .
hV 'P‘USIf! 11:01,}... K .
.1 )v “5 .. y. . .
.\ . . :53. t 2.3. 15.3.“; II. .
Ii“: . "WM
:3 . .35; TI
. I?! II Vi.

 

 

{33%;
i. .. {$9.353
H‘ II 1 (i. .v

 

 

. (it.
gain”: 2.
“5.03.3.” in.
.5 t . .‘r

 

{11.1.1}!!!
:1 .2. 1.17;.
.v in: I.-.

a. $5.1}
. 3.... 3:52»:
I [ASL-I‘ rfiigl!
.1
(I 2)....
Eros...
1...!
1......

Iclln

 

7/. fl. .. .rubu. . . , . , ., . . . . V. ‘ JIL‘JHH‘WJE.
JJW‘MWW E1: . . 1%..»va

 

.\ :1. y
in». gang .p. . . . .. . .. «9....

2“!

 

LIBRARY
Michigr *7 State
University

 

 

 

This is to certify that the
dissertation entitled

ELECTROSPINNING CELLULOSE BASED NANOFIBERS
FOR SENSOR APPLICATIONS

presented by

Steven Nartker

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Chemical Engineering

 

 

 

Major Professor’s Signature}
/ 21' 0 X V 01043?

Date

 

MSU is an Affirmative Action/Equal Opportunity Employer

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProlecc&Pres/ClRC/DateDuo.indd

ELECTROSPINNING CELLULOSE BASED NANOFIBERS FOR SENSOR
APPLICATIONS

By

Steven Nartker

A DISSERTATION

, Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Chemical Engineering

2009

ABSTRACT

ELECTROSPINNING CELLULOSE BASED NANOFIBERS FOR SENSOR
APPLICATIONS

By

Steven Nartker

Bacterial pathogens have recently become a serious threat to the food and
water supply. A biosensor based on an electrochemical immunoassay has been
developed for detecting food borne pathogens, such as Escherichia coli (E. coli)
01572H7. These sensors consist of several materials including, cellulose,
cellulose nitrate, polyaniline and glass fibers. The current sensors have not been
optimized in terms of microscale architecture and materials. The major problem
associated with the current sensors is the limited concentration range of
pathogens that provides a linear response on the concentration conductivity
chart. Electrospinning is a process that can be used to create a patterned fiber
mat design that will increase the linear range and lower the detection limit of
these sensors by improving the microscale architecture. Using the
electrospinning process to produce novel mats of cellulose nitrate will offer
improved surface area, and the cellulose nitrate can be treated to further improve

Chemical interactions required for sensor activity.

The macro and micro architecture of the sensor is critical to the performance
of the sensors. Electrospinning technology can be used to create patterned
architectures of nanofibers that will enhance sensor performance. To date
electrospinning of cellulose nitrate has not been performed and optimization of
the electrospinning process will provide novel materials suitable for applications
such as filtration and sensing. The goal of this research is to identify and
elucidate the primary materials and process factors necessary to produce
cellulose nitrate nanofibers using the electrospinning process that will improve
the performance of biosensors.

Cellulose nitrate is readily dissolved in common organic solvents such as
acetone, tetrahydrofuran (T HF) and N,N dimethylformamide (DMF). These
solvents can be mixed with other latent solvents such as ethanol and other
alcohols to provide a solvent system with good electrospinning behavior. Using
cellulose nitrate in biosensor materials provides excellent antibody binding
characteristics that are resistant to pH changes.

Sensors will be constructed of electrospun materials and compared to
existing materials. The main advantage of electrospinning fiber mats is‘the
increased surface area, and controllable morphology, which ultimately affects
biosensor performance. Characterization tools will include Environmental
Scanning Electron Microscopy (ESEM), BET N2 adsorption, X-Ray Photoelectron

Spectroscopy (XPS), Dynamic Mechanical Analysis (DMA) and AC impedance.

ACKNOWLEDGEMENTS

My first acknowledgement goes to Professor Lawrence T. Drzal. I would like
to thank him for providing the research tools, personal support, encouragement,
mentoring which made my research enjoyable. I would also like to thank my
Ph.D. committee members, professors llsoon Lee, Ramani Narayan, and
Evangelyn Alocilja, and to the Department of Chemical Engineering and
Materials Science.

I would like to extend a special thank you to Dr. Manju Misra, Dr. Hazel-Ann
Hosein, Dr. Per Askeland, Brian Rook, Ed Drown, and Mike Rich in the
Composite Materials and Structures Center for providing a safe, professional,
and instructive work environment, and for their friendship along the years. I
would especially like to thank Dr. Askeland for his support, expertise and
guidance. I would also like to thank Sara and Michael Wiederoder for their
assistance. I would also like to thank all the graduate students that I shared an
office with.

I am very grateful for the unlimited support of my family through this process.

I truly appreciate Amy who provided encouragement and support.

 

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................... viii
LIST OF FIGURES ................................................................................ x
CHAPTER 1
INTRODUCTION AND BACKGROUND .............................................................. 1
Problem Statement .................................................................................... 1
Research Goals ......................................................................................... 3
Research Approach ................................................................................... 4
CHAPTER 2
BACKGROUND AND LITERATURE REVIEW ................................................... 6
Abstract ..................................................................................................... 6
Introduction ................................................................................................ 7
Electrospinning .......................................................................................... 8
Modeling Of The Electrospinning Process .......................................... 11
Predicting Fiber Diameter Using Solution Viscosity ............................ 13
Fiber Diameter Predictions Using Force Balances ............................. 20
Dimensional Analysis Approach ......................................................... 25
Modeling Summary ............................................................................. 34
Biosensors ............................................................................................... 34
Acoustic .............................................................................................. 37
Optical ................................................................................................ 39.
Electrochemical .................................................................................. 41
Immunoassay ..................................................................................... 43
Electrochemical Immunoassay ........................................................... 44
Summary ................................................................................................. 51
References .............................................................................................. 52
CHAPTER 3
ELECTROSPINNING POLY(VINYL ALCOHOL) PVOH .................................... 60
Abstract ................................................................................................... 60
Introduction To Electrospinning Poly(Vinyl Alcohol) ................................ 61
Experimental ............................................................................................ 63

V

Materials ............................................................................................. 64

Morphology ......................................................................................... 64
Thermal Properties ............................................................................. 65
Mechanical Properties ........................................................................ 65
Results And Discussion ........................................................................... 66
Morphology ............................................................................ 66
Fiber Size Distribution ......................................................................... 72
Effect Of Electric Field Strength .......................................................... 73
Fiber Diameter Summary .................................................................... 77
Alignment Distribution ......................................................................... 83
Effect Of Polymer Flow Rate On Morphology ..................................... 87
Effect Of Solution Viscosity On Morphology ....................................... 90
Effect Of Electrospinning Environment On Morphology ...................... 96
Effect Of Solution Conductivity On Morphology .................................. 96
Effect Of Solvent Volatility On Morphology ......................................... 97
Effect Of Solution Rheology On Morphology ...................................... 98
Thermal Properties ................................................................................ 101
Mechanical Properties ........................................................................... 102
Conclusions ........................................................................................... 107
References ............................................................................................ 1 1 1
CHAPTER 4
ELECTROSPINNING CELLULOSE NITRATE FOR BIOSENSOR
APPLICATIONS ............................................................................................. 115
Abstract ................................................................................................. 1 15
Introduction ............................................................................................ 1 17
Background ........................................................................................... 1 18
Experimental .......................................................................................... 122
Materials ........................................................................................... 122
Morphology ....................................................................................... 123
Surface Chemistry Properties ........................................................... 123
Contact Angle Measurements .......................................................... 123
X-Ray Photoelectron Spectroscopy .................................................. 124
Plasma Treatment ............................................................................ 124
Fourier Transform Infrared Spectroscopy ......................................... 125
Thermal Properties - Thermal Gravimetric Analysis ........................ 125
Electrical Properties - Electrical Impedance Spectroscopy .............. 125
Mechanical Properties — Dynamic Mechanical Analysis ................... 126
Results And Discussion .......................................................................... 126
Morphology ....................................................................................... 126
Effect Of Electric Field Strength ........................................................ 129
Effect Of Solution Viscosity On Morphology ..................................... 131
Effect Of Solvent Volatility On Morphology ....................................... 132
Surface Properties ............................................................................ 135
Plasma Treatment ............................................................................ 135
vi

_J.

 

Contact Angle ................................................................................... 1 36

X-Ray Photoelectron Spectroscopy .................................................. 145
Fourier Transform Infrared Spectroscopy ......................................... 150
Thermal Properties - Thermal Gravimetric Analysis ........................ 155
Electrical Properties .......................................................................... 157
Mechanical Properties - Dynamic Mechanical Analysis ................... 157
Conclusions ........................................................................................... 159
References ............................................................................................ 162
CHAPTER 5
BIOSENSOR PERFORMANCE ...................................................................... 165
Abstract ................................................................................................. 165
Introduction ............................................................................................ 165
Experimental .......................................................................................... 168
Nitrocellulose Fiber Mat Preparation ................................................ 168
Antibody Attachment To Capture Pad .............................................. 169
Preparing Conjugate Solution ........................................................... 171
Preparation Of Test Solution ............................................................ 172
Test Strip Construction ..................................................................... 173
Results And Discussion ......................................................................... 175
References ............................................................................................ 175
CHAPTER 6
CONCLUSIONS .............................................................................................. 178
Future Work ........................................................................................... 182

vii

LIST OF TABLES

Table 2.1
Table 2.2
Table 2.3
Table 2.4
Table 2.5

Table 3.1

Table 3.2

Table 3.3

Table 3.4

Table 3.5
Table 3.6

Table 3.7

Table 4.1

Table 4.2

Table 4.3
Table 4.4

Daga model data set .............................................................. 16
McKee data table ................................................................... 18
Fridrikh model data set. .......................................................... 24
Helgeson data table ............................................................... 32
Summary of electrospun material used for biosensor
applications ............................................................................ 50
Effect of increasing the electrospinning variables on the
resultant fiber morphology. ..................................................... 63

Separation distance affects the fiber diameter and alignment.
Applied voltage 10 kV, syringe pump rate 0.1 ml/h, 10 wt%

PVOH in water ........................................................................ 78
The effect of variables on the diameter and variability (standard
deviation) of electrospun PVOH fibers. .................................. 80
Summary of various electric field strengths and the resultant

fiber morphology ..................................................................... 82
Summary of alignment distributions. ...................................... 85
Summary of processing parameters and their effects on the

fiber diameter. Results avg. 2 SD, n = 40. .............................. 95
Mechanical properties of electrospun PVOH materials. MSU is
a reference to this work. ....................................................... 104

Fiber diameter and alignment changes as a result of solution

viscosity ................................................................................ 128
The effect of separation distance on the fiber diameter and

alignment .............................................................................. 131
Solvent volatility effect on fiber diameter and alignment ....... 134
Contact angle measurements ........................................ I ....... 143

viii

 

Table 4.5

Table 4.6

Table 4.7

Table 4.8

Water contact angle data showing the effect of plasma
treatment .............................................................................. 144

XPS summary table — relative ON and 0 concentrations are
constant with plasma treatment, but functionalities have
changed ............................................................................... 150

FTIR summary table of major absorbencies and the associated
chemical group assignment .................................................. 154

Mechanical properties of electrospun cellulose nitrate fibrous
materials ............................................................................... 158

LIST OF FIGURES

Images in this dissertation are presented in color.

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 2.6

Figure 2.7

Figure 2.8
Figure 2.9
Figure 3.1

Figure 3.2

Figure 3.3

Electrospinning schematic showing A) the syringe, B) the
Taylor Cone, C) the stable region, D) the instability region and
E) the rotating drum style collector. .......................................... 9

Daga data chart showing a good fit for PEO water, but a poor
fit for CN and PVOH. .............................................................. 17

McKee model chat showing good correlation for PEO at low
viscosity values and a relatively poor fit for CN and PVOH. 19

Fridrikh model chart showing correlations for PCL, PEG and

PAN. ....................................................................................... 25
Helgeson chart showing good correlation for almost all data

sets ......................................................................................... 33
Biosensor schematic .............................................................. 35

Biosensor consisting of an antibody as the biological element,
conducting polymer chain as the transducer and electrospun

fibers as the matrix. ................................................................ 36
Classification of biosensors .................................................... 37
Conductive nanofiber biosensor schematic. ........................... 45

Electrospinning schematic showing (A) the polymer solution in
the syringe, (B) the Taylor Cone at the needle tip, (C) the
stable region of the charged jet, (D) the unstable region of the
charged polymer jet and (E) the C-shaped void gap collector 67

Various collectors used in the electrospinning of PVOH. A) A
flat plate of aluminum, B) a rotating steel drum with a slot milled
in the surface, C) a parallel set of copper electrodes. ............ 68

ESEM comparison of electrospun PVOH. A) Stationary
collector, B) Rotating drum collector with slot, C) Stationary
parallel electrode collector ...................................................... 69

Figure 3.4 ESEM micrograph of electrospun 10 wt% PVOH fibers. 10 cm
Separation distance, 500 X magnification, scale bar shown is
100 micrometers ..................................................................... 71

Figure 3.5 High magnification (2,000X) ESEM image of electrospun
PVOH. The scale bar is 25 um. ............................................. 73

Figure 3.6 Fiber size as a function of electric field strength and
concentration for 54,400 MW PVOH. Increases in electric field
strength reduce fiber diameters. ............................................. 76

Figure 3.7 ESEM images of electrospun PVOH fibers, showing variation
with separation distance A) 8 cm, B) 10 cm, and C) 12 cm.
Applied voltage 10 kV, pump rate 0.1 ml/h, 10 wt% in water. 79

Figure 3.8 Fiber diameter comparison for 10 wt% PVOH fibers
electrospun from separation distances of 8 cm, 10 cm and 12
cm. ......................................................................................... 83

Figure 3.9 Angle distribution of 10 wt% electrospun PVOH fibers from
separation distances of 8 cm, 10 cm, and 12 cm. .................. 87

Figure 3.10 ESEM images of electrospun PVOH fibers, showing the
variation with syringe pump flow rate A) 0.01 ml/h, B) 0.1 ml/h,
C) 0.2 ml/h, and D) 0.3 ml/h. Applied voltage 10 kV, 10 cm
separation distance, 10 wt% PVOH in water. ......................... 90

Figure 3.11 Viscosity chart of PVOH, indicating the viscosity range in which
fibers can be readily formed. .................................................. 93

Figure 3.12 ESEM images of electrospun PVOH fibers, showing the
variation with concentration A) 8 wt%, B) 10 wt%, and C) 12
wt%. Applied voltage 10 kV, syringe pump rate 0.1 ml/h, 10 cm
separation distance ................................................................ 94

Figure 3.13 ESEM image of electrospun PVOH fibers, showing the
optimized morphology. Applied voltage 10 kV, syringe pump
rate 0.1 ml/h, 10 wt% PVOH in water, 10 cm separation
distance. ................................................................................. 95

Figure 3.14 PVOH Silica solution. Image shows that silica is agglomerated
in solution. .............................................................................. 99

Figure 3.15 TEM image of electrospun PVOH with silica. Image shows that
silica aggregations are still present in final fiber. .................. 101

Figure 3.16 Thermal analysis of electrospun PVOH and silica loaded
PVOH. .................................................................................. 102
xi

Figure 3.17 Stress strain chart for electrospun PVOH. Ultrasonic mixing of

silica provides the best improvement. .................................. 105

Figure 3.18 Modulus comparison for the electrospun PVOH and silica

reinforced PVOH. ................................................................. 106

Figure 3.19 Relative increases in fiber diameter as a function of

Figure 4.1

Figure 4.2

Figure 4.3

Figure 4.4

Figure 4.5

Figure 4.6

Figure 4.7

Figure 4.8

Figure 4.9

electrospinning variables. ..................................................... 109

ESEM of electrospun cellulose nitrate comparing the alignment
capabilities. A) Fibers were collected using a stationary parallel
electrode collector design. B) Fibers were collected using a

rotating drum. ....................................................................... 122

FESEM image showing that single fiber sizes are available less
than 100 nm in diameter ....................................................... 127

ESEM image of patterned electrospun fibers using a stationary
collector. ............................................................................... 129

ESEM of electric field strength variations, A) 12 cm (0.9
kV/cm), B) 10 cm (1.1 kV/cm), C) 8 cm (1.4 kV/cm) and D) 6
cm (1.8 kV/cm). .................................................................... 130

ESEM images of 6,8 and 10 wt% cellulose nitrate electrospun
fibers. ................................................................................... 132

The effect of solvent volatility on fiber morphology. Variation in

solvent vapor pressure results in various volatility A) 100%

Acetone, B) 70:30 Ethanolecetone, C) 50:50 Ethanolecetone.
134

Electrospun cellulose nitrate shows an open fiber mat with air
occupying the interfibrillar spaces. A) Fibers electrospun from
plastic syringe, B) glass syringe C) image (A) after contrast
enhancement and threshold screening, D) image (B) after
processing. ........................................................................... 139

Apparent contact angles for electrospun fibers with various
diameters and fiber spacing. ................................................ 142

Comparison of water on the surface of electrospun cellulose
nitrate fibers. The non-wetting behavior is shown in A) neat
electrospun fibers and after plasma treatment B) the water
droplet wets the surface. ...................................................... 144

Figure 4.10 Comparison of C13 spectra. An addition of carbonyl and/or

carboxylate signal (289 eV) is evident. ................................. 146
xii

Figure 4.11 N1s XPS spectra, the reduction in nitrate ester signal (408 eV)
is evident. ............................................................................. 148

Figure 4.12 Full FTIR scan of neat electrospun cellulose nitrate and 90 s
and 360 s plasma treated cellulose nitrate fibers. ................ 151

Figure 4.13 FTIR analysis of the nitro functional group absorbencies. 152

Figure 4.14 FTIR analysis of hydroxyl and carbonyl peaks, showing an
increase in absorbance for plasma treated samples. ........... 153

Figure 4.15 TGA therrnogram of the electrospun fiber networks prepared
from two different solvent systems, 60:40 THFzDMF and 70:30

Ethanol Acetone. .................................................................. 156
Figure 4.16 ESEM showing fiber coverage differences amongst A) 8 wt%,

B) 10 wt% and C) 12 wt% solutions. .................................... 159
Figure 5.1 Schematic of the biosensor device ....................................... 167

Figure 5.2 Schematic showing a cross section of the capture pad on a test

strip. ..................................................................................... 168
Figure 5.3 CLSM image of neat cellulose nitrate fibers and fibers with

fluorescent-tagged antibodies attached ................................ 170
Figure 5.4 Non-Conductive biosensor design ........................................ 174

Figure 5.5 Preliminary biosensor performance data shows that the sensor
works and is sensitive for several dilutions of bacteria. ........ 175

xiii

CHAPTER 1

Introduction

Problem Statement

Bacterial pathogens have recently become a serious threat to the food and
water supply. A biosensor based on an electrochemical immunoassay has been
developed for detecting food borne pathogens, such as Escherichia coli (E. coli)
O157:H7. These sensors consist of several materials including, cellulose,
cellulose nitrate, polyaniline (PANi) and glass fibers. The major problem
associated with the current sensors is the limited concentration range of
pathogens that provides a linear response on the concentration conductivity
chart. The current sensors have not been optimized in terms of microscale

architecture and materials.

The macro and microscale architecture of the sensor is critical to the
performance of the sensors. The commercial capture pad materials that are
used do not optimize surface area or capillary flow properties, which are
essential for efficient biosensor performance. Electrospinning technology can be
used to create patterned architectures of nanofibers that will enhance sensor

performance. Electrospinning is a process that can be used to create nanoscale

diameter fibers with large surface areas in a patterned design that will increase
the linear range and lower the detection limit of these sensors by improving the
microscale architecture. To date electrospinning of cellulose nitrate and cellulose
nitrate functionalized with PANi has not been performed and optimization of the
electrospinning process will provide novel materials suitable for applications such
as filtration and sensing. Electrospinning is the preferred method for producing
nanoscale fibrous materials since it can be controlled to produce various
morphological structures within the overall capture pad size constraint of 20 mm

X5mm.

Electrospinning to produce novel mats of cellulose nitrate has the potential to
provide substrates with increased surface area and a preferred alignment to
enhance pathogen transport and detection. Post-treatment of the electrospun
cellulose nitrate fibers will be performed to further improve chemical interactions,
and capillary flow properties required for sensor activity. Cellulose nitrate can
serve as a core polymer and conductive polymers, such as PANi or polypyrrole
(PPy) can be polymerized on the surface and the resultant core sheath fibers
have potential for use as biosensor materials. Addition of biologically active
materials to the cellulose nitrate membrane will provide a novel approach to
creating biosensor materials with improved functionality. Controlling the
electrospinning process to produce aligned homogenous nanoscale fibrous mats
that offer optimal pore size, surface area, capillary flow and chemical properties
will increase the detection limit and expand the concentration range inherent in

current biosensor devices.

Research Goals

The goal of this research is to identify and elucidate the primary material
variables and process parameters necessary to produce cellulose nitrate, and
cellulose nitrate functionalized with PANi or PPy nanofibers using the
electrospinning process in order to improve the performance of biosensors.
Ultimately all the material and processing parameters will be evaluated to
determine the optimal conditions to consistently produce nanofiber films that will

be utilized as capture pad materials in biosensor devices.

Electrospinning cellulose nitrate has not been reported in the literature.
Cellulose nitrate is readily dissolved in common organic solvents such as
acetone, N, N’ dimethylformamide (DMF) and tetrahydrofuran (THF). These
solvents can be mixed with other latent solvents such as ethanol and other
alcohols to provide a solvent system with good electrospinning behavior. The
importance of this research then is to understand the conditions necessary to
produce aligned homogenous cellulose nitrate nanofibers. Since cellulose nitrate
has the potential to provide excellent antibody binding characteristics that are
resistant to pH changes in biosensor devices, the initial emphasis will focus on
electrospinning of pure cellulose nitrate. Following success in this aspect of the
research, a core sheath fiber consisting cellulose nitrate with polyaniline or

polypyrrole will be used to obtain conductive fiber networks.

Research Approach

Initial efforts to understand and model the electrospinning process will be
conducted using a well-characterized water-soluble polymer system of poly (vinyl
alcohol) (PVOH). The choice of PVOH as a model was dictated by its easy
dissolution in deionized water and its documented predictable electrospinning
behavior. The electrospinning processing parameters of the electric field
strength, solution flow rate, viscosity and collector design will be investigated for
their relationship to changes in the fiber diameter and morphology. Once the
general electrospinning trends are understood, the system will be changed to the

cellulose nitrate polymer solution.

In the case of cellulose nitrate, two solvents are necessary to provide the
most stable electrospinning behavior. These solvents can be adjusted to
balance the surface tension, viscosity, and volatility of the solution. Cellulose
nitrate and cellulose nitrate functionalized with PANii or PPy will be characterized
for morphological electrical, mechanical, thermal, surface area, capillary flow,
surface chemistry and biosensor performance properties. The fiber morphology
will be characterized first, using Environmental Scanning Electron Microscopy
(ESEM). The ESEM will provide fiber size distribution data, fiber alignment
properties as well as an optical verification of pore size distribution.
Quantification of the surface area and pore size distribution will be accomplished
by nitrogen adsorption coupled with analysis using the Brunauer, Emmett, and
Teller (BET) equation. The electrospun fibers will also be investigated through

the use of thermal gravimetric analysis (TGA) to quantify the polymer

degradation temperature, verify the PANi or PPy loading and quantify any
residual solvents. Bulk chemical compositional information will be obtained by
Fourier Transform infrared spectroscopy (FTIR) and the fiber surface chemical
analysis will be performed using X-Ray Photoelectron Spectroscopy (XPS).
Electrical characterization will be completed using AC impedance
measurements. Physical properties such as capillary flow and water contact
angle will also be measured. Methods such as plasma treatment will be
employed to alter the capillary flow and water contact angle properties. The

effects of post-treatment on the fiber morphology will also be understood.

Ultimately the electrospun nanofibrous materials will be fabricated into
biosensors and compared to the commercially available existing materials. As a
result of this research, the primary material variables and process parameters
necessary to produce cellulose nitrate, and cellulose nitrate functionalized with
PANi or PPy nanofibers using the electrospinning process in order to improve the
performance of biosensors will be identified and utilized to optimize biosensor

performance.

CHAPTER 2

BACKGROUND AND LITERATURE REVIEW

Abstract

This chapter provides a description of the electrospinning process, and the
development of sensor devices, in particular biosensor devices using the
electrospinning technique. The nanofibers that are produced generally improve
the quality and performance of the device by producing an increased surface
area, controllable architecture, and specific chemical incorporation that only
electrospinning can provide. The development of models has allowed
researchers to eliminate costly steps in creating new electrospun technologies.
The electrospinning process has allowed for the creation of biosensors with lower
detection limits, increased stability, improved specificity, and enlarged linear
response regions. The electrospun biosensor is still in its infancy, but the future
is full of promise as researchers focus on specific details of the process and its

effects on the biosensor device performance.

Introduction

Nanotechnology is the study and development of materials at nanoscale
levels. It is a rapidly growing scientific discipline due to its enormous potential in
creating novel materials that have advanced applications, such as high value
sensors. This technology has tremendously impacted many different science
and engineering disciplines, such as electronics, materials science, and polymer
engineering. Nanofibers, due to their high surface area and porosity, have
applications as filter media, adsorption layers in protective clothing, and
reinforcing materials. Electrospinning has been found to be a viable technique to

produce nanofibers in an efficient manner.

Cellulose is the most abundant organic material on the surface of earth. It is
the main constituent of paper and plant cell walls. Cellulose consists of linear
glucose polymer chains, which form hydrogen bonded supramolecular structures.
Cellulose is a renewable resource, but is not soluble in water and most organic
solvents. Functionalized cellulose such as cellulose nitrate (CN) is soluble in
many common organic solvents and is important for many industries with

applications in packaging, coatings, adhesives and sensors.

Biosensors are an important application for electrospun cellulose based
materials since they offer improved surface area over existing technologies. The
improved surface area is responsible for lower the detection limit, decreasing the
detection time and improving the sensitivity of the sensors. Surface nitro (N02)
functional groups are essential for binding specific antibodies used in the

biosensor construction.

Electrospinning

In textile and fiber science related scientific literature, fibers with diameters in
the range 100 nm - 500 nm are generally referred to as nanofibers.1
Electrospinning is a unique approach using electrostatic forces to produce
nanoscale fibers. The advantages of the electrospinning process are its technical
simplicity and its easy adaptability. In general electrospun fibers have small pore
size and high surface area. There is also evidence that the relatively large static
charges in electrospun fibers can be effectively handled to produce three-

dimensional structures.2

The apparatus used for electrospinning is simple in construction, consisting of
a high voltage electric source with positive or negative polarity, a syringe pump
with capillaries or tubes to carry the solution from the syringe or pipette to the
spinneret, and a conducting collector such as aluminum. The collector can be
made of any shape according to the product requirements, such as a flat plate,
rotating drum, or patterned collector. Fibrous structures can be produced as
random, aligned or patterned networks, depending on the processing equipment
used; in particular the collector choice is crucial for producing patterned
electrospun architectures. Polymers, composites, ceramics and metals have all
been created using either solution based or melt based electrospinning.3 A

schematic of the electrospinning process is shown in Figure 2.1.

   

Syringe Pump

 

 

Figure 2.1 Electrospinning schematic showing A) the syringe, B) the Taylor
Cone, C) the stable region, D) the instability region and E) the
rotating drum style collector.

A polymer solution or melt is forced through a needle from a syringe pump to
form a pendant drop of the polymer at the tip of the capillary. A high voltage
potential is applied to the polymer solution inside the syringe through an attached
electrode, thereby inducing free charges into the polymer solution. These
charged ions move in response to the applied electric field towards the electrode
of opposite polarity, thereby transferring tensile forces to the polymer liquid.4 At
the tip of the capillary, the pendant hemispherical polymer drop forms in the
presence of an electric field; this is known as the Taylor cone.5 When the applied

potential reaches a critical value required to overcome the surface tension of the

liquid, a jet of liquid is ejected from the cone tip.‘ After the initiation from the
cone, the jet undergoes a chaotic motion or bending instability and is field
directed towards the oppositely charged collector, which collects the charged
fibers.6 As the jet travels through the atmosphere, the solvent evaporates,
leaving behind a dry fiber on the collecting device. For low viscosity solutions, the
jet breaks up into droplets, while for solutions with appropriate viscosity, it travels

to the collector as a charged fiber jets.2

Anton Forrnhals first demonstrated production of synthetic filaments using
electrostatic forces in 1934.7 The use of electrostatic forces to create fibers from
a polymer melt or solution was first known as electrostatic spraying. This name
was condensed in the 1990’s to the current name of electrospinning. It has been
shown recently that the electrospinning process is capable of producing fibers in
the submicron range.‘5 Electrospinning has increased in popularity over the last
ten years due to its flexibility to produce numerous polymeric fibers and the
consistency and ease of producing fibers in the nanoscale. These fibers
generally have higher surface area than regular fibers, and have applications in

filtration, catalysis, biomedicine, protective clothing and sensors.

Research activity on the electrospinning of nanofibers has been successful in
spinning submicron range fibers from different polymeric solutions and melts.
Variations of the collector size and shape are crucial for developing the desired
final fiber arrangement. Recently groups have used flat solid surfaces to obtain
random arrays of fibersa, while wire meshesg, rotating drums10 and patterned

electrodes11 can be used to create patterned or aligned fiber networks. In

10

conventional fiber spinning methods, the fiber is subjected to a group of tensile,
gravitational, aerodynamic, rheological, and inertial forces.4 In electrospinning,
the tensile forces created in the axial direction of the flow of the polymer by the
induced charges in the presence of an applied electric field accomplish the
production of fibers. The collector design and electric field parameters are

essential for obtaining the desired morphology of fibers.

Electrospinning is a relatively low volume operation, which typically produces
materials with a small geometric footprint; therefore the desired products should
be utilized in high value applications. The focus of this review is the exploitation
of electrospinning technology for the development of biosensor devices.
Reviews of the general electrospinning literature are readily available12‘15. A
review of nanofiber materials for biosensors is also available‘e'". lnforrnation
regarding the polymers that can be used, the processing equipment parameters

and specifically the use of these fibers in biosensor devices will be discussed.

Modeling of the Electrospinning Process

The electrospinning process can be used to produce nanoscale fibers. These
fibers have large aspect ratios, and high surface area per unit volume, which is
important for many applications, such as sensors. The electrospinning process
is very complex and the resultant fiber size is influenced by numerous processing
parameters, material variables and equipment designs. The majority of the

information regarding the electrospinning process has been obtained empirically,

11

but the complexity of the process makes empirical determination of the
parameter effects very difficult, costly and often impractical. By using a Suitable
theoretical model, the effects of parameters on the fiber diameter can be

evaluated in a more efficient manor.

The modeling data can be used to determine which parameters have the
greatest influence on the resultant fiber diameter. Special concentration on the
most influential factors can be conducted using empirical studies to determine
relationships for the resultant fiber size. Empirical observations indicate that the
smallest fiber diameters can be obtained at low flow rates, and low polymer
concentrations, due to the small amount of material available. It has been
determined that solution concentration is primarily responsible for final fiber
diameter,2 but this approach is limited to a narrow window of spinnable solution

concentrations, and it is further reduced when defect free fibers are required.

Using a model that considers the stretching of a viscous charged fluid in the
presence of an electric field could lead to a quicker approach for determining the
final jet diameter. The model predicts the final diameter of the fluid arises from a
force balance between the surface tension of the fluid and electrostatic charge
repulsion. Several approaches based on stretching have been reported.”21
Simple scaling laws have also been proposed, but they are not as universally
applicable as other more rigorous modeling techniques.”24 There have also

been efforts to predict fiber alignment based on the applied electric fields.25

In the electrospinning process, a polymer solution is in contact with a large

electric potential at the tip of a small capillary. The electrical charges that

12

develop and the fluids free surface interact with the external electric field,
resulting in the emission of a steady fluid jet that thins as it accelerates downfield
as shown in Figure 2.1. The charged jet experiences a whipping instability,
leading to bending and stretching of the jet, which is observed as loops of
increasing size as the instability grows.4 The whipping jet thins dramatically as it
travels towards the collector. The solvent evaporates during the whipping

process, and solid fibers are collected on the collector surface.

The goal of predictive models is to ascertain the approximate solution
properties and electrospinning equipment variables before entering the
laboratory. This method will reduce the amount of time, money and energy
necessary to produce electrospun fibers. To date, the electrospinning modeling
literature) is somewhat limited. There are sources, which define basic scaling
relationships for the resultant fiber diameters, describe the bending and
stretching of the jet, and dimensionless analysis to correlate known properties to
fiber diameters and morphologies. A brief review of several of the more common
models is given below. In each case a table of the data will be shown followed
by a graphical representation. A concise discussion of the advantages and

drawbacks of each model will conclude each section.

Predicting Fiber Diameter using Solution Viscosity

Information regarding the solution viscosity is required as an initial basis to

determine the resultant fiber diameter. There are numerous relations in the

13

literature for viscosity and its effect on fiber diameter.25'28 The most common is a
power law relation.” 3° More sophisticated approaches include understanding
the force balances and relating solution to electrical forces. The final approach is

to use dimensionless analysis.

The most basic predictive models have been proposed to correlate the fiber
diameter with the solution viscosity or concentration. There is a minimal
concentration necessary to produce homogenous fibers. Viscosity and
concentration are related and increasing either variable tends to produce larger
diameter fibers. If the viscosity of the solution is too low, beaded morphology will

dominate, and if the viscosity is too high, fibers will tend to agglomerate.

Most viscosity values that are reported in the literature are zero-shear values.
The experimental data show a strong dependence on viscosity for fiber
morphology.” 31' 32 Nearly every referenced work concluded that increasing the
zero-shear viscosity, using either a higher MW polymer or higher concentration,
would increase the resulting fiber radius. The only exceptions to this rule would
be in cases where the nozzle becomes partially plugged or flow is somehow
decreased so that a full size droplet is not available at the needle tip. The initial
elongational viscosity is weakly related to the zero-shear viscosity. lt
characterizes the stretching of the polymer solution, which is a transition from the
tip of the Taylor cone to the beginning of the jet. The higher the initial
elongational viscosity means a stronger stretching force must be applied to

create charged fibers. This increased force is related to the applied voltage.33

14

Investigations of viscosity as it relates to fiber diameter are numerous.29
There is a desire to quantify the relationships that connect the resultant fiber
morphology and fiber diameter to the solution’s rheological and electrical
properties and the processing parameters. The goal of Daga’s model was to
determine if there is a correlation between the zero shear viscosity and
electrospun fiber diameter for aqueous polymer solutions. Equation 2.1 shows
the power law dependence for the diameter (D) based on the zero shear

viscosity (no).

D = (127 i 6)n00.16:0.02

(Equation 2.1)

The data set that was used to evaluate the accuracy of the model is shown as
Table 2.1. The data set includes values that were determined at the average
values in Equation 2.1, that is the actual equation that was used in the modeling

is shown as Equation 2.2.

D ___ (127)n00.16

(Equation 2.2)

15

Table 2.1 Daga model data set

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Zero Daga Model
Reference Polymer Wt% Solvent VShear Actual Fiber Model Fiber Percent
rscosnty Radius (m) Radius (m) comparison
(kg/ms)
This Work CN 12 6:4 THF:DMF 12.85 2.25E-07 9.48.E-08 137%
This Work CN 8 6:4 TH FzDMF 8.45 9.50E-08 8.86.E-08 7%
This Work CN 6 7:3 EtOh:Ace 6.50 1.75E-07 8.50.E-08 106%
This Work CN 10 7:3 EtOh:Ace 15.20 6.00E-07 9.74.E-08 516%
This Work CN 8 7:3 EtOh:Ace 11.65 4.50E-07 9.33.E-08 382%
This Work CN 6 7:3 EtOh:Ace 9.25 1.65E-07 8.99.E-08 83%
This Work PVOH 10 water 5.80 1.50E-07 8.35.E-08 80%
This Work PVOH 10 water 5.80 1 .40E-07 8.35.E-08 68%
Daga, 2006 PEO 2.52 water 0.18 4.20E-08 4.79.E-08 -12%
Daga, 2006 PEO 3.74 water 0.91 6.65E-08 6.21 .E-08 7%
Daga, 2006 PEO 4.5 water 2.65 7.65E-08 7.36.E-08 4%
Daga, 2006 PEO 5.5 water 6.44 8.90E-08 8.49.E-08 5%
Daga, 2006 PEO 6.5 water 15.60 9.55E-08 9.78.E-08 -2%
Deitzel, 2001 PEG 7 water 4.00 1 .13E-07 7.86.E-08 43%
Deitzel, 2001 PEO 10 water 16.00 2.00E-07 9.82.E-08 104%

 

 

A graphical representation of the data set is shown as Figure 2.2. The solid
model prediction line shows a good fit with the PEO water data. The cellulose
nitrate data points do not fit very well, with errors of actual fiber diameter as
compared to model predicted diameters larger that 100%. The polyvinyl alcohol
data points show a closer fit, with errors that are less than 100%. The general

trend is observed for all data, that is, as the zero shear viscosity is increased, the

fiber diameter increases.

16

Daga Model

| CN 0 PVOH 0 PEG —Mode| Prediction
7.0E-O7

 

6.0E-07 .. _.__.._..___... ___ MM..__. _M ._

 

 

5.0E—07 ‘-—-—— -————— — -———-—-—— -_-_______ .__ --._.__--__ -M -MM MMMM-M

4.05.07 . M._. M .-.-..M- -MM --__ MM_ MM,

 

 

3.oeo7«— WW -- - -_

 

 

 

 

Fiber Radius (m)

2.05-07 W —-— —-

 

 

ll

 

. 1.05-07 ~~ ' - -—-—— - — -

 

 

0.0E+OO

 

0.0 5.0 10.0 15.0 20.0
Zero Shear Viscosity

Figure 2.2 Daga data chart showing a good fit for PEO water, but a poor fit for
CN and PVOH.

Another model relating the zero shear viscosity to the final fiber diameter was
proposed by McKee.30 As shown Equation 2.3, this model uses much stronger
power law dependence. The data table that was used to test the validity of this

model is shown in Table 2.2.

D = (0.05)n0(0'8) (Equation 2.3)

17

Table 2.2 McKee data table

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

, , Zero McKee Model
0 Electric Field Shear Actual
Reference Polymer Wt /o Solvent "(Strength Viscosity Fiber Model Fiber Percent
9 WA 53) Radius (m) . ,
(kg/ms) Radius (m) comparison
This Work CN 12 6:4 THF:DMF 1.67805 12.85 2.25E-07 4.E-07 -36%
This Work CN 8 6:4 THF:DMF 1.33E+05 8.45 9.50E-08 3.E-07 -62%
This Work CN 6 7:3 EtOh:Ace 1.25E+05 6.50 1.75E-07 2.E-07 -14%
This Work CN 10 7:3 EtOh:Ace 1.00E+05 15.20 6.00E-07 4.E-07 49%
This Work CN 8 7:3 EtOh:Ace 1.00E+05 11.65 4.50E-07 3.E-07 39%
This Work CN 6 7:3 EtOh:Ace 1.00E+05 9.25 1.65E-07 2.70E-07 -39%
This Work PVOH 10 water 8.00E+04 5.80 1.50E-07 1.86E-07 -19%
This Work PVOH 10 water 1.10E+05 5.80 1.40E-07 1.86E-07 -25%
Daga, 2006 PEO 2.52 water 1.54E+05 0.18 4.20E-08 1.15E-08 264%
Daga, 2006 PEO 3.74 water 1.54E+05 0.91 6.65E-08 4.22E-08 58%
Daga, 2006 PEO 4.5 water 1.54E+05 2.65 7.65E-08 9.93E—08 -23%
DaMOOG PEO 5.5 water 1.54E+05 6.44 8.90E-08 2.02E-07 -56%
Daga, 2006 PEO 6.5 water 1.54E+05 15.60 9.55E-08 4.1OE-07 ~77%
Deitzel, 2001 PEG 7 water 4.24E+04 4 .00 1.13E—07 1.38E—07 -18%
Deitzel, 2001 PEO 10 water 4.24E+04 16.00 2.00E-07 4.18E-07 -52%
Lee, 2003 PVC 15 THF 1.00E+05 1.00 3.00E-06 4.55E-08 6491 %
Lee, 2003 PVC 15 6:4 THF:DMF 1.00E+05 1.00 4.00E-07 4.55E-08 779%
Lee, 2003 PVC 15 4:6 THF:DMF 1.00E+05 1.00 5.00E-07 4.55E-08 999%
Lee, 2003 PVC 15 DMF 1.00E+05 1.00 2.00E-07 4.55E-08 339%

 

 

A graphical representation of the data set is shown as Figure 2.3. The model
shows a decent fit for all data points. The trend of increasing zero shear
viscosity leads to increased fiber diameter is observed. This particular equation

is a more accurate fit than the model proposed by Daga.29

18

McKee Model

' CN 0 PVOH 0 PEG —Model Prediction
7.0E-07

 

6.0E-07 W . W W— —. -_____,_M, . _. ____________ -

5.0E-07 «

4.0E-07 *7

3.05-07 ~

Fiber Radius (m)

ZOE-O7 ~

 

 

1.0E-07 —

 

 

 

0.0E+00

 

0.0 5.0 10.0 15.0 20.0
Zero Shear Viscosity

Figure 2.3 McKee model chat showing good correlation for PEO at low
viscosity values and a relatively poor fit for CN and PVOH.

Correlations for the final fiber diameter as a function of zero-shear viscosity
have been shown. In addition to the relationship shown above Equation 2.3, it
was proposed that the viscosity dependence of the fiber diameter was due to the
influence of polymer chain entanglements on the rheology of the solutions, and
thus the fiber diameter correlated well with the polymer concentration scaled by

the critical concentration for entanglements as shown in Equation 2.4.

19

9 (Equation 2.4)

Further work by McKee et al showed that for solutions with associative
interactions, such as hydrogen bonding, the fiber diameter increases more

strongly with polymer concentration than expected for nonassociative polymers.34

The advantage of these models is that it is relatively accurate and easy to
calculate, rely solely upon the zero shear viscosity of the solution. In cases
where crude estimations are necessary, this model would provide the appropriate
results. The disadvantage is that the data set used is small, and does not vary
much in terms of processing and solution variables. Most electric field strengths
used in the data set are similar. This model may not be applicable to filled

solutions, solutions with different conductivity, or different processing variables.

Fiber Diameter Predictions Using Force Balances

The model proposed by Fridrikh predicts a fiber diameter resulting from a
force balance between electrostatic charge repulsion and surface tension
forces.35 This is shown as Equation 2.5 below. This model relies on parameters
and data that are very difficult to obtain experimentally. For example the radius

of whipping X is only shown to hold for polymer solutions with sufficient

20

conductivity. Overall, it is clear that the model does not accurately predict the
fiber diameter using electrostatics and solution rheology alone. The assumptions

of Q/I scaling and the x value are too severe and limited to specific cases.

The model considers the whipping jet a slender viscous object. The
equations that describe the motion of the whipping jet have been successfully
applied to the linear analysis of electrostatically driven jets.” 37 These equations
are only relevant to the initial stages of the whipping phenomenon. To determine
the final diameter of the charged jet, the nonlinear instability region needs to be
modeled. The model solves equations of motion for the jet as a function of the
material properties (conductivity (K), dielectric permittivity (a), dynamic viscosity
(u), surface tension (y) and density (p)), as well as processing parameters (flow

rate (Q), electric field strength (E00) and electric current (l)). The current is

determined by fluid, equipment and processing parameters. The model assumes
that the fluid is Newtonian and neglects elastic effects due to drying of the jet.
The outside medium is assumed to be air and has a dielectric constant e, and
provides a uniform external pressure. The model also assumes that the elastic
effects and fluid evaporation are negligible. The model assumes that there is
minimal jet thinning after the initiation of the whipping instability. The thinning jet
is characterized by the jet diameter h and radius of curvature R. Analysis of the
thinning jet shows that the characteristic length scale of the thinning is the
contour length of the jet L, which is typically much larger than the radius of
curvature R of whipping, therefore h/L << 1 and NR << 1, which is generally

observed in the lab.35 These assumptions lead to the terminal radius of the jet as

21

defined by Equation 2.5 below. Spivak et al used a similar approach with similar

results.” 39

 

_ 2 3
h, = (ye Q2 2 J
I ”(2 lnX - 3) (Equation 2.5)

In Equation 2.5 above, 7 is the surface tension (g/sz), e is the permittivity
constant (8.85x10-12 (C2*s2 / kg*m3)), Q is the volumetric flow rate (ma/s), | is the
electric current (C/s) and x is a characteristic length scale defined as the radius
of curvature / jet diameter (dimensionless). Since X is difficult to measure, it is
often assumed to have a value of 100. It has been observed to vary from 10 to

1000, but the In of X varies slightly over this entire range.35

Using Equation 2.5 above, a decision can be made to either combine the Q/l
term into a constant, which typically varies as 2/3, or to substitute the values of Q
and I. In most cases since Q/l is not known and is very difficult to measure, the
2/3 assumption is reasonable. Q is simply the volumetric flow rate, but to obtain a
value for l, electrical properties of the solution must be known, as well as the fiber

jet velocity. Equation 2.6 is the relationship for electric current I (C/s).

22

I = 21r00hv + JrEkh2

(Equation 2.6)

In Equation 2.6, do is the electrical charge (C/mz), h is the jet diameter (m), U
is the jet velocity (m/s), E is the applied electric field (V/m) and K is the
conductivity of solution (6% / kg*m3). In order to solve for the electrical charge I,
the permittivity of the solution so, must be known. This is difficult to measure and
is not commme reported in the literature. The speed of the jet has been
reported in the literature but it is not common.40 Therefore this model is limited by

the assumption of the 2/3 scaling. The data set used to verify this model is

shown as Table 2.3.

23

Table 2.3

Fridrikh model data set.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fridrikh Model
Surface
Tension Actual Fiber Model Fiber Percent
Reference Polymer Wt% Solvent _(kg/s2) Radius (m) Radius (m) Comparison
This Work CN 8 6:4 THF:DMF 0.0034 8.00E-08 2.600.E-07 -69%
This Work CN 8 6:4 TH FzDMF 0.0034 9.50E-08 2.600.E-07 -63%
This Work CN 8 6:4 THF:DMF 0.0034 9.50E-08 2.600.E-07 -63%
This Work CN 8 6:4 TH F:DMF 0.0034 8.00E-08 2.600.E-07 -69%
This Work CN 8 6:4 THF:DMF 0.0034 5.50E-08 2.600.E-07 -79%
This Work CN 8 6:4 THF:DMF 0.0034 6.00E-08 2.600.E-07 -77%
This Work CN 6 7:3 EtOh:Ace 0.0026 7.00E-07 1.965.E-07 256%
This Work CN 6 7:3 EtOh:Ace 0.0026 2.00E-07 1.965.E-07 2%
This Work CN 6 7:3 EtOh:Ace 0.0026 1.75E-07 1.965.E-07 -11%
This Work CN 6 7:3 EtOh:Ace 0.0026 1.65E-07 1.965.E-07 -16%
This Work CN 6 7:3 EtOh:Ace 0.0026 1.85E-07 1.965.E-07 -6%
This Work CN 6 7:3 EtOh:Ace 0.0026 2.25E-07 1.965.E-07 15%
This Work CN 6 7:3 EtOh:Ace 0.0026 2.10E—07 1.965.E-07 7%
This Work CN 6 7:3 EtOh:Ace 0.0026 2.10E-07 1.965.E-07 7%
This Work CN 6 7:3 EtOh:Ace 0.0026 1.65E-07 1.96E-07 -16%
This Work PVOH 10 water 0.0053 1.50E-07 4.08E-07 -63%
This Work PVOH 10 water 0.0053 1.40E-07 4.08E-07 -66%
Fridrikh, 2003 PAN 8 DMF 0.0037 2.50E-07 2.83E-07 -12%
Fridrikh, 2003 PAN 10 DMF 0.0037 4.00E-07 2.83E-07 42%
Fridrikh, 2003 PCL 8 3:1 CHCl3tMeOH 0.0025 7.00E-07 1.91E-07 267%
Fridrikh, 2003 PCL 12 3:1 CHCl31MeOH 0.0028 1.10E-06 2.15E-07 413%
Fridrikh, 2003 PEG 2 water 0.0062 2.50E-07 4.73E-07 -47°/o
Daga, 2006 PEO 2.52 water 0.0075 4.20E—08 5.73E-07 -93%
Daga, 2006 PEO 3.74 water 0.0075 6.65E-08 5.73E-07 -88%
Daga, 2006 PEO 4.5 water 0.0075 7.65E-08 5.73E-07 -87%
Daga. 2006 PEO 5.5 water 0.0075 8.90E-08 5.73E-07 -84°/o
Daga, 2006 PEO 6.5 water 0.0075 9.55E-08 5.73E-07 -83%
Deitzel, 2001 PEG 7 water 0.0043 1.13E-07 3.28E-07 -66%
Deitzel, 2001 PEO 10 water 0.0037 2.00E-07 2.83E-07 -29%
Lee, 2003 PVC 15 THF 0.0035 3.00E-06 2.71 E-07 1005%
Lee, 2003 PVC 15 6:4 THF:DMF 0.0032 4.00E-07 2.47E-07 62%
Lee, 2003 PVC 15 4:6 THF:DMF 0.0036 5.00E-07 2.74E-07 82%
Lee, 2003 PVC 15 DMF 0.0043 2.00E-07 3.33E-07 -40%

 

 

The advantage of the Fridrikh model is that it predicts a terminal jet diameter,
which is a consequence of balance between normal stresses due to surface
tension and surface charge repulsion and can be determined from knowledge of
the flow rate, electric current, and the surface tension of the fluid. The 2/3 scaling

with the inverse volume charge density Q/l predicted by the model can be

24

confirmed experimentally using several concentrations and several various types
of polymer solutions, from hydrogen bonding aqueous systems of PVOH and
PEG to DMF solvent systems of PCL, PAN and ON. These provide convincing
evidence for the correctness of the model by the reasonable prediction of the dry

fiber diameter for PCL, PEG, and PAN as shown in Figure 2.4.

Fridrikh Model

- CN 0 PVOH - PAN . PCL 0 PEG x PVC -Model Prediction
3.5E-06

 

3.0E-06 "~WWWWWW . -WWWW~WW-_-L”+C_C. .,,L___~_

2.0E-06 ~——~~~W——WW W WW W W _ , ,2 ..-.__.Wfi___,..

1.5E'm _ 2--.“... __ _ __ . __s_e. *.__._ .-m on--. __...-. .__ ._.-___.~_.. _,, W

Fiber Radius (m)

1.0E-06 ~»~—~—~—-——‘ -W - __ WW ,H-u2.__ _ . ._ ____

 

 

 

5.0E—07 . -~ - , WWW WW+WW W - 2___.-..-__--_____. . -. #_

 

 

o.oe+oo . ' W s . 4
0.0020 0.0030 0.0040 0.0050 0.0060 0.0070 0.0080
Surface Tension

 

Figure 2.4 Fridrikh model chart showing correlations for PCL, PEG and PAN.

Dimensional Analysis approach

Several analytical models for electrospinning based on slender body
electrohydrodynamic theory have been previously proposed that attempt to

predict jet behavior.41 Stability analysis of electrostatically driven jets by led to the

25

development of operating maps over which different operating regimes could be
predicted, such as: whipping instability, varicose instability, and stable jetting.” 37
There have also been more detailed analyses reported that attempt to predict the
final fiber diameter of electrospinning jets.19’35'39 The main disadvantage of these
models is that they all involve internal parameters that are impossible to know a
priori and are very difficult to measure in practice. For example, Spivak et al.39
and Feng19 developed models to predict the asymptotic behavior of the jet radius
in terms of an initial jet radius, which is not known. Similarly, Fridrikh et al.35
proposed a force balance model for the final diameter of electrospun fibers, but it
implicitly contains X. the radius of whipping, which has proven difficult to measure
or predict, and introduces further approximations for mass transfer to ultimately
compare with experimental data. This model also assumes a polymer flow to
electrical current scaling estimation. As a result, these models and analyses do
not allow for a priori prediction of electrospun nanofiber morphology. They do
provide a theoretical foundation for the development of a simple scaling analysis
relating fiber morphology to spinning solution properties and controllable process

parameters.

Several empirical relationships have been proposed for the prediction of
electrospun fiber morphology from one fluid property. For example, McKee et al.
showed that the diameter of electrospun polymer fibers displays a uniform power
law scaling with the zero-shear viscosity.” 34 Shenoy et al.27 identified the
important need to be above the critical entanglement concentration of the

polymer to form uniform fibers.

26

However, later results by McKee et al.34 and Daga et al.29 show that this
result is not universal, especially for fibers spun from hydrogen-bonding systems
as well as aqueous polyethylene oxide (PEO). When comparing fiber diameter
versus zero shear viscosity for various neat polymer-solvent systems at different

operating conditions, a large variation in behavior is observed.

Feng42 identified the following dimensionless groups as governing parameters
for electrospinning in a model developed from conservation equations for mass,
momentum, and electrical charge applied to an electrospinning jet of a

Newtonian fluid:

The Peclet number (Fe) is a ratio of the electrical conduction to convective
time scales and represents the motion of the free electrical charges relative to

the jet motion.

 

(Equation 2.7)

The jet Reynolds number (Re) is a ratio of the inertial to viscous stresses and

represents the kinetic energy of the jet relative to the viscous dissipation.

27

RC = pVORO

"0 (Equation 2.8)

The Weber number (We) is a ratio of the inertial to surface stresses.

Mam

(Equation 2.9)

The dimensionless field strength (II!) is a ratio of the electrostatic to inertial

Stresses.

8E3
pvo

(Equation 2.10)

These dimensionless groups contain internal variables that cannot be directly

controlled or readily measured, namely, R0 and v0. It is standard engineering

practice to eliminate these internal variables in order to develop useful

28

correlations between dimensionless groups that contain only measurable
spinning fluid properties and measurable electrospinning parameters.” The

dimensionless stress that drives jet elongation, H1, is a ratio of the electrostatic

to the viscous stresses. This dimensionless group contains only solution

variables that are measurable and controllable processing parameters.

’2
28 ¢§
l = —2
knoL (Equation 2.11)

A second dimensionless group arises by considering a dimensionless

electrospun fiber radius Rjet/RO- Elimination of the initial jet radius using the

dimensionless groups yields the dimensionless electrospun fiber diameter.

pyR'et -
=—T’—=0h2

770

112
(Equation 2.12)

Where Oh is the Ohnesorge number, which can be used to characterize the

stability and morphology of electrospinning jets.

29

= 770
l/2
(pl/R jet) (Equation 2.13)

 

The characteristic wet radius Rjet of the electrospinning jet can be calculated
knowing the dry fiber radius Rf and the initial concentration of the polymer in

solution Ws.

 

1
Rjet = f W;
(Equation 2.14)
—2
28 (1)2
HIOh = kL2( R0 )1/2
pl’ jet (Equation 2.15)

These dimensionless groups are functions of the dielectric permittivity of

atmosphere (3); conductivity (K), density (p), zero-shear viscosity (no), and
. surface tension (y) of the electrospinning fluid; the applied electric field (E0); and
characteristic jet radius (R0) and velocity (v0) scales. The characteristic electric
field strength E0 is taken here to be the average field defined as the voltage

applied to the capillary tip divided by the working distance between the capillary

30

tip and collecting target, (Do/L. Table 2.4 below shows the values required for the

dimensionless analysis, or Helegeson's Model. The values of conductivity of
cellulose nitrate solutions were measured using a simple handheld conductivity
meter. For samples consisting of 8 wt% ON in a 60 : 40 mixture of THF : DMF,
the conductivity was determined to be 7.1 its/cm. The density of cellulose nitrate
was measured by weighing known volumes of solutions. This data was plotted to
accurately determine the density of the particular solution. It was determined that
the 8 wt% CN in a 60:40 mixture of THF:DMF had an average density of 0.995
g/cm3. Zero shear viscosity was obtained from plotting the rotational viscosity
versus concentration and extrapolating back to the y intercept. Measuring the 8
wt% ON in 60:40 mixture of THF:DMF resulted in a zero shear viscosity of 8.65

Pa*s.

31

Table 2.4 Helgeson data table

Actual
Fiber
Radius (m)

Zero Shear
Viscosity

(kg/ms)

Fiber

Solvent Radius

Reference Polymer Wt%

Percent
Pl 1

7' EtOh:Ace
7:3 EtOh:Ace

7:3
water
water
water
water
water
water
2001 water
water
Lee 2003 THF
2003 6:4 .
4:6 THF:DMF

5

6.5000

15.2000
1

9.2500

.1

-08
1.75E-07
6.00E-07

1 .65E-07
1
1.40E-07
4.20E-08
6.
7 -08
-08
9.55E-08
1.13E-07
2.
3.00E-06
4. -07
5.00E-07
00 E-O7

. -1
5.3E-11
1.5E-11
1. -
2.4511
1. -1
3 ..
W09
-10
1 -1
4. -11
1.9541
5.4542
1.1 -1
1.6E-09
1.6E-09
1.6E-09
1.6E-09

1
5163.36
7408.84

1
7567.25

3608.
127.
566.65
1611

3817

2958.18
10459.34
191.2383

550.07
467.11

1 .
1.37.E-07
5.38.E-08

9.02.E-08

1.69
1.14
1.04.
9.31

. -08
8.01.E-09
5.
1.50.E-07
4.31. -07

 

671.08

The advantage of the Helgeson dimensionless analysis model is that the
elimination of internal variables in a dimensional analysis of the electrospinning
of a Newtonian fluid provides a strong correlation between the dimensionless
fiber radius, as characterized by the Ohnesorge number, and a new
dimensionless group I11, which relates controllable operating parameters and
measurable solution spinning properties as shown in Figure 2.5. A stable jet-
operating regime has been defined above a critical Ohnesorge number of 100.
Above this value, the fibers are typically homogenous with uniform diameter,
since the operation is stable. Below Ohnesorge numbers of 100, the process is

much less stable, as and fiber inhomogenieties including beading are common.

32

Since this is a relatively recent publication (2007) the applicability and accuracy

has yet to be challenged.

Helgeson Model
I CN 0 PVOH 0 PEG x PVC —Model Prediction

 

1.E+05

1.E+04

 

1.E+03 4» -

 

.c
O

 

1.5+02

 

1.E+01

 

 

1.E+00 . T
1.0E-1 2 1.0E-11 1 .OE-10 1.0E-09 . 1.0E-08 1.0E-07

Pil

 

Figure 2.5 Helgeson chart showing good correlation for almost all data sets

The disadvantage to using the Helgeson dimensional analysis model is that
several solution parameters must be known. The values of conductivity, zero
shear viscosity and surface tension are relatively simple to obtain, but they are
often not reported in the literature. Once these values are known, the model
does provide relatively good accuracy, especially when using the Helgeson data

set.43

33

Modeling Summary

A crucial factor for producing nanoscale fibers using the electrospinning
apparatus involves knowing a priori the expected fiber diameter. Several models
have been proposed to calculate the fiber diameter based on solution and
processing variables. The models vary in complexity and accuracy. The McKee
model is a simple zero shear viscosity relation that proves to be accurate for the
data set chosen. A more detailed dimensional analysis provides a more accurate
approach to determining the fiber diameter and operating conditions through its
definition of the Ohnesorge number. Using the models presented can quickly
eliminate experimental steps and provide expected results for the solution
chosen. The modeling shows that concentration or solution viscosity is an
important variable to consider, but other factors, such as electric field strength

and conductivity must be considered to accurately estimate fiber diameters.

Biosensors

Biosensors are composed of three major parts: the sensitive biological
element, the transducer or detector element and an electronic interface. Figure

2.6 shows a schematic illustration of a biosensor device.

34

3.75

- Membrane
, Display

. , . L

 

Figure 2.6 Biosensor schematic

The biological element is a biologically derived or biomimic compound that
has highly specific interactions for the biological molecules to be detected. This
specificity makes the biosensor a powerful tool when a high selectivity is
required. The biological element is also known as the bioreceptor. Figure 2.7
shows an antibody as an example of the biological element in a fiber based
biosensor device. The biological element is closely associated with a transducer
that converts the physicochemical signal occurring during the molecular
recognition phenomenon into an electrical signal. The electrical signal is then
transferred to a device that allows for quantification of the interactions. The ability
of biomolecules to react with very low concentrations of substances allows
biosensors to be used in various applications such as the monitoring of impurities
in water, air, soil, foods, and in the detection of medically important molecules

such as hormones, sugars, and peptides in body fluids.

35

Antigen

  

, ,, Complex
Conducting
Polymer

       
 

Conducting path

 

Affixed
Antibody

Figure 2.7 Biosensor consisting of an antibody as the biological element,
conducting polymer chain as the transducer and electrospun fibers
as the matrix.

There are many applications and classifications of biosensors. The main
design criteria for the biosensor device include low cost disposable material,
reliable identification of a target molecule, availability of a suitable biological
recognition element, and high sensitivity. Detection mechanisms include,

photometric, piezoelectric, electrochemical and immunoassay. Immunoassay

36

detection mechanisms are most prevalent in the electrospinning literature and
are often detected electrochemically. Figure 2.8 shows an illustration of the

biosensor classifications.

Biosensors l

l

 

 

 

lemunoassay J [Electrochemical] | Optical l L Acoustic j
Potentiometry Fluorescence Piezoelectric
Voltammetry Infra-red Ultrasonic
Coulometry Photoionization
Conductimetry Fiber Optics

 

 

 

 

 

Figure 2.8 Classification of biosensors

Acoustic

Piezoelectric sensors utilize crystals, which undergo a phase transformation
when an electrical current is applied to them. An alternating current (AC)
produces a standing wave in the crystal at a characteristic frequency. This
frequency is highly dependent on the surface properties of the crystal, such that if
a crystal is coated with a biological recognition element the binding of a large
target analyte to a receptor will produce a change in mass which in turn alters the

resonant frequency, which gives the binding signal.

37

Polymer thin films are used extensively in acoustic biosensors. Nanoflber
films can be used as a sensing interface for thickness shear mode (TSM)
piezoelectric sensors. TSM sensors coated with nanofiber films made of poly-
lactic acid—co-glycolic acid (PLAGA) polymers were studied under various
ambient conditions and were reported to possess better sensitivities than their
thin film counterparts 44. For TSM resonators, the resolution varies linearly with
the surface area of the sensing interface. It was concluded that electrospun
polymer nanofibers would be an ideal material for this purpose. Another
publication from this group showed that piezoelectric biological sensors that were
electrospun from PLAGA offered a unique advantage 45. The PLAGA
membranes consisted of fibers with diameters of approximately 500 nm. These
fibers were deposited on the gold surface of a thickness shear mode (TSM)
micro viscoelastic sensor operating at 10 MHz. The NF thin film coated TSM
sensors were exposed to gaseous and liquid media under various ambient
conditions. The most interesting result was the different response of the NF
coated TSM sensor to hydrophilic and hydrophobic liquid loading. The aerial
density of water is larger than that of propanol, but the NF coated TSM sensors
showed more loading from propanol. This effect is very unique and presents a
very advantageous feature of NF thin films as a chemical or biochemical-sensing

interface.

38

Optical

Photometric biosensors are optical devices that are based on the
phenomenon of surface plasmon resonance. Specifically, gold and other
materials absorb laser light when thinly coated on the surface of glass. The
absorption of laser light produces electron waves called surface plasmons. This
only occurs at a specific wavelength of incident light and at a specific angle.
When target analytes are bound to the surface of gold, the angle of light changes
thus producing a measurable signal. The principle of this sensor is that any
change in the concentration of its surroundings produces a change in its optical
properties. Therefore, there is a noticeable change in the amount of radiation
reflected or emitted. Infrared, UV, and fluorescence are the spectra commonly

used. Fluorescence is the most applied method for optical biosensors.

Optical sensors are relatively new and not much work has been carried out in
this field. From first principles, the sensitivity of an optical sensor would depend
on the amount of fluorescence emitted, which in turn is proportional to the
available area of the interface. Thus by change in morphology, if porosity could
be increased, the sensitivity would also increase proportionately. Electrospun
nanofibers can promote an increase in porosity, especially if porous fibers are
created. Electrospun nanofibers would be the optimal choice in architecture and

materials for the fabrication of optical sensors.

The design and fabrication of the electrospun nanofibers is crucial for the
preparation of highly sensitive optical sensors 46. Combining the techniques of

electrospinning and electrostatic layer-by-layer adsorption, highly sensitive

39

fluorescent sensors were fabricated. A fluorescent probe was electrostatically
assembled onto the surface of electrospun nanofibrous membranes whose
fluorescence could be quenched on using extremely low concentrations of
aqueous methyl viologen (MV2+) and cytochrome C solutions. The high
sensitivity is attributed to the high surface area-to-volume ratio of the electrospun
membrane and efficient interaction between the fluorescent conjugate polymer

and the analytes 47.

Another combination technique utilized both wet spinning and electrospinning
to create virus-based microscale and nanoscale fibers. When M13 viruses were
blended with polyvinyl pyrolidone PVP fibers, an intact structure of the viruses
was formed. These fibers demonstrated that novel biomaterials can be fabricated
from a programmed organism to extend the dimension of engineered viruses into
fibers providing useful biological functions and highly sensitive catalytic functions

in the future for biomedical biosensor applications 48.

Optical chemical sensors were fabricated by electrospinning a mixture of
FAA-PM and thermally cross-linkable polyurethane latex in one solution. These
sensors showed high sensitivities due to the high surface area-to-volume ratio of
the nanofibrous membrane structures. Electrospun nanofiber membranes with
individual fiber diameters of 400 nm were prepared. A fluorescent polymer
poly(acrylic acid) - poly(pyrene methanol) (PAA-PM) was employed for the
detection of metal ions (Fe3+ and Hg2+) and 2,4-dinitrotulene (DNT). The
electrospun fibrous material exhibited a sensitivity of almost three orders of

magnitude higher than that of films of the same material ‘6.

40

A variation of the multiple component electrospinning technique utilizes
multiple syringes or core and shell nozzles to deliver different solutions, which
result in the formation of a composite fiber network. A novel dual syringe
electrospinning method was employed to prepare a composite fiber network of a
semiconducting, electroluminescent polymer MEH-PPV and mesoporous SBA-15
silica 49. The idea behind using two different solutions was to make an intenNoven
network of fibers containing a polymer and a molecular sieve material and study
the optical properties of the resulting composites. The composite fibers obtained
by electrospinning MEHPPV and SBA-15 silica together exhibit a blueshift in the
emission, which can also be altered by introducing organic moieties into the
inorganic material. The shift in emission is mainly attributed to non-aggregation of
the polymer chains, which occurs due to the polymer chain conformation. The
silica mesoporous fibers acted as a nanospacer thereby reducing interchain
interactions, resulting in the blueshift in the emission. Hence, dual syringe
electrospinning is a way to alter the optical properties of the fluorescent polymers

in fibrous form, which will have interesting applications in molecular electronics.

Electrochemical

In the electrochemical method the conductance of a solution or fibrous
network is measured, using inert electrodes, alternating current, and an

alternating null current. However, it is noteworthy that there is no electric current

41

or potential gradient and the concentration of the analyte is measured as a

function of the conductance of ions.

Recent efforts were undertaken in the production of nanofibers for
electrochemical sensors using electrospinning as shown in Table 2.5. Metal ions
such as Pd, Pt, and Au can be coated or electrodeposited onto the surface of the
fibers to improve the conductivity and resolution. The literature shows that
polymers such as polyaniline, polypyrrole, and polyamic acid have been
electrospun and successfully used as sensing interfaces 5°. For electrochemical
sensors, the resolution varied in proportion to the conductivity of the interface; in
this case, it is the number of electrons transferred that governs the sensitivity.
Therefore, nanofibers that have a very high surface area would be idealistic for

electrochemical biosensors as well.

Recently, many kinds of nanomaterials such as gold and Si02 nanoparticles
have been widely used in constructing electrochemical biosensors 5"55. Bharathi
and co-workers have demonstrated that gold nanoparticles can self-assemble
both inside the network and on the surface of the silica gel 56. The gold
nanoparticles immobilized by a silica gel three-dimensional network can act as
tiny conducting centers and facilitate the electron transfer 55. It has also been
reported that gold nanoparticles not only promote the electron transfer between
the analyte and the electrode surface but could also act as a conducting base for
the assembly of biomolecules 57. They can provide a favorable microenvironment
for the active immobilization of glucose oxidase (GOx) due to their excellent

biocompatibility.

42

immunoassay

Immunoassay biosensors are a general classification of sensors that embody
antibodies as their selective binding components. They take advantage of the
high selectivity provided by the molecular recognition of antibodies.
Immunoassay biosensors can be divided in principle into two categories: non-
labeled and labeled immunoassay biosensors. Non-labeled immunoassay
biosensors are designed so that the immunocomplex that is the antigen—antibody
complex is directly determined by measuring the physical changes induced by
the formation of the complex. In contrast, in a labeled immunoassay biosensor, a

sensitively detectable label is incorporated.

Microelectronics has sufficiently minimized the instrumentation and made the
sensing device portable, but the sensing interface is still the most crucial
component of the biosensor device. Therefore, the available surface area of the
biosensor should be optimized in order to maximize the number of active sites.
The design approach is to adapt such a morphology that is small yet possesses
a large surface area for interaction. The choice to use electrospun nanofibers is
obvious since they can provide these advantages with ease. Nanofibers have
been recently tested for biomedical applications as well as for fabrication of
nanofiber-based immunoassay biosensors 58. Specific target molecules such as
antibodies can be tailored on the surface of the nanofibers at high

concentrations. It is also possible to attach more than one antibody on the

43

nanofiber mesh, which would create a basis for multiple analyte detection

capability.

Electrochemical Immunoassay

The electrochemical immunoassay biosensor combines the reactivity of the
antibody and its antigen with the detection method of an electrochemical sensor.
Electrochemical immunoassays are biosensors constructed with antibodies as
biological elements, attached to an electrochemical transducer 59. Conductive
materials used to carry out electrochemical events include silver, gold, and

organic polymers 60’ 6‘

. Biosensors based on electrochemical immunoassays
using conductive polymers such as polyaniline, polypyrrole, polyacetylene, and
polythiophene have gained industrial acceptance 62. The conductive polymer
acts as the electrochemical transducer to convert the biological signal into an
electrical signal 63. Polyaniline in particular, has been one of the most extensively

investigated conducting polymers, due to its excellent stability in liquid form,

promising electronic properties 5“, and strong biomolecular interactions 65.

These biosensors only have two electrodes and are extremely sensitive,
robust and accurate. There are two basic fiber structures used for the production
of these biosensors, conductive fiber networks, and non-conductive fiber
networks. Figure 2.7 above shows the non-conductive fiber network and Figure

2.9 shows the conductive fiber alternative.

44

Antigen ‘

   
    

AntibodyV‘

Conductive Electrospun Fibers

Figure 2.9 Conductive nanofiber biosensor schematic.

A sensing material of hydrolyzed poly(2-(3-thienyl) ethanol butoxy
carbonylmethyl urethane) (H-PURET) was recently assembled onto the surface
of electrospun nanofibrous cellulose acetate membrane 66. The electrospun
fibers had diameters from 100 — 400 nm. High sensitivity in detecting extremely
low concentration (ppb) of methyl viologen (MV2+) and cytochrome c (cyt c) was
reported. It is believed that the high surface area to volume ratio of electrospun
nanofibrous membrane and efficient interaction between sensing material and

detected substance are responsible for this significant improvement.

There are several technologies proposed for urea detection through use of

immobilized urease. Most of them offer a narrow detection limit in either low or

45

high concentrations of urea. A urea biosensor utilizing urease immobilized on
gelatin beads was reported 67. This system was successful at achieving long
storage stability, with a half-life of 240 days. The sensors’ drawbacks were a
detection limit from 0.8 to 23 mM and a response time of 6 minutes. Another
sensing system proposed was on porous glass beads showed a linear relation
between current and urea concentration within a narrow range of 0—600 uM urea
solutions 68. A chemical field effect transistor with a pH-sensitive gate was
proposed based on the photosensitive layers of PVA. The PVA was cast using a
spin coating technique and was capable of detecting urea at concentrations of 5
to 25 mM 69. Composite electrically conductive PAn—PBMA homogenous films by
were prepared by casting, these films offered the detection limits of 3—20 mM 7°.
Urea has been immobilized on Si02 film as a pH-sensitive layer in light
addressable potentiometric sensors (LAPS) and once again had a narrow range
of detection from 5 to 15 mM. A detailed discussion of longevity for the

mentioned product was not provided".

Nanocomposite fibers of urease and PVP were prepared by the
electrospinning technique 72. The non-woven mat formed with fibers from 7 to
100 nm in diameter, has potential as a urea biosensor. The immobilized enzyme
remained active inside the polymer solution, and more importantly the reactivity
was maintained inside the electrospun non-woven mats. The electrospun
membrane acted as catalyst in the hydrolysis of various urea solutions. The
biggest improvement over prior technology is the fast response time, sensitivity

to low concentrations of urea, and a more versatile design.

46

The electrospinning method can be both convenient and efficient to produce
enzymatic electrodes for biosensors. Biocomposite nanofibers were prepared by
electrospinning glucose oxidase (GOD) with Poly(vinyl alcohol) PVA solution 73.
The immobilized glucose oxidase enzyme remained active inside the
biocomposite (PVA/GOD) membrane. The glucose biosensors fabricated using
the electrospun PVA/GOD exhibited a rapid response time (1 s) and a higher
output current (IA level) to glucose in the normal and diabetic level. The linear
response range from 1 to 10 mM and the lower detection limit (0.05 mM) of the
sensor can meet the demand in the detection of the glucose in medical

diagnosis.

Nanofibrous glucose electrodes have also been fabricated by using an
electrospun composite membrane consisting in part of polymethylmethacrylate
74. The nanofiber electrodes showed an excellent detection limit, wide linear
range response, and operational stability and were free from electrochemical
interference of co-existing species. These features provide advantages for
utilizing the electrospinning technique to obtain structures suitable for the
immobolization of biomolecules in the process of fabricating new and unique
biosensor devices. Electrospinning has also been used to prepare the
nanofibrous membranes composed of poly(vinylidene flouoride) and
poly(aminophenylboronic acid) 75. The nanofibrous membrane displayed an
excellent linear response to the detection of glucose for the concentration range

of 1 to 15mM and a very rapid response time of less than 6 seconds. The

electrospun nanofibrous membranes were subjected to interference studies

47

using uric acid, ascorbic acid, acetaminophen, fructose, and mannose. It was
determined that the interferents did not give significant overlapping current signal
during the determination of glucose. The electrospun nanofibrous mats also

provide improved reproducibility toward glucose detection and storage stability.

Nanoporous silica nanofibers have been prepared using tetramethyl
orthosilicate (TMOS) as a silica precursor, poly (vinyl alcohol) (PVA) and fi-D-

glucose 76.

The enzyme immobilized inside the mesoporous silica fibers was
horseradish peroxidase (HRP). The PVA aids in the formation of fibers since it
increases viscosity of the solution. Fibers were formed with small diameters from
100 to 200 nm. The resultant nonwoven mats exhibited extreme porosity of 2 to
4 nm, most likely due to the glucose template present in solution. The fibers with
mesoporosity showed a fourfold increase in activity as compared to the
conventional nontemplated silica samples, and threefold increase in activity as
compared to the HRP immobilized silica powders. The improved surface area,
mechanical flexibility, thermal stability, reusability and freedom of encapsulating

various enzymes make porous silica nanofibers excellent candidates for

biosensor substrates.

Biotin has been successfully incorporated into PLA nanofibers through
electrospinning without significantly changing the morphology and size of the
resulting nanofibers 77. The fibers were approximately 200 nm in diameter. The
nanoflber membranes were characterized with Electron Probe Microanalyzer
(EPMA) to confirm biotin incorporation in the fibers. Although biotin is dispersed

throughout the nanofiber membranes the distribution is heterogeneous and

48

regions of locally high biotin concentration are evident. Confocal microscopy
analysis indicates that both specific and non-specific binding occur between
streptavidin and the biotin incorporated PLA nanofiber membranes. With blocking
reagent, non-specific binding can be suppressed effectively. Fluorescence
labeled streptavidin binds along the surface of PLA-biotin fibers providing
evidence that biotin incorporated via electrospinning is available at the fiber
surface. Biosensor assay experiments confirm that the PLA nanofiber
membranes can successfully transport analyte solutions via wicking. A
biotinylated DNA probe was successfully captured by immobilized streptavidin in
a preliminary biosensor assay using an electrospun PLA nanofiber membrane as

substrate.

Nylon-6 nanofibers have been electrospun, impregnated with Au nanoclusters
to increase the conductivity and immobilized with glucose oxidase 16. The results
show that electrospun nanofibers can offer a quick response time within 130
seconds. The new biosensors can selectively and accurately detect glucose
levels with resolution of 0.113—uV/ mM. The sensors were also stable; the
membrane with immobilized enzyme gave results with less than 5% absolute

error after two months.

49

 

Table 2.5

Summary of electrospun material used for biosensor applications

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ChloroformzAcetone

 

 

 

Electric Field
Strength Flow Rate Fiber Size (nm)/
Spinning Solution (kV/cm) (mL/h) Pattern Reference
0 .
5 Wt /° PLAGA '" 1.33 -- 500 / Aligned Kwoun 2001
DMF
9 wt% CA in 2:1
AcetonezDM AC 1.25 - 1.33 -- 100 - 400/ Random Wang 2004
M13 W“ + PVP '" 2.00 0.18 - 0.36 100 - 200 / Random Lee 2004
aqueous buffer
.740 Wt.% MEH'PPV 0.83 - 0.91 -- 200 / Random Madhugiri 2003
"1 1,2 dichloroethane
18.6 wt% PAA-PM +
36.5 wt% PU in DMF 1.00 -—- 100 - 400 / Random Wang 2002
3:7 ”reasePVP '" 1.00 0.9 7 - 100 / Random Sawicka 2005
Ethanol
8 wt% PVA + GOD in
PBS Buffer 1.00 0.2 70 - 250 I Random Ren 2006
PWA +
MWCNT(PDDA) in 1.67 10 200 - 400 / Random Manesh 2008
7:3 DMFzAcetone
PVDF + PAPBA in '
7:3 DMF: Acetone 1.67 10 150/ Random Manesh 2007
Glucose+PVA+HRP
in PBS Buffer 1.33 -- 200 / Random Patel 2006
'5-3 wt% PLA +Biotin
in 3:1 1.25 - 1.5 0.3 - 0.6 150 - 5,000 / Random Li 2006

 

50

 

 

Summary

This chapter gives a brief description of electrospinning, and the development
of biosensor devices using the electrospinning technique. The nanofibers that
are produced generally improve the quality and performance of the device.
These improvements are due to the increased surface area, controllable
architecture, and specific chemical incorporation that only electrospinning can
provide. The electrospinning process has allowed for the creation of biosensors
with lower detection limits, increased stability, improved specificity, and enlarged
linear response regions. The electrospun biosensor is still in its infancy, but the
future is full of promise as researchers focus on specific details of the process

and its effects on the biosensor device performance.

51

 

References

10.

MacDiarmid, A. G.; Jones, W. E.; Norris, I. D.; Gao, J.; Johnson, A.
T.; Pinto, N. J.; Hone, J.; Han, 3.; K0, F. K.; Okuzaki, H.; Llaguno,
M., Electrostatically-generated nanofibers of electronic polymers.
Synthetic Metals 2001, 119, (1-3), 27-30.

Deitzel, J. M.; Kleinmeyer, J.; Harris, 0.; Tan, N. C. B., The effect of
processing variables on the morphology of electrospun nanofibers
and textiles. Polymer 2001,42, (1), 261-272.

Lyons, J.; Li, C.; Ko, F ., Melt-electrospinning part I: processing
parameters and geometric properties. Polymer 2004, 45, (22),
7597-7603.

Reneker, D. H.; Chun, |., Nanometre diameter fibres of polymer,
produced by electrospinning. Nanotechnology 1996, 7, (3), 216-
223.

Taylor, G., Disintegration of water drops in electric field.
Proceedings of the Royal Society of London Series a-Mathematical
and Physical Sciences 1964, 280, (138), 383.

Koombhongse, S.; Liu, W. X.; Reneker, D. H., Flat polymer ribbons
and other shapes by electrospinning. Journal of Polymer Science
Part B-Polymer Physics 2001, 39, (21 ), 2598-2606.

Formhals, A. Process and apparatus for preparing artificial threads.
1975504, 1934.

Kim, J. S.; Reneker, D. H., Polybenzimidazole nanofiber produced
by electrospinning. Polymer Engineering and Science 1999, 39, (5),
849-854.

Katta, P.; Alessandro, M.; Ramsier, R. 0.; Chase, G. G., .
Continuous electrospinning of aligned polymer nanofibers onto a
wire drum collector. Nano Letters 2004, 4, (11), 2215-2218.

Sundaray, B.; Subramanian, V.; Natarajan, T. S.; Xiang, R. 2.;

Chang, C. C.; Fann, W. S., Electrospinning of continuous aligned
polymer fibers. Applied Physics Letters 2004, 84, (7), 1222-1224.

52

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

Li, D.; Ouyang, G.; McCann, J. T.; Xia, Y. N., Collecting electrospun
nanofibers with patterned electrodes. Nano Letters 2005, 5, (5),
913-916.

Tan, 8.; Huang, X.; Wu, 8., Some fascinating phenomena in
electrospinning processes and applications of electrospun
nanofibers. Polymer Intemational 2007, 56, 1330-1339.

Li, D.; Xia, Y. N., Electrospinning of nanofibers: Reinventing the
wheel? Advanced Materials 2004, 16, (14), 1151 -1 170.

Huang, 2. M.; Zhang, Y. Z.; Kotaki, M.; Ramakrishna, S., A review
on polymer nanofibers by electrospinning and their applications in

nanocomposites. Composites Science and Technology 2003, 63,
(15), 2223-2253.

Teo, W. E.; Ramakrishna, S., A review on electrospinning design
and nanofibre assemblies. Nanotechnology 2006, 17, (14), R89-
R106.

Ramakrishna, S.; Lala, N.; Garudadhwaj, H.; Ramaseshan, R.;
Ganesh, V. K., Polymer Nanofibers for Biosensor Applications. In
Molecular Building Blocks for Nanotechnology, 2007; pp 377-392.

Mohanty, S. P. K., E., Biosensors: a tutorial review. 2006, 25, (2),
40.

Wan, Y. Q.; Guo, 0.; Pan, N., Thermo-electro-hydrodynamic model
for electrospinning process. International Journal of Nonlinear
Sciences and Numerical Simulation 2004, 5, (1 ), 5-8.

Feng, J. J., The stretching of an electrified non-Newtonian jet: A
model for electrospinning. Physics of Fluids 2002, 14, (11), 3912-
3926.

Reneker, D. H.; Yarin, A. L.; Fong, H.; Koombhongse, 8., Bending
instability of electrically charged liquid jets of polymer solutions in
electrospinning. Journal of Applied Physics 2000, 87, (9), 4531 -
4547.

Shin, Y. M.; Hohman, M. M.; Brenner, M. P.; Rutledge, G. 0.,

Experimental characterization of electrospinning: the electrically
forced jet and instabilities. Polymer 2001, 42, (25), 9955-9967.

53

 

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

Wan, Y. 0.; He, J. H.; Yu, J. Y., Experimental verification of scaling
law between current and applied voltage in electrospinning. Iranian
Polymer Journal 2006, 15, (3), 265-268.

He, J. H.; Wane, Y. Q., Allometric scaling for voltage and current in
electrospinning. Polymer 2004, 45, (19), 6731 -6734.

Carroll, C. P.; Joo, Y. L., Electrospinning of viscoelastic Boger
fluids: Modeling and experiments. Physics of Fluids 2006, 18, (5).

Bunyan, N. M; Chen, J.; Chen, |.; Farboodmanesh, S., Electrostatic
effects on electrospun fiber deposition and alignment. Polymeric
Nanofibers 2006, 918, 106-120.

Doshi, J.; Reneker, D. H., Electrospinning Process and Applications
of Electrospun Fibers. Journal of Electrostatics 1995, 35, (2-3),
1 51-160.

Shenoy, S. L.; Bates, W. D.; Frisch, H. L.; Wnek, G. E., Role of
chain entanglements on fiber formation during electrospinning of
polymer solutions: good solvent, non-specific polymer-polymer
interaction limit. Polymer 2005, 46, (10), 3372-3384.

Koski, A.; Yim, K.; Shivkumar, 8., Effect of molecular weight on
fibrous PVA produced by electrospinning. Materials Letters 2004,
58, (3-4), 493-497.

Daga, V. K.; Helgeson, M. E.; Wagner, N. J., Electrospinning of
neat and laponite—filled aqueous poly(ethylene oxide) solutions.
Journal of Polymer Science Part B-Polymer Physics 2006, 44, (11),
1 608-161 7.

McKee, M. G.; Wilkes, G. L.; Colby, R. H.; Long, T. E., Correlations
of Solution Rheology with Electrospun Fiber Formation of Linear
and Branched Polyesters. Macromolecules 2004, 37, (5), 1760-
1767.

Deitzel, J. M.; Kleinmeyer, J. D.; Hirvonen, J. K.; Tan, N. C. B.,
Controlled deposition of electrospun poly(ethylene oxide) fibers.
Polymer 2001, 42, (19), 8163-8170.

Fong, H.; Chun, l.; Reneker, D. H., Beaded nanofibers formed
during electrospinning. Polymer 1999, 40, 4585-4592.

54

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

Drew, 0; Wang, X.; Samuelson, L. A.; Kumar, J., The Effect of
Viscosity and Filler on Electrospun Fiber Morphology. Journal of
Macromolecular Science, Part A 2003, 40, (12), 1415 - 1422.

McKee, M. G.; Hunley, M. T.; Layman, J. M.; Long, T. E., Solution
rheological behavior and electrospinning of cationic
polyelectrolytes. Macromolecules 2006, 39, (2), 575-583.

Fridrikh, S. V.; Yu, J. H.; Brenner, M. P.; Rutledge, G. C.,
Controlling the fiber diameter during electrospinning. Physical
Review Letters 2003, 90, (14).

Hohman, M. M.; Shin, M.; Rutledge, G.; Brenner, M. P.,
Electrospinning and electrically forced jets. l. Stability theory.
Physics of Fluids 2001, 13, (8), 2201-2220.

Hohman, M. M.; Shin, M.; Rutledge, G.; Brenner, M. P.,
Electrospinning and electrically forced jets. ll. Applications. Physics
of Fluids 2001, 13, (8), 2221-2236.

Spivak, A. F .; Dzenis, Y. A., Asymptotic decay of radius of a weakly
conductive viscous jet in an external electric field. Applied Physics
Letters 1998, 73, (21), 3067-3069.

Spivak, A. F .; Dzenis, Y. A.; Reneker, D. H., A model of steady
state jet in the electrospinning process. Mechanics Research
Communications 2000, 27, (1 ), 3742.

Leon, M. B.; Harold, G. C.; Juan, P. H., Direct measurement of fluid
velocity in an electrospinning jet using particle image velocimetry.
Journal of Applied Physics 2007, 102, (9), 094308.

Saville, D. A., ELECTROHYDRODYNAMICS:The Taylor-Melcher
Leaky Dielectric Model. Annual Review of Fluid Mechanics 1997,
29, (1), 27-64.

Feng, J. J., Stretching of a straight electrically charged viscoelastic
jet. Journal of Non-Newtonian Fluid Mechanics 2003, 116, (1 ), 55-
70.

Helgeson, M. E.; Wagner, N. J., A correlation for the diameter of
electrospun polymer nanofibers. Aiche Jouma/ 2007, 53, (1), 51-55.

Kwoun, S. J. L., R.M.; Han, B. and K0, F.K. In Polymer Nanofiber

Thin Films for Biosensor Applications, IEEE 27th Annual Northeast
Bioengineering Conference, 2000; 2000; pp 9-10.

55

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

Kwoun, S. J. L., R.M.; Han, B. and Ko, F.K., A Novel Polymer
Nanofiber Interface for Chemical and Biochemical Sensor
Applications. In Modeling and Simulation of Microsystems, 2001; pp
338-341.

Wang, X.; Drew, C.; Lee, S. H.; Senecal, K. J.; Kumar, J.;
Samuelson, L. A., Electrospun Nanofibrous Membranes for Highly
Sensitive Optical Sensors. Nano Letters 2002, 2, (11), 1273-1275.

Wang, X. K., Y-G.; Drew,C., Ku,B-C., Kumar,J., Samuelson L.A.,
Electrostatic Assembly of Conjugated Polymer Thin Layers on
Electrospun Nanofibrous Membranes for Biosensors. Nano Letters
2004, 4, (2), 331 -334.

Lee, S. W.; Belcher, A. M., Virus-Based Fabrication of Micro- and
Nanofibers Using Electrospinning. Nano Letters 2004, 4, (3), 387-
390.

Madhugiri, S.; Dalton, A.; Gutierrez, J.; Ferraris, J. P.; Balkus, K. J.,
Electrospun MEH-PPV/SBA-15 Composite Nanofibers Using a
Dual Syringe Method. Journal of the American Chemical Society
2003, 125, (47), 14531-14538.

Jianrong, C.; Yuqing, M.; Nongyue, H.; Xiaohua, W.; Sijiao, L.,
Nanotechnology and biosensors. Biotechnology Advances 2004,
22, (7), 505-518.

Qhobosheane, M.; Santra, S.; Zhang, R; Tan, W. H., Biochemically
functionalized silica nanoparticles. Analyst 2001, 126, (8), 1274-
1278.

Hilliard, L. R.; Zhao, X. J.; Tan, W. H., Immobilization of .
oligonucleotides onto silica nanoparticles for DNA hybridization
studies. Analytica Chimica Acta 2002, 470, (1 ), 51-56.

Yu, A.; Liang, Z.; Cho, J.; Caruso, F ., Nanostructured
Electrochemical Sensor Based on Dense Gold Nanoparticle Films.
Nano Letters 2003, 3, (9), 1203-1207.

Xiao, Y.; Ju, H. X.; Chen, H. Y., Hydrogen peroxide sensor based
on horseradish peroxidase-labeled Au colloids immobilized on gold
electrode surface by cysteamine monolayer. Analytica Chimica
Acta 1999, 391, (1), 73-82.

56

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

Chen, Z. J.; Ou, X. M.; Tang, F. 0.; Jiang, L., Effect of nanometer
particles on the adsorbability and enzymatic activity of glucose
oxidase. Colloids and Surfaces B-Biointerfaces 1996, 7, (3-4), 173-
1 79.

Bharathi, S.; Nogami, M.; lkeda, 8., Novel electrochemical
interfaces with a tunable kinetic barrier by self-assembling
organically modified silica gel and gold nanoparticles. Langmuir
2001,17,(1),1-4.

Zhang, S.; Wang, M; Niu, Y.; Sun, C., Immobilization of glucose
oxidase on gold nanoparticles modified Au electrode for the
construction of biosensor. Sensors and Actuators B: Chemical
2005, 109, (2), 367-374.

Huang, J.; Virji, 8.; Weiller, B. H.; Kaner, R. B., Polyaniline
Nanofibers: Facile Synthesis and Chemical Sensors. Journal of the
American Chemical Society 2003, 125, (2), 314-315.

Tahir, Z. M.; Alocilja, E. C.; Grooms, D. L., Polyaniline synthesis
and its biosensor application. Biosensors & Bioelectronics 2005,
20, (8), 1690-1695.

Brunelle, S., Electroimmunoassay technology for foodbome
pathogen detection. Detection Technology 2001, 55-62.

kim, J. H., Cho, J.H., Cha, G.S., Conductimetric membrane strip
immunosensor with polyaniline-bound gold colloids as signal
generator. Biosensors & Bioelectronics 2000, 14, 907-915.

Hoa, D. T. K., T.N.; Punekar, N.S.; Srinivasa, R.S.; Lal, R. and
Contractor, A.Q., Biosensor Based on Conducting Polymers.
Analytical Chemistry 1992, 64, 2645-2646.

Chiang, J. C.; Macdiarrnid, A. G., Polyaniline - Protonic Acid Doping
of the Emeraldine Form to the Metallic Regime. Synthetic Metals
1986, 13, (1-3), 193-205.

Syed, A. A. D., M. K., Polyaniline - A novel Polymeric Material.
Talanta 1991, 38, (8), 815-837.

Imisides, M. D. J. R. W., G.G., Microsensors based on conducting
polymers. Chemtech1996, 26, (5), 19-25.

Wang, X., Kim,Y-G., Drew,C., Ku,B-C., Kumar,J., Samuelson, L.A.,
Electrostatic Assembly of Conjugated Polymer Thin Layers on

57

67.

68.

69.

70.

71.

72.

73.

74.

75.

Electrospun Nanofibrous Membranes for Biosensors. Nano Letters
2004, 4, (2), 331-334.

Srivastava, P. K.; Kayastha, A. M.; Srinivasan, Characterization of
gelatin-immobilized pigeonpea urease and preparation of a new
urea biosensor. 2001, 34, (Pt 1), 55-62.

Yoneyama, K.; Fujino, Y.; Osaka, T.; Satoh, l., Amperometric
sensing system for the detection of urea by a combination of the
pH-stat method and flow injection analysis. Sensors and Actuators
B-Chemical 2001, 76, (1 -3), 152-157.

Sant, W.; Pourciel, M. L.; Launay, J.; Do Conto, T.; Martinez, A.;
Temple-Boyer, P., Development of chemical field effect transistors
for the detection of urea. Sensors and Actuators B—Chemical 2003,
95, (1-3), 309-314.

Castillo-Ortega, M. M.; Rodriguez, D. E.; Encinas, J. C.;
Plascencia, M.; Mendez-Velarde, F. A.; Olayo, R., Conductometric
uric acid and urea biosensor prepared from electroconductive
polyaniline-poly(n-butyl methacrylate) composites. Sensors and
Actuators B-Chemical 2002, 85, (1-2), 19-25.

Seki, A.; lkeda, S.; Kubo, l.; Karube, l., Biosensors based on light-
addressable potentiometic sensors for urea, penicillin and glucose.
Analytica Chimica Acta 1998, 373, (1), 9-13.

Sawicka, K.; Gouma, R; Simon, S., Electrospun biocomposite
nanoflbers for urea biosensing. Sensors and Actuators B-Chemical
2005, 108, (1 -2), 585-588.

Ren, G. L.; Xu, X. H.; Liu, 0.; Cheng, J.; Yuan, X. Y.; Wu, L. L.;
Wan, Y. Z., Electrospun poly(vinyl alcohol)/glucose oxidase
biocomposite membranes for biosensor applications. Reactive &
Functional Polymers 2006, 66, (12), 1559-1564.

Manesh K.M., K. H. T., Santhosh P., Gopalan A.l., Lee K.P., A
novel glucose biosensor based on immobilization of glucose
oxidase into multiwall carbon nanotubes-polyelectrolyte-loaded

electrospun nanofibrous membrane. Biosensors & Bioelectronics
2008,23, (6), 771-779.

Manesh, K. M.; Santhosh, P.; Gopalan, A.; Lee, K. P., Electrospun
poly(vinylidene fluoride)/poly(aminophenylboronic acid) composite
nanofibrous membrane as a novel glucose sensor. Analytical
Biochemistry 2007, 360, (2), 189-195.

58

76.

77.

Patel, A. C.; Li, S. X.; Yuan, J. M.; Wei, Y., In situ encapsulation of
horseradish peroxidase in electrospun porous silica fibers for
potential biosensor applications. Nano Letters 2006, 6, (5), 1042-
1046.

Li, D. P.; Frey, M. W.; Baeumner, A. J., Electrospun polylactic acid

nanofiber membranes as substrates for biosensor assemblies.
Journal of Membrane Science 2006, 279, (1-2), 354-363.

59

CHAPTER 3

ELECTROSPINNING POLY(VINYL ALCOHOL) PVOH

Abstract

Submicron poly(vinyl alcohol) (PVOH) fiber mats were prepared by
electrospinning of aqueous PVOH solutions in 6— 12 wt% concentration. Fumed
silica was added to the electrospinning solution as nanoscale reinforcements for
the electrospun fibers. Fiber morphology was observed using electron
microscopy and effects of electrospinning parameters including electric voltage,
separation distance, flow rate and solution parameters such as concentration
were evaluated. Results showed that, when PVOH was electrospun at larger
separation distances, the fiber diameter variation increased. Altering the solution
flow rate can also change the morphological structure. At high voltages above
10kV, electrospun PVOH fibers exhibited a broad diameter distribution. When the
solution concentration was increased, the morphology was changed from beaded
fiber to uniform fiber and the average fiber diameter could be increased from 250
nm to 1200 nm. It was also found that additions of silica had significant effects on

the fiber diameter and the morphology of electrospun PVOH fibers because of

60

the different solution conductivity, surface tension and viscosity. Thermal and

mechanical properties of the electrospun PVOH mats were also measured.

Introduction to Electrospinning Poly(vinyl alcohol)

Poly(vinyl alcohol) (PVOH) is a semi-crystalline, hydrophilic polymer with
good chemical and thermal stability.1 A number of grades of PVOH are
commercially available, which can be divided into two types: fully hydrolyzed and
partially hydrolyzed PVOH, which depend on the amount of acetate groups left in
the backbone.2 PVOH is highly biocompatible, is non-toxic and it dissolves easily
in water. 3 These properties have led to the use of PVOH in a wide range of
applications in medical, cosmetic, food, pharmaceutical and packaging
industries. PVOH has been electrospun for applications such as drug delivery
systems, contact lenses and artificial organs. It is also used in more traditional
applications, such as adhesives, paper coatings, and stabilizers.4 Recently, Ding
et al. 5 have used electrospinning as a processing technique to produce
nanoscale fibrous aggregates. The molecular weight of the PVOH has significant
effects on the solution properties, including the viscosity, surface tension and
electrical conductivity. 6 These solution properties can have drastic impacts on
the electrospinning process. 7 Experiments have been conducted at several
molecular weights and solution concentrations to highlight the range of structures

that may be produced by electrospinning." 8

61

Electrospinning PVOH over a range of pH values has also been performed.9
PVOH can also be mixed with inorganic materials such as silver to make
composite nanofibers with antimicrobial activity.10 PVOH can also be mixed with

silica for use as a silica wire precursor or as composite fibers.” 12

The need for nanofibrous materials with high surface area, small pore size,
and the production of patterned three - dimensional structures has increased the
interest in nanofibers. While various groups have conducted theoretical
modeling studies of the electrospinning process, other groups have focused on
efforts to characterize the structure and morphology of nanofibers as a function
of process parameters and material characteristics. The successful production of
nanofibers by electrospinning resulted from a balance of the electrostatic forces

and the viscoelastic behavior of the polymer.

The structure and properties of electrospun fibers were influenced by
processing parameters, such as solution feed rate, applied voltage, nozzle to
collector distance, spinning environment; and material properties, such as
solution concentration, viscosity, surface tension, conductivity, and solvent vapor
pressure. Substantial work has been done to characterize the properties of fibers
as a function of processing and material parameters. The goal of this
experimental work was to compare the results obtained varying the solution and
processing variables to published reports. PVOH dissolved in water served as a
simple model system to understand the electrospinning process. Varying the
parameters in a controlled pattern provided results similar to the trends shown in

Table 3.1. Increasing viscosity and flow rate provided more polymeric material to

62

undergo the electrospinning process, resulting in increased fiber diameters.
When solvent volatility, solution conductivity and electric field strength were
increased, the fiber diameters decreased. Upper and lower limits exist for all
solution and processing variables and these values will be explained in further

detail.

Table 3.1 Effect of increasing the electrospinning variables on the resultant

 

 

 

 

 

 

 

 

 

 

 

fiber morphology.
Variable Increasing Fiber Diameter Morphology
Electric Field Strength Decrease Less Homogenous
Polymer Flow Rate Increase Less Homogenous
Solution Viscosity Increase More Homogenous
Solution Conductivity Decrease More Homogenous
Solvent Volatility Decrease Less Homogenous
Experimental

The driving force of the charged jet was provided by two high voltage
sources. The high voltage sources were used that were capable of generating up
to 30 kV DC of potential difference, and the setup was designed to be able to run
on positive or negative polarity or a combination. Typically a total of 8 — 12 W
was applied over a distance of 8 to 12 cm to produce uniform fibers. This
research proved a combination of positive polarity at the syringe and negative
polarity at the collector yielded the most predictable fiber morphology. The

negative polarity at the collector typically maintained at a constant 1 kV.

63

Adjusting the flow rate of the fluid and the magnitude of the electric field
controlled the spinning rate, which influenced the resultant fiber size. The
syringe pump rate was varied from 0.01 mI/hr up to 4.0 mI/hr. The needle
diameter could have also been altered, which would change the initial fiber jet
size, which would ultimately affect the final fiber diameter, but for these

experiments a 20—gauge needle tip was employed.

Materials

Poly(vinyl alcohol) (PVOH) was donated by the Celanese Corporation. 10 wt
% solutions of Celvol 425 PVOH were prepared in deionized water by vigorous
mixing for 12 hours at 90°C. This PVOH solution had 96.5 % degree of
hydrolysis and 580 cps viscosity, which was determined using a Brookfield
viscometer. Pendant drop shape analysis of this solution, in this research work,
yielded an average surface tension of 53.4 mN/m. Fumed silica (Cab-O-Sil M5)

was purchased from Cabot Corporation.

Morphology

An Electroscan 2020 Environmental Scanning Electron Microscope (ESEM)
was used to obtain sample images. The samples were first mounted on ESEM
pedestals, then sputter coated with gold. The samples were imaged at 15 W

from 500 x to 10,000 x magnification to obtain alignment and size distribution
64

 

information. Alignment efforts were made so that the electrospun fibers could be
used as fillers in composite materials. Preliminary experiments have shown that

a high degree of alignment is possible.

Transmission electron microscope (TEM) (JEOL 2100FEF 200 kV field
emission) was used to obtain images for the silica-loaded samples. The goal
was to determine how effective the silica loading was, the dispersion of silica

particles can be seen in the electrospun fibers.

Thermal Properties

A TA Instruments 2950 TGA was used to evaluate the thermal properties of
the electrospun materials. The goal of the thermal analysis was to determine the
degradation temperature, if any residual water was present in the system and the

residue values.

Mechanical Properties

A TA instruments 2980 Dynamic Mechanical Analyzer (DMA) was used to
evaluate the mechanical properties of the electrospun materials. The goal was to
determine the modulus for the neat electrospun PVOH and silica filled
electrospun PVOH fibers. To date, accurate and reliable mechanical property
analysis has not been fully realized. The initial measurements obtained in this

research were compared to published literature values.
65

Results and Discussion

Morphology

The electrospinning setup was tailored to provide nanofiber mats that were
either aligned or randomly oriented. Aligned mats could be produced in several
ways, including using high speed rotating collectors, a parallel array of collectors,
or auxiliary electric fields.13 One of the objectives of this research was to be able
to reproducibly produce aligned electrospun fibers of uniform diameter to serve
as the basis of the sensors. Electrospun fibers are highly charged and undergo a
chaotic ‘whipping’ motion between the needle tip and the collector due to their
interaction with the electric field. A series of experiments were organized in
which the collector and the location of the field were varied in order to reduce the
motion of the fibers and collect an ordered array. This involved collection on a
flat plate, rotating drum and parallel electrode “C" shaped collector. Aligned fibers
were demonstrated by harvesting the fibers that spanned the gap of a “C”
shaped collector as shown in Figure 3.1. In Figure 3.1, the polymeric solution
was loaded in a glass syringe fitted with a metal needle, which was attached to
the syringe pump. The syringe pump supplied a constant flow of solution through
the needle. At the needle, a high voltage (positive) connection is made to the
high voltage power supply. The positively charged fibers were deposited on a
collector that was either grounded, or held at negative potential. Collecting the
fibers across a void gap enabled the fibers to be easily gathered for future

processing and evaluation.
66

Syringe Pump

 

Figure 3.1 Electrospinning schematic showing (A) the polymer solution in the
syringe, (B) the Taylor Cone at the needle tip, (C) the stable region
of the charged jet, (D) the unstable region of the charged polymer
jet and (E) the C-shaped void gap collector

The collector can be replaced with various geometrical designs. The designs
and orientations of the collectors affect the fiber deposition process. A schematic

illustration of the collectors used is shown as Figure 3.2.

67

 

 

 

 

 

 

 

Figure 3.2 Various collectors used in the electrospinning of PVOH. A) A flat
plate of aluminum, B) a rotating steel drum with a slot milled in the
surface, C) a parallel set of copper electrodes.

Using a flat plate (Figure 3.2 (A)) as a collector provided a simple collection
device that could be used to provide morphology information easily. The flat
plate collector was used to produce a random array of fibers. This technique
provided a facile method to understand the effects of altering the solution and
electrospinning parameters on the fiber morphology. Once a satisfactory set of
parameters was chosen, a rotating drum was added to improve the fiber
alignment (Figure 3.2 (B)). Difficulties were encountered using the rotating drum
to obtain aligned fibers. It was thought that the high speed generated turbulence
in the air, which forced the fibers into an undesirable unaligned pattern. As a

result, the parallel electrode design (Figure 3.2 (C)) was employed. This design
68

 

allowed the fibers to span the gap between the top and bottom electrode, which
resulted in a high degree of alignment. A comparison of fibers collected from a
various styles of collectors is shown in Figure 3.3. The solution parameters were
adjusted to final values of 10 wt% PVOH in water. The electrospinning setup
was operated with 10 kV potential difference, a separation distance of 8.0 cm

(between the needle tip to collector plate) and a syringe pump rate of 1.0 ml/h.

 

Figure 3.3 ESEM comparison of electrospun PVOH. A) Stationary collector,
B) Rotating drum collector with slot, C) Stationary parallel electrode
collector.

As Figure 3.3 shows, a drastic difference in alignment was observed by
changing the collector design. The stationary flat plate (Figure 3.3 (A)) provided
a random distribution of electrospun fibers. Alignment of fibers was improved in
the rotating drum setup (Figure 3.3 (B)), but the highest degree of alignment was

obtained in the parallel electrode setup (Figure 3.3 (C)).

69

The goal of fiber alignment was essential since the fibers were to be
incorporated into composites. A 90 mm rotating drum that was capable of 2500
rpm was used as the rotating collector. This provided a speed of 22 m/s at the
face of the collector. It was determined that the rotating drum collector provided
the highest throughput of electrospun fibrous material, but the collection of
material was difficult. To alleviate this problem, a 25 mm slot was milled in the
surface of the drum. This allowed fibers to span the gap as the drum was
rotating. The gap allowed fibers to be easily transferred to a different backing for
testing and evaluation. The gap also produced a fan effect, which caused too
much interference and instabilities in the air to obtain truly aligned fibers. As the
lightweight polymer jet approached the surface of the rotating drum, it was forced
into a random patter as a result of the wind. The platform that was used to
mount the motor, axle and collector was also vibrating erratically at high speed,

which was thought to further increase the instability of the process.

From the idea of the slotted rotating drum, the parallel electrode evolved. The
idea was to use the open gap that was advantageous to the rotating drum, but
hold the device stationary. The parallel electrode produced very highly aligned
fiber mats, but the throughput was extremely limited. At this point it was
determined that operating the collector at a slight negative potential, the fiber
morphology and throughput would improve. This was due to the fact that the
positively charged fibers that deposited on the collector built up charge and
would repel incoming positively charged fibers. By adding a negative charge at

the collector the positive charge build up would be eliminated and the aligned

70

network could be realized. The morphology of the optimized aligned parallel

electrode setup is shown in Figure 3.4.

 

Figure 3.4 ESEM micrograph of electrospun 10 wt% PVOH fibers. 10 cm
Separation distance, 500 X magnification, scale bar shown is 100
micrometers

lmagePro software 1‘ was utilized to aid in the fiber diameter and alignment
analysis. ESEM images were used with lmagePro to count and size the
individual fibers. The classical method of viewing the samples from the top down
was chosen and it was assumed that the fibers were circular in cross section. 15
The purpose of the top down method is for easy alignment characterization and

fiber size estimation.

71

Fiber Size Distribution

An important advantage of the electrospinning process is that the device
parameters in conjunction with the solution properties could be tailored to
accurately produce fibers of the desired size. These experiments have shown
that altering the solution concentration was an effective method to alter the fiber
diameters. Several variables have been explored to determine their effects on
the resultant fiber size. it was expected that increasing the electric field strength
would induce more instability, thus reducing the fiber diameter. Increasing the
polymer flow rate would increase the amount of material available and increase
the fiber diameter. Increases in viscosity or concentration would also provide
larger diameter fibers. The morphology of the fibers is also dependent upon the
volatility of the solvent. Highly volatile solvents, such as THF are difficult to
continuously electrospin, due to rapid solvent evaporation, but have been known
to result in highly porous fiber structures.16 In order to accurately determine the
fiber sizes, a higher magnification image is required. An example of a high

magnification image used for fiber size analysis is shown in Figure 3.5.

72

 

Figure 3.5 High magnification (2,000X) ESEM image of electrospun PVOH.
The scale bar is 25 pm.

Effect of Electric Field Strength

Baumgarten observed that increasing the applied voltage resulted in a
doubling of fiber length with small changes in fiber diameter.17 Megelski et al.
investigated the fiber diameter dependence on applied voltage using polystyrene '
(PS). The PS fiber size decreased from about 20 pm to 10 pm with an increase
in voltage from 5 KV to 12 KV, while there was no significant change observed in
the pore size distribution.18 These results concur with the interpretation by
Buchko et al., who observed a decrease in the fiber diameter with an increase in
the applied field while spinning silk -like polymer fiber (SLPF) with fibronectin
functionality.19 Generally, it has been accepted that an increase in the applied

voltage increases the deposition rate due to higher mass flow from the needle tip.

73

Jaeger et al. used a two-electrode setup for electrospinning by introducing a
ring electrode in between the nozzle and the collector.20 The ring electrode was
set at the same potential as the electrode immersed in the polymer solution. This
setup was thought to produce a field-free space at the nozzle tip to avoid
changes in the shape of the jet-initiating surface due to varied potential. Although
this setup reduced the unstable jet behavior at the initiation stage, bending
instability is still dominant at later stages of the process, causing a chaotic motion
of the jet before depositing itself as a non-woven matrix on the collector. Deitzel
et al. experimented with a new electrospinning apparatus by introducing eight
copper rings in series in between the nozzle and the collector for dampening the
bending instability.7 The nozzle and the ring set were subjected to different
potentials (ring set at a lower potential) of positive polarity, while the collector
was subjected to a negative polarity. The idea behind this setup was to change
the shape of the macroscopic electric field from the jet initiation to the collection
target in such a way that the field lines converge to a center line above the
collection target by the applied potential to the ring electrodes. The authors have
suggested that the bending instability of fibers was dampened by the effect of the
converging field lines producing straight jets. Controlled deposition helps to
produce specific deposition patterns and also yarn like fibers. Deitzel et al.
investigated the spinning of polyethylene oxide (PEO) in aqueous solution using
multiple electric fields, which resulted in fibers deposited over a reduced area
due to the dampening of bending instability. The multiple field technique was also

shown to produce fibers of smaller diameter than the conventional

74

electrospinning method, which consists only of a positive needle tip and

grounded collector.

The structure and morphology of electrospun fibers was easily affected by the
nozzle to collector distance because of their dependence on the deposition time,
evaporation rate, and whipping or instability interval. Buchko et al. examined the
morphological changes in SLPF and nylon electrospun fibers with variations in
the distance between the nozzle and the collector screen. They showed that
regardless of the concentration of the solution, decreased nozzle to collector
distance produced wet fibers and beaded structures. SLPF fiber morphology
changed from round to flat with a decrease in the nozzle collector distance from 2
cm to 0.5 cm.19 This result showed the effect of the nozzle to collector distance
on fiber morphology. The work also showed that aqueous polymer solutions
require more distance for dry fiber formation than systems that use highly volatile
organic solvents. Megelski et al. observed that decreasing the nozzle to collector
distance resulted in bead formation, while preserving the ribbon shaped

morphology of PS fibers.18

Several studies specifically focused on using PVOH and understanding the
effect of electric field strength.2' 5' 21 Lee et al conducted a very thorough study of
the PVOH system and concluded that the diameter of the electrospun fibers
decreased with an increase in the electric field strength, and the diameters also
decreased with decreasing the solution concentration.8 A summary of these
results is shown in figure 6. The data shown in Figure 3.6 was generated using a

lower molecular weight PVOH of 54,400 grams per mole as compared to 65,000

75

grams per mole PVOH used in this study. It was determined that higher

molecular weight samples follow the same trend.8

Final Fiber Radius vs Concentration (54,400 MW)

 

 

 

 

 

 

360.00
340.00 .. -
//

320.00 . W . W~
E 300.00 « //
m
.2 E’//
3 280.00 «I _ W ~ W-
a:
3 260.00 W - ~ W W W W W ‘—
u_ ____

240.00

220.00 ..

200.00 . . f

7% 8% 9% 10% 11 % 12% 13%
Concentration PVOH (wt%)

Figure 3.6 Fiber size as a function of electric field strength and concentration
for 54,400 MW PVOH. Increases in electric field strength reduce
fiber diameters.

76

Fiber Diameter Summary

Three different electric field strength experiments have been employed to
examine the effect on fiber morphology. The applied voltage was maintained
constant at 10 kilovolts (kV), while the separation distance was varied from 8 to
12 cm. In all cases the applied voltage was 10 W, the syringe pump rate was
0.1 ml/h, the collection time was 1 hour, and the solution was 10 wt% PVOH in
water. Variation in the electric field strength resulted in fiber size and relative
fiber alignment differences. There is a broad range of fiber diameters for the 8
and 12 cm separation distance and the smallest variation of fiber diameter
occurred at a separation distance of 10 cm. A summary of the results is shown in
Table 3.2. As expected, the 8 cm separation distance produced fibers with the
largest diameter, but on average, the 10 cm separation distance produced the
smallest diameter fibers. The 12 cm separation distance was expected to
produce fibers with the smallest diameters, but they were found to be similar in
size to the 10 cm. The fibers collected at 12 cm had undergone the most
instability, resulting in a wide range of fiber diameters. The alignment of fibers
collected at 12 cm indicated that they are the most random, showing that the
instability region was not controlled well; therefore fiber size did not agree with

expected results.

77

Table 3.2

in water.

Separation distance affects the fiber diameter and alignment.
Applied voltage 10 kV, syringe pump rate 0.1 ml/h, 10 wt% PVOH

 

 

 

 

 

Separation Distance 8 cm 10 cm 12 cm
Average Diameter (nm) 599 355 380
Standard Dev. Dia. (nm) 116 43 73
Fibers within 10 deg. Of 93 80 69
average (%)

Standard Dev. Angle 6.5 9.5 14
(deg)

 

 

 

 

 

 

The fiber morphology was only slightly different when comparing the ESEM
images. As shown Figure 3.7, the 10 cm separation distance produced the
highest number of fibers in the same collection time. The 8 cm separation
distance produced the largest diameter fibers. This was due to the limited space
for the whipping phenomenon to occur. The whipping phenomenon was the part
of the process responsible for the rapid fiber thinning. The 12 cm separation
distance produced the widest range of fiber diameters with a lower relative ‘
alignment. This is a result from collecting fibers from a separation distance that
is too large. The whipping phenomenon, which is also the instability region,

The fibers

dominates and the fibers collected are various in size and shape.

that were collected from a separation distance of 10 cm were the most uniform in

78

size and shape due to a proper balance of electric field strength and solution

properties.

   

Figure 3.7 ESEM images of electrospun PVOH fibers, showing variation with
separation distance A) 8 cm, B) 10 cm, and C) 12 cm. Applied
voltage 10 kV, pump rate 0.1 ml/h, 10 wt% in water

The average fiber diameter of electrospun 10 wt % PVOH ranges from 300 to
450 nm. This was determined by viewing samples in the ESEM coupled with
image analysis. Higher magnification is crucial for accurately determining the
fiber size and alignment. A series of high magnification images were required to

obtain statistically significant measurements.

An experimental protocol was developed to understand the effect of each
variable (separation distance, electric field strength and solution flow rate) on the
resultant fiber diameter. In order to ascertain which variable has the most
significant effect on fiber morphology, all variables were held constant except the
variable of interest. A larger separation distance at a fixed applied voltage will

result in a lower electric field strength, which should have increased the fiber

79

diameter. Samples were prepared at separation distances of 8, 10 and 12 cm. A

summary table of the variables and the resultant fiber diameters is shown as

Table 3.3. The variable of interest for each data set is highlighted.

Table 3.3

deviation) of electrospun PVOH fibers.

The effect of variables on the diameter and variability (standard

 

 

 

 

 

 

 

 

 

 

Egctric Field Syringe TJiameter
Concentration Strength Pump Rate Diameter Std. Dev.
(wt%) (lecm) (ml/h) (rim) (nm)
8 1 .00 0.1 500 260
10 1 .00 0.1 280 60
12 1.00 0.1 1100 960
10 0.83 0.1 380 80
10 1 .00 0.1 360 40
10 1 .25 0.1 600 120
10 1 .00 0.1 800 240
10 1 .00 0.2 780 400
10 1.00 0.3 1180 720

 

 

 

 

 

 

 

In the first set of experiments, the concentration was varied from 8 wt% to 12
wt%. This represented a 50% increase in the concentration, yielding a 120%
increase in the fiber diameter as compared to the 8 wt% data. This variable
provided the most significant change in fiber diameter. The upper limit of
concentration (12 wt%) was not stable and produced a very large distribution of
fiber diameters as evidenced by the standard deviation. The electric field

strength was 1.0 kV/cm, (8 W, 8 cm) the syringe pump flow rate was 0.1 ml/hr

through a 2 inch 20 gauge blunt tip needle.
80

in the second set of experiments, all of the experimental parameters were
maintained constant from the 10 wt% trial, and the electric field strength was
varied. Using a separation distance of 12 cm with an applied voltage of 10 kV
resulted in an electric field strength of 0.83 kV/cm. When the electric field
strength was increased to 1.0 kV/cm (10 kV, 10 cm) the fiber diameter did not
change significantly, but the variability (standard deviation) was reduced. A
separation distance of 8 cm with an applied voltage of 10 kV yielded an electric
field strength of 1.25 kV/cm. This relatively large field strength increased fiber
diameter variability and resulted in larger diameter fibers. The fiber diameter
increased with increasing field strength, which is not expected. The fiber
diameter should decrease with increased electric field strength due to more
electric forces to induce charged fiber whipping. The fact that fiber diameter
increased is explained by the short distance (8cm) between the needle tip and
collector. There was a geometric limitation, and the expected whipping

phenomenon, which greatly reduces fiber diameter, could not occur.

Thethird set of experiments varied the syringe pump rate while maintaining
all other variables constant. In general, increasing the polymer flow rate resulted
in larger diameter fibers. The lower limit 0.01 ml/h and even the middle value of
0.2 ml/h produced fibers with a large amount of variation in their diameters. The
flow rate chosen for the other experiments (0.1 ml/h) was the best choice for

homogenous fiber production.

Another set of electric field strength experiments was run using a lower

applied electric voltage of 8 kV. Table 3.4 shows the statistics for the three
81

different separation distances. Since the applied voltage is constant for the three

different cases, as the distance changes, the electric field strength is altered.

Table 3.4 Summary of various electric field strengths and the resultant fiber

morphology.

 

 

 

 

 

 

 

 

 

 

8 cm 10 cm 12 cm
(1.0 kVIcm) (0.8 kVIcm) (0.67 kVIcm)
Average Diameter (mm 600 355 380
Standard Deviation (nm) 115 40 70
Diameter Range (nm) 450 170 380

The expected trend of higher electric field strength producing smaller
diameter fibers is not apparent in this data set either, as shown in Table 3.4. The
trend is slightly present when comparing the 10 cm (0.8 kVIcm) and 12 cm (0.67
kVIcm) data. The decreased electric field strength resulted in larger diameter
fibers. The large variability in the 8 cm separation distance (1 kVIcm) data was

the reason for the inconsistencies.

Figure 3.8 shows the percentage of analyzed fibers that are within a 50 nm
diameter range. The 0.8 kV/cm data shows that 80% of the fibers are within the
range of 200-300 nm and furthermore 92% of the fibers are within 100 nm from
the mean diameter. The goal of creating uniform fibers was for the intended
application of composite materials. Uniform materials also aid in estimating
individual fiber properties based on bulk sample data. The data shows that

uniform fiber shape was attainable with this experimental setup.
82

Separation Distance shows Variations in Fiber Widths

I12cm I10cm I8cm'

150 200 250 300 350 400 450 500 550 600
Width (nm)

Frequency
A—LNNww-h-b
omomomomom

 

Figure 3.8 Fiber diameter comparison for 10 wt% PVOH fibers electrospun
from separation distances of 8 cm, 10 cm and 12 cm.

Alignment Distribution

Collection of aligned fibers has been shown to be important for composite and
biomedical applications. Aligned electrospun fibers can significantly improve the
mechanical properties of the composite. The importance of a high degree of

alignment was essential to fully realize the reinforcing effect of the materials.22

83

Aligned fibers are important in biomedical applications, since the morphology can

control tissue growth rates.23

Previous research has shown that not using the proper setup, and in our case
the negative potential collector, resulted in poor fiber alignment. This was likely
due to charge buildup on the collector, which repelled the incoming charged jet.
The number of fibers that were within 10 degrees of the average fiber angle was
chosen as a standard for comparison and fiber angles were determined using
lmagePro software“. If the degree of alignment was larger than 80%, the

network was considered aligned.

The 8 cm separation distance, (1 kVIcm) data had the lowest range and the
least amount of standard deviation. 92% of the fibers from this sample were
within 10 degrees of the average. The mean angle parameter was arbitrary; it
depended upon how the sample was loaded in the ESEM for analysis. As
expected, the fibers collected at 8 cm were the most aligned, since they had
traveled less distance through the instability region. As Table 3.5 shows,
decreasing the electric field strength, broadened the range and lowered the
relative amount of alignment. At 10 cm, (0.8 kVIcm), 79% of the fibers are still
within 10 degrees of the average fiber. At 12 cm (0.67 kV/cm), only 69% of the
fibers are within 10 degrees of the average angle of alignment. The fibers
collected at 12 cm experience the greatest instability and were therefore the most
random in size and alignment. These fibers appeared nearly parallel in most

regions examined with the ESEM.

84

The alignment distribution results indicate that closer distances between
source and collector result in a smaller number of stray fibers. The increased
field strength forces the charged fibers to be aligned by the collector rather than
going astray. It has also been shown that the void gap plays a significant role in
how the charged fiber behaves near the collector.24 Void gap was defined as the
distance between the electrodes of the collector. Separation distance was
defined as the distance between the collector and the needle tip, or origin of the

electrospinning process.

 

 

 

 

 

Table 3.5 Summary of alignment distributions.
8 cm 10 cm 12 cm
(1.0 kVIcm) (0.8 kVIcm) (0.67 kVIcm)
Average Angle (deg) 43 121 30
Standard Deviation (deg) 6 9 13
Range (deg) 22 52 54
AN (10) 92 79 69

 

 

 

 

 

 

Table 3.5 shows a summary of the alignment distributions for fibers that were
electrospun from 8, 10 and 12 cm separation distances. The mean angle was an

arbitrary value, which was just the average angle of the fibers observed in the

85

ESEM. The standard deviation and range values were measures of the
statistical variation of the alignment. The 8 cm separation distance provided the
smallest range and the smallest standard deviation. AN(10) is a value defined as
the percentage of fibers within 10 degrees of the average fiber alignment. In the
case of the 8 cm separation distance, 92% of the fibers were within 10 degrees
of the average, which is considered highly aligned. As Figure 3.9 shows, the
alignment distribution proves that closer distances result in less stray fibers. In
this chart, the majority of the fibers in all electric field strength scenarios are

within 5 degrees from the average fiber angle.

86

Separation Distance shows variation in Alignment
Distribution

E38cm l10cm I120m

I f f 2 1

I

I 1

l I I l

I- -,-l,._ 1-2-2.1.?
5 10 15 20 25

Angie (+/- deg)

70

Frequency
—|~ N 00 A 01 O)
O O O O O O

O

 

Figure 3.9 Angle distribution of 10 wt% electrospun PVOH fibers from
separation distances of 8 cm, 10 cm, and 12 cm.

Effect of Polymer Flow Rate on Morphology

The flow rate of the polymer from the syringe was an important process
parameter since it influenced the jet velocity and the material transfer rate. In the
case of PS fibers, Megelski et al. observed that the fiber diameter and the pore
diameter, the open areas on the individual fibers, increased with an increase in
the polymer flow rate.18 As the flow rate increased, fibers had pronounced

beaded morphologies and the mean pore size increased from 90 to 150 nm.

87

The conservation of mass for the electrospun jet states that volumetric
flow rate (Q) was based on the initial jet radius (R) and the constant axial fluid

velocity v as shown in Equation 3.1.

Jrsz = Q
(Equation 3.1)

Increasing the polymer flow rate increased the jet size, which ultimately
increased the size of the collected fiber?"5 increasing the flow rate increased the
velocity of the jet and the size of the jet, and if the separation distance was
maintained constant, the time for solvent evaporation was reduced. This lead to
structures that contained more entrapped solvent molecules and were more

agglomerated than fibers produced at lower syringe flow rates.

The effect of polymer flow rate, varied by the syringe pump rate, on the fiber
morphology was studied for the 10 wt% PVOH in water system. As the flow rate
increased, the fibers became less uniform. At low flow rates of 0.01 mllh, the
amount of fibers formed over a 1-hour time span were inconsequential. A flow
rate of 0.1 mllh resulted in homogeneous fiber formation. The flow-rate was
increased to 0.2 ml/h and 0.3 ml/h. All other processing parameters were kept
constant: 10 kV applied voltage, 10 cm separation distance, and 1-hour collection
time. The results of the experiments are shown in Figure 3.10. Figure

3.105howed a broad range of syringe pump rates. In panel A of Figure 3.10, the

88

syringe pump rate was too low (0.01 mllh) and there was insufficient material
available to be electrospun into fibers. The density of material collected was
much lower as compared to the other flow rates examined. Due to the low flow
rate, large variations in shape and size were observed. In panel B of Figure
3.10, the syringe pump rate (0.1 mllh) was ideal. The fibers appeared
homogenous in shape and size, with very few imperfections. The density of the
fiber mat was much greater than the mat obtained using the lower flow rate. In
panel C of figure 10, the flow rate was slightly too large. The fibers appeared to
be agglomerating, resulting in shape and size in homogeneities. In panel D of
Figure 3.10, the flow rate was too high, which resulted in numerous
agglomerations and fiber imperfections. It is clear that faster flow rates lead to
agglomerated fiber structures since the solvent cannot be volatilized sufficiently

from the larger mass of material constrained by the process parameters.

89

 

Figure 3.10 ESEM images of electrospun PVOH fibers, showing the variation
with syringe pump flow rate A) 0.01 mllh, B) 0.1 ml/h, C) 0.2 ml/h,
and D) 0.3 mllh. Applied voltage 10 kV, 10 cm separation distance,
10 wt% PVOH in water.

Effect of Solution Viscosity on Morphology

The viscosity had a dramatic effect on the electrospinning process. Highly

viscous solutions could not be successfully electrospun due to needle clogging
90

issues and insufficient pump capacity. Solutions that had extremely low viscosity
could not be electrospun since they were too dilute and had a much lower
probability for the polymer molecules to connect in the fiber formation process.
Solution concentration determines the limits for the formation of electrospun
fibers due to its relationship with viscosity and surface tension.7 Low
concentration solutions form droplets as a result of surface tension, while higher
concentration prohibits fiber formation due to higher viscosity. In the case of a
PEO/water system, a bimodal distribution in fiber diameter was observed at
higher concentrations. In the PEO system, Dietzel et al. related the average fiber
diameter and the solution concentration by a power law reiationship7. They
interpreted the variations in the fiber diameter and morphology to the shape of
the jet-originating surface, which concurred with the observations of Zong et al.26
Undulating morphologies in fibers were attributed to the delayed drying and the
stress relaxation behavior of the fibers at lower concentrations. As is evident from

the discussions, the concentration of the polymer solution influences the spinning

of fibers and controls the fiber structure and morphology.

As shown in Equation 3.2, the viscosity was directly related to the fiber
diameter. The actual value of the exponent varies for each polymer system, and
is also dependent on other solution and process variables. To date, there is not
a reliable, accurate equation of state relating viscosity to fiber diameter, but as a

general estimate Equation 3.2 is acceptable.

91

d = 7)
(Equation 3.2)

Molecular weight and viscosity are related, and their effects on morphology
are comparable. In general when a polymer with a higher molecular weight is
dissolved in a solvent, the viscosity is higher. Molecular chain entanglements are
necessary to produce continuous fibers. The entanglement of the molecular
chains prevents the electrically driven jet from breaking up, thus maintaining a
continuous fiber jet. Buchko observed that monomeric solutions did not produce
fiber structures when using the electrospinning process.19 When molecular
weight and viscosity increased, the fiber morphology changed from beads on a

string, to smooth continuous fibers to agglomerated structures.

Varying the concentration of the polymer solution directly affected the
viscosity of the solution and showed a non-linear dependence. Fibers could only
be formed in a specific viscosity range of 200 cps (8 wt% ) to 900 cps (12 wt%)

as shown in Figure 3.11.

92

Viscosity versus Weight Percent of PVA

 

 

 

 

 

 

1400
Optimal Fiber
1200 ~« W W W . ,WFonnation
Region
1000
a .II‘ 1
3, 800 W- . , WW W W W ,1
E:
g 600+
S 400 r , ,
200
0 , _ _ _ ___ _
0 2 4 6 8 10 12 14
Weight Percent PVA

Figure 3.11 Viscosity chart of PVOH, indicating the viscosity range in which
fibers can be readily formed.

Figure 3.12 compared ESEM images of electrospun PVOH at various
concentrations. The low concentration (8 wt%) resulted in bead-on-string
morphology, 10 wt% produced nearly homogeneous fibers and 12 wt% resulted
in larger fibers with some agglomerations and some non-fibrous structures. It
was determined that 10 wt% was the optimal concentration for homogeneous

fibers.

93

 

Figure 3.12 ESEM images of electrospun PVOH fibers, showing the variation
with concentration A) 8 wt%, B) 10 wt%, and C) 12 wt%. Applied
voltage 10 kV, syringe pump rate 0.1 ml/h, 10 cm separation
distance

The results of the experiments showed that electrospinning 10 wt% PVOH at
a syringe pump rate of 0.1 ml/h, a separation distance of 10 cm and an applied
voltage of 10 kV resulted in the most homogeneous fiber structure. The
optimized processing parameters lead to fibers shown in Figure 3.13. These
fibers were optimized for alignment and homogeneous fiber diameter. The fiber
diameter was determined to be 360 nm with a very small deviation of only 40 nm.
In this case, 97% of the fibers are aligned within 10 degrees of the average fiber.
There are only 3% of the fibers that are not considered aligned. The fiber mat
appears uniaxially aligned and homogenous in diameter with very few

imperfections. The detailed specifications are summarized in Table 3.6.

94

 

Figure 3.13 ESEM image of electrospun PVOH fibers, showing the optimized
morphology. Applied voltage 10 kV, syringe pump rate 0.1 ml/h, 10
wt% PVOH in water, 10 cm separation distance.

Table 3.6 Summary of processing parameters and their effects on the fiber
diameter. Results avg. :1: SD, n = 40.

Field Strength

Pump Rate

Concentration

 

95

Effect of Electrospinning Environment on Morphology

Environmental conditions around the spinneret, like the surrounding air, its
relative humidity (RH), vacuum conditions, surrounding gas, etc., also influenced
the fiber structure and morphology. Baumgarten observed that acrylic fibers spun
in an atmosphere of relative humidity more than 60% did not dry properly and get
entangled on the surface of the collector.27 The breakdown voltage of the
atmospheric gases is said to influence the charge retaining capacity of the fibers.
Srinivasan et al. proposed a new mechanism for pore formation by evaporative
cooling called “breathe figures”. Breathe figures are pores that occurred on the
fiber surfaces due to the imprints of condensed moisture droplets caused by the
evaporative cooling of moisture in the air surrounding the spinneret. Megelski et
al. investigated the pore characteristics of PS fibers at varied RH and
emphasized the importance of phase separation mechanisms in explaining the

pore formation of electrospun fibers.18

Effect of Solution Conductivity on Morphology

Most polymers are slightly conductive, with the exception of insulating
materials. The charged ions in the polymer increase the charge-carrying capacity
of the jet, thereby subjecting it to higher elongational tension with the applied
electric field. Baumgarten showed that the jet radius varied inversely as the cube
root of the electrical conductivity of the solution.17 Zong et al. added ionic salts to

the electrospinning solution to increase conductivity. They found that poly (D,L-
96

lactic acid) PDLA fibers with added ionic salts such as KHzPO4, NaHzPO4, and
NaCl produced bead-less fibers with relatively smaller diameters ranging from

200 to 1000 nm.28

Effect of Solvent Volatility on Morphology

Solvent volatility plays a major role in the formation of nanostructures.
Electrospinning involves rapid solvent evaporation and phase separation due to
jet thinning. Solvent vapor pressure critically determined the evaporation rate
and the drying time. Bognitzki et al. found that the use of highly volatile solvents
like dichloromethane yielded poly (L — lactic acid) PLLA fibers with pore sizes of
100 nm in width and 250 nm in length along the fiber axis.29 Lee et al. evaluated
the effect of volume ratio of the solvent on the fiber diameter and morphology of
electrospun PVC fibers.30 Average fiber diameters decreased with an increase in
the amount of DMF in the THF/DMF mixed solvent. Lee et al. found the
electrolytic nature of the solvent to be an important parameter in electrospinning.
Megelski et al. studied the characteristics of electrospun fibers with respect to the
physical properties of solvents. The influence of vapor pressure was evident
when PS fibers were spun with different THF/DMF ratios. The vapor pressure of
THF is 129 mm Hg, while DMF is only 2.7 mm Hg. A large amount of THF
resulted in a porous film being formed, while a large amount of DMF resulted in

smooth homogenous fibers to be formed

97

Effect of Solution Rheology on Morphology

The rheological parameters of the solution are most strongly affected by the
entanglement characteristics of the polymer. Addition of nanoscale fillers, such
as silica particles, altered the entanglement characteristics of the polymer and
can significantly alter the viscoelastic parameters of the electrospinning solution.
The filler-polymer interactions resulted in a change in the entanglement density
and distribution. It was found that adding untreated silica particles offered more
hydrogen bonding opportunities that influenced the rheology of solutions. This
phenomenon was similar to increasing the molecular weight or increasing the
degree of polymerization in unfilled systems. The critical processing parameter
was the applied electric field and how it affected the diameter of the fiber jet. A
TEM image of the PVOH with silica in solution is shown in Figure 3.14. This
image of silica shows that it is still agglomerated in solution. The individual silica
particles are not well dispersed. This leads to fiber imperfections after the

solution has undergone the electrospinning process.

98

 

Figure 3.14 PVOH Silica solution. Image shows that silica is agglomerated in
solution.

The mixing method had a significant impact on the dispersion of silica
particles in the solution and ultimately in the electrospun fiber. A simple stir-bar
mixing at maximum level for 3 hours did not provide enough energy to fully
disperse the silica particles. Ideally the particles would have been 20 — 30 nm in
size, with minimal agglomerations. Ultrasonication was necessary to improve the
dispersion of the particles in solution. A Cole-Parmer 750 watt utrasonicator
operated at power level of 20 for 3 hours with continuous stir-bar mixing was
used as the Ultrasonication treatment. Dispersion of silica particles in PVOH
solution can be verified using surface tension and viscosity measurements.
Improved dispersion increases the hydrogen bonding opportunities, which

increase the surface tension and viscosity of the solutions. As an example, 10
99

wt% PVOH in water solution was prepared with a surface tension of 53 mN/m
and a viscosity of 580 cps. The addition of 10 wt% silica (as a percentage of
PVOH content) was mixed using a simple magnetic stir bar method. The stir bar
mixed solution had a surface tension of 55 mN/m and a viscosity of 600 cps. To
improve the dispersion, ultrasonication was used on the same solution (10wt%
Silica) and the surface tension was 60 mN/m and the viscosity was 740 cps. It is
clear that the addition of silica increases solution surface tension and viscosity
and the increase can be amplified with the ultrasonication mixing method. A
TEM image of the electrospun fiber is shown in Figure 3.15. The agglomerations

are still present in the fiber after the electrospinning process.

100

 

Figure 3.15 TEM image of electrospun PVOH with silica. Image shows that
silica aggregations are still present in final fiber.

Thermal Properties

Thermal gravimetric analysis (Figure 3.16) showed the silica-modified fibers
contained approximately 14 wt% residuals, as compared to 2 wt% residuals for
the neat PVOH. It was concluded that ultrasonication alone might not be
sufficient to obtain individual particles in solution. The use of a surfactant to limit

aggregation may be required for this system.

101

TGA Comparison Neat PVOH and Silica Loaded PVOH

 

 

   
  

 

 

 

 

 

W float" FVA’ “ W16Wt% Stir ea; Mixed Silica in PVA '
— 10 wt% Ultrasonicated Silica in PVA
100 a

90

8O

70
°\° 60 2%.... .. ..
:5, 50 I.
o J .
3 40 14.5 % Ultrasonicate

30 1 W 14.2% Stir-Bar Mixed

2.6 % Neat PVOH
20
10 .. z 2
0 . ,. ._ - .. ,, _ .. . .. -._-_._.. ._ ,2, . . . ”fa—--. “mush,“ ___-_. _.__~-._ __._.
0 200 400 600 800 1000

Temperature C

Figure 3.16 Thermal analysis of electrospun PVOH and silica loaded PVOH.

Mechanical Properties

Mechanical properties were measured to characterize the strength of the
electrospun fiber mats relative to the bulk material. These values were
generated from a collection of aligned fiber samples. The data produced was

then compared to literature data.

A table of tensile strength and modulus is shown as Table 3.7. There is a
wide variation in the values reported, due in part to the wide variation in fibers

produced using the electrospinning process and variations in starting material.

102

In general, it is expected that increasing the molecular weight of the PVOH would
improve the mechanical behavior. This was evident in the work of Lee et al,
PVOH with a molecular weight of approximately 75,000 g/mole exhibited a
tensile modulus and yield strength of 1.0 GPa and 16 MPa, respectively, and
were increased to 1.2 GPa and 18 MPa for a PVOH with a molecular weight of
175,000 g/mole.8 These two data points were produced from PVOH that was
polymerized to form an atactic polymer. This was the first publication that
measured the mechanical properties of electrospun PVOH fibers, so the test
method may have resulted in values that were higher than actual values. All

other tensile strength values reported in table 7 are similar to each other.

The molecular weight trend cannot be extended to the work of Jirawut,
which showed that a similar molecular weight of 75,000 g/mole produced a
tensile modulus and yield strength of 4.3 MPa and 35.3 MPa.31 It should be noted
that the fiber size in Jirawut’s work is much larger than Lee et al., and the degree
of hydrolysis was not listed for Lee et al, but was 98% in Jirawut’s work. The
tensile modulus values for electrospun fiber mats were significantly lower than
the reported value of 1.9 GPa for solvent cast PVOH films.32 The results showed
a high degree of variation, since mechanical testing of electrospun materials is
still a new procedure. The alignment method employed in the present study
resulted in increased tensile modulus as compared to randomly oriented fiber
networks. The tensile strength was comparable to other electrospun PVOH fiber

materials as shown in Table 3.7.

103

Table 3.7 Mechanical properties of electrospun PVOH materials. MSU is a
reference to this work.

 

 

 

 

 

 

 

 

 

Molecular Fiber Tensile Tensile

Weight Diameter Strength Modulus
PVOH Material [Ref] (g/mol) (nm) (MPa) (MPa)
Random Electrospun Fibers [5] 65,000 250 7.2
Random Electrospun Fibers [3‘] 75,000 2000 4.3 35.3
Random Electrospun Fibers [33] 78,000 130 2.6 83
Random Electrospun Fibers [3‘] 74,800 270 16 1000
Random Electrospun Fibers [3‘] 176,000 280 18 1200
Aligned Electrospun Fibers [MSU] 65,000 300 5.2 225
Solvent Cast PVOH Film [32] 186,000 NA 83 1900

 

 

 

 

 

Tests were also performed to determine the effect of silica reinforcement,
using two separate mixing methods, simple stir-bar mixing and ultrasonication.
The dispersion was characterized by TEM microscopy as shown in Figure 3.15
above. In this micrograph, individual silica particles can be seen, but were still
present in large groups. The agglomeration was much reduced as compared to
the simple stir-bar mixing method. The stress strain behavior of the electrospun

materials is shown in Figure 3.17.

104

 

Neat PVA Epoxy mounted Stress Strain Data

‘— Neat Electrospun PVOH
-— Electrospun PVOH + 10% Silica Ultrasonicatedl
——- Electrospun PVOH +10% Silica Stir Bar Mixed

 

12

 

Stress (MPa)

 

 

 

0 0.01 0.02 0.03 0.04 0.05
Strain

Figure 3.17 Stress strain chart for electrospun PVOH. Ultrasonic mixing of
silica provides the best improvement.

The summary of the silica reinforcement is shown in Figure 3.18. It was clear
that silica provided a reinforcement effect and this effect was maximized with the

better stirring method of ultrasonication.

105

Modulus Comparison (MPa)

700

 

600 ~ ~WWWW~

500 W. .-- » ,, ,m

400

300

Modulus (MPa)

200

 

  

 

 

Neat PVOH 10 wt% Silica in PVOH 10 wt% Silica in PVOH
(Stir Bar) (Ultrasonicated)

Figure 3.18 Modulus comparison for the electrospun PVOH and silica
reinforced PVOH.

Electrospinning resulted in mechanical property values that were lower
than expected, and since the dispersion of silica was relatively poor, the
reinforcement effect was also limited. When perfectly dispersed and wetted in the
matrix, filler particles can maximize the possible attachment points with the
polymer, that is the apparent crosslink density of the matrix, and then maximize
the hardness and modulus.35 The surface interaction with the polymer matrix was
critical with these fillers since their high-surface area could have lead to a very

large number of attachment points with the matrix, and the tensile properties

106

were expected to increase with the filler volume fraction.36 The addition of silica
particles to the electrospinning solution has produced fibers with increased
elastic forces. Specifically in this experiment; the addition of silica particles
increased the hydrogen bonding opportunities, which allowed the material to
recover better from deformation after stress has been released. The modulus
data is shown in figure 18. It was determined that ultrasonication provided a more
uniform dispersion of the silica particles, which is proved by improving the
modulus. Ultrasonication improved the dispersion of the silica particles as

compared to simple stir-bar mixing. The modulus data confirmed this effect.

Conclusions

Aligned nanofibers are essential for developing high strength
nanocomposites. Aligned fibers with a narrow width distribution were obtained
through careful control of the experimental parameters. The use of PVOH as a
model compound has allowed the variables affecting the electrospinning process
to be systematically varied and the process to be evaluated for consistency.
Aligned fibers were demonstrated by harvesting the fibers that spannedhthe gap
of a parallel electrode collector arranged with a positive voltage at the needle tip
and negative voltage at the collector. Maintaining the collector at negative
potential rather than at ground aids the fiber collection efficiency; fewer stray
fibers are produced. Collecting the fibers across a void gap causes a high

degree of alignment. The conductive strips of the collector enable the fibers to

107

be easily gathered for future processing and evaluation. As a result, the general
trends discovered using this polymer system could be applied to other polymer

solutions.

A summary of the most significant variables affecting the resultant fiber
diameter is shown in Figure 3.19, using data from Table 3.6. The figure shows
that concentration, also denoted by solution viscosity was the most dominant
variable in this data set. Increases in concentration provided a positive effect by

increasing fiber diameters significantly.

108

 

Summary of Major Variables and their Effect on Fiber Diameter

 

Solution Flow Rate
I Electric Field Strength
I Solution VIscosity

 

 

 

Relative Increase in Fiber Diameter

 

 

 

Figure 3.19 Relative increases in fiber diameter as a function of electrospinning
variables.

Applied voltage was determined to give a positive response, which is
counterintuitive. Theoretically, an increase in applied voltage, or electric field
strength should reduce the fiber diameter. In this case, fiber diameters were
found to increase. This can be explained by the large variability in the fiber

diameters, as a result of a non-stable electrospinning process.

Increases in solution flow rate provided relatively low increases in fiber

diameter. The actual diameters of the fibers increased slightly and since the

109

increased flow rate produced numerous fiber imperfections and agglomerations,

the actual increase in diameters was exaggerated.

110

References

Koski, A.; Yim, K.; Shivkumar, 8., Effect of molecular weight on
fibrous PVA produced by electrospinning. Materials Letters 2004,
58, (3-4), 493-497.

Zhang, C. X.; Yuan, X. Y.; Wu, L. L.; Han, Y.; Sheng, J., Study on
morphology of electrospun poly(vinyl alcohol) mats. European
Polymer Journal 2005, 41, (3), 423-432.

Hodge, R. M.; Edward, G. H.; Simon, G. P., Water absorption and
states of water in semicrystalline poly(vinyl alcohol) films. Polymer
1996, 37, (8), 1371-1376.

Li, J. K.; Wang, N.; Wu, X. S., Poly(vinyl alcohol) nanoparticles
prepared by freezing-thawing process for protein/peptide drug
delivery. Journal of Controlled Release 1998, 56, (1-3), 117-126.

Ding, 8.; Kim, H. Y.; Lee, S. C.; Shao, C. L.; Lee, D. R.; Park, S. J.;
Kwag, G. B.; Choi, K. J., Preparation and characterization of a
nanoscale poly(vinyl alcohol) fiber aggregate produced by an
electrospinning method. Journal of Polymer Science Part B:
Polymer Physics 2002, 40, (13), 1261 -1 268.

Mishra, R.; Rao, K. J., Electrical conductivity studies of
poly(ethyleneoxide)-poly(vinylalcohol) blends. Solid State Ionics
1998, 106, (1-2), 113-127.

Deitzel, J. M.; Kleinmeyer, J.; Harris, D.; Tan, N. C. B., The effect of
processing variables on the morphology of electrospun nanofibers
and textiles. Polymer 2001, 42, (1 ), 261-272.

Lee, J. S.; Choi, K. H.; Do Ghim, H.; Kim, S. S.; Chun, D. H.; Kim,
H. Y.; Lyoo, W. 8., Role of molecular weight of atactic poly(vinyl
alcohol) (PVA) in the structure and properties of PVA nanofabric
prepared by electrospinning. Journal of Applied Polymer Science
2004, 93, (4), 1638-1646.

111

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Son, W. K.; Youk, J. H.; Lee, T. 8.; Park, W. H., Effect of pH on
electrospinning of poly(vinyl alcohol). Materials Letters 2005, 59,
(12), 1571-1575.

Hong, K. H., Preparation and properties of electrospun poly (vinyl
alcohol)lsilver fiber web as wound dressings. Polymer Engineering
and Science 2007, 47, (1 ), 43-49.

Shao, C. L.; Kim, H.; Gong, J.; Lee, D., A novel method for making
silica nanofibres by using electrospun fibres of
polyvinylalcohol/silica composite as precursor. Nanotechnology
2002, 13, (5), 635-637.

Shao, C. L.; Kim, H. Y.; Gong, J.; Ding, B.; Lee, D. R.; Park, S. J.,
Fiber mats of poly(vinyl alcohol)lsilica composite via
electrospinning. Materials Letters 2003, 57, (9-10), 1579-1584.

Teo, W. E.; Ramakrishna, S., A review on electrospinning design
and nanofibre assemblies. Nanotechnology 2006, 17, (14), R89-
R106.

Image-Pro Software - MediaCybemetics -
http://wwwmediacycoml.

Doshi, J.; Reneker, D. H., Electrospinning Process and Applications
of Electrospun Fibers. Journal of Electmstatics 1995, 35, (2-3),
1 51 -1 60.

Park, J. Y.; Han, B. W.; Lee, I. H., Preparation of electrospun
porous ethyl cellulose fiber by THF/DMAc binary solvent system.
Journal of Industrial and Engineering Chemistry 2007, 13, 1002-
1 008.

Baumgarten, P. K., Electrostatic spinning of acrylic microfibers.
Journal of Colloid and Interface Science 1971 , 36, (1 ), 71-79.

Megelski, S.; Stephens, J. 8.; Chase, D. B.; Rabolt, J. F., Micro-
and nanostructured surface morphology on electrospun polymer
fibers. Macromolecules 2002, 35, (22), 8456-8466.

Buchko, C. J.; Chen, L. C.; Shen, Y.; Martin, D. 0., Processing and

microstructural characterization of porous biocompatible protein
polymer thin films. Polymer 1999, 40, (26), 7397-7407.

112

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

Jaeger, R.; Bergshoef, M. M.; Batlle, C. M. I.; Schonherr, H.;
Vancso, G. J., Electrospinning of ultra-thin polymer fibers.
Macromolecular Symposia 1998, 127, 141-150.

Zhang, C. X.; Yuan, X. Y.; Wu, L. L.; Sheng, J., Properties of
ultrafine fibrous poly(vinyl alcohol) membranes by electrospinning.
Acta Polymerica Sinica 2006, (2), 294-297.

Bergshoef, M. M.; Vancso, G. J., Transparent nanocomposites with
ultrathin, electrospun nylon-4,6 fiber reinforcement. Advanced
Materials 1999, 11, (16), 1362-1365.

Xu, C. Y.; Inai, R.; Kotaki, M.; Ramakrishna, S., Aligned
biodegradable nanofibrous structure: a potential scaffold for blood
vessel engineering. Biomaterials 2004,25, (5), 877-886.

Li, D.; McCann, J. T.; Xia, Y. N., Uniaxial alignment of electrospun
nanofibers. In Polymeric Nanoflbers, Amer Chemical Soc:
Washington, 2006; Vol. 918, pp 319-329.

Colman, P. C.; Yong Lak, J., Electrospinning of viscoelastic Boger
fluids: Modeling and experiments. Physics of Fluids 2006, 18, (5),
053102.

Zong, X. H.; Kim, K.; Fang, D. F.; Ran, S. F.; Hsiao, B. S.; Chu, 8.,
Structure and process relationship of electrospun bioabsorbable
nanofiber membranes. Polymer 2002, 43, (16), 4403-4412.

Srinivasan, G.; Reneker, D. H., Structure and Morphology of Small-
Diameter Electrospun Aramid Fibers. Polymer International 1995,
36, (2), 195-201.

Zong, X. H.; Ran, S. F.; Fang, D. F.; Hsiao, B. S.; Chu, B., Control
of structure, morphology and property in electrospun poly ‘

(glycolide-co-lactide) non-woven membranes via post-draw
treatments. Polymer 2003, 44, (1 7), 4959-4967.

Bognitzki, M.; Frese, T.; Wendorff, J. H.; Greiner, A.,
Submicrometer shaped polylactide fibers by electrospinning.
Abstracts of Papers of the American Chemical Society 2000, 219,
U491 -U491 .

Lee, K. H.; Kim, H. Y.; La, Y. M.; Lee, D. R.; Sung, N. H., Influence

of a mixing solvent with tetrahydrofuran and N,N-
dimethylforrnamide on electrospun poly(vinyl chloride) nonwoven

113

31.

32.

33.

34.

35.

36.

mats. Journal of Polymer Science Part B-Polymer Physics 2002,
40, (19), 2259-2268.

Jirawut, J.; Ratana, R.; Pitt, 8., Fabrication of &a|pha;-chitin
whisker-reinforced poly(vinyl alcohol) nanocomposite nanofibres by
electrospinning. Nanotechnology 2006, (17), 4519.

Zhang, X.; Liu, T.; Sreekumar, T. V.; Kumar, S.; Moore, V. C.;
Hauge, R. H.; Smalley, R. E., Poly(vinyl alcohol)lSWNT Composite
Film. Nano Lett. 2003, 3, (9), 1285-1288.

Wang, X.; Chen, X.; Yoon, K.; Fang, D.; Hsiao, B. S.; Chu, B., High
Flux Filtration Medium Based on Nanofibrous Substrate with
Hydrophilic Nanocomposite Coating. Environ. Sci. Technol. 2005,
39, (19), 7684-7691.

Lee, J. S.; Choi, K. H.; Do Ghim, H.; Kim, S. S.; Chun, D. H.; Kim,
H. Y.; Lyoo, W. 3., Role of molecular weight of atactic poly(vinyl
alcohol) (PVA) in the structure and properties of PVA nanofabric

prepared by electrospinning. Journal of Applied Polymer Science
2004, 93, (4), 1638-1646.

Westermann, S.; Kreitschmann, M.; Pyckhout-Hintzen, W.; Richter,
D.; Straube, E.; Farago, B.; Goerigk, G., Matrix Chain Deformation
in Reinforced Networks: a SANS Approach. Macromolecules 1999,
32, (18), 5793-5802.

Guyard, A.; Persello, J.; Boisvert, J-P.; Cabane, B., Relationship
between the polymer/silica interaction and properties of silica
composite materials. Joumal of Polymer Science Part B: Polymer
Physics 2006, 44, (7), 1134-1146.

114

CHAPTER 4

ELECTROSPINNING CELLULOSE NITRATE FOR BIOSENSOR

APPLICATIONS

Abstract

Cellulose is the most abundant organic material on the surface of earth. It is
the main constituent of paper and plant cell walls and is a renewable resource.
Cellulose consists of linear glucose polymer chains, which form hydrogen
bonded supramolecular structures. Cellulose is not soluble in water and most
organic solvents but functionalized cellulose such as cellulose nitrate is soluble in
many common organic solvents and is important for many industries with

applications in packaging, coatings, adhesives, biochemistry and sensors.

Electrospinning is a method of producing nanoscale fibers from apolymer
solution using high voltage electric forces. This technique is capable of producing
fibers in the nanometer to micron diameter range. Electrospinning has been
extensively investigated as a simple method to prepare fibers from polymer
solutions or melts. The electrospinning process allows for alignment and

patterning of the fibers through careful control of the processing parameters,

115

choice of suitable collector materials, and geometry. Electrospun fibers can be
utilized in various applications including biosensors, filtration media, biomedical
devices, and composites. The ability to control the alignment of fibers is
essential for producing high flow membranes with a large surface area to volume

ratio, which is crucial for biosensor applications

Cellulose nitrate has been electrospun into fibers with controllable diameters
in the range of 100 to 1,000 nm. Collecting fibers with controllable size and
orientation is essential for preparing biosensor devices. A unique rotating
collector and a stationary parallel electrode void gap collector have been
designed for collection of the fibers. The controllable size and alignment allow
for properties such as pore volume, surface area and capillary flow to be

managed.

Neat electrospun cellulose nitrate scaffolds were found to be hydrophobic in
nature and unsuitable for capillary flow biosensor devices. This limitation was
overcome by plasma treating the surface of the fibers. The surface chemistry
modification allowed for improvements in capillary flow by altering the surface
functional groups to ones that were more hydrophilic. This technique is ideally
suited to electrospun scaffolds because, unlike other physical and chemical
modification techniques, it did not compromise the bulk architecture of the
material. Plasma treatments of electrospun cellulose nitrate scaffolds resulted in
a significant increase in wettability due to the incorporation of oxygen-containing

groups (hydroxyls, carbonyls, carboxyls), which were detected by XPS and FTIR.

116

The thermal, electrical, and mechanical properties of the electrospun fiber
mats were examined. Thermal analysis proved that solvents were not present in
the electrospun fiber mats and the onset of thermal degradation was 205 °C.
Electrical characterization showed that the neat fibers were insulating, as
expected. The mechanical behavior of the fiber mats was lower than films, due
to the large amount of void spaces in the fiber mats, with tensile modulus values

in the 10 — 13 MPa range.

Introduction

Cellulose is the most common organic polymer, and is considered an
almost inexhaustible source of raw material for environmentally friendly and
biocompatible products.1 Many applications are available for cellulose, including

the production of cellulose composites with synthetic and bio-based polymers.

Cellulose nitrate is prepared from cellulose through a nitration reaction
involving nitric and sulfuric acid. Cellulose nitrate esters, also known as
nitrocellulose, are commercially available with various degrees of esterification
up to the maximum substitution degree of 3 nitrate ester groups per monomer
(14% nitrogen). The degree of esterification influences its solubility properties.
Cellulose nitrate can be dissolved in ether, aldehydes and some alcohols.
Solutions of cellulose nitrate are commonly referred to as collodion. Collodion
has been used extensively as a wound dressing material and a carrier for topical

medications. Cellulose nitrate is also commonly used for immobilizing nucleic

117

acids in biochemistry applications. Cellulose nitrate also has the ability to
immobilize proteins for applications in Western blots, due to its non-specific
affinity for amino acids. Cellulose nitrate is also widely used in diagnostic tests
where antigen-antibody binding occurs, such as pregnancy test and Albumin

tests.

Background

Cellulose acetate has been the most widely researched and electrospun
cellulose derivative.2 The cellulose acetate can be deacetylated, resulting in a
regenerated cellulose membrane. Lu et al., prepared cellulose acetate (CA)
fibers using the electrospinning technique with acetone, acetic acid and
dimethylacetamide (DMAC) as the solvents.2 The source of cellulose for these
experiments was cellulose acetate with a DS of 2.45 and a number average
molecular weight of 30,000 Da (obtained from Aldrich). CA concentrations were
varied from 10 wt% to 25 wt% depending on the solvent. Several mixtures of
DMAC with either acetone or acetic acid produced suitable solvent systems for
electrospinning CA. Fifteen percent CA in acetonezDMAC mixtures at ratios
between 2:1 and 10:1 could be continuously electrospun into fibrous membranes.
The highest concentration to yield homogeneous fibers was 20 wt% CA in a 2:1
ratio of acetone to DMAC. The 2:1 aCetone: DMAC mixture was the most
versatile mixing ratio because it allowed continuous electrospinning of CA
solutions at concentrations between 12.5 and 20 wt%. The 12.5 and 20 wt% CA

in 2:1 acetonezDMAC corresponded to a viscosity in the range of 1.2 to 10.2
118

poise and surface tension of 26 dyne/cm. Viscosities lower than 1.2 poise lead to
beaded fiber formation, whereas a viscosity higher than 10.2 poise caused
instability in electrospinning. Smooth fibers of uniform diameters from 100 nm to
over 1,000 nm were generated using solutions with the appropriate viscosity.
Fiber sizes were observed to decrease with decreasing CA concentrations, which

was expected.

Ma et al. prepared cellulose acetate solutions suitable for electrospinning
using a mixed solvent of acetone Idimethylformamide / trifluoroethylene (3/1/1).3
The cellulose acetate source was from Fluka, Mr = 29,000 and 40% acetyl
groups. The electrospinning solution was approximately 16 wt% CA in the mixed
solvent. Fibers were formed in the range of 200 nm to 1,000 nm. The fiber
network was subjected to a heat treatment, and then the resultant fiber network
was characterized. It was determined that the tensile strength of the membranes
was 7.6 Mpa. These membranes consisted of non-aligned fibers with a large

diameter range and variation.

Son et al. electrospun ultrafine cellulose acetate fibers with various
concentrations of silver nitrate (AgN03) for antimicrobial activity.4 A 10 wt% CA
solution, with an acetyl content of 40% and molecular weight of 30,000 in
acetone water (80:20) was used as a baseline solution with 0.05, 0.3 and 0.5
wt% AgN03 added for functionalization. The addition of AgN03 Increased the
charge density in CA solutions and thus stronger elongational forces were

imposed on the ejected jets under the electrical field, resulting in substantially

119

straighter and thinner CA fibers. The resultant fibers were 600 nm to 1,600 nm in

diameter.

Ethyl cellulose (EC) has been successfully electrospun by Wu et al. using a
mixed solvent. EC with Mn=72,000 and a substituted degree of ethyl of 2.2 was
dissolved in various ratios of tetrahydrofuran (THF) and DMAC.5 As the content
of THF is increased in the solvent system, the fiber diameters increased. The
increase in fiber diameter was due to an increase in the solution viscosity. In
general concentrations from 7 up to 27 wt% of EC can be prepared and these

solutions produced fibers that were 200 nm to 3,000 nm.

Hydroxypropyl cellulose (HPC) fibers have been electrospun by Shukla et al.
from solutions of 2-propanol and ethanol.6 Two separate spinning solutions were
prepared, 15 wt% HPC in 2- propanol yielded homogeneous fibers without bead
on string morphology, while the 15 wt% HPC in ethanol showed larger fiber
variations including bead on string morphology. The purpose of these fibers is to
provide a template for constructing tin oxide fibers. The fiber diameters ranged

from submicron (>100 nm) to true nanofibers (< 100 nm).

In this research work cellulose nitrate fibers have been produced with
diameters ranging from 100 nm to 1,000 nm. Highly aligned fibers are obtained
using a patterned stationary collector, and partially aligned fiber networks can be
obtained using a rotating drum collector. Literature suggests that high speed
(>10 m/s) rotating collectors, gas flow, and electrical or mechanical shielding can
also be used to increase alignment of fiber networks, but similar experiments

were not attempted.7 The rotating collector works well to obtain a larger yield of
120

aligned fibers in a shorter amount of time,8 but the rotational speed must be fast
enough to compare with the speed of the jet.9 The current device provides fiber
mats with 65% of the fibers within 10 degrees of the desired orientation. An
aligned sample must consist of at least 90% of the fibers within 10 degrees of the
desired orientation. A stationary collector with parallel electrodes can also be
used to easily harvest aligned fibers, but is limited to a small area of coverage.10
The stationary collector can provide alignment with greater than 90% fibers within
10 degrees of the desired orientation. A comparison of cellulose nitrate fiber
morphology obtained using the stationary collector and rotating drum collector is
shown as Figure 4.1. Typically a 4-hour collection time was used on the
stationary collector, while the same fiber mat thickness could be produced in 2
hours on the rotating drum collector. This was most likely due to the fact that
stray fibers were eliminated using the rotating drum collector. Stray fibers can
collect on the electrode surface or other grounded objects near the
electrospinning setup. Since the entire drum Surface was the collector, there
were virtually no stray fibers. Using the parallel electrode device, the top and

bottom electrodes collect a large amount of material that is considered waste.

121

 

Figure 4.1 ESEM of electrospun cellulose nitrate comparing the alignment
capabilities. A) Fibers were collected using a stationary parallel
electrode collector design. B) Fibers were collected using a
rotating drum.

Experimental

Materials

Cellulose nitrate was purchased from Scholle Corporation (College Park, GA).
Cellulose nitrate had a nitrogen content of 12% which corresponds to a degree of
substitution of 2.25. Solutions of cellulose nitrate were prepared in two different
solvent systems: a 70:30 (wtzwt) mixture of ethanol : acetone and a 60:40 (wtzwt)
mixture of tetrahydrofuran : N,N-dimethylformamide. All solvents were purchased
from Sigma Aldrich (St. Louis, MO). Several different solutions were prepared by

dissolving cellulose nitrate in each of the solvent systems by magnetic stir bar
122

mixing for 12 hours. The electrospinning unit was purchased from Kato — Tech
Co (Kyoto, Japan) and was capable of 40 kV, syringe pump rates of 0.03 to 3.0
cm/min, rotating drum speeds of 0.01 to 9.0 m/min and collection times up to 4

hours.

Morphology

The cellulose nitrate was examined using Environmental Scanning Electron
Microscopy (ESEM), and ImagePro image analysis software.11 An Electroscan
2020 ESEM was used to obtain sample images. The samples were first
mounted on ESEM pedestals, then sputter coated with gold. The samples were
imaged at 15 W from 500 to 10,000 times magnification to obtain morphological

and size distribution information.

Field emission scanning electron microscopy (FESEM) was also used to aid
in morphological characterization of electrospun cellulose nitrate. A JEOL 6300F

FESEM was utilized to obtain images up to 100,000X magnification.

Surface Chemistry Properties

Contact angle measurements

123

Water contact angles were measured on electrospun fiber surfaces using a
Kruss DSA10. The water droplet was placed on the surface of the material and
an immediate image was captured. The captured image was used to calculate
the contact angle. The samples were measured ten times and averaged.

Samples were run as processed and after a plasma treatment procedure.

X-ray Photoelectron Spectroscopy

A Perkin Elmer Phi 5400 ESCA system was used for experimentation with a
magnesium Ka X-ray source. Samples were analyzed at pressures between 10'9
and 1043’ torr with a pass energy of 29.35 eV and a take-off angle of 45°. The
spot size was roughly 250 um". Atomic concentrations were determined using
previously determined sensitivity factors. All peaks were referenced to the

signature 018 peak for adventitious carbon at 284.6 eV.

Plasma Treatment

Plasma treatments were carried out in Plasma Sciences system with a gas
pressure of 0.29 torr and at 300W. The samples were presented for plasma
treatment in the form of electrospun fiber bundles mounted on polycarbonate
window frames on the outer surface to aid in handling. The samples were
approximately 1 in x 1 in. These samples were laid flat in the plasma reactor.

Plasma treatments were performed for 0, 90 and 360 seconds

124

 

Fourier Transform Infrared Spectroscopy

A Perkin Elmer Spectrum One FTIR was used to analyze the bulk chemistry
of the electrospun cellulose nitrate materials. The bulk chemistry was important
to analyze to ensure high quality samples were being produced without
transformations or impurities. The FTIR spectra obtained were compared to neat

cellulose nitrate materials.

Thermal Properties - Thermal Gravimetric Analysis

A TA Instruments 2950 was used to determine the amount of residual solvent
in the electrospun fiber mat, and the thermal degradation temperature of the
cellulose nitrate fiber bundle. The electrospun fibers will be used at room
temperature, however thermal characterization is important to understand for

heat assisted drying steps that may be necessary.

Electrical Properties - Electrical Impedance Spectroscopy

Electrical Impedance Spectroscopy was used to verify that the electrospun
cellulose nitrate fiber bundles were non-conductive. A Gamry Instruments 300
series electrical impedance spectrometer was used to determine the resistance

of electrospun samples at 1 Hz.

125

 

Mechanical Properties - Dynamic Mechanical Analysis

A TA instruments 0800 Dynamic Mechanical Analyzer (DMA) was used to
evaluate the mechanical properties of the electrospun materials. The goal was to
determine the modulus for various fiber sizes, which correspond to different fiber
packing densities of electrospun cellulose nitrate. As the fiber packing density is

increased, the tensile modulus is expected to increase.

Results and Discussion

Morphology

The concentration of cellulose nitrate in solution was found to be the most
dominant factor controlling fiber size. Increasing the concentration resulted in
increased fiber diameters. This is due to increased viscosity causing more
entanglements leading to a structure that is more resistant to elongation under

the external whipping and stretching forces.

Microscopic images were viewed perpendicular to the fiber matjfor easy
alignment characterization and fiber size estimation. It was assumed that the
fibers were circular in cross section.10 A typical FESEM micrograph is shown in
Figure 4.2. The FESEM was used for high magnification images to accurately

determine the fiber diameters.

126

 

 

 

Figure 4.2 FESEM image showing that single fiber sizes are available less
than 100 nm in diameter.

The electrospinning parameters were 8 cm separation distance, 0.02 mm/min
syringe pump rate, a stationary collector and 30-minute collection time. The
images verified that increasing the solution viscosity lead to larger diameter
fibers. A 6 wt% (300 cps) cellulose nitrate solution produced fibers with an
average diameter of 330 nm, while fibers electrospun from 8 wt% (500 cps)
solution were on average 900 nm in diameter and from a 10 wt% (2800 cps)
solution were approximately 1200 nm as shown in Table 4.1. The alignment is in
terms of percentage within 10 degrees of the average angle when observed

using electron microscopy.

127

Table 4.1 Fiber diameter and alignment changes as a result of solution

 

 

 

 

viscosity.
Concentration (wt%) Viscosity (cps) Fiber Alignment
Diameter (nm) (%)
6 280 350 i 50 82
8 510 900 t 85 51
10 2750 1200 :l: 80 68

 

 

 

 

 

 

Collector design coupled with process control was used to control fiber
patterns.12 Figure 4.3 displays an aligned cross-pattemed design using a
perpendicular electrode array requiring rotating the collected fiber mat 90 ° after
a fixed amount of time. The advantage to collecting fibers in this fashion was
that the procedure provided a fiber network structure with adjustable porosity.
The layered morphology can be controlled by collecting fibers for longer times in

each direction to build up fiber layer thickness.

128

 

Figure 4.3 ESEM image of patterned electrospun fibers using a stationary
collector.

Effect of Electric Field Strength

The electric field strength can be controlled by altering the separation
distance, the distance between the needle tip and collector, or by adjusting the
applied voltage. Typically the electric field strength was maintained near 1
kV/cm. Figure 4.4 shows the effect of varying the electric field strength while
maintaining a constant applied voltage of 10 W. A slight negative charge (1 W)
was maintained at the collector to reduce to residual charging phenomenon and
promote thicker electrospun fiber mats with fewer stray fibers. In this experiment,

the solution was 6 wt% cellulose nitrate in a 70 : 30 ethanol : acetone mixed

129

solvent system. The solution flow rate was maintained at 0.5 ml/h through a 22
gauge needle and collection times were 2 minutes. Fibers were harvested from
a parallel void gap electrode with a 2.5 cm distance between top and bottom

electrode.

Figure 4.4 shows that there was a range of electric field strengths that
could be used to produce homogenous aligned fiber networks. As the electric
field strength values increased, there was a slight decrease in fiber diameter, but

the fiber diameter trend is not statistically significant. The electric field strength

was determined to be an ineffective means to control the fiber diameter.

 

Figure 4.4 ESEM of electric field strength variations, A) 12 cm (0.9 kVIcm), B)
10 cm (1.1 kV/cm), C) 8 cm (1.4 kV/cm) and D) 6 cm (1.8 kV/cm).

Electric field strength can be used to alter the fiber alignment. At longer
separation distances, the instability region was more prominent and resulted in a
lower number of aligned fibers. As shown in Table 4.2, the alignment improved
from 38% at 12 cm to nearly 90% at 6 cm. The alignment is defined as the
percentage of fibers that are within 10 degrees of the average fiber angle as

determined using ESEM images and ImagePro image analysis software. The
130

general trend of improved alignment as the separation distance was decreased
was verified. Limiting the distance between source and collector and thereby
decreasing the instability improved the bulk fiber alignment. The higher electric
field strength would likely increase the polymer chain orientation on the molecular

level since it provides higher elongation forces.”

Table 4.2 The effect of separation distance on the fiber diameter and

 

 

 

 

 

alignment.
Sample Separation Electric Field Fiber Alignment (%)

Distance (cm) Strength Diameter within 10

(kVIcm) (nm) degrees)

A 12 0.9 1405 1 950 38

B 10 1.1 4002110 81

C 1.4 340 1- 85 82

D 1.8 350 1 10 90

 

 

 

 

 

 

 

Effect of Solution Viscosity on Morphology

As the viscosity increased, the polymer chains became more entangled
and were more resistant to the stretching phenomenon common in the
electrospinning process. This resulted in fibers with larger diameters. There was
also more charge repulsion in the larger fibers, which lead to more randomly
distributed fiber networks. Figure 4.5 shows the ESEM images for 6,8 and 10
wt% cellulose nitrate dissolved in a mixed solvent system of 60:40 THF:DMF.

131

The electric field strength was 1.8 kV/cm and the flow rate was 1.0 ml/h. Figure
4.5 shows that increasing the viscosity of the solution increased the fiber

diameter and also decreased the alignment of the fiber mat.

 

Figure 4.5 ESEM images of 6,8 and 10 wt% cellulose nitrate electrospun
fibers.

Table 4.1 shows that when the viscosity is increased, the fiber diameter
increases. There is also a trend of increasing randomness, as the fiber diameter
gets larger. The fiber diameter trend was modeled with the Helgeson model
(Chapter 2). This model provided a reasonable prediction for the resultant fiber

diameters often with errors less than 50%.

Effect of Solvent Volatility on Morphology

Cellulose nitrate is shipped in 100% acetone and although this can be
electrospun directly, the process is discontinuous, resulting in many fiber mat

deformations. This is due to the rapid evaporation of acetone. Using acetone as

132

the only solvent results in clogged needle tips and makes the electrospinning
process difficult to control. Ethanol was added to acetone to reduce the volatility
of the solvent system. Ethanol cannot be used as the only solvent for cellulose
nitrate due to solubility limitations. It was determined that a 70 wt% ethanol, 30
wt% acetone solvent system provided the most controllable process. This
solvent was named 70:30 Ethanol:Acetone based on the weight composition of
each solvent. Figure 4.6 shows the difference in morphology when using
ethanol and acetone blends. The processing conditions were not optimized in
this study; the purpose was to determine what solvent blend ratio provided the
most uniform fiber diameters. As Figure 4.6 shows, 100% acetone is too volatile
and results in agglomerated fibers and clogged needle tips. 70:30
Ethanol:Acetone provided the longest time between clogged needles, greater
than 6 hours. As Figure 4.6 and Table 4.4 show, the 50:50 Ethanol:Acetone
appears to provide the least fiber imperfections, but it still resulted in clogged
needle tips. The 70:30 blends were optimized in other experiments to yield fiber
mats with little or no inhomogeneities. The processing conditions were 10 wt%
cellulose nitrate, 8 kV applied voltage, 8 cm separation distance (1.0 kV/cm) and

a 2.0 ml/h flow rate.

133

 

 

Figure 4.6 The effect of solvent volatility on fiber morphology. Variation in
solvent vapor pressure results in various volatility A) 100%
Acetone, B) 70:30 Ethanol:Acetone, C) 50:50 Ethanol:Acetone.

The reduction in vapor pressure generally reduced the fiber diameter as
shown in Table 4.3. The goal of altering the solvent vapor pressure was to
improve the electrospinning process, making it continuous, without interruption
by eliminating the problem of clogged needle tips. It was determined that a 70:30
ethanol:acetone blended solvent provided the longest time between needle

clogs, while still producing homogenous fiber networks.

Table 4.3 Solvent volatility effect on fiber diameter and alignment

 

134

Surface Properties

Plasma Treatment

The surfaces of the electrospun cellulose nitrate fibers are composed of alkyl
(C-C) and ether (C-O-C) functional groups. Plasma treatments are necessary to
increase the hydrophilic nature fibers by forming groups such as carbonyls
(C=O), and carboxyl (COOH) functional groups. The nitrogen containing
functional groups are also affected by plasma treatment; the surface nitrate
(N03) functional groups are converted to nitroso (N=O) and amine (N-N). The
exact functional groups cannot be identified from XPS alone. The changes to the

fibers from the plasma treatment allowed for capillary flow to be realized.

Plasma chemistry, even of a simple gas such as oxygen, is extremely
complex. When plasma contacts an organic surface, the number of possible
reactions that may occur is very large. There are two main reactions that occur:
(1) etching reactions that result in the removal of the surface material; (2)
reactions that result in modification of the surface chemistry. Plasma treatment
was chosen over solution—based methods due to fiber mat strength limitations. It
was found that repeated washing cycles damaged the morphology of the fiber

mat.

Etching results from either energy deposition from the plasma into the

polymer surface, or from reactions between the polymer surface and the reactive

135'

 

species in the plasma. There is sufficient energy in glow-discharge plasma to
cause bond breakage, if enough of this energy is deposited from the plasma into

the surface, bond breakage will occur.

Energy transfer from the plasma to the cellulose nitrate surface occurs by two
mechanisms: (1) direct energy transfer, from the inelastic collisions of high-
energy particles with the surface; (2) radiative energy transfer. The reactions of
cellulose nitrates in argon plasmas have been used since argon cannot
chemically react with the surface like oxygen can. The effect of energy transfer
without the complication of chemical reactions by reactive species in the plasma
was studied. Argon plasma treatments were shown to induce considerable
changes in the surface chemistry of cellulose nitrate.14 However, little or no
weight loss in the cellulose nitrates was observed. It can therefore be assumed
that at about 100 W, energy transfer from the glow discharge plasmas was not
primarily responsible for etching; this was most likely to have arisen from
oxidation at the surface, resulting in the formation of small volatile fragments that
were pumped away in the plasma gas. There were significant weight changes,

or etching when using oxygen plasma.15

Contact Angle

Cellulose nitrate films and fibers were subjected to contact angle
measurements in order to measure their surface energy and change in surface

energy with surface treatment. Water based solutions must be able to wet the

136

 

fiber surface in order to achieve capillary flow in the biosensor device. A contact
angle less than 90 ° is required for wetting. As the contact angle decreases, the

wetting increases and capillary flow is generally increased.

The contact angles for the cellulose nitrate film and electrospun cellulose
nitrate fibers from both glass and plastic lab were is shown in Table 4.4. The film
was cast in a glass Petri dish directly from the stock solution and the electrospun
fibers were prepared using either a plastic syringe assembly or glass syringe
assembly. It was determined that using plastic syringe produced contact angles
higher than expected as a result of plastic contaminants being extracted and
mixing with the cellulose nitrate solutions. Plastic syringes were replaced with
glass to eliminate the contamination effect. The neat electrospun cellulose
nitrate contact angle was still much larger than 90 °, this effect can be explained

using the Cassie equation and an areal density parameter.

For a Cassie drop, the contact angle is an average between the value of air
(180°) and on the solid (9), which can be calculated for geometric textures such
as posts, fibers or dots.16 Denoting cps as the fraction of solid in contact with the

liquid (cps is dimensionless, but smaller than unity), we find:

cos 6* = —1 + ¢s(1+ cos 6) (Equation 4.1)

Cellulose nitrate solvent cast films exhibited a contact angle of 65°, which is
the value used for 9. Examining the electrospun fiber morphology using ESEM

and lmagePro image analysis (Figure 4.7), the fraction of solid in contact with the

137

water droplet (4),.) was determined. Using ImagePro software, the images were
first contrast normalized using the best fit contrast enhancement procedure,
which assigns the bottom 3% of the pixels to the shadow value (0) and the top
3% to the highlight value (255). The remaining pixels are evenly distributed,
providing the contrast enhancement. Finally a threshold value was set (128 for
plastic and 102 for glass) to yield the images shown in Figure 4.7. These images
represent the areas of the fiber mats expected to make contact with the water
droplet in the contact angle experiments. The plastic syringe produced a more
dense fiber mat as compared to the glass syringe, with all other experimental
parameters being held constant. The plastic syringe produced a fiber coverage
cps of 0.37, which means that 37% of the area under the water droplet during
contact angle analysis is solid cellulose nitrate fiber. Using a glass syringe
yielded (p5 of 0.32. Substituting these values Equation 4.1 yielded an apparent
contact angle of 118° for the plastic syringe and 122° for the glass syringe. These
values were relatively close to the experimentally measured data (Table 4.4).
The plastic syringe measured value was for electrospun cellulose nitrate fiber
mats. This was consistent with values determined experimentally. Similar

results have been observed in the literature using nylon as the material.

138

 

 

Figure 4.7 Electrospun cellulose nitrate shows an open fiber mat with air
occupying the interfibrillar spaces. A) Fibers electrospun from
plastic syringe, B) glass syringe C) image (A) after contrast
enhancement and threshold screening, D) image (B) after
processing.

According to Rutledge et al, electrospun nylon fiber mats exhibited an initial
contact angle of 135°. This value was significantly higher than the contact angle

of 69° measured for a smooth solvent cast nylon film.17 This was the result of the

139

metastable Cassie state, in which air was trapped in the interfibrillar spaces.18 It
has been suggested that roughness promotes the trapping of air between the
water droplet and fiber mat, which in turn prevents the water droplet from

t.19

penetrating the ma This corresponds with the data obtained for the cellulose

nitrate system.

Superhydrophobicity obtained from an intrinsically hydrophilic material has
been proposed to be the consequence of a robust metastable state that can be

” 2° or “re-entrant" structures. ”'23 In both cases, it was

obtained with “overhang
shown that the apparent contact angles could be larger than 150 degrees as long
as the structure of the surface was such that the intrinsic contact angle (>0 deg)
was realized locally.24 Significantly, cylindrical objects like fibers arranged with
their axes within the plane of the surface, as is the case for electrospun
nonwovens, are examples of re-entrant structures. Analysis of the nonwoven
fiber geometry reveals two important design parameters. The first is the scaled
distance between fibers, s/d, where d is radius of the fibers and s is the half-
‘distance between fibers, as shown previously”; this parameter is indicative of the
areal density of fibers. Large (hydrophobic) contact angles are realized when the

electrospun fiber mat has s/d >> 1. From fliis, the apparent contact angle, 0*,

can be calculated using Equation 4.2, where x=d/s.

.. x(.7t—0)C

xsin6
c056 = —

056

1 ‘
x + 1 x + 1 (EquatIon 4.2)

140

 

Figure 4.8 shows how these theoretical apparent contact angles for
electrospun cellulose nitrate cylinders vary with d/s. The contact angle decreases
as the separation distance decreases while the fiber diameter is fixed. Figure 4.8
also shows that the droplet sitting on spheres has a higher contact angle than
that sitting on cylinders of comparable radius. Both of the predictions are
consistent with the experimental data.25 No attempt is made here to fit the data
precisely since the random orientation of fibers in the real materials, as well as
the three-dimensional nature of the mats imaged, complicates the determination

of an appropriate value for the separation distance between fibers.

141

 

-—100nm —200 nm ——300 nm —-400 nm ——500 nm

160
150
- 14o
- - W- - , , ., - , , -r----- 130

, v ,, - - ------ - , , 120
,,, -- - WW 11o

- . ~ - - , , 100

 

- , so
,, , 70

Contact Angle (deg)

20

 

 

 

500 400 300 200 100 0
Fiber Separation Distance (nm)

Figure 4.8 Apparent contact angles for electrospun fibers with various
diameters and fiber spacing.

The knowledge obtained from ESEM images can be coupled with changes in
processing conditions to produce fiber mats with improved wettability. The goal
is to use Figure 4.8 as a guide in fiber preparation. A tightly packed network of
fibers will reduce the water contact angle, but may have porosity limitations. The
Cassie equation can be used as a general guide to estimate the contact angle of
the various fiber morphologies, but in the case of the electrospun cellulose nitrate

fibers, there may be differences in surface chemistries that would limit its use.

142

Surface treatments can also be used to further enhance the capillary flow

properties.

To reduce the contact angle to an acceptable level that would allow for
capillary flow, the fibrous mats were subjected to an oxygen plasma surface

treatment.

Table 4.4 Contact angle measurements

 

 

 

 

 

 

 

 

 

 

 

Sample Name Contact Angle
Solvent Cast Cellulose Nitrate Film (0) 65
Plastic Electrospun Cellulose Nitrate Fibers 131
Glass Electrospun Cellulose Nitrate Fibers 110
Plasma Treated Plastic Electrospun Cellulose Nitrate Fibers 64
Plasma Treated Glass Electrospun Cellulose Nitrate Fibers 12
Solvent Cast Nylon Film [Rutledgel 69
Electrospun Nylon Fibers [Rutledgi 135

 

The variation in contact angle after plasma treatment can be explained by
differences in the surface chemistry of the materials. Aligned electrospun
cellulose nitrate fiber mats were exposed to radio frequency (RF) generated 02
plasma (13.6 MHz, 120 W, 250 mTorr 02 for 903). The plasma treatment's effect
on contact angle is shown in Figure 4.9. This drastic difference in contact angle
was explained by the oxidation of hydroxyl groups on the cellulose backbone to
form carboxylate and/or carbonyl functional groups. This transformation did not
alter the relative concentrations of atoms present at the surface, but it did alter

the type of bonding, which will be explained further in the XPS results section.

143

 

 

 

Table 4.5 shows that treating both sides of the electrospun fiber mat had little

difference on the resultant water contact angle. This data was obtained using the

plastic lab ware.

 

 

 

Figure 4.9 Comparison of water on the surface of electrospun cellulose nitrate
fibers. The non-wetting behavior is shown in A) neat electrospun
fibers and after plasma treatment B) the water droplet wets the
surface.

Table 4.5 Water contact angle data showing the effect of plasma treatment

one side (903)

both sides (90$)

 

X-Ray. Photoelectron Spectroscopy

Chemical surface modifications resulted from direct oxidation of the cellulose
nitrate surface by the oxygen plasma, or from chemical reactions induced on the
cellulose nitrate surface by energy transfer from the plasma. Surface modification
arising from oxidation is not thought to be particularly important. In the Cis
envelope a broadening and increase in high energy counts were noted. This
indicates that the carbon atoms of the cellulose nitrate backbone structure have

been oxidized to form carbonyl or carboxyl functional groups.

Figure 4.10 shows the corrected C1s core level spectra for electrospun
cellulose nitrate fiber mats that have been plasma treated for 0, 90 and 360
seconds. From these spectra it can be seen that considerable changes in the
surface chemistries of the cellulose nitrates has taken place with plasma
treatment. There was a reduction in the signal from the C-OH, which appeared
at approximately 287 eV. The appearance of a new peak at 289 eV indicates
that new carbon functionalities have been introduced with the plasma treatment.
The exact identity cannot be determined using XPS, but it is most likely due to
the formation of carbonyl (aldehydes and ketones) and carboxylate (acids and

esters).

145

 

0 Sec

 

‘ ,
I

360 Sec g/M,

Figure 4.10 Comparison of C15 spectra. An addition of carbonyl and/or
carboxylate signal (289 eV) is evident.

146

The changes observed in the NIS core level cannot be attributed to surface
oxidation. The appearance of new nitrogen functionalities at lower binding energy
is indicative of surface reduction, not oxidation. Very similar changes in this core
level were reported in the argon plasma treatment of cellulose nitrates.14 Here,
direct energy transfer was shown to result in the removal of some nitrate esters
from the cellulose nitrate surface, and in the conversion of nitrate esters into new
nitrogen functionalities. Surface modification by direct energy transfer was shown

to be limited to the outermost 10 angstroms of the surface of polymeric materials.

In the NIS core level there was a reduction in the signal from the nitrate ester,
which appeared at about 408 eV, and the appearance of new peaks at 402 and
401 eV. The width of the new nitrogen peaks indicates that at least three new
nitrogen functionalities were created by the plasma treatment. Their identity
cannot be determined from the spectra alone; however, from a consideration of
the binding energy range, possible assignments are N=O and NR4+ at 402 eV
and N-N, -C-N, NR4“ at 401 eV. The shift in nitrogen peaks is shown in Figure

4.11.

147

 

 

408 ev

   
  

 

360 sec

Figure 4.11 N1s XPS spectra, the reduction in nitrate ester signal (408 eV) is
evident.

148

Radiative energy transfer has been shown to be responsible for the
conversion of nitrate esters into the new nitrogen functionalities.“ This occurred
on a much slower time-scale than direct energy transfer, but occurred to a much
greater depth into the material. In oxygen plasma both direct and radiative
energy transfer from the plasma to the surface will occur, and energy transfer
brought about changes observed by XPS in the NIS core level spectra. In argon
plasma, cellulose nitrates do not etch. However, the modified surface resulting
from energy transfer in oxygen plasma was being continually etched. The new
surface presented to the plasma, as a result of etching, was also removed by the

plasma.15

The XPS results are summarized in Table 4.6 below. The relative
amounts of carbon, nitrogen and oxygen are not significantly changing with
plasma treatment time, but the surface chemistry of the cellulose nitrate is
changing. The nitrate functional group is being reduced as evidenced by the shift
to lower binding energies. The cellulose nitrate backbone Is also being oxidized

to form esters, aldehydes or carbonyl structures.

149

 

 

Table 4.6 XPS summary table — relative C,N and 0 concentrations are
constant with plasma treatment, but functionalities have changed

 

"_Plasma At mic Concentration Percentage
Treatment N-N N=O N03 I C-C C-Q C=O, COOH I

Time 401ev 4OZev 408ev 285ev 287 ev COC 289%: C18 N15 015
OS 0 0 100 44.23 55.77 0 0 _ 46.3 10.14 43.52

905 8.96 0 91.04 34.19 40.94 16.64 8.23 . 46.8 8.43 44.73
3608 7.71 7.54 84.75 38.55 37.41 19.84 4.2 .' 47.3 8.18 44.58

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fourier Transform Infrared Spectroscopy

 

The purposes of FTIR experiments were to confirm that the bulk chemistry of
the cellulose nitrate has not been altered and to validate chemical changes from
plasma treatment. Figure 4.12 shows the FTIR spectra for neat electrospun
cellulose nitrate and samples that were plasma treated for 905 and 3605. The
only significant difference in absorbance occurs with the hydroxyl groups at 3515

cm'1 and 3200 cm‘1 and the carbonyl peak at 1750 cm".

150

-— Neat CN -- — 90 5 Plasma — 360 5 Plasma

 

 

 

 

 

 

N02 N 32
8
C
a
3 N02
2
c-o-c
OH CH2 c=o
3600 3100 2600 2100 1600 1100 600

Frequency cm-1

Figure 4.12 Full FTIR scan of neat electrospun cellulose nitrate and 90 s and
360 3 plasma treated cellulose nitrate fibers.

The absorption bands for nitrate groups were strongest. Two very strong
absorption bands at 1640 and 1275 cm"1 were evident due to the asymmetric and
symmetric N02 vibration frequencies. Other absorption bands related to the
nitrate functional groups are found at 830, 740 and 670 cm“. Figure 4.13 shows
that there is not a significant difference in nitro group absorbance for the neat

and plasma treated samples.

151

 

I0 5 90s I360s
0.12

 

110%
0.1

103%

0.08 1

   
  

0.06

N02 asymmetric stretching
N02 symmetric stretching

   

O-N02 stretching

   

0.04

0.02 ,

 

 

1640 1275 830

Figure 4.13 FTIR analysis of the nitro functional group absorbencies.

The CH stretching region (2800-3000 cm") of Figure 4.12 is characterized by
two bands of smaller intensity near 2900 and 2950 cm". Based on IR analysis of
cellulose the 2950 cm'1 band can be assigned to CH2 stretching mode of
nitrocellulose and the 2900 cm‘1 band to the CH stretching mode.26 The small
peak at 1750 cm'1 is due to carbonyl (C=O) stretching. These differences are

shown in Figure 4.14 below.

152

 

.05 905 .3605

 

0.008
0.007
0.006
0.005
0.004
0.003 1

0 002 CH stretching OH stretching C=O stretching

0.001

 

 

 

3515 3200 1750

Figure 4.14 FTIR analysis of hydroxyl and carbonyl peaks, showing an increase
in absorbance for plasma treated samples.

The very strong 1275 cm-1 band is mainly due to the N02 vibration. However,
a band of medium intensity is observed at 1365 cm", which can be assigned to
the CH bending mode. In the 1000-1200 cm'1 range a strong IR band is
observed at 1150 cm". The same band is also observed in IR spectra of
cellulose at 1162 cm'1 and of collodion at 1158 cm". This band has been found
in most polysaccharides and is believed to be characteristic of the ring structure,
C-O stretching, asymmetric stretching of C1-O-C4' bridge and C-OH bending

modes. Another strong band appears in this region at 1060 cm". This band, in
153

the spectrum of native cellulose has been assigned to the C-0 stretching mode
in the C1-O-C4' bridge. The nitrate absorption bands have a strong intensity in
the 740-950 cm'1 regions and cover most other vibration bands, which may occur
in this region. However, a weak band still appears near 910 cm'1 and is probably
also associated with the ring skeleton: this absorption band is generally related to

a(1 -4)-D-glycosidic linkages. All major absorption peaks are shown in Table 4.7.

Table 4.7 FTIR summary table of major absorbencies and the associated
chemical group assignment

nment
OH stretching

CH

C=O
bend
CH
Si-CH3
deformation
CH deformation
deformation

C-O stretching

 

154

 

The pattern observed for plasma treatment using XPS is confirmed with FTIR.
The plasma treatment reduces the nitro group and oxidizes the cellulose
backbone structure. This is most easily seen with the drastic increase in the
carbonyl peak with plasma exposure. FTIR has also been used to verify the

cellulose nitrate bulk chemistry and provide a baseline for future work.

Thermal Properties — Themial Gravimetric Analysis

Thermal analysis identified the presence on water in the resultant fiber
network. A solvent system using ethanol and acetone resulted in uniform fibers,
but the processing was hindered by the fact that the solvents were too volatile,
which resulted in excessive needle clogging, which hindered the throughput of
the process. Ethanol has a vapor pressure of 59 mm Hg and a boiling point of
78°C. It has been shown that ethanol is a suitable solvent for electrospinning.6
Acetone has a vapor pressure of 229 mm Hg and a boiling point of 56°C. The
mixture of ethanol and acetone is very volatile and does not result in residual

solvent within the electrospun fibers as shown in Figure 4.15.

Another solvent system consisted of a mixture of THF and DMF. THF has a
relatively high vapor pressure of 129 mm Hg and a relatively low boiling point of
66 °C. It is expected that all THF will be rapidly evaporated during the
electrospinning process. DMF has a low vapor pressure of 3 mm Hg and a high
boiling point of 153°C. Due to the high boiling point and low vapor pressure of

DMF, it is thought that residual DMF will remain entrapped in the electrospun

155

 

fiber network. As Figure 4.155hows, there is only a small amount of water

present in fibers that were electrospun from both solvent systems. The onset of

degradation occurs at the same temperature, 205°C and the residual value of 1.5

wt% is the same for each solvent system.

110%

100% 1
90% ~
80% -
70% 1
60% 1

Wt%

50%

40% <
30% ~
20% 1

1 0%
0%

TGA 10 wt% Cellulose Nitrate

-- -- 60:40 THF:DMF - - 70:30 Ethanol:Acetone

 

 

 

W

 

60:40 THF:DMF - 1.5%
170:30Eth'Ace-15%

 

 

 

 

 

100 200 300 400 500 600 700 800 900
Temperature (Deg. C)

Figure 4.15 TGA thennogram of the electrospun fiber networks prepared from
two different solvent systems, 60:40 THF:DMF and 70:30 Ethanol

Acetone.

156

Electrical Properties

Electrical properties of the electrospun fibers were measured to establish a
baseline for creating conductive polymer networks. Electrical Impedance
Spectroscopy was used to measure the resistance of the fiber network at 1 Hz.
This value will be used as a baseline value for the preparation of conductive
fibers. It was determined that the neat electrospun cellulose nitrate had a
resistance of 5.26E+10 ohms and a conductivity of 1.90E-09 S/cm. As expected,

the cellulose nitrate was not conductive.

Mechanical Properties - Dynamic Mechanical Analysis

Mechanical properties were measured to characterize the tensile modulus of
the electrospun fiber mats prepared from 8, 10 and 12 wt% solutions of cellulose
nitrate in a 60:40 THF:DMF solvent mixture. These values were generated from
a collection of five aligned fiber mat samples that were clamped in the DMA using
a polycarbonate window frame. The data produced was then compared to

literature data, which was only available for film samples.

A summary of tensile modulus as it relates to fiber diameter and more
importantly fiber network density is shown as Table 4.8. There is a large
difference in the modulus values, due to the considerable variance in fiber
packing density. As Figure 4.16 shows, there is a considerable difference in fiber
coverage amongst the samples analyzed. In general, it is expected that

decreasing the concentration of starting solution from 12 wt% to 8 wt% would

157

 

result in more fibers per unit volume and would therefore improve the mechanical
behavior. Film samples exhibit tensile modulus values of 1300 MPa, which are
much higher than their electrospun fiber counterparts.27 This is due in part to the
volume and density of the fiber mat; the mechanical properties are relatively low

due to the discontinuous, microporous nature of the nonwoven fibers

Table 4.8 Mechanical properties of electrospun cellulose nitrate fibrous

 

 

 

 

materials
Solution Tensile Fiber Diameter Fiber Coverage (%)
Concentration (wt%) Modulus (M Pa) (nm)
8 13.4 180 79
10 13.9 220 79
12 10.2 945 19

 

 

 

 

 

 

158

 

Figure 4.16 ESEM showing fiber coverage differences amongst A) 8 wt%, B) 10
wt% and C) 12 wt% solutions.

Electrospinning resulted in mechanical property values that were lower than
films of cellulose nitrate. This was expected since the fiber network is open and
porous, and since the bulk density of the electrospun fiber samples were much
lower than the films. In summary, it was evident that smaller fibers maximized
the space within the volume used for analysis and resulted in the highest
modulus. To prepare biosensor capture pads, the fibers must be made to
withstand multiple handling, washing and treatment steps, while still providing

improvements in terms of surface area.

Conclusions

Cellulose nitrate was electrospun to form fiber mats with individual fiber
diameters in the range of 100 to 1,000 nm. A unique rotating collector was
designed that allowed for collection of the fibers without sustaining any damage

to the fiber mat. Establishing a procedure to control fiber size and alignment will

159

allow for properties such as pore volume, surface area and capillary flow to be
managed. The concentration of cellulose nitrate in solution was found to be the
most dominant factor controlling fiber size. Increasing the concentration resulted
in increased fiber diameters due to increased viscosity causing more
entanglements, leading to a structure that was more resistant to external fiber
whipping and stretching forces. Electrospun cellulose nitrate nanofibers are
suitable for biosensor applications because they provide a large available surface

area and cellulose nitrate is a material with great protein binding ability.

The electrospun cellulose nitrate fiber mats were found to be hydrophobic in
nature, due to morphology issues, and were therefore unsuitable for capillary
flow biosensor devices. This limitation was overcome by plasma treating the
surface of the fibers. A 90 second plasma treatment was found to be most
effective for creating surface functional groups that were hydrophilic, thus
reducing the water contact angle and improving the capillary flow properties. The
plasma treatment process allowed for the formation of carbonyl (aldehydes and
ketones) and carboxylate (acids and esters) that were deteced using XPS and
FTIR. The surface chemistry modification allowed for improvements in capillary
flow by altering the surface functional groups to ones that were more hydrophilic.
This technique is ideally suited to electrospun scaffolds because, unlike other
physical and chemical modification techniques, it did not compromise the bulk

architecture of the material.

A correlation between fiber packing density and mechanical properties was

determined. In the case of 8 and 10 wt% cellulose nitrate solutions, the fiber

160

mats had similar densities and therefore resulted in similar mechanical
properties. The 12 wt% solution had larger fibers, but a much more open porous
structure and resulted in fiber mats with lower tensile modulus. The mechanical
behavior of the fiber mats was lower than films, due to the large amount of void

spaces in the fiber mats, with tensile modulus values in the 10 - 13 MPa range.

Thermal analysis was performed to determine if any residual solvent
remained in the fiber mat, which may be detrimental. It was determined that the
fiber mats did not contain any residual solvent, had a low water composition of
2wt% and the onset of thermal degradation was 205 °C. As expected, the

electrical characterization showed that the neat fibers were insulating.

Cellulose nitrate nanofibers were electrospun from a mixed solvent system of
60 wt% THF and 40 wt% DMF. The resulting fiber network was composed of
fibers that were approximately 95 nm in diameter and collected on a rotating
drum with a PVDC film coating, allowing for easy fiber harvesting and random
fiber morphology. These fibers will be used in a biosensor device that requires
capillary flow. The neat electrospun fibers were hydrophobic; therefore a plasma
treatment was developed to produce a surface that was hydrophilic. The plasma
treatment introduced new surface functional groups, such as aldehydes and
carbonyl structures on the cellulose nitrate backbone structure to increase

wettability and capillary flow.

161

References

Dieter Klemm, B. H. H.-P. F. A. B., Cellulose: Fascinating
Biopolymer and Sustainable Raw Material. Angewandte Chemie
International Edition 2005, 44, (22), 3358-3393.

Liu, H. 0.; Hsieh, Y. L., Ultrafine fibrous cellulose membranes from
electrospinning of cellulose acetate. Journal of Polymer Science
Part B-Polymer Physics 2002, 40, (18), 2119-2129.

Ma, 2. W.; Kotaki, M.; Ramakrishna, S., Electrospun cellulose
nanofiber as affinity membrane. Journal of Membrane Science
2005, 265, (1-2), 115-123.

Son, W. K.; Youk, J. H.; Lee, T. S.; Park, W. H., Preparation of
antimicrobial ultrafine cellulose acetate fibers with silver

nanoparticles. Macromolecular Rapid Communications 2004, 25,
(18), 1632-1637.

Wu, X. H.; Wang, L. G.; Yu, H.; Huang, Y., Effect of solvent on
morphology of electrospinning ethyl cellulose fibers. Journal of
Applied Polymer Science 2005, 97, (3), 1292-1297.

Shukla, 8.; Brinley, E.; Cho, H. J.; Seal, 8., Electrospinning of
hydroxypropyl cellulose fibers and their application in synthesis of
nano and submicron tin oxide fibers. Polymer 2005, 46, (26),
12130-12145.

Zhou, W.; Li, Z.; Zhang, Q.; Liu, Y.; Wei, F.; Luo, G., Gas Flow-
Assisted Alignment of Super Long Electrospun Nanofibers. Journal
of Nanoscience and Nanotechnology 2007, 7, 2667-2673.

Tong, H.-W.; Wang, M., Electrospinning of Aligned Biodegradable
Polymer Fibers and Composite Fibers for Tissue Engineering
Applications. Journal of Nanoscience and Nanotechnology 2007, 7,
3834-3840.

G. Mathew, J. P. H. J. M. R. D. J. L. C. N., Preparation and
anisotropic mechanical behavior of highly-oriented electrospun
poly(butylene terephthalate) fibers. Journal of Applied Polymer
Science 2006, 101, (3), 2017-2021.

162

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Li, D.; Ouyang, G.; McCann, J. T.; Xia, Y. N., Collecting electrospun
nanofibers with patterned electrodes. Nano Letters 2005, 5, (5),
913-916.

Image-Pro Software - MediaCybemetics -
http://wwwmediacycoml.

Gu, B. K.; Sohn, K.; Kim, S. J.; Kim, S. l., Fabrications of
Nanofibers as Crossed Arrays by Electrospinning. Journal of
Nanoscience and Nanotechnology 2007, 7, 4202-4205.

Lee, H.; Yoon, H.; Kim, G., Highly oriented electrospun
polycaprolactone micro/nanofibers prepared by a field-controllable
electrode and rotating collector. Applied Physics A: Materials
Science & Processing.

Short, R. D.; Munro, H. 8.; Ward, R. J.; McBriar, l.; Chan, H. S. 0.;
Tan, K. L., An xps investigation of me radiation-induced
degradation reactions of cellulose nitrate .1. Argon plasma induced
surface-reactions. Polymer Degradation and Stability 1991 , 31, (2),
21 1-218.

Short, R. D.; Munro, H. 3.; Ward, R. J.; Chan, H. S. 0.; Tan, K. L.,
An x-ray photoelectron-spectroscopy investigation of the radiation-
induced degradation reactions of cellulose nitrate .2. Oxygen
plasma-induced reactions. Polymer Degradation and Stability 1994,
43, (1), 67-74.

Lafuma, A.; Quere, D., Superhydrophobic states. Nature Materials
2003, 2, (7), 457-460.

Ma, M. L.; Gupta, M.; Li, Z.; Zhai, L.; Gleason, K. K.; Cohen, R. E.;
Rubner, M. F.; Rutledge, G. C., Decorated electrospun fibers
exhibiting superhydrophobicity. Advanced Materials 2007, 19, (2),
255.

Zhai, L.; Cebeci, F. C.; Cohen, R. E.; Rubner, M. F., Stable
superhydrophobic coatings from polyelectrolyte multilayers. Nano
Letters 2004, 4, (7), 1349—1353.

Shirtcllffe, N. J.; McHale, G.; Newton, M. l.; Chabrol, G.; Perry, C.
C., Dual-scale roughness produces unusually water-repellent
surfaces. Advanced Materials 2004, 16, (21), 1929.

Herminghaus, S., Roughness-induced non-wetting. EPL
(Europhysics Letters) 2000, (2), 165.

163

21.

22.

23.

24.

25.

26.

27.

Nosonovsky, M., Multiscale Roughness and Stability of
Superhydrophobic Biomimetic Interfaces. Langmuir 2007, 23, (6),
3157-3161.

Tuteja, A.; Choi, W.; Ma, M.; Mabry, J. M.; Mazzella, S. A.;
Rutledge, G. C.; McKinley, G. H.; Cohen, R. E., Designing
Superoleophobic Surfaces. Science 2007, 318, (5856), 1618-1622.

Jian-Lin, L.; Xi-Qiao, F.; Gangfeng, W.; Shou-Wen, Y., Mechanisms
of superhydrophobicity on hydrophilic substrates. Journal of
Physics: Condensed Matter 2007, (35), 356002.

Ma, M.; Hill, R. M.; Rutledge, G. C., A Review of Recent Results on
Superhydrophobic Materials Based on Micro- and Nanofibers.
Journal of Adhesion Science and Technology 2008, 22, 1799-1817.

Ma, M.; Mao, Y.; Gupta, M.; Gleason, K. K.; Rutledge, G. C.,
Superhydrophobic Fabrics Produced by Electrospinning and
Chemical Vapor Deposition. Macromolecules 2005, 38, (23), 9742-
9748.

Sekkal, M.; Dincq, V.; Legrand, P.; Huvenne, J. P., Investigation of
the glycosidic linkages in several oligosaccharides using FT-IR and
FT Raman spectroscopies. Journal of Molecular Structure 1995,
349, 349-352.

In Modern Plastics Encyclopedia, 1999; pp B158 - B216.

164

CHAPTER 5

BIOSENSOR PERFORMANCE

Abstract

This chapter describes the development of electrospun cellulose nitrate fiber
mats for use as a conductimetric biosensor for detecting Escherichia coli
O157:H7. The biosensor consists of two components: an immunosensor that is
based on electrochemical sandwich immunoassay, and a reader for signal
measurement. The architecture of the immunosensor utilizes a lateral flow
system that allows the liquid sample to move from one pad to another. The
biosensor provides a specific, sensitive, low volume, and near real-time detection
mechanism. Results are presented to highlight the performance of the biosensor

for Escherichia coli O157:H7, which is of concern for health and safety issues.

Introduction

Biosensors are a potentially new application for electrospun cellulose based
materials since they offer improved surface area over existing technologies. The
large available surface area produces higher sensitivity, decreasing the detection

time and improving the detection limit of the sensors.

165

Biosensor development and testing using the electrospun nanofibers were
conducted by collaborating students at the Nana-Biosensors Lab. Methods and
data on these experiments are presented here to show feasibility of the

electrospun nanofibers for use in biosensors.

The performance of the electrospun cellulose nitrate fibers was evaluated
using a simple detection mechanism of monitoring changes in resistance over
time using a digital multimeter following methods and approaches by
Muhammad-Tahir and Alocilja (2004), Pal et al. (2008) and Pal and Alocilja
(2009).“3 E. coli O157:H7 was used in conjunction with the appropriate
antibodies and conductive polyaniline complex to evaluate the effectiveness and

sensitivity of the electrospun cellulose nitrate materials.

The conductive biosensor operates by using an electrical signal to determine
the presence and concentration of a target analyte. The conductimetric
biosensor consists of three porous membranes, referred to as the application,

capture, and absorption pads (Figure 5.1).1

166

Analyte Nanofibers

    
 
  
 
   

KY , tfl
-1. {It
95471
_ '3 .g.

. . . '94:“ 1: ”R7
Appllcatlon Pad gisfigfi? /
13,1303 I!» Silver Electrodes

 

Capillary Flow Direction
Absorption Pad

Figure 5.1 Schematic of the biosensor device

To begin detection, a fluid sample containing the antigen (Ag) is mixed with a
conjugate of a specific antibody (Ab) and an electrically active polyaniline coated
magnetic particle in a phosphate buffered saline solution.2 Then, the Ag-Ab-
polymer complex is re-suspended in sterile deionized water for testing. This
solution is applied to the application membrane and flows across the capture pad
by capillary action through the membrane pores. Finally, on the capture
membrane, an Ab—Ag-Ab-polymer sandwich forms when the Ag-Ab-polymer
complex binds to an immobilized Ab on the capture pad (Figure 5.2).2 The
conductive polymer creates an intercalated network bridging the gap between the
electrodes, thus completing the circuit. The resistance of an electrical signal
applied to the sensor can then be measured, with the resistance being inversely

proportional to the Ag concentration of the sample.

167

Analyte

’ Nitrocellulose Nanofibers

Disconnected Electrodes Connected Electrodes

Gold or Silver Electrode

    

 

 

Nitrocellulose Nanofibers

 

 

  

 

 

Figure 5.2 Schematic showing a cross section of the capture pad on a test
strip.

Experimental

Nitrocellulose Fiber Mat Preparation

Cellulose nitrate was purchased from Scholle Corporation (College Park,
GA). Solutions of cellulose nitrate were prepared in a 60:40 (wtzwt) mixture of
tetrahydrofuran (THF):N,N-dimethylformamide (DMF). Solvents were purchased
from Sigma-Aldrich (St. Louis, MO). Suitable solutions were made by dissolving 8

wt% (cellulose nitrateztotal solvent) by magnetic stir bar mixing for 2 hours.

The electrospinning unit was purchased from Kato Tech Co (Kyoto,
Japan) and operated with electric field strength of 10 kV, syringe pump rate of
0.01 mm/min, rotating drum speeds of 9 m/min, and collection times up to 4
hours. The collector was a rotating drum type collector wrapped in polyvinylidene

dichloride (PVDC) to allow for easy fiber harvesting and resist fiber charge build

168

up. The separation distance was maintained at 10 cm between the needle tip

and drum surface.

The PVDC backed electrospun fiber mats were plasma treated in a
Plasma Sciences system for 90 seconds with a gas pressure of 0.29 torr and at a
power level of 300W. The plasma treatment reduced water contact angles from
the hydrophobic range of approximately 110 degrees to very hydrophilic around
12 degrees. This resulted in a vast improvement in capillary flow. The plasma
treated samples were mounted on thin frame made of polycarbonate using spray
adhesive. The samples were then cut into rectangular shapes approximately 2

inches by 5 inches for ease of capture pad preparation.

Electrodes were made of colloidal silver ink; which was applied to the capture
pads using an airbrush operated at 28 psi. The airbrush deposition of conductive
material was found to provide the most uniform electrodes with the least amount
of fiber substrate damage. A mask was utilized to create electrodes 2.5 cm long
and 1 cm wide with a 0.5 cm wide channel in between for testing. Individual test
strips were cut from the capture pad with electrodes so that each test strip was
2.5 cm long with a narrow electrode on each side of the 0.5 cm wide channel.

The finished test strips were stored in a sealed Petri dish.

Antibody Attachment To Capture Pad

Gluteraldehyde was used as a linking agent to attach the antibodies to the

electrospun fiber mats. The plasma treated electrospun fiber mats were rinsed

169

 

 

 

with 100 pL of distilled water then let dry for 45 minutes in the biosafety cabinet.
Then 20 uL of 0.1% glutaraldehyde (CsHaOz) solution per 1.0 in2 of fiber mat was
applied with a hand pipette using a polycarbonate coverslip. This was allowed to
dry in the biosafety cabinet for 30 minutes. The antibody attachment was
confirmed using fluorescent-tagged antibodies and confocal scanning laser

microscopy (CLSM) and the results are shown in Figure 5.3.

 

Figure 5.3 CLSM image of neat cellulose nitrate fibers and fibers with
fluorescent-tagged antibodies attached.

A thin polycarbonate coverslip was placed on top of the fiber membrane to
evenly disperse solution of the entire surface. The coverslip was removed and
the fiber mat was allowed to dry for 30 minutes in the biosafety cabinet. After
the fiber mat was dried, 20 uL of 0.5mg/ml per 1.0 In2 of fiber mat of polyclonal

Eco/i O157:H7 antibodies diluted in a phosphate buffer solution were applied to
170

the fiber membranes. A new clean polycarbonate cover slip was placed on top of
the fiber mat to ensure even distribution of the antibody solution and promote
capillary flow. The fiber mat was then sealed in a Petri dish with Parafilm and

incubated for 1 hour at 37 °C to allow the antibodies to bind to the nanofibers.

The antibody coated fiber mats were washed with 20 pL of tris (C4H11N03)
buffer with 0.1% tween-20 (C53H11402e) per 1.0 in2 of fiber mat. A new clean
polycarbonate cover slip was placed on top of the fiber mat to ensure even
distribution of the buffer solution. The fiber mats were sealed in Petri dishes and
incubated for 45 minutes at 37 °C. The fiber mats were dried for 30 minutes in
the fume hood and can be renamed capture pads. The capture pads were

stored in a 4 °C refrigerator in a sealed Petri dish when not in use.

Preparing Conjugate Solution

In order to obtain a signal from the test strips, a magnetic conductive
polymer must be used to form a bridge between the antibodies and the
airbrushed electrodes. The method by Pal et al. (2008) was used to prepare the
magnetic conductive polyaniline (PANi) solution.2 Specifically, 100 pg of
magnetic iron oxide (FE203) and polyaniline (PANi) conjugate were added to a
1.5 mL test tube. Then 1 mL of phosphate buffer solution (PBS) was added to
test tube and sonicated for 10 minutes to disperse the particles. The appropriate
amount of antibody was added to make the concentration of the solution 0.5

mg/ml. It was incubated in a hybridization oven at 25 °C at 30 rpm for 1 hour.3

171

After incubation, the test tube was vortex mixed and placed in a magnetic
separator. After 2 minutes, the supernatant was removed using a micropipette to
remove any unbound antibodies. Then 1 mL of tris buffer solution with 0.1%
casein was added to the test tube. The test tube was then vortex mixed again,
placed in the magnetic separator for 2 minutes, then the supernatant solution
was removed with a micropipette. Then 1 mL of tris buffer solution with 0.1%
casein was added to the test tube again. After the final rinse, the solution was
resuspended in 1 mL PBS. This solution was stored in a 4 °C refrigerator until

needed for test solution preparation.

Preparation of Test Solution

In order to obtain test solutions, the bacteria solution was serially diluted in
sterilized water. 100 uL of bacteria solution, 800 pL of water, and 100 uL of
conjugate solution were added to a 1.5 mL test tube. This solution was vortex
mixed and allowed to react for 30 minutes at 20 rpm and 25 °C in a hybridization
oven to increase bacteria antibody bonding. The tube was then vortex mixed
again and placed in the magnetic separator for 2 minutes. The supernatant
containing unbound bacteria was removed using a micropipette and replaced
with 1 mL of PBS with 0.1% tween-20. The test tube was vortex mixed and
placed in the magnetic separator for 2 minutes. The supernatant was removed
and replaced with 1 mL of sterilized water. The final solution was vortex mixed

and applied to the test strip by dispensing 50 uL onto strip or until flow across the

172

 

 

channel was deemed sufficient. The resistance of the test strip between

electrodes was continually monitored using a multimeter.

Test Strip Construction

Test strips were constructed using the electrospun fiber mats as capture
pads and commercially available Kim-wipe materials for application and
absorption pads. The test strips were cut to the appropriate length and mounted
onto the sensor-testing platform with double-sided tape. The Kim-wipe materials
were cut and placed on the platform for application and absorption pads at the
top and bottom of the capture pad. The airbrushed electrodes were connected to
the test platform electrodes using colloidal silver ink and allowed to dry for 10
minutes. The application pad was attached to the platform and capture pad, and
the absorption pad was constructed by placing the Kim-wipe material under the
coverslip and a small section of commercially available nitrocellulose membrane

material was placed on the Kim-wipe.

The completed test strip was covered with a polycarbonate sheet and held
in place using binder clips on a glass slide to provide uniform pressure on the
sensor channel. 50 [IL of test solution was dispensed onto the application pad.
Resistance was continually monitored using an impedance meter with 1-second
intervals for 8 minutes and recorded and plotted. Test solution was continually
monitored to ensure that the entire area of the test strip was being utilized. Once

the lateral flow across the test strip was achieved, the resistance of the device

173

 

 

was stable and ready for analysis. The tests were performed in triplicate and the
data was averaged. When the device was operational, the polyaniline complex
became bound to the antibody present on the fiber and conductivity increased via

a percolated network as shown in Figure 5.4.4

Antigen

Antibode
"va‘n

Conducting
Polymer

 
  
    

Conducting path

    
  

Affixed
Antibody

 

I"
Electrospun Fiber

Figure 5.4 Non-Conductive biosensor design.

174

The preliminary data shows that the sensor works and that it is sensitive to
the bacteria concentration (Figure 5.5). The data shows that as storage time
increases, the antibodies will denature and be less effective. Using an

immediate testing procedure provides the highest sensitivity.

Biosensor Performance

 

Average Resistance (kOhm)

 

 

 

10M 10"-2 10"-3 104-4 10"-5 10A-6 10"-7 Ctrl
Bacteria Dilution

Figure 5.5 Preliminary biosensor performance data shows that the sensor
works and is sensitive for several dilutions of bacteria.

Results and Discussion

Figure 5.5 shows the preliminary data as a bar graph. It can be seen that

further improvements would help to distinguish between the control sample and

175

the dilutions. Improvements need to be realized in terms of storage time and
testing time. Improvements are needed to increase the resistance of the control
solution. This will allow for further dilutions to be accurately detected. A
procedure needs to be developed to understand and limit the denaturing effect of

the antibodies and provide reliable reproducible data sets.

176

References

Muhammad-Tahir, Z.; Alocilja, E. C., A Disposable Biosensor for
Pathogen Detection in Fresh Produce Samples. Biosystems
Engineering 2004, 88, (2), 145-151.

Pal, 8.; Setterington, E. B.; Alocilja, E. C., Electrically active
magnetic nanoparticles for concentrating and detecting Bacillus
anthracis spores in a direct-charge transfer biosensor. Ieee
Sensors Journal 2008, 8, (5-6), 647-654.

Pal, 8.; Alocilja, E. C., Electrically active polyaniline coated
magnetic (EAPM) nanoparticle as novel transducer in biosensor for
detection of Bacillus anthracis spores in food samples. Biosensors
& Bioelectronics 2009, 24, (5), 1437-1444.

Yuk, J. 8.; Jin, J. H.; Alocilja, E. 0.; Rose, J. B., Performance

enhancement of polyaniline-based polymeric wire biosensor.
Biosensors & Bioelectronics 2009, 24, (5), 1348-1352.

177

CHAPTER 6

CONCLUSIONS

Understanding the processing conditions and solution variables that affect the
electrospinning process is essential for creating uniform homogenous nanoscale
fiber mats. Post treatment processes allow the fiber mats to take advantage of
the surface chemistry and capillary flow properties required for sensor
applications. Combining the knowledge gained from altering the processing
conditions in combination with post-treatment procedures allows for construction

of sensor substrates with tunable properties.

It was determined through a series of experiments that varied solvents,
solvent ratios and quantities that a mixed solvent system of THF and DMF
provided the best results for the electrospinning process in terms of homogenous
fiber production, limited operator interaction (no needle clogging) and ease of
solution preparation. Specifically the optimal solution was determined to be 8
wt% cellulose nitrate dissolved in a 60:40 (by weight) THF:DMF solvent system.
This solution allowed the electrospinning operation to be run continuously,

without interruption for up to 12 hours.

The ideal solution could be electrospun onto any type of collector, depending
on the desired fiber morphology. A split electrode (c-shaped) collector allowed

178

the production of aligned fiber mats. A rotating drum that was wrapped in PVDC
reduced the alignment, but provided a dramatic improvement in production rate.
It was determined that the optimum operating conditions for the Kato-Tech
Nanofiber Electrospinning Unit were 10 kV applied to the syringe needle, a
needle tip to collector spacing of 10 cm a syringe pump flow rate of 0.01 mm/min
and when using the rotating drum, a rotating drum speed of 9 m/min. These
processing conditions allowed for uniform fiber mats 2.5 inches wide and 28
inches long to be produced in 4-hour intervals. The use of PVDC on the rotating
drum allowed for the fiber mat to be easily transferred from the collector without
damaging the fiber network. Increasing the concentration of the cellulose nitrate
in solution produced larger diameter fibers, which decreased the surface area of
the fiber mat. Slowing the rotation rate of the drum decreased the fiber
alignment. Altering the size and morphology of the fibers is important to modify
the performance of the sensing device. After removing the PVDC film from the

collector, the fibers were subjected to a plasma treatment process.

Plasma treatment was performed on the electrospun fiber mats to increase
the capillary flow. Capillary flow is required for adequate biosensor performance.
It was determined that 90 seconds of oxygen plasma with a gas pressure of 0.29
torr and power level of 300W provided the optimal capillary flow properties while
maintaining good antibody adhesion characteristics. The changes in surface
chemistry were observed using XPS and contact angle measurements. Contact
angle measurements proved that the surface was becoming more hydrophilic

with plasma treatment. The water contact angle was decreased from 110

179

degrees to 12 degrees after the plasma treatment. The electrospinning
equipment had an effect on the contact angle. Plastic syringes were found to
produce fiber mats with higher, more hydrophobic, contact angles. This was due
to the strong solvents THF and DMF leaching materials from the plastic Iabware.
XPS proved that oxidation was responsible for decreasing the contact angles.
The relative amounts of carbon, nitrogen and oxygen were not significantly
changing with plasma treatment time, but the surface chemistry of the cellulose
nitrate was changing. The nitrate functional group was being oxidized as
evidenced by the shift to lower binding energies. The cellulose nitrate backbone
was also oxidized to form esters, aldehydes or carbonyl structures. It is
recommended that glass lab ware be used to limit contaminants. Plasma
treatment should be performed on the PVDC backing to maintain a rigid fiber mat

with minimal damage to the cellulose nitrate fibers.

An airbrush was operated with a line pressure of 28 psi to deposit colloidal
silver particles on the plasma treated cellulose nitrate fibers. The fiber mats were
mounted to polycarbonate sheets using spray adhesive and masked using the
rectangular steel mask with 0.5 mm width. It was determined that 4 to 5 inch
distance of airbrush gun to fiber sample produced the most uniform paint
coverage. The mask was then placed on a large magnet with the sample
between the magnet and the mask. This masking method allowed for half of the
mask to be used at one time, but significantly limited the stray silver particles in
the sensor channel. After electrode deposition, the samples were transferred

back to Biosystems Engineering department for treatment and testing.

180

The procedure was optimized to produce the highest sensitivity and best
signal to noise ratio. The samples were rinsed with deionized water twice and
allowed to dry for 30 minutes. Then 20 micro liters per square inch of 0.1%
gluteraldehyde solution was applied and dried with a coverslip for 30 minutes.
Then 20 micro liters per square inch of antibody solution was applied and
incubated at 37 ° C for 1 hour. Then 20 micro liters per square inch of tris buffer
solution was applied and incubated at 37 ° C for 45 minutes. Silver electrodes

were applied to the nanofiber mats using an airbrush.

The next step in fiber mat preparation was to cut individual test strips
using a razor blade in a guillotine fashion. The test strips were attached to the
testing platform with double sided tape and hand painted silver electrodes. A
polycarbonate coverslip was used to ensure proper flow on test solution. The
test solution was prepared by adding magnetic polyaniline-antibody solution to
control dilution (deionized water), in 102, and 10‘1 dilutions of original culture.
These were put in a hybridization oven for 30 minutes at 30 rpm and 25 ° C to
increase bacteria antibody bonding. They were then separated three times to

concentrate bacteria solution and remove other materials.

Measuring the voltage across the channel continuously for a period of 8
minutes provided the test results. A small Kim wipe at top and bottom below
cellulose fiber pads was added to improve sensor performance. A polycarbonate
coverslip was placed over the Kim wipes and capture pad regions. Then 50 pL
of the test solution was placed onto application pad (cellulose pad cut in half) and

results were logged and recorded.
181

It was determined that the sensor was sensitive to various E. coli O157:H7
bacteria concentrations, but improvements still need to be realized. The
sensitivity and testing/storage time limitations need to be improved. Reducing
the time between sensor pad constructions and testing can realize improvements

in sensitivity.
Future Work

Electrospun fiber mat production rate can be improved by increasing the
number of syringes within the electrospinning unit. Glass syringes should be
added with a minimal spacing of 5 cm between the needle tips to limit the
interactions. In order to improve alignment using the rotating drum, higher
rotational speeds must be realized. This is only possible with a higher speed
motor or altering the pulley ratio on the electrospinning unit. Fiber sizes can be
increased by increasing solution concentration and by changing fiber sizes in
conjunction with processing variables; changes in fiber mat porosity can be

realized. Alterations of porosity may have an impact on the sensor sensitivity.

Understanding the time degradation behavior of test solutions and the
electrospun fiber materials will allow for testing procedure optimization. By
improving the testing procedure, sensor performance improvements can be
achieved. A study of various fiber sizes, which will lead to fiber mat porosity
changes and capillary flow differences, should be studied to understand the

optimal conditions necessary to realize the most sensitive device.

182

 

Ill

RI
Bllll
Ull

"
Y”.
T”
S"
E”
V”
Nm

"

fl

3 03063 1125

311

l

 

lllllllllllll