Raj Balaji

download Raj Balaji

of 68

Transcript of Raj Balaji

  • 8/6/2019 Raj Balaji

    1/68

    UNIVERSITY OF CINCINNATI

    Date:

    I, ,

    hereby submit this original work as part of the requirements for the degree of:

    in

    It is entitled:

    Student Signature:

    This work and its defense approved by:

    Committee Chair:

    11/18/2009 321

    29-Oct-2009

    Balaji Raj

    Master of Science

    Electrical Engineering

    Advanced Aqueous Solutions for Low Voltage and Electrolysis-Free

    Electrowetting Devices

    Jason Heikenfeld, PhD

    Punit Boolchand, PhD

    Ian Papautsky, PhD

    Jason Heikenfeld PhD

    Punit Boolchand PhD

    Ian Pa autsk PhD

    Bala i Ra

  • 8/6/2019 Raj Balaji

    2/68

    Advanced Aqueous Solutions for Low Voltage and

    Electrolysis-Free Electrowetting Devices

    A thesis submitted toThe Division of Research and Advanced Studies

    of theUniversity of Cincinnati

    in partial fulfillment of requirement for the degree of

    MASTER OF SCIENCE

    In theDepartment of Electrical and Computer Engineering

    of the College of EngineeringOctober 2009

    By

    Balaji Raj

    B.S. in Computer Engineering,University of Cincinnati

    Committee Chair - Dr. Jason Heikenfeld

    Committee Members - Dr. Boolchand and Dr. Papautsky

  • 8/6/2019 Raj Balaji

    3/68

    ABSTRACT

    Electrowetting has come a long way since the concept was first introduced by

    Gabriel Lippmann in 1875. Since the emergence of Bruno Berges concept of

    electrowetting on dielectrics in the 1990s, electrowetting has seen an explosion of

    interest. Recent development in electrowetting applications has led to the need for low

    operating voltages for many applications. Besides a low voltage requirement, larger

    contact angle modulation, reliable, and electrolysis free electrowetting are important for

    all electrowetting applications.

    The basis of this thesis is to compare and introduce ways to attain low voltage

    electrowetting on a two-layer dielectric system. Two dielectrics are investigated, Al2O3

    and Si3N4. Surfactants are then tested in the form of sodium dodecyl sulfate (water

    soluble) and triton X-15 (oil soluble) to observe how the interfacial surface tension

    affects electrowetting effect. Once low voltage electrowetting is observed, the

    investigation of electrolysis free and reliable electrowetting is discussed. Various

    surfactants are tested for dielectric failure and a catanionic surfactant is synthesized.

    The catanionic surfactant is also tested for dielectric failure. Once those tests are

    performed, water is replaced as a medium for electrowetting by propylene glycol. With

    larger molecules for propylene glycol, dielectric failure is substantially reduced.

    The results of the work described herein show the promise for low voltage andreliable electrowetting.

  • 8/6/2019 Raj Balaji

    4/68

  • 8/6/2019 Raj Balaji

    5/68

    ACKNOWLEDGEMENT

    I wish to express gratitude towards my academic advisor, Dr. Jason Heikenfeld,

    for having the confidence in me as a student in the Novel Device Laboratory. His

    continued effort to guide, support, and encourage my research endeavors here at the

    University of Cincinnati pushed me towards new challenges, and multidisciplinary

    understanding of new concepts, and techniques associated with new technologies have

    pushed me through to attain this thesis.

    I would like to thank the members of my lab for their continuous support,

    creativity, advice, and assistance during my thesis work.

    In addition to this, I would like to express my gratitude to Dr. Punit Boolchand and

    Dr. Ian Papautsky for taking the time to be part of my thesis committee.

  • 8/6/2019 Raj Balaji

    6/68

    i

    TABLE OF CONTENTS

    CHAPTER 1: INTRODUCTION.......................1

    1.2 Background related to this Thesis...3

    1.3 In This Thesis............3

    CHAPTER 2: ELECTROWETTING BASICS AND THEORY ...5

    2.1 Contact angle on planar surfaces...........5

    2.1.1 Interfacial Surface Tension5

    2.1.2 Liquid Droplet on planar surface in air ..........6

    2.1.3 Liquid Droplet on planar surface in oil ambient ...............8

    2.2 Electrowetting on planar surfaces...9

    2.2.1 In Air or Oil ...........................9

    2.3 Limitations in electrowetting...10

    2.4 Low voltage electrowetting .......12

    2.5 Review of past work in Low Voltage Electrowetting .....14

    CHAPTER 3: LOW VOLTAGE ELECTROWETTING DEVICES .................18

    3.1 Introduction...................................18

    3.2 Fabrication.................................20

    3.3 Experimental results....22

  • 8/6/2019 Raj Balaji

    7/68

    ii

    CHAPTER 4: ION AND LIQUID DEPENDENT DIELECTRIC FAILURE IN

    ELECTROWETTING SYSTEMS..27

    4.1 Introduction...................................27

    4.2 Background and experimental results .......................................................28

    4.3 Choice, Preparation and Synthesis of Test Liquids...32

    4.4 Dielectric Fabrication and Preparation 35

    4.5 Experimental Setup for EW and Dielectric Failure Tests .36

    4.6 Results...37

    4.7 Discussion and Conclusion49

    CHAPTER 5: DISCUSSION AND FUTURE WORK.51

    5.1 Summary of this work .51

    5.2 Achievement in the context of others work..52

    5.3 Future work...53

    REFERENCES55

  • 8/6/2019 Raj Balaji

    8/68

    iii

    LIST OF FIGURES

    Figure Page

    1 Figure 1.1 Electrowetting applications that are being developed at

    the (a) University of Cincinnati [12], (b) Philips Research [13], (c)

    University of California Los Angeles [14], and (d) Duke University

    [5].

    2

    2 Figure 2.1 An aqueous droplet on a planar surface in air with

    Youngs angle.

    7

    3 Figure 2.2 An aqueous droplet on a planar surface in silicone

    oil with Youngs angle of 180.

    9

    4 Figure 2.3 Microscopic view of droplet showing three phase

    state during electrowetting.

    10

    5 Figure 2.4 Common limitations in electrowetting (a) dielectric

    charging (b) oil charging and (c) electrolysis [17].

    11

    6 Figure 2.5 Electrowetting responses for varying wt. % of SDS

    in 0.1 M NaCl solution [21].

    16

    7 Figure 3.1 Diagrams and photographs of example

    electrowetting optical devices that can be implemented in

    arrayed format on glass substrates [23].

    19

    8 Figure 3.2 (a,b) Diagrams and (c) photographs of basic

    electrowetting droplet characterization in an oil ambient [23].

    20

  • 8/6/2019 Raj Balaji

    9/68

    iv

    9 Figure 3.3 Electrowetting data on samples processed with

    varied CYTOP 809M wt. % in the spin coated fluorosolvent

    solution. The final approximate thickness, based on wt. % of the

    CYTOP 809M in the spin coated solution, is also labeled on the

    plot. Two samples were tested for each curve and the results

    averaged [23]. 1 wt. % sodium dodecyl sulfate (SDS) in water

    was used as the surfactant for this approach.

    23

    10 Figure 3.4 Diagrams of dielectric charging and its effect on

    contact angle saturation [23].

    24

    11 Figure 3.5 Electrowetting data with ~50 nm CYTOP 809M

    deposited on 150 nm Si3N4 films. All experiments utilized 0.1

    wt. % Triton X-15 in the dodecane. AC vs. DC and the addition

    of KCl vs. SDS (sodium dodecyl sulfate) were also included in

    the experiments. Two samples were tested for each curve and

    the results averaged [23].

    25

    12 Figure 4.1(a) Initial state of a sessile droplet in oil ambient. (b)

    Proper charge buildup and electromechanical alteration of the

    wetting response. (c, d, e) Non-ideal behavior due to dielectric

    charging, oil charging, or electrolysis at dielectric defects [17].

    30

    13Figure 4.2 Conductivity (S/cm) and interfacial surface tension

    (IFT, mN/m) of aqueous DTA-OS catanionic surfactant solution

    vs concentration (wt %). The cmc value for DTA-OS was

    determined to be ~0.065 wt% [17].

    34

  • 8/6/2019 Raj Balaji

    10/68

    v

    14 Figure 4.3(a) Plot of contact angle vs. electrowetting voltage

    for conventional surfactant solutions; the wt% used are above

    the cmc points [17].

    38

    15 Figure 4.3(b) Electrowetting (EW) contact angle response for

    several wt% of catanionic surfactant (DTA-OS) solution [17].

    39

    16 Figure 4.4 Dielectric failure current vs. voltage for various

    aqueous solutions on a fluoropolymer/Al2O3 dielectric stack.

    Every test, and individual positive and negative polarity voltage

    sweeps, were measured at independent dielectric sites. The

    data are presented without any averaging, allowing one to

    appreciate that the data are fairly repeatable even for dielectric

    failure tests [17].

    41-43

    17 Figure 4.5 Dielectric failure current vs voltage for(a) pure

    propylene glycol and (b) propylene glycol with 1 wt. % SDS. For

    comparison with (b) a water:SDS solution of similar conductivity

    is plotted in (c) [17].

    46-47

    18 Figure 4.6 Measured contact angle relaxation (charge injection)

    vs. time for several aqueous solutions biased at 28 V. Voltage

    was applied at ~0.5 s, and the contact angles captured with a

    VCA Optima system [17].

    49

  • 8/6/2019 Raj Balaji

    11/68

    vi

    LIST OF TABLES

    Table Page

    1 Table 4.1. Table of surfactants utilized herein. 32

    2 Table 4.2. Conductivity measurements for various liquids tested. 44

  • 8/6/2019 Raj Balaji

    12/68

    1

    CHAPTER 1

    INTRODUCTION

    Electrowetting traces its origins back to 1875 through the work of Gabriel

    Lippmann [1]. Lippmann introduced the theory of electrocapillarity [1] from which

    electrowetting has come to be. In his works, Lippmann found that capillary depression

    of mercury can be varied by the application of a voltage between the mercury and

    electrolyte solution (where the mercury was in contact with the electrolyte solution). For

    practical applications, one major obstacle was found; the electrolytic decomposition of

    water after the application of voltage in the range of only 10-3

    volts [2]. In the early

    1990s, however, B. Berge introduced the concept of using a thin insulating layer to

    separate a conductive liquid from the metal electrode [3]. The purpose of this was to

    eliminate the causes of electrolysis. This concept is now widely accepted or known as

    electrowetting on a dielectric (EWOD) [3], and has caused enormous interest in

    electrowetting devices.

    Today, electrowetting [2] is the common term used to describe the

    electromechanical [4] reduction of a liquid contact angle on a hydrophobic

    dielectric/electrode substrate. A reversible contact angle change as much as >100 can

    be achieved in oil ambient. Fast contact line velocities (>10 cm/s) and low electrical

    energy (~1s to 10s mJ/m2) are also typical. Few other micro or electrofluidic

    technologies can match this performance. There has been a global effort to use

    electrowetting for applications such as lab-on-chip [5, 6, 7], optics [8, 9], and displays

    [10, 11], to name a few. Some of these applications have also been developed here at

    the University of Cincinnati, Philips Research Eindhoven, University of California Los

  • 8/6/2019 Raj Balaji

    13/68

    2

    Angeles, and Duke University (Figure 1.1). The creation of new device platforms has

    outpaced the development of improved electrowetting materials. This thesis will attempt

    to provide new insight and improved performance for electrowetting solutions and

    dielectrics. In particular, it is a goal of this thesis to show improved low-voltage

    electrowetting.

    Figure 1.1 Electrowetting applications that are being developed at the (a) University of

    Cincinnati [12], (b) Philips Research [13], (c) University of California Los Angeles [14], and (d)Duke University [5].

  • 8/6/2019 Raj Balaji

    14/68

    3

    1.2 Background related to this Thesis

    A brief overview will be made on some recently published research on low

    voltage electrowetting. It will be discussed in more detail in chapter 2. Low voltage

    electrowetting has been achieved, as published by Berry et al. and Moon et al. Berry et

    al. showed voltages as low as ~5 V with very thin dielectric. However, the materials they

    use may not be feasible for commercial devices that are fabricated on glass or plastic.

    Moon et al. also achieved low voltage electrowetting, ~15 V, but the electrowetting

    response was marginal. Most importantly, neither group investigated the important topic

    of reliability and degradation.

    1.3 In this thesis

    Low voltage electrowetting is desirable for virtually all applications including lab-

    on-chip [5,6,7], optics [8,9], and displays [10,11]. Most electrowetting devices are

    arrayed which requires the need of thin-film-transistors [15] driven at < 15 V. In order to

    achieve this, high capacitance will be required which would reduce the overall

    electrowetting operating voltage. Therefore, herein dielectric materials and hydrophobic

    fluoropolymer surfaces will be investigated to obtain high capacitance by making dual-

    layered dielectric on a substrate. Keeping the requirements of high capacitance in mind,

    variables such as thickness of the dielectric, relative permittivity of the dielectric, and

    material of the electrode will be looked at. Surfactants will also be explored to reduce

    interfacial surface tension between the aqueous/oil interface which can further reduce

    the required electrowetting voltage.

    However, it is also known that some surfactants in solution can pose a problem

    during the electrowetting operation. The most common problems associated with

  • 8/6/2019 Raj Balaji

    15/68

    4

    electrowetting are dielectric charging, oil charging and electrolysis. Some

    fluoropolymers have pores in the nano scale and possibly the micro scale. Ions from

    these surfactants may be small enough to form a liquid wire through the dielectric layers

    causing dielectric failure in the material. This would halt or reduce the electrowetting

    effect. Therefore, also explored herein is the type of surfactant that can be used such

    that low voltage electrowetting is achieved without dielectric failure. Observed in the

    latter chapters, polar liquids that contain ions that are physically larger in size than other

    counter ions show trends of a less likely onset of dielectric failure. Even though the work

    presented herein is not entirely comprehensive, it does provide clear guidance on howcarefully selecting various liquids and ionic content leads to improved electrowetting

    performance.

  • 8/6/2019 Raj Balaji

    16/68

    5

    CHAPTER 2

    ELECTROWETTING BASICS AND THEORY

    This chapter will give a better understanding of electrowetting theory. Contact

    angle on planar surfaces will be described first followed by electrowetting on planar

    surfaces.

    2.1 Contact angle on planar surfaces

    First, the interfacial surface tensions of an aqueous liquid droplet on a planar

    substrate will be discussed. This droplet will be placed in an air ambient and a non-polar

    liquid ambient and the corresponding equilibrium state of the interfacial surface tension

    at the 3 phase contact line will be discussed.

    2.1.1 Interfacial Surface Tension

    Interfacial surface tension is the basis for contact angle of a liquid droplet on any

    given surface whether it is planar or structured, or hydrophobic or hydrophilic. By law of

    thermodynamics, when a system is in equilibrium it tries to attain the lowest energy

    state. For instance, when a liquid droplet is placed on a surface the droplet attains a

    form of semi-spherical shape. This is due to interfacial surface tension forces acting on

    it. Interfacial surface tension plays an important role in determining the shape of a

    droplet due to the various molecular interactions that take place at the interface. Liquids

    with a high surface tension tend to exhibit a stronger spherical shape, like water for

    instance, which has a high surface tension ~75 mN/m.

  • 8/6/2019 Raj Balaji

    17/68

    6

    Surface tension can be expressed as follows [16]

    A

    W

    (1)

    which is the work done per unit increase in surface area where (N/m) is surface

    tension, W is work done or change in energy and A is change in surface area. In this

    thesis, some liquids will contain salts which do not change the surface tension of the

    water. However, with the addition of surfactants, surface tension of the liquids can be

    decreased by as much as an order of magnitude.

    2.1.2 Liquid Droplet on planar surface in air

    Most polar liquids, such as water, exhibit a hydrophobic contact angle if placed

    on a non polar surface. Due to this hydrophobic surface and the high surface tension of

    the liquid, a water droplet is placed on this non polar surface and exhibits a spherical

    cap shape. The droplet contact angle is determined by the Youngs equation and thisequation is used to determine the thermodynamic equilibrium of the three interfacial

    surface tension vectors. The three interfacial surface tension vectors are the FA

    (fluoropolymer/air), WA (water/air) and WF (water/fluoropolymer), and the common

    vertex of those vectors is commonly referred to as the triple point or contact line. Figure

    2.1 provides a graphical representation of the triple phase contact point. Like any

    system at equilibrium, the vectors must be balanced. Balancing all the vectors in place

    we obtain Youngs equation [8].

  • 8/6/2019 Raj Balaji

    18/68

    7

    Figure 2.1 An aqueous droplet on a planar surface in air with Youngs angle.

    YWAFAWF cos- (2)

    where Y is the Youngs angle. Generally, liquid droplets that exhibit contact angles

    larger than 90, then the planar surface is considered hydrophobic. If Youngs angle is

    lower than 90, the surface is generally classified as being hydrophilic. However, for

    electrowetting applications, generally a large Youngs angle is preferred and therefore

    the planar surface is made such that it is hydrophobic. Most non-polar surfaces that

    contain either hydrocarbons or fluorocarbons are hydrophobic in nature. Non-polar

    surfaces containing fluorocarbons, however, exhibit higher degree of hydrophobicity

    [16]. Therefore, fluoropolymers are commonly used and contain C-F bonds that are

    covalent and are electronically screened by the presence of a large number of fluorine

    atoms in the polymer chain [16]. Contact angles for water droplets on fluoropolymer

    surfaces are generally between 110 and 120 in air ambient which is desirable for

    electrowetting.

  • 8/6/2019 Raj Balaji

    19/68

    8

    2.1.3 Liquid Droplet on planar surface in oil ambient

    If the air ambient was replaced by a non-polar liquid in the form of oil, assuming

    the same non-polar (hydrophobic) surface, a higher Youngs angle will be observed.

    This is because the oil has a surface tension closer to the hydrophobic surface than the

    air. In fact, if the interfacial tension between the oil and the fluoropolymer is close to

    zero, the Y will be close or equal to 180. This is the same case when using many

    silicone oils. For alkane oils Youngs angle is closer to ~160.

    Figure 2.2 provides a graphical representation of the Youngs angle in oil

    ambient. If the oil is chosen to have a larger density than water, the water droplet will

    sink to the fluoropolymer surface. However, there will be a very thin film of oil between

    the fluoropolymer surface and the water droplet. Oil has other effects including

    reduction in hysteresis, effects of gravity, interfacial surface tension, and also slowing

    evaporation.

    2.2 Electrowetting on planar surfaces

    In this section we describe electrowetting on planar surfaces for an air or oil

    ambient system.

    2.2.1 In Air or Oil

    Electrowetting on dielectrics (EWOD), as described by B. Berge [3], is an effect

    caused due to the induced attraction of a liquid, via application of a voltage, on a

    dielectric surface. The electrowetting system entails the use of a thin dielectric layer

  • 8/6/2019 Raj Balaji

    20/68

    9

    Figure 2.2 An aqueous droplet on a planar surface in silicone oil with Youngs angle of 180.

    (fluoropolymer) placed between two conducting electrodes, the first is the water droplet

    and the second is a conducting substrate like silicon or metalized glass. As soon as

    voltage is applied to this system, capacitive charging takes place and there exists

    charge accumulation at the water-fluoropolymer surface. The fluoropolymer layer is well

    insulated so this prevents any effects of electrolysis. Between the fluoropolymer layer

    and the substrate, an equal and opposite charge is induced as can be seen in Fig. 2.3.

    The electrowetted angle is governed by the Youngs-Lippmann equation which

    can be further simplified as the electrowetting equation. Youngs-Lippmann [2] equation

    is as follows

    WA

    YV.d

    Vcos

    2r..

    2

    1cos 0

    (4)

    where V is the electrowetted contact angle, Y is the Youngs angle at V = 0, V is the

    voltage applied, d is the thickness, 0 is the dielectric constant of free space, r is the

  • 8/6/2019 Raj Balaji

    21/68

    10

    Figure 2.3 Microscopic view of droplet showing three phase state during electrowetting.

    dielectric constant and WO is the water/air or water/oil interfacial surface tension.

    The device structure, between the water droplet and the electrode, acts like an

    equivalent capacitor and the capacitance is dependent on the thickness and dielectric

    constant of the dielectric material. When a voltage is applied between this electrowetting

    system, charge accumulation takes place close to the water droplet-dielectric interface.

    This charge accumulation is directly proportional to the applied voltage and capacitance

    per unit area of the dielectric. Due to this charge accumulation, there is an

    electromechanical force in the horizontal in-plane that alters the balance of the

    interfacial surface tension forces. This additional force is labeled in Figure 2.3.

    Increased wetting is due to this electromechanical force in the horizontal plane and is

    the component responsible for electrowetting.

    2.3 Limitations in Electrowetting

    In an ideal setup, electrowetting would generally work without any flaws based on

    theoretical calculations as explained in the previous section. However, this is not the

    case as there are limitations in electrowetting. These limitations usually occur

    depending on the material choice one uses for the dielectric or the type of oil used in an

  • 8/6/2019 Raj Balaji

    22/68

    11

    Figure 2.4 Common limitations in electrowetting (a) dielectric charging (b) oil charging and (c)

    electrolysis [17].

    electrowetting test or even the aqueous solution used. The most common limitations are

    dielectric charging, oil charging, and electrolysis. Figure 2.4 gives a better depiction of

    this.

    Figure 2.3 shows an ideal electrowetting experiment where applied voltage will

    drop across the entire dielectric and accumulation of applied charge will only take place

    in the aqueous phase. Charge injection is something that needs to be avoided for

    desirable electrowetting to take place. This can be observed in Figure 2.4a. Dielectric

    charging can cause fatigue in the fluoropolymer [18] and often provides unpredictable

    electrowetting response. More importantly, this can lead to contact angle saturation and

    increased contact angle hysteresis [19].

    Another limitation is charge injection in the oil and not within the dielectric layer.

    Oil with low-breakdown field or containing some form of soluble charged particles, seen

    in Figure 2.4b can show signs of electrowetting saturation or droplet relaxation.

    Dodecane or tetradecane oil seems to be the best choice of oil for electrowetting and

    show no signs of saturation or droplet relaxation. If there are pin-holes or other major

  • 8/6/2019 Raj Balaji

    23/68

    12

    defects that exist in the dielectric, electrolysis can occur and at low voltages, formation

    of gas bubbles can be observed in the aqueous phase, as seen in Figure 2.3c.

    In the upcoming chapters, these limitations will try to be reduced to achieve the

    best electrowetting response for various applications. In particular, it is the goal of this

    thesis to develop low voltage electrowetting but with materials such that the limiting

    phenomena in Figure 2.4 are avoided.

    2.4 Low voltage electrowetting

    Low voltage electrowetting is desired for most electrowetting applications. Fromthe electrowetting equation, equation (5), the following conditions would generally be

    desired for low voltage electrowetting:

    decreasing the dielectric thickness, d;

    decreasing the interfacial surface tension between the aqueous ambient and

    other ambient like air or oil;

    choosing an appropriate dielectric such that is has a high-K dielectric constant.

    The general trend of achieving low voltage electrowetting is to decrease dielectric

    thickness. Usually a single thin dielectric is coated on a substrate followed by a

    hydrophobic fluoropolymer layer. Both these layers have capacitance in series and it

    results in an equivalent capacitance which is generally dominated by the lower value ofcapacitance. Capacitance can be increased in accordance to the following equation:

  • 8/6/2019 Raj Balaji

    24/68

    13

    d

    .A.C

    r0 (7)

    Here C is the capacitance, A is the area of the dielectric, 0 is the dielectric permittivity,

    r is the dielectric constant and d is the thickness of the dielectric.

    However, with this approach, there are some disadvantages that could hamper

    the complete potential of increasing capacitance. Dielectric failure tends to increase for

    thinner dielectrics. Charges tend to accumulate at the fluoropolymer surface and inject

    into the fluoropolymer [20]. This charge injection is one of the reasons for electrowettingsaturation. This has been explained in the Limitations of Electrowetting section.

    The next variable that would have to varied in the electrowetting equation in

    order to obtain low-voltage electrowetting would be the interfacial surface tension

    between the water/air interface or water/oil interface. This can be carried out by the

    introduction of a surfactant in the water (water soluble surfactant) or even in oil (oil

    soluble surfactant). The effects of surfactants on interfacial surface tension will be

    observed in the latter chapters. Reduction in the interfacial surface tension between

    water/air or water/oil interface will result in lowering operating voltage for electrowetting.

    However, the oil density must be matched to the water density to reduce the effects of

    vibration or gravity on the system. With lower surface tension there is also a longer

    duration for a water droplet to rise back to its original state when applied voltage is

    removed. This can be explained by the water/oil interfacial surface tension which results

    in less force to pull back the water droplet after applied voltage is removed.

  • 8/6/2019 Raj Balaji

    25/68

    14

    One more area that could be explored to achieve low voltage electrowetting is

    the high-rof the dielectric material chosen. Most fluoropolymers or other hydrophobic

    materials that show a large initial contact angle (Youngs angle) usually have a smaller

    dielectric constant, in the range of ~2-3. Most high-rdielectric layers are not

    hydrophobic as they have strong polar domains which results in high surface energy

    and wetting of polar liquids such as water. Therefore the use of stand-alone high- r

    dielectrics for electrowetting are not suitable, resulting in the requirement of an upper

    hydrophobic fluoropolymer. Therefore the best practice is to use a multiple layer

    dielectric to achieve low voltage electrowetting. This will be seen in the next section andchapters that follow.

    2.5 Review of past work in Low Voltage Electrowetting

    In brief, some important findings from the past will be looked at here and ways to

    overcome some of the problems of low voltage electrowetting will be addressed in this

    thesis. Berry et al. [18,21] had reported very low voltage results (< 5V) using

    fluoropolymer CYTOP on thermally grown SiO2. A surfactant was used in the form of

    sodium dodecyl sulfate (SDS), in water, to lower the interfacial surface tension. This

    would allow for lower electrowetting voltages and greater contact angle modulation.

    Their group varied the thicknesses of the both the SiO2 and the CYTOP layer. The

    oxide thicknesses that were tested were 11 nm, 30 nm and 100 nm. The thicknesses of

    the CYTOP were 6 nm, 12 nm, 20 nm, 28 nm, and 50 nm. On top of these variations in

    thicknesses, the SDS was tested with varying wt. % for each sample. It was observed

    that operating voltages were as high as 14 V and as low as 3 V. The surfactants were

  • 8/6/2019 Raj Balaji

    26/68

    15

    tested with various wt. % in order to change in the interfacial surface tensions that

    would affect the operating voltage and electrowetting response. It was also briefly

    mentioned that contact angle saturation was observed and due to fluoropolymer

    charging or dielectric failure. The best results were obtained for the following scenario:

    6 nm CYTOP spin coated on 11 nm SiO2 with varying wt. % (0.03, 0.1, 1, 2) with

    operating voltages as low as 3 V. Figure 2.5 will provide a clearer picture of

    their results.

    Although the results seem promising, there still exists contact angle saturation due to

    surface charging or dielectric failure. Also, the use of SiO2 as a thermal oxide is not a

    good option for low cost fabrication on glass [18,21] and thereby reduces it potential use

    for commercial electrowetting applications such as optical arrayed devices. Furthermore

    this group reported no results on reliability, and as will be seen in this thesis it is unlikely

    the materials they use could result in reliable electrowetting devices.

    Moon et al. [22] had reported electrowetting as low as 15 V. They had tested four

    various cases. One included using a single layer of amorphous fluoropolymer (Teflon

    AF, DuPont) with thickness of 100 nm on a silicon substrate. The second case involved

    coating 12 m parylene on a silicon substrate followed by 20 nm Teflon AF. The next

    case was depositing a barium strontium titanate film (~70 nm) using metal organic

    chemical vapor deposition (MOVCD) on a Pt/Ti/Si wafer followed by a thin layer (20 nm)

    of Teflon AF. Finally, SiO2 layer (~ 100 nm) was grown on a silicon substrate followed

    by a layer of 20 nm Teflon AF. The following were observed for the various cases:

  • 8/6/2019 Raj Balaji

    27/68

    16

    Figure 2.5 Electrowetting responses for varying wt. % of SDS in 0.1 M NaCl solution [21].

    Case 1 - ~ 15 for ~100 nm Teflon AF. Dielectric failure observed at larger

    potentials due to electrolysis.

    Case 2 - ~ 40 for ~12 m Parylene and ~100 nm Teflon AF but voltages up

    to 200 V required.

    Case 3 - ~ 40 observed at 15 V for ~70 nm barium strontium titanate and ~20

    nm Teflon AF dielectric layer but titanates have been shown to have tendencies

    for increased chance of charge trapping leading to hysteresis or even

    delamination [23].

  • 8/6/2019 Raj Balaji

    28/68

    17

    Case 4 - ~ 40 observed at ~20 25 V after which contact angle saturation

    was observed for ~100 nm SiO2 and ~20 nm Teflon AF.

    The third case showed the lowest electrowetting voltage (15 V) and largest

    contact angle modulation (~ 40). However, the reported contact angle modulation is

    insufficient for optical devices such as micro prisms [22]. Also, as will be shown herein,

    other factors must be considered if reliable low voltage electrowetting is to be achieved.

    In this thesis, low voltage electrowetting devices on substrates are achieved on

    films that are fully compliant for devices such as electrowetting optics [8,9]. In addition

    to that, dielectric, liquid, and ionic surfactants were studied and optimized for greatly

    improved reliability.

  • 8/6/2019 Raj Balaji

    29/68

    18

    CHAPTER 3

    LOW VOLTAGE ELECTROWETTING DEVICES

    3.1 Introduction

    In the previous chapter, it was highlighted that low voltage electrowetting is

    required for most commercial applications. Typically such voltage should be < 15 V,

    especially for arrayed devices that are operated with thin-film-transistors [15]. Figure 3.1

    shows some examples of arrayed devices developed at the University of Cincinnati. It

    was therefore hypothesized that multi-layered dielectrics and surfactants would be

    required to decrease the operating voltage of an electrowetting device.

    A basic electrowetting system consists of a liquid/dielectric/electrode capacitor.

    Here, electrowetting is performed via application of a voltage across this capacitor. The

    dielectric layer is itself hydrophobic or coated with a hydrophobic fluoropolymer. This

    hydrophobic nature provides a liquid with a high surface tension, such as water, a

    Youngs angle that is > 110 in an air ambient and a Youngs angle > 160 in oil

    ambient. The preference of oil ambient compared to an air ambient is for the following

    reasons: (1) it reduces any effect of gravity, (2) alleviates contact angle hysteresis, and

    (3) provides the largest range of contact angle modulation. Youngs angle as mentioned

    in chapter 2 is given by the following equation:

    Y

    WOFOWFcos (2)

    TheYoungs angle is determined by the in-plane balance of interfacial surface tensions

    (, mN/m) of the water (W), oil (O), and fluoropolymer (F) phases.

  • 8/6/2019 Raj Balaji

    30/68

    19

    Figure 3.1 Diagrams and photographs of example electrowetting optical devices that can be

    implemented in arrayed format on glass substrates [23].

    In this electrowetting system, when a voltage is applied it creates an

    electromechanical force per unit length (mN/m) near the three-phase point. Figure 3.2

    shows the resulting decrease in the liquid contact angle due to electrowetting. This

    device structure and oil ambient of Figure 3.2 will be the structure used for low-voltage

    electrowetting.

  • 8/6/2019 Raj Balaji

    31/68

    20

    Figure 3.2 (a,b) Diagrams and (c) photographs of basic electrowetting droplet characterization

    in an oil ambient [23].

    3.2 Fabrication

    In general, electrowetting films can be fabricated on silicon, glass and even

    plastic substrates such as polyethylene-naphthalene as long as the maximum

    processing temperatures are kept between ~ 120 to 180 C. The main substrate layer

    can be an electrode such as Silicon (Si) or for applications related to optics, a

    transparent Indium Tin Oxide (ITO, In2O3:SnO2) can be deposited on a glass or plastic

    substrate. In order to achieve low-voltage electrowetting, equation (5) in chapter 2, it

    can be understood that energy needs to be stored in the dielectric capacitor. It is known

    that even the best fluoropolymers are insufficient for low voltage electrowetting

  • 8/6/2019 Raj Balaji

    32/68

    21

    whatever the thickness may be. As fluoropolymer thickness decreases, the thickness

    can reach the size of a defect such as a pore or other electrically conductive pathway.

    So in its standalone state, fluoropolymers will not achieve the desired high-capacitance

    for low voltage. Therefore, it was hypothesized that using a multi-layered hydrophobic

    dielectric would achieve low-voltage electrowetting.

    For this work, ~100 nm Aluminum Oxide (Al2O3, ~9) and ~150 nm Silicon

    Nitride (Si3N4, ~7) were investigated, and they both have breakdown fields that exceed

    2 MV/cm. In the past, other materials have been utilized such as Silicon Dioxide (SiO2)

    [18,22] and Barium and Strontium Titanate (BaTiO3, SrTiO3) [20,22]. However, this workand past observation have shown that SiO2 can suffer from electrochemical attack

    (blistering) and the titanates showed increased occurrence for charge trapping

    (hysteresis) or even delamination. For applications such as lab-on-chip, these problems

    can be overlooked, but for optics it could be problematic due the requirement of > 106

    device cycles. Therefore further reliability testing may be required for all materials used

    here. Two approaches were taken using the same fluoropolymer.

    The first approach involved using an inorganic dielectric in the form of Aluminum

    Oxide (Al2O3) which was deposited on a Si substrate with atomic layer deposition

    (Cambridge Nanotech Savannah 100 atomic layer deposition tool). Excellent

    dielectric properties with ~100 nm Al2O3 were observed with deposition temperatures

    ranging to as low as 120 C. This was followed by spin coating and thermal curing of

    various wt. %, 0.5, 1.0 and 2.0 wt. % of CYTOP 809M in fluorosolvent. The varied wt. %

    of CYTOP 809M was explored to determine how variation in fluoropolymer thickness

    with a constant single dielectric layer of Al2O3 would affect low-voltage electrowetting.

  • 8/6/2019 Raj Balaji

    33/68

    22

    To obtain optimal surface smoothness for CYTOP 809M, annealing >160 C for ~15 min

    was determined. Better smoothness reduces the effect of electrowetting hysteresis.

    However, no substantial degradation was observed in electrowetting response at lower

    annealing temperatures. Dodecane oil was used for all electrowetting experiments and

    the electrowetting liquid was an aqueous solution that will be discussed later.

    The second approach involved using a different inorganic dielectric, Si3N4 and

    ~150 nm Si3N4 was deposited on a glass/ITO electrode by plasma enhanced chemical

    vapor deposition (external commercial source). This was followed by depositing a thin

    layer of a fluoropolymer, CYTOP 809M ( ~2), via spin coating and thermally cured at ~180 C on the substrate. ~ 50 nm of 1 wt % solution of CYTOP 809M was spin coated

    and annealed for this second approach. Both the first and second approaches resulted

    in a high-capacitance double-layered dielectric.

    3.3 Experimental Results

    The aqueous solution tested was 1 wt. % sodium-dodecyl-sulfate, a surfactant, in

    water. This solution lowers the interfacial water/oil surface tension (eq. 2) (WO) by ~10X

    to ~5 mN/m [17,23] thereby reducing required voltage predicted by eq. 2. Figure 3.3

    shows electrowetting results with varied CYTOP 809M thickness as deposited on Al2O3

    layer. A DC potential was applied for these electrowetting tests. From the observed

    data, it shows that an optimal thickness for CYTOP 809M is ~50nm for desired low

    voltage electrowetting. Anything above or below 1 wt. % of CYTOP 809M appeared to

    result in charging of the fluoropolymer surface. The effects of charging are diagrammed

    in Figure 3.4 and can be experimentally observed from the data in Figure 3.3. In

    summary:

  • 8/6/2019 Raj Balaji

    34/68

  • 8/6/2019 Raj Balaji

    35/68

    24

    Figure 3.4 Diagrams of dielectric charging and its effect on contact angle saturation [23].

    Based on the experimental observations of Figure 3.3, one may ponder as to

    why the thicker 100 nm CYTOP 809M does not show charging until higher voltages, yet

    also did not attain the minimum contact angle. One way of explaining it could be that the

    fluoropolymer CYTOP 809M has a low relative permittivity r ~2, and at ~100 nm the

    CYTOP 809M limits the overall capacitance compared to the ~100 nm Al2O3 (r ~9).

    This results in a lower electromechanical force that would drive the electrowetting

    process. Since electrowetting has d1/2 dependence, based on the electrowetting

    equation discussed in chapter 2, and because ideally CYTOP 809M breakdown voltage

    has a ddependence, thicker dielectrics (fluoropolymer CYTOP 809M in this case) with

    lower permittivity should exhibit lower final contact angles at higher voltages. However

    this is not the case as seen from Figure 3.3. The results in Figure 3.3 were found to be

    repeatable, and the reason for the poorer performance of the CYTOP 809M (~100 nm)

    is not understood at this time.

    The next sets of experiments were performed on the glass/ITO/Si3N4 substrate,

    the CYTOP 809M was fixed at ~50 nm thickness and various surfactants tested. Figure

    3.5 shows the results of the experiments. All experiments used 0.1 wt. % Triton X-15 oil

  • 8/6/2019 Raj Balaji

    36/68

    25

    Figure 3.5 Electrowetting data with ~50 nm CYTOP 809M deposited on 150 nm Si3N4 films. All

    experiments utilized 0.1 wt. % Triton X-15 in the dodecane. AC vs. DC and the addition of KCl

    vs. SDS (sodium dodecyl sulfate) were also included in the experiments. Two samples were

    tested for each curve and the results averaged [23].

    soluble surfactant in dodecane, and either 1 wt. % sodium dodecyl sulfate or 1 wt. %

    KCl. The experiments included the comparison of AC vs. DC as well as the addition of

    SDS vs. KCl with respect liquid conductivity. The validity of the AC tests were ensured

    by using ionic KCl in the absence of SDS. Therefore electrowetting capacitor charging

    would not be RC time-constant limited. As seen from the Figure 3.5, lower contact

    angles are observed with AC bias. There are two main reasons for this:

    At 1 kHz, AC bias, there isnt sufficient time for the CYTOP 809M layer to charge

    and cause saturation.

  • 8/6/2019 Raj Balaji

    37/68

    26

    At AC bias, there is a lower chance of hysteresis since at the microscopic level,

    the contact line remains in continuous motion and therefore has a higher chance

    to surpass microscopic defects that cause hysteresis (contact line pinning).

    Another observation can be made in Fig. 3.5, where combining SDS surfactant in

    aqueous phase and Triton X-15 in oil reduces the final electrowetting contact angle to

    60 with only 8V operation. This is due to the reduction in WO and therefore less surface

    tension force to counter act the electrowetting effect. In conclusion, CYTOP 809M

    thickness was optimized and surfactants explored to achieve electrowetting well below

    15 V. However, reliability of such systems was not certain, and is the topic of the next

    chapter.

  • 8/6/2019 Raj Balaji

    38/68

    27

    CHAPTER 4

    ION AND LIQUID DEPENDENT DIELECTRIC FAILURE IN

    ELECTROWETTING SYSTEMS

    4.1 Introduction

    In order to attain low voltage electrowetting, based on the electrowetting equation

    [2], it is understood that lowering the interfacial surface tension between the oil and

    aqueous phases can help achieve this. The use of ionic surfactants can help achieve

    lower interfacial surface tension; however, the ionic content present can increase

    electrolysis at dielectric defects and possibly cause electrochemical corrosion of

    electrodes. Another way to obtain low voltage electrowetting would be to decrease the

    thickness of the dielectric thereby increasing the capacitance. This approach, however,

    increases the risk of device failure via current flow due to the increased electric field

    across the dielectric. It is therefore not surprising that the first commercial electrowetting

    products used high voltage and thick dielectrics, such as the high voltage ~60 VAC

    liquid lenses (Varioptic SA) with a thick (~3-5 m) Parylene dielectric. For reliable low

    voltage electrowetting, a better understanding of the nature of the dielectric failure in

    electrowetting systems is required. Dielectric failure modes in electrowetting might be

    distinct from failure in solid-state systems. It was suspected that using liquid electrodes,

    due to their unique ability to propagate through complex dielectric defects, could result

    in increased dielectric failure. In this chapter it will be seen that the larger the physical

    size of the polar liquid molecule, the lower the chances of dielectric failure. Similar

    results will be seen for cations and anions that are of distinct size. Positive and negative

  • 8/6/2019 Raj Balaji

    39/68

    28

    voltage sweeps will be implemented to reveal any trends in dielectric failure. The work

    herein, although not comprehensive, provides initial guidance, and shows careful

    selection and synthesis of liquid and ionic content and how it leads to improved

    electrowetting reliability.

    4.2 Background and experimental results

    Choice of materials and experiments will be first reviewed via some background.

    It should be noted that the use of a polar liquid such as deionized water is not a

    practical option for electrowetting in areas that include bioapplications such as lab-on-

    chip as the ionic content is crucial to the health of proteins, cells, DNA, etc. and in

    electrowetting optics an AC voltage is preferred due to minimization of contact angle

    hysteresis [25]. The DC electrowetting response is poor for low-conductivity liquids such

    as propylene glycol and to properly accumulate charge near the contact line, ionic

    content must be added. An assumption can be made that for all or most real-world

    electrowetting applications, ionic content is a requirement. Another assumption that will

    be made is the need for a two-layer dielectric stack (Figure 4.1a) for low-voltage

    electrowetting devices due to the naturally porous nature of the fluoropolymer film. This

    two layer approach allows any conductive pores in the stand-alone thin fluoropolymer to

    terminate against the strong insulation properties of the inorganic dielectric. However,

    even these inorganic oxide or nitride dielectrics show marginal improvement in dielectric

    reliability. Any possibility of self-healing failure [26] is barred for a system using a liquid

    electrode. Close attention must therefore be given for selection of ionic content to allow

    reliable, low voltage electrowetting. Other than dielectric failure, the electrowetting effect

    can degrade for several other reasons. This has been briefly discussed in chapter 2. In

  • 8/6/2019 Raj Balaji

    40/68

    29

    the ideal case of electrowetting, it is assumed that all the ions or charges remain in the

    liquid and solid electrodes. Figure 4.1 shows that ions that appear anywhere else but

    the solid or liquid electrodes can cause a decline in electrowetting response. It can be

    observed in Figure 4.1c that fluoropolymer charging [19] can cause contact angle

    saturation (relaxation). The same can be said about inadequate insulation [27] of the oil,

    as seen in Figure 4.1d, which results in a degraded response to electrowetting [27].

    Electrolysis, seen in Figure 4.1e, can also degrade electrowetting response by

    damaging the electrowetting surface or by causing a voltage drop across the liquid

    electrode. It was therefore postulated that in order to decrease the effects ofelectrowetting degradation, larger size ions should be used as they will generally have a

    lower chance of penetrating the insulating materials such as the dielectric or oil phase.

    Tests will be performed for both dielectric failure and dewetting (saturation) for a variety

    of ionic solutions. All tests will be performed with only DC voltage. Once reliable tests

    are performed for DC voltages, AC voltages can be tested to reduce propagation of ions

    through the dielectric layer.

  • 8/6/2019 Raj Balaji

    41/68

    30

    Figure 4.1(a) Initial state of a sessile droplet in oil ambient. (b) Proper charge buildup and

    electromechanical alteration of the wetting response. (c, d, e) Non-ideal behavior due to

    dielectric charging, oil charging, or electrolysis at dielectric defects [17].

  • 8/6/2019 Raj Balaji

    42/68

    31

    4.3 Choice, Preparation and Synthesis of Test Liquids

    Herein two polar liquids were tested, deionized (DI) water and propylene glycol. It

    is known that H2O partially dissociates into H3O+ (hydronium) and OH- (hydroxide) at

    300 K. Under an applied electric field, further dissociation occurs [28]. Even though

    deionized water can be marginally conductive for DC electrowetting, it is often

    considered impractical in many devices or devices biased with AC voltage. Propylene

    glycol was also tested as it contains larger size molecules. Small inorganic ions

    (hydrodynamic radii ~1-2 ) were also tested using inorganic salts such as NaCl and

    KCl. Larger ions were explored using an anionic surfactant, sodium dodecyl sulfate

    (SDS) and a cationic surfactant dodecyltrimethylammonium chloride (DTAC). These can

    be compared in Table 4.1. Nonionic surfactants were not tested due to their unlikely

    impact on dielectric. Any aqueous surfactant was also chosen that contains a large

    molecule for both the anion and the cation. The following process was used to custom-

    synthesize a catanionic surfactant, dodecyltrimethylammonium octanesulfonate (DTA-

    OS). 1g of DTAC and 1g of sodium octanesulfonate (SOS) were each separately

    dissolved in methanol (99.8%, M-TediaCorp.) to make two clear solutions containing

    0.867 g (3.787 x 10-3 mol) of docecyltrimethylammonium cation and 0.894 g (4.6296 x

    10-3 mol) of the octanesulfonate anion. These two solutions were then combined at

    room temperature and a final clear solution was formed.

  • 8/6/2019 Raj Balaji

    43/68

    32

    Table 4.1. Table of surfactants utilized herein.

    Surfactant Structure

    SDS

    SOS

    DTAC

    DTA-OS

    The ion-exchange reaction is instantaneous and goes to completion. This solvent

    was then evaporated leaving behind ~2 g of solid residue which consisted of inorganic

    salt (NaCl) and catanionic surfactant (DTA-OS). The residue was then dissolved in ~ 75

    mL of DI water and placed in a dialysis tube (Fisherbrand) having a porosity of 12000

    14000 Da and dry cylinder diameter of 28.6 mm. This porosity should allow organic

    ions, to a lesser degree, to diffuse through it compared to other individual inorganic ions

    in solution, and micelles or multimolecule phases to an even lesser degree of diffusion.

    Conductivity measurements were then measured using an electrode of a conductivity

    meter (OAKTONCON6/TDS6 measured in S/cm, 0.5%). This electrode was placed

    in the dialysis tube, and the tube was sealed and submerged in a continuous flow of DI

    water. Conductivity was measured at 30 min intervals until a time of 6 hrs at which time

    the conductivity became constant which indicated that most of the inorganic salt had

    diffused out of the tube. The remaining solution containing DTA-OS was removed from

  • 8/6/2019 Raj Balaji

    44/68

    33

    the dialysis tube and heated to evaporation. 0.968 g of solid residue was obtained after

    evaporation. From earlier tests performed, it was determined that NaCl and sodium

    octane sulfonate do not dissolve in chloroform (99.9%, M-Tedia Corp.). To verify the

    complete removal of NaCl and identify any excess unreacted sodium octane sulfonate,

    ~0.2 g of the residue was redissolved in chloroform. The resulting solution was cloudy

    signaling the presence of NaCl salt and/or sodium octanesulfonate. The cloudy

    precipitate was removed by passing it through a filter paper. The remaining filtrate was

    heated to evaporate the chloroform and obtain ~0.135 g of solid DTA-OS surfactant. To

    further confirm the purity of the DTA-OS, a 0.05 wt. % aqueous solution was tested forCl- ions by adding ~ 5 drops of 0.1 M silver nitrate (AgNO 3) in ~20 mL of DTA-OS

    solution. The test was negative i.e. no precipitate of silver chloride (AgCl) was observed.

    Since the critical micelle concentration (cmc) for DTA-OS was not available, it had to be

    experimentally determined by measuring the conductivity vs. concentration (wt %) of the

    DTA-OS solution. The graph of Figure 4.5 are conductivity (, S/cm) and surface

    tension ( in air, mN/m) vs concentration (wt. %) of the DTA-OS solution. Due to

    reduced electrophoretic mobility for larger micelles, lamellar sheets, and precipitates

    [29,30,31], conductivity increases more slowly above the cmc point. The cmc was found

    to be ~ 0.065 wt. % (1.54 mM). This is lower than the cmc for SDS (~0.24 wt. %, 8.32

    mM) which is expected from a catanionic surfactant as both the component ions (DTA-

    and OS-) are amphiphilic. The cmc for DTA-OS was further confirmed by measuring thesurface tension vs. concentration (wt. %) of the DTA-OS solution (Figure 4.2). In order

    to measure the surface tension of the DTA-OS solution, the pendant drop function of the

    VCA Optima contact angle measurement system was utilized.

  • 8/6/2019 Raj Balaji

    45/68

    34

    Figure 4.2 Conductivity (S/cm) and interfacial surface tension (IFT, mN/m) of aqueous DTA-

    OS catanionic surfactant solution vs concentration (wt. %). The cmc value for DTA-OS was

    determined to be ~0.065 wt.% [17].

    4.4 Dielectric Fabrication and Preparation

    As mentioned earlier, thicker the dielectric materials like Parylene C (~3-5 m)

    can exhibit excellent resistance to dielectric failure. However, this comes at the cost of

    requiring a high operating voltage (~50 - 100 V). Therefore, for low voltage (

  • 8/6/2019 Raj Balaji

    46/68

    35

    aluminosilicate glass substrates that are coated with a transparent conducting

    electrode. Indium tin oxide (SnO2:In2O3, 100 /sq, ~50 nm thickness, display grade)

    seemed to be the most suitable transparent electrode and was purchased from PG&O

    Inc. 100 nm Al2O3 (r ~ 9.1) was coated on this transparent conductive electrode via

    atomic layer deposition using a Cambridge Nanotech Savannah 100 ALD System at

    250 C. Two precursors were used namely trimethylaluminum (Sigma-Aldrich) and DI

    water. There was also a precursor pulse time and N2 purge time used. They were 0.015

    s and 8 s respectively. A thin layer of fluoropolymer was then spin-coated. The

    fluoropolymer used was a 1 wt % solution of Asahi CYTOP 809M in fluorosolvent Ct.Solv. 180. The spin cycle consisted of a 500 rpm spread for 15 s and a 1000 rpm spin

    for 45 s. The next step was an annealing step and the sample was annealed at 180 C

    for 30 min. The resulting fluoropolymer thickness was ~50 nm. Another fluoropolymer in

    the form of DuPont Teflon AF was not investigated due to past observations that

    showed the surface to suffer from charge injection for thin films (

  • 8/6/2019 Raj Balaji

    47/68

    36

    viewed through the transparent side walls of the acrylic box. A tungsten cat whisker

    probe tip (0.5 m in diameter) was inserted into the polar droplet in order to provide

    electrical bias to solution and the other end of the probe tip was connected to a Trek

    amplifier (model 603 A) which was further coupled to a Tektronix AFG 310 function

    generator. 1 V/s step of voltage was applied to the system by using a LabVIEW

    program and the droplet profile was taken by synchronized video capture. The

    SnO2:In2O3 was connected to electrical ground.

    The second set of experiments was the dielectric failure tests. These tests were

    performed in air (not in oil) as the Youngs contact angle (Y) of the polar droplet is

    160 in oil. The reason for a lower Youngs angle is due to the

    reduction in the electrical influence of increasing capacitor area with voltage

    (droplet/dielectric/SnO2:In2O3). Another possibility is using a completely fixed area

    capacitor [32]. This was not explored as it eliminates high-field generation near the

    sharp contour of the contact line and this effect must be included if dielectric failure

    measurements are to be considered as meaningful. For each dielectric failure test, the

    voltage was swept from 0 to 30 V at 1 V/s. Both the positive and negative polarity

    voltage sweeps were independently measured at fresh sample locations, else failure in

    one polarity could damage the dielectric and influence failure in the other polarity. The

    dimension of the contacting droplet was ~2 mm in diameter for each experiment. The

    tests were carried out three times for each polarity and for each solution. Three current-

    voltage plots were made and plotted independently in order to show any variations in

    the voltage for dielectric failure.

  • 8/6/2019 Raj Balaji

    48/68

    37

    4.6 Results

    First, electrowetting data were plotted. These can be observed in Figure 4.3 and

    they confirm that the ability of surfactants to decrease ao and the required

    electrowetting voltage. In order to obtain the ao, the pendant drop method was used in

    oil ambient. The ao of DI water was measured to be ~53 mN/m and 1 wt. % SDS was

    measured as ~5.7 mN/m. It can be deduced from Figure 4.3b that the higher the wt % of

    DTA-OS, the lower the initial contact angle (Y). This is probably due to the lower

    interfacial surface tensions at the aqueous solution-fluoropolymer interface and the

    aqueous solution-oil interface.

    The plots of Figure 4.4 show the results of the dielectric failure tests performed

    with various solutions. In Figure 4.4a, no dielectric failure is observed on the DI water

    for both positive and negative polarities of applied voltage. If there was a dielectric

    defect present on the surface, due to the low conductivity of DI water (Table 2) any

    measurable electrolysis could be considered irrelevant. It would be more meaningful if a

    measurement was made by adding small inorganic ions (salt) to DI water. As seen in

    Figure 4.4b, dielectric failure is observed when an aqueous 1 wt. % solution of KCl is

    tested. Since it is known that atomic layer deposition produces high-density, pin-hole

    free Al2O3 films, the dielectric failure observed could not be from the result of ion

    conduction through large defects such as pinholes. With standard metal electrodes, the

    dielectric failure electric field (Ebd) is >30 V/100nm for Al2O3. It is therefore clear that

    with a liquid electrode, the small inorganic ions and water are able to propagate through

  • 8/6/2019 Raj Balaji

    49/68

    38

    Figure 4.3(a) Plot of contact angle vs. electrowetting voltage for conventional surfactant

    solutions; the wt.% used are above the cmc points [17].

    even the smallest pathways in the dielectric. A brief discussion can be about the polarity

    dependence of dielectric failure in Figure 4.4b.

    In general, it is accepted that mono-valent anions such as Cl - and OH- typically

    adsorb at the neutral polymer surface. It has even been validated that there exists

    preferential anion adsorption for electrowetting on fluoropolymers [33]. The adsorbed

    ions generally do not move easily under very high electric field and if they do move, they

    are not as mobile as the ions are positioned further away from the surface. Via Debye

    screening, the surface adsorbed anions locally deplete additional anions. Therefore, this

    could be a

  • 8/6/2019 Raj Balaji

    50/68

    39

    Figure 4.3(b) Electrowetting (EW) contact angle response for several wt.% of catanionic

    surfactant (DTA-OS) solution [17].

    source of explanation as to why stand-alone fluoropolymer film exhibits water

    electrolysis at lower voltages for positive bias. Positive ions comprise the mobile Debye

    sheath inside a small pore. This might help explain the dependence of failure in Figure

    4.4b where is shows clearly that the K+ ions are more easily driven into the pathway for

    dielectric failure. However, it must be clarified that in the experiments, afluoropolymer/Al2O3 stack was used and the Al2O3 favors surface adsorption of cations

    instead of anions [34], and this could alter the conclusions made earlier. Further

    exploration of this was not performed herein and could be an involved investigation in

  • 8/6/2019 Raj Balaji

    51/68

    40

    itself. From further observation of the next set of data, it will be determined that size of

    ions has a much stronger effect on dielectric failure than ion polarity. In this present

    work, concluding on ion-size dependence of failure is the primary goal.

    A commonly used anionic surfactant, SDS, was tested. It has been consistently

    seen that most low voltage devices at the University of Cincinnati [9] show electrolysis

    when a positive bias is applied to SDS solution but not observed when a negative bias

    is applied. Therefore, to better understand this phenomenon, a more thorough

    investigation is presented herein. Figure 4.4c shows that small inorganic Na+ ions

    penetrate the dielectric easily, under a positive bias, whereas the larger dodecyl sulfate

    (DS-) ion is clearly less able to penetrate the dielectric, under negative bias. One might

    wonder if this is a manifestation of the Stern layer and Debye screening effect proposed

    for the polarity dependence of failure for the KCl solution (Figure 4.4b). In order to

    confirm ion size dominance for dielectric failure, a cationic DTAC surfactant was tested.

    Figure 4.4d shows that the failure trend is similar to but opposite to that of Figure 4.4c,

    where in the positive bias the large dodecyltrimethylammonium (DTA+) group is unable

    to penetrate the dielectric. In the negative bias, it can be seen that the Cl- ions penetrate

    the dielectric surface more easily. From the comparison of current-voltage results for

    SDS and DTAC, both of which are 1 wt. % in solution and have similar conductivity

    (~1000 S/cm, Table 4.2), it supports the argument that ion size is indeed a dominant

    factor for dielectric failure in electrowetting devices.

  • 8/6/2019 Raj Balaji

    52/68

    41

  • 8/6/2019 Raj Balaji

    53/68

    42

  • 8/6/2019 Raj Balaji

    54/68

    43

    Figure 4.4 Dielectric failure current vs. voltage for various aqueous solutions on a

    fluoropolymer/Al2O3 dielectric stack. Every test, and individual positive and negative polarity

    voltage sweeps, were measured at independent dielectric sites. The data are presented without

    any averaging, allowing one to appreciate that the data are fairly repeatable even for dielectric

    failure tests [17].

    The next solution explored was the catanionic DTA-OS surfactant. It was

    explored as it may provide the essential electrical conductivity for electrowetting devices

    but have a greatly reduced chance of dielectric failure. The main factors that could help

    support this are that this surfactant contains both cation and anion that are large due to

    the long alkane chain attached to them, and a lower cmc point making the solution more

    electrically resistive. The dielectric failure results shown in Figure 4.4e do indeed

    confirm that the DTA-OS is free from failure in both the positive and negative polarities.

  • 8/6/2019 Raj Balaji

    55/68

  • 8/6/2019 Raj Balaji

    56/68

    45

    exhibit least drag for ion flow. Since it may be difficult to design a material that would

    have the ability to completely prevent water penetration, an alternate approach will have

    to be taken. It was then postulated that another solvent be used in the form of propylene

    glycol ( 99.5%, Fisher Chemical) due to its larger liquid molecule size. This would

    reduce the chance of liquid penetration into the dielectric. Pure propylene glycol was

    found to be very electrically resistive, and from initial tests to even allow any form of

    electrowetting to take place with DC voltage. Therefore no dielectric failure could be

    observed. This can be observed in Figure 4.5a. When a surfactant, in the form of 1 wt.

    % SDS, was added to propylene glycol the conductivity increased to ~31S/cm. There

    was also strong DC electrowetting response (143 to 57 for 0 to 16 V bias). It should

    also be noted that for the propylene glycol:SDS solution, no dielectric failure was

    observed in the positive bias (Figure 4.5b) unlike that which was observed for the

    water:SDS (Figure 4.4c). But lack of dielectric failure here could be due to the lower

    conductivity of the propylene glycol:SDS solution therefore a similar conductivity of

    water:SDS solution was made and tested (Figure 5.5c). Dielectric failure was still

    observed in the positive bias for water:SDS. It is therefore likely that propylene glycol is

    less likely to create a liquid wire compared to water through a dielectric. Therefore for

    increased reliability of some electrowetting devices (non-aqueous), propylene glycol

    might be a more useful liquid. However, an additional co-solvent (medium size

    molecule) will be required to reduce the liquid viscosity for fast fluid motion [36]. Otherfactors affecting liquid wire formation include surface adsorption along the walls of a

    pore and it could be

  • 8/6/2019 Raj Balaji

    57/68

    46

  • 8/6/2019 Raj Balaji

    58/68

    47

    Figure 4.5 Dielectric failure current vs voltage for(a) pure propylene glycol and (b) propylene

    glycol with 1 wt. % SDS. For comparison with (b) a water:SDS solution of similar conductivity is

    plotted in (c) [17].

    different for propylene glycol as compared to water. But this will not be explored and

    may be of interest in future studies.

    Most of the tests presented thus far primarily deal with dielectric failure of

    electrowetting devices. It was later decided to explore ion-size relationship towards the

    onset of electrowetting saturation (charge injection) [25]. One way to observe saturation

    due to charge injection is to look at the time-dependent contact angle saturation. The

    time-dependent contact angle relaxation for DI water and SDS can be observed in

    Figure 4.6. Applied voltages are well beyond the saturation (28 V). The data presented

  • 8/6/2019 Raj Balaji

    59/68

    48

    shows an ion dependence on the contact angle relaxation. The least relaxation is shown

    by the case that drives the largest ion (dodecyl sulfate, C12SO4 at -28 V) into the

    dielectric. Other factors will have to be considered but this preliminary study supports

    previous conclusions made. Relaxation behavior can be affected by interfacial tension

    and time-dependent interfacial tension [37]. It can also be said that OH- ions in the SDS

    solution should generally have the ability to inject into the dielectric unless the dodecyl

    sulfate somehow screens similarly charged ions at the pore entrances. After comparing

    the data from Figure 4.5c and Figure 4.6, it can be suggested that ion-dependent

    dielectric failure may have similar implications on understanding contact anglerelaxation due to charge injection. The similarities between dielectric failure and

    saturation (charge injection) are not unexpected.

  • 8/6/2019 Raj Balaji

    60/68

    49

    Figure 4.6 Measured contact angle relaxation (charge injection) vs. time for several aqueous

    solutions biased at 28 V. Voltage was applied at ~0.5 s, and the contact angles captured with a

    VCA Optima system [17].

    4.7 Discussion and Conclusion

    Based on all the data presented here on electrowetting failure mode, it can be

    deduced that it is highly dependent on ion size. This work provides initial guidance to

    electrowetting practitioners, even though the materials tested are far from

    comprehensive. The dielectric failure voltages reported in these findings far exceed the

    actual required electrowetting voltage. It should not be assumed that without dielectric

    failure, reliable and long-lived electrowetting operation is always guaranteed. It is nearly

    always the case that the further one operates from the dielectric failure voltage of a

    dielectric (for solid and liquid electrodes), the less the electrical degradation of the

  • 8/6/2019 Raj Balaji

    61/68

    50

    dielectric or electrode. Therefore, it has been suggested that the use of large ions in

    electrowetting liquids can help improve reliability as the ions eliminate the root problem

    (charging penetration) rather than simply delaying visible charging effects. Other factors

    must be considered, however, such as electrochemical reactivity of the dielectric and

    electrode. In this work, no such degradation was observed as a result of the small

    electrical currents measured but prior aging studies [23] have shown radially

    propagating electrochemical degradation that usually starts at the point of electrical

    failure. Other areas that could be investigated in the future include surface adsorption

    and Debye screening inside dielectric pores. With the work reported here, there provesto be positive commercialization prospects for low-voltage (

  • 8/6/2019 Raj Balaji

    62/68

    51

    CHAPTER 5

    DISCUSSION AND FUTURE WORK

    5.1 Summary of this work

    The basis of this work was to attain low voltage electrowetting by choosing the

    appropriate dielectrics for active-matrix driven electrowetting applications developed at

    the University of Cincinnati. The first aim was to obtain electrowetting at low DC

    voltages, specifically under 10 V, by choosing the appropriate thicknesses for the

    dielectric. Two types of dielectric were used, Al2O3 and Si3N4. Both dielectrics included

    ~50 nm CYTOP 809M and exhibited good electrowetting response under 10 V. Fromthe data compiled, the dielectrics and surfactants used do indeed show low voltage

    electrowetting but in the DC regime, saturation (via charge penetration) takes place.

    Therefore, the electrowetting solution needed to be investigated further.

    Once the type of dielectric to be used was established, which was the Al2O3 due

    to the high quality of atomic layer deposition, the next step was to observe the effects of

    various surfactant solutions under different polarities of voltage. Positive and negative

    voltage sweeps were applied to various aqueous solutions to observe any trends in

    dielectric failure. After testing the various surfactant solutions, dielectric failure was not

    observed for the synthesized catanionic surfactant. Reason being that catanionic

    surfactants contain long alkyl chains for both the cations and anions which means that

    the surfactant contained ions that were too large to cause dielectric failure. Therefore,

    from the data compiled and results observed, it has been suggested that the use of

    large ions in electrowetting liquids can help improve reliability as the ions eliminate the

    root problem (charging penetration) rather than simply delaying visible charging effects.

  • 8/6/2019 Raj Balaji

    63/68

    52

    After testing the surfactants, a proposal was made to use liquid electrodes due to their

    unique ability to propagate through complex dielectric defects. The use of propylene

    glycol was suggested as a medium compared to water due to its larger molecular size.

    From the results, propylene glycol showed no dielectric failure as a standalone or with

    SDS surfactant solution. This provides further evidence that relatively smaller

    dissociated water molecules (H3O+) and (OH-) also contribute to the formation of a

    conductive pathway through the dielectric. From this, it is safe to say that propylene

    glycol could be a better alternative to water for electrowetting as it does not exhibit

    dielectric failure.

    5.2 Achievement in the context of others work

    From the data gathered here it is fair to say that dielectric failure can be avoided

    with the dielectrics used and the surfactants tested herein. We shall now briefly

    overview the benefits of the findings here for low voltage electrowetting in comparison to

    the work of other groups that are trying to experimentally attain low voltage

    electrowetting. In brief, the Al2O3 is a better dielectric atomic layer deposition reduces

    the chances of having any pores in the dielectric layer. Unlike the option of thermally

    grown SiO2 [18,21] as a dielectric layer which is not a feasible option for low cost

    commercial applications, or the barium strontium titanate [22] grown via metal organic

    chemical vapor deposition, the Al2O3 is more viable. Titanates have been shown to

    increase the chances of charge trapping and at times even fluoropolymer delamination

    [23]. Other groups, such as Berry et al. and Moon et al., have not focused their work on

    effects of saturation due to surfactants for low voltage electrowetting. Even though

    these groups make mention of electrowetting saturation, no further details were

  • 8/6/2019 Raj Balaji

    64/68

    53

    provided for the cause of it. Given the results of the work herein, we do see the

    dielectric failure effects for SDS surfactant solution with a positive DC bias. Therefore a

    better alternative in the form of a catanionic surfactant, such as the DTA-OS that was

    synthesized, one would have to assume that it would be better for reliable electrowetting

    since the surfactant contains long alkyl cation and anion chains. However, synthesizing

    this surfactant is a time consuming process and requires experimental precision. There

    are very few, if any, commercial catanionic surfactants in the market. Therefore for

    industry, this might not be the most desirable option available. In turn, the use of

    propylene glycol can be suggested as a medium instead of water for electrowetting as itis easily attainable and showed no signs of dielectric failure, with or without SDS.

    5.3 Future Work

    From the knowledge of surfactants and dielectrics, it can be claimed that they

    can certainly be applied to most electrowetting applications for low voltage

    electrowetting. However, there is always scope for improvement when it comes down to

    material characterization for low voltage electrowetting. One area that could be studied

    in order to enhance reliability and longevity of dielectric layers is using composite

    dielectrics instead of the standard double-layered approach that was proposed in this

    thesis. Using the composite dielectric approach, the following is assumed. Most

    dielectrics have pores in the micro- or nano- scale. By using composite dielectrics, if

    there exists some form of charge penetration through the first layer, the dielectrics that

    follow will have a mismatch therefore not allowing charge penetration to the metal

    electrode since the pores wont follow through the entire composite dielectric. An

    example of this process would be using Al2O3 and a photo-resist with good dielectric

  • 8/6/2019 Raj Balaji

    65/68

    54

    properties such as SU-8. This is currently being studied at the University of Cincinnati,

    where 4 layers of these dielectrics are deposited or spin coated on a substrate. First, a

    thin layer, ~50 nm, of Al2O3 is deposited on a metal substrate or metal electrode

    substrate. This is followed by spin coating SU-8 to attain a thickness of ~100 nm.

    Finally, a hydrophobic dielectric layer, such as CYTOP, is spin coated on the composite

    dielectric layer. If this particular setup shows no dielectric failure for any surfactant

    used, be it cationic or anionic, then electrowetting response tests will have to be

    performed on them. Once desirable low voltage electrowetting response is observed,

    this might be an area to transition to.

  • 8/6/2019 Raj Balaji

    66/68

    55

    References

    [1] G. Lippmann, Relations entre les phenom`enes electriques et capillaries, Ann.Chim. Phys. 5 494, 1875

    [2] F. Mugele and J. Baret, "Electrowetting: From basics to applications," J. Phys.-Condens. Mat., vol. 17, pp. R705-R774, 2005

    [3] B. Berge, "Electrocapillarity and wetting of insulator films by water," ComptesRendus De L Academie Des Sciences Serie II, vol. 317, pp. 157-163, 1993

    [4] T. B. Jones, "An electromechanical interpretation of electrowetting," J. Micromech.Microengin., vol. 15, pp. 1184-1187, 2005

    [5] R.B. Fair, Digital microfluidics: is a true lab-on-a-chip possible?, Microfluidics andNanofluidics, vol. 3, pp. 245 281, 2007

    [6] H. Moon, A.R Wheeler, R.L Garrell,J. A. Loo, C.-J. Kim, An integrated digitalmicrofluidic chip for multiplexed proteomic sample preparation and analysis by MALDI-MS,Lab on a Chip, vol. 6, pp. 1213-1219, 2006

    [7]I. Barbulovic-Nad, H. Yang, S. Philip, and A.R. Wheeler, Digital microfluidics for cell-based assays,Lab on a Chip, vol. 4, pp. 519-526, 2008

    [8] B. Berge and J. Peseux, Variable focal lens controlled by an external voltage: Anapplication of electrowetting, The Eur. Phys. J. E, vol. 3, pp. 159-163, 2000

    [9] J. Heikenfeld, N. Smith, M. Dhindsa, K. Zhou, M. Kilaru, L. Hou, J. Zhang, E. Kreit,and B. Raj, Recent Progress in Electrowetting Optics, Opt. Phot. News, vol. 20, pp.20-26, 2009

    [10] R. A. Hayes and B. J. Feenstra, "Video-speed electronic paper based onelectrowetting," Nature, vol. 425, pp. 383, 2003

    [11] J. Heikenfeld, K. Zhou, E. Kreit, B. Raj, and S. Yang, Electrofluidic displays usingYoungLaplace transposition of brilliant pigment dispersions, Nature Photonics, vol. 3,pp. 292-296, 2009

    [12] J. Heikenfeld, Electrowetting Optics on target for record optical performance,

    SPIE Newsletter, pp. 19862 (invited), 2008

    [13] S. Kuiper and B. H. W. Hendriks, Variable-focus liquid lens for miniature cameras,App. Phys. Lett., vol. 85, pp. 1128-1130, 2004

    [14] U.C. Yi and C.J. Kim, "Soft printing of droplets pre-metered by electrowetting,Sensors and Actuators, vol. 114, pp. 347-354, 2003

  • 8/6/2019 Raj Balaji

    67/68

    56

    [15] R. van Dijk, B.J. Feenstra, R.A. Hayes, I.G.J. Camps, R.G.G. Boom, M.M.H.Wagemans, A. Giraldo, B.v.d. Heijden, R. Los, and H. Feil, "Gray scales for videoapplications on electrowetting displays," SID Int, Symp. Digest Tech., vol. 37, pp. 1926-1929, 2006

    [16] K. Bhat, Electrowetting Textiles - A New Paradigm for Tuning of TextileWettability, M.S. Thesis, U. of Cincinnati, 2007

    [17] B. Raj, M. Dhindsa, N. Smith, R. Laughlin, and J. Heikenfeld, Ion and LiquidDependent Dielectric Failure in Electrowetting Systems, Langmuir, vol. 25, pp. 12387-12392, 2009

    [18] S. Berry, J. Kedzierski, and B. Abedian, Irreversible electrowetting on thinfluoropolymer films, Langmuir, vol. 23, pp. 12429-12435, 2007

    [19] H.J.J. Verheijen, and M.W.J Prins, Reversible electrowetting and trapping ofcharge: model and experiments, Langmuir, 1999, vol. 15, pp. 66166620, 1999

    [20] M. Kilaru, Hydrophobic Dielectrics of Fluoropolymer/Barium TitanateNanocomposites for Low-Voltage and Charge Storing Electrowetting Devices, M.S.Thesis, U. of Cincinnati, 2007

    [21] S. Berry, J. Kedzierski, and B. Abedian, Low voltage electrowetting using thinfluoroploymer films, J. of Coll. and Inter. Science, vol. 303, pp. 517-524, 2006

    [22] H. Moon, S.K. Cho, R.L. Garrell, and C. -J. Kim, Low voltage electrowetting-on-dielectric,

    J. of App. Phys., vol. 92, pp. 4080-4087,

    2002

    [23] B. Raj, N.R. Smith, L. Christy, M. Dhindsa, and J. Heikenfeld, CompositeDielectrics and Surfactants for Low Voltage Electrowetting Devices, IEEE UGIMSYMPOSIUM, pp. 187-190, 2008

    [24] M. Dhindsa, and J. Heikenfeld, Electrowetting on Superhydrophobic Surfaces:Present Status and Prospects, J. of Adh. Science and Technology, vol. 22, pp 319-3342008

    [25] F. Li, and F. Mugele, How to make sticky surfaces slippery: Contact anglehysteresis in electrowetting with alternating voltage, App. Phys. Lett., vol. 92, pp.244108, 2008

    [26] Y. A. Ono, Electroluminescent Displays, World Scientific Publishing, vol. 1, 1995

    [27] K. Zhou, J. Heikenfeld, K.A Dean, E.M. Howard, and M.R. Johnson, A fulldescription of a simple and scalable fabrication process for electrowetting displays, J.Micromech. Microeng., vol. 19, pp. 065029065041, 2009

  • 8/6/2019 Raj Balaji

    68/68