Title of Thesis: CHARACTERIZATION AND FABRICATION OF
NOVEL RUTHENIUM OXIDE-ZINC BATTERIES
Author: Mahsa Dornaja , Master of Science, 2010
Directed by: Professor Christopher Davis
Department of Electrical and Computer Engineering
University of Maryland - College Park
Ultrathin galvanic cells that can comply with a variety of form factors and electronic
system packages are of tremendous technological importance. They show promise for
application in  exible electronic systems. This Thesis describes a  exible, high energy
density galvanic cell, based on hydrated ruthenium (IV) oxide-zinc electrodes utilizing
RuO2 nH2O nanoparticles in amounts that are not cost-prohibitive. This battery is
suitable for use in a number of environments. It is non-toxic, environmentally friendly,
and, based on aqueous electrolytes, it is safe in operation. It can be optimized for volume
manufacture at low cost. As it functions at much lower cell voltage than Li-ion batteries,
it can be recharged remotely, at conveniently low voltage, by harvesting, for example,
radio-frequency (RF) energy. At present this cell demonstrates a speci c capacity of over
50 mAh cm 2. This is the largest capacity reported for thin- lm cells. Also, its cycle life
of up to 200 charge-discharge cycles is very promising for use as a secondary galvanic cell.
This Thesis focuses on (i) the optimization of anode and cathode materials, (ii) the
identi cation of the composition of the electrolyte, (iii) the choice of the most suitable cell
separator, and (iv) the elaboration of the most e cient assembly procedures.
CHARACTRIZATION AND FABRICATION OFNOVEL RUTHENIUM OXIDE-ZINC
BATTERIES
by
Mahsa Dornaja 
Thesis submitted to the Faculty of Graduate School of the
University of Maryland, College Park, in partial ful llment
of the requirement for the degree of
Master of Science
2010
Advising Committee:
Professor Christopher Davis (Chair)
Professor Martin Peckerar
Professor Neil Goldsman
Copyright c by
Mahsa Dornaja 
2010
Dedication
I dedicate this dissertation to my parents and my brothers for all their support and love
which was my greatest inspiration in life.
You always believed in me.
I Love you all
ii
Acknowledgement
The accomplishment reported in this Thesis would not have been possible without signi cant
help, contribution, and moral support received from my colleagues, friends, and family.
I would like to express my indebtedness and gratefulness to Professor Martin Peckerar
for his guidance and assistance over that past 2 years. Also, I would like to thank my dis-
sertation committee, Professor Christopher Davis, the Chair, and Professor Neil Goldsman.
I feel privileged for having collaborated with numerous talented individuals at FlexEl,
LLC, and I would like to acknowledge my colleagues: Dr. Bob Proctor, Dr. Zeynep Dilli,
Karina Drees, Vincent De Chacus, and Benjamin Krupsaw.
My special thanks go to Dr. Daniel Lowy for his sharing his insight in the chemical and
electrochemical aspects of our battery system, for his support and encouragements. It was
a real pleasure learning from him as a student, and working with him, as a colleague.
As a member of the Analog and mixed-signal Laboratory at University of Maryland, I
would like to recognize my colleagues and friends: Kwangsik Choi, Filiz Yesilkoy, and Scott
Bauman. I would like to thank the entire promise family at the Graduate School O ce,
University of Maryland.
And last but not least, I would like to express deep gratitude to my family. I thank my
father for all his support, love, and care. I thank my mother, Marjaneh, for all those times
when she stood by me, for having been my strength when I was weak, for her tireless love
and support.
I would like to thank my brothers, Hamid Reza and Saeed. You are the best brothers
that anyone could have. You are the reason that I can go on in my life.
I would like to thank my friend, Dr. Ali Izadi, for his support and care over my under-
graduate years.
Thanks to Almighty God for giving me strength, health, and mental fortitude to suc-
cessfully complete this Master?s journey.
Sincerely,
Mahsa Dornaja 
iii
CONTENTS
1. MOTIVATION AND BACKGROUND : : : : : : : : : : : : : : : : : : : : 1
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Goals and Accomplishment of This Thesis . . . . . . . . . . . . . . . . . . . 4
1.3 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1 Thin-Film Lithium and Lithium-Ion Batteries . . . . . . . . . . . . . 4
1.3.2 Flexible Zinc-Carbon Batteries . . . . . . . . . . . . . . . . . . . . . 7
1.3.3 Flexible Hybrid Power Sources . . . . . . . . . . . . . . . . . . . . . 10
1.3.4 Thin-Film Hydrated RuO2 Electrodes . . . . . . . . . . . . . . . . . 10
2. THE ELECTROCHEMISTRY OF GALVANIC CELLS : : : : : : : : : 13
2.1 Brief description of battery operation . . . . . . . . . . . . . . . . . . . . . . 13
2.2 Thermodynamic Considerations . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.1 Oxidation-Reduction Reactions . . . . . . . . . . . . . . . . . . . . . 14
2.2.2 Electrochemical Potential . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3 Kinetics of a Battery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3.1 Overpotential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4 Chemical and Physical Considerations of a Galvanic Cell . . . . . . . . . . . 23
2.4.1 Electrolyte E ects . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.4.2 Separators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3. FABRICATION AND CHARACTERIZATION OF NOVEL ZINC-RUTHENIUM
OXIDE CELLS : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 29
3.1 Chemical Considerations for Novel Zinc-Ruthenium Oxide Cell . . . . . . . 29
3.1.1 Close Look at the Cathode Side . . . . . . . . . . . . . . . . . . . . . 29
iv
3.1.2 Close Look at the Anode Side . . . . . . . . . . . . . . . . . . . . . . 32
3.1.3 Close Look at the Cell Capacity . . . . . . . . . . . . . . . . . . . . 33
3.1.4 Initial Electrolyte Utilized in Zn  RuOx Galvanic Cells . . . . . . . 33
3.2 Physical Consideration and Fabrication of a Novel RuO2 Cell . . . . . . . . 35
3.2.1 Initial Separator Utilized in Zn  RuOx Galvanic Cell . . . . . . . . 37
3.2.2 Thin Separators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2.3 Cathode Current Collectors . . . . . . . . . . . . . . . . . . . . . . . 37
3.2.4 Anode Current Collectors . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2.5 Anode in Zn  RuOx Galvanic Cell . . . . . . . . . . . . . . . . . . 39
4. EXPERIMENTAL TECHNIQUES : : : : : : : : : : : : : : : : : : : : : : : 45
4.1 ELECTROCHEMICAL Instrumentation . . . . . . . . . . . . . . . . . . . 45
4.1.1 Electrochemical Techniques Performed in 3-Electrode Geometry . . 47
4.1.2 Electrochemical Techniques Performed in 2-Electrode Geometry . . 47
4.2 Electrochemical Techniques and Testing Procedures . . . . . . . . . . . . . 48
4.2.1 Cyclic Voltammetry (CV) . . . . . . . . . . . . . . . . . . . . . . . . 48
4.2.2 Open Circuit Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.2.3 Charge /Discharge Studies . . . . . . . . . . . . . . . . . . . . . . . 49
4.2.4 Shelf Life/Self-Discharge . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.2.5 Internal Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
5. MEASUREMENTS AND TESTING RESULTS : : : : : : : : : : : : : : 61
5.1 Electrolyte E ects Measurements . . . . . . . . . . . . . . . . . . . . . . . . 61
5.1.1 Original Electrolyte . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
5.1.2 Overview of Electrolytes of Di erent pHValue . . . . . . . . . . . . . 62
5.2 Cathode Current Collector . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.3 Separator Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.4 Anode Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.4.1 Zinc-RuOx Batteries . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.4.2 Al  RuOx Battery . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
v
5.5 Monitoring Hydrogen Discharge and Water Decomposition by Cyclic Voltam-
metry in Open Cell Experiments . . . . . . . . . . . . . . . . . . . . . . . . 70
5.6 XPS Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.6.1 Analysis on the Cathode Side . . . . . . . . . . . . . . . . . . . . . . 74
5.6.2 Analysis on the Anode Side . . . . . . . . . . . . . . . . . . . . . . . 75
5.6.3 Further XPS Measurements and Analysis . . . . . . . . . . . . . . . 76
6. MANUFACTURING AND COST ANALYSIS OF Zn RuO2 BATTER-
IES : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 81
6.1 Current Stage of the Novel Ruthenium(IV) Oxide Cells . . . . . . . . . . . 81
6.1.1 Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.1.2 Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.1.3 Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
6.1.4 Cycle Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
6.1.5 Safety and Sustainability . . . . . . . . . . . . . . . . . . . . . . . . 84
7. ACCOMPLISHMENTS AND FUTURE CONSIDERATIONS : : : : : 85
7.1 Summary of Main Accomplishments . . . . . . . . . . . . . . . . . . . . . . 85
7.2 Future Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
7.2.1 Anode Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
7.2.2 Cathode Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . 87
7.2.3 Electrolyte Optimization . . . . . . . . . . . . . . . . . . . . . . . . 88
vi
LIST OF TABLES
3.1 The anodic material used in batteries. . . . . . . . . . . . . . . . . . . . . . 39
5.1 Denomination of di erent electrolytes with di erent pHvalue and the charge
capacities enabled by them. All batteries were made under similar conditions
and testing environment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.2 Di erent current collectors. . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.3 Nominal continuous current for di erent cathode current collectors . . . . . 66
5.4 Di erent separators with di erent thickness and their performance in capac-
ity, number of cycles, and internal resistance. . . . . . . . . . . . . . . . . . 66
5.5 Potential windows for various electrolytes. . . . . . . . . . . . . . . . . . . 73
6.1 The Capacity Comparison of Non-Rechargeable Flexible Batteries. . . . . 82
vii
LIST OF FIGURES
3.1 Constant load discharge of a 4cm2 cell through 1k load. . . . . . . . . . . 34
3.2 (a)Top view of the battery. (b) Cross sectional view of battery. . . . . . . 35
3.3 Piece of a separator under the microscope. . . . . . . . . . . . . . . . . . . 41
4.1 Schematic of the potentiostat: the equivalent circuit of a 3-electrode cell
(Ei is the input voltage, Rm - current-measuring resistor, and CA - control
ampli er). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.2 Principle of cyclic voltammetry. In the left bottom panel the applied potential
is displayed vs. time, and the current response vs. time is shown above it,
while the right panel shows the output, the voltammogram in which current
is plotted relative to the applied potential. . . . . . . . . . . . . . . . . . . 49
4.3 European sign convention. . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.4 Constant current charging with 240 A current. . . . . . . . . . . . . . . . 51
4.5 Constant current charging with 50 A current. . . . . . . . . . . . . . . . . 52
4.6 Constant voltage charge current of a Zn RuO2 cell, when charged at 1:4V
for 1h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.7 Constant load discharge of a battery through 1k external load, providing a
charge capacity of over 10mAh=cm2. . . . . . . . . . . . . . . . . . . . . . 53
4.8 Voltage vs. time discharge recorded by CP for calculating the internal resis-
tance of a cell. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.9 100 cycles accomplished from a 4cm2 Zinc-RuO2 battery with 1k load and
1:4V charging intervals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.10 The capacity plot drawn from a 4cm2 Zn-RuO2 battery . . . . . . . . . . 55
4.11 800 cycles accomplished by the Zn  RuO2 battery. . . . . . . . . . . . . . 56
viii
4.12 Three-month discharge of a cell with 16cm2 of active area, operating with
electrolyte of near neutral pH. . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.13 Two months discharge of a 5V cell with 4cm2 of active area and acidic pH. 58
4.14 The equivalent circuit of a battery. . . . . . . . . . . . . . . . . . . . . . . 59
4.15 Calculating the internal resistance of a cell by using CP technique; it yields
the value of 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.1 Plot of battery capacity versus pH for di erent electrolytes. Note that con-
necting point is just for visual clarity. . . . . . . . . . . . . . . . . . . . . . 62
5.2 Capacity versus di erent pHvalues of the electrolytes that belong to the
same family; they all contain di erent concentrations of zinc chloride and
ammonium chloride. Note that connecting point is just for visual clarity. . 64
5.3 Capacity versus the thickness of a separator obtained from Zinc-RuO2 bat-
tery system. Note that connecting point is just for visual clarity. . . . . . . 67
5.4 Number of cycles versus the thickness of a separator obtained from the Zn 
RuO2 battery system. Note that connecting point is just for visual clarity. 67
5.5 Internal resistance of a cell versus the di erent separators thickness. Note
that connecting point is just for visual clarity. . . . . . . . . . . . . . . . . 68
5.6 Discharge curve for Zn RuO2 battery with pH2:48 electrolyte over 1k load 69
5.7 Approximately 100 cycles accomplished with Zn RuO2 battery through 1k 
load with charging cycle in between two discharges, at an applied voltage of
1:4V. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.8 (a). Discharge curve for the Al RuO2 battery with pH6:2 electrolyte. (b).
Discharge curve for the Al  RuO2 battery with pH0:2. . . . . . . . . . . 70
5.9 CV scans recorded for electrolyte E10A5, at 100mV=s: inner (blue) scan:
from  1:05V to +1:22V vs. Ag(RE), while the external curve was obtained
in the potential range from  1:05V to +1:22V vs. Ag (RE). . . . . . . . . 72
5.10 The Si(2p) peak obtained with XPS. . . . . . . . . . . . . . . . . . . . . . 74
5.11 Full XPS spectrum for cathode-side samples from undischarged (top) and
discharged (bottom) batteries. . . . . . . . . . . . . . . . . . . . . . . . . . 77
ix
5.12 Top: Zoom in on the C1s/Ru3d peak; left: new battery; right: discharged
battery. Bottom: Zoom in on the Ru3p split peak; left: new battery; right:
discharged battery. Of the four peaks visible in the bottom left graph, from
left to right (higher-to-lower binding energy) the peaks are ZnAuger, Ru1p,
ZnAuger, the other Ru1p. In the discharged battery graph, only the Znpeaks
remain clearly visible. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.13 Full XPS spectrum for anode-side (Zn) samples from undischarged (top) and
discharged (bottom) batteries. . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.14 Zoom in on the Zn2p peaks from the new (left) and discharged (right) bat-
teries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.15 Piece of a separator under the microscope. . . . . . . . . . . . . . . . . . . 80
x
Chapter 1
MOTIVATION AND BACKGROUND
\Imagine a world without batteries! A teenager walks outside
wearing headphones, tethered to home by a lengthy extension
cord. An old man winds his pacemaker like a pocket watch In
thousands of ways, large and small, batteries have changed our
daily lives."
Mary Ellen Bowden, Chemistry is Electric! (Chemical Heritage
Foundation, Philadelphia, 1997), p.26.
1.1 Motivation
Modern life has become progressively more mobile, and therefore dependent on the avail-
ability of portable power sources. In recent years there has been a rapid increase in the
use of battery operated devices, such as cell phones, Blackberries, laptops, iPods, remote
car lock controls, wall clocks, toys, and remote controls for TV, DVD, or CD players [1, 3].
According to statistical accounts, at present an average person in an industrialized country
consumes as many as 35 batteries a year. In the United States the market for batteries
grows at 6% annually, and was expected to reach $15 billion in 2009[4, 5]. According to
forecasts, the global battery demand will increase 4.8% annually through 2012. China will
record the largest gains, and will likely surpass the U.S. as the largest market. Consumer
battery demand will outperform the market as a whole; non-lead-acid secondary battery
market gains will outpace the demand for primary and lead-acid secondary batteries [4].
Rising demand for portable nanoelectronic devices triggers today?s unprecedented devel-
opment of dimensionally small batteries [2, 6, 7]. In order to maintain adequate operation
1
time of portable devices, which are becoming increasingly powerful and power hungry, the
consumer market demands small size batteries with high energy density as well as a rea-
sonable runtime and longevity [8, 9, 10].
By de nition, a battery is an assembly of single electrochemical or voltaic cells, connected
in series or parallel, which convert energy evolved in chemical reactions into low-voltage,
direct current electricity. It consists of a positive electrode (the cathode) and a negative
electrode (the anode), separated by an electrolyte, which is generally a liquid solution
containing a salt dissolved in a solvent, and is imbedded in a felt (the separator). All these
components are involved in electrochemical charge transfer reactions, which proceed at
the interface between the electrodes and the electrolyte. These reactions convert chemical
energy into electric energy, so that they can power various loads placed in the external
circuit. As most charge transfer reactions are reversible, secondary cells can be recharged
by electricity supplied from an external power source [11, 12]. Going into more detail,
at the anode electrons are generated to do external work. In lithium-ion batteries the
anode contains lithium, commonly held within graphite. The cathode is the electronegative
electrode to which electrons migrate through the external electrical circuit, while positive
ions migrate inside the cell. Hence, the role of the electrolyte is to allow the  ow of ions
(e.g., Li+ ions, protons, hydroxide ions, etc.) from one electrode to another. One important
requirement for the electrolyte is to be stable in the presence of both electrodes. Inert
current collectors allow the transport of electrons to and from the electrodes. They are
typically metals, which do not react with the electrode materials; in most cases, copper is
used for the anode and aluminum for the cathode [13].
Existing thin- lm battery technologies have severe cost and performance limitations,
which hinder the e cient use of the device. Future utilization of electrical energy, with
a greater emphasis on renewable energy sources, depends on the development of the next
generation of batteries. As of today, limited energy storage capacity of individual battery
cells and the lack of fast recharge cycles with long cell lifetimes represent additional restric-
tions. In order to improve the energy storage capacity of a given cell signi cantly, one needs
to increase the cell voltage and/or the amount of charge stored reversibly per unit mass
and volume. These requirements call for the exploration of new materials with nanoscale
2
features that could enhance reversible charge storage [5].
Over the past two decades solid-state thin- lm lithium and lithium-ion batteries have
been developed. Many of these systems were rechargeable, and exhibit long cycle lives.
They have been devised with a variety of electrode materials and in di erent cell con g-
urations, on thin ceramic, metal, or silicon substrates [14]. The goal of this Thesis is to
show how to build fully  exible and robust electrochemical cells. This requires multiple
components with speci c electrochemical and interfacial properties to be integrated into
single units. These basic components, i.e., the electrodes, separator, and electrolyte, can all
be integrated into single contiguous nano-composite units, which can then serve as building
blocks for a variety of thin mechanically  exible energy storage devices [15]. In previous
works, such batteries have been less than 15  m thick, and have important applications in
a variety of consumer and medical products. Cells with crystalline LiCoO2 cathodes can
deliver up to 30% of their maximum capacity between 4.2 and 3 V at discharge currents
of 10mA=cm2 At lower discharge-charge cycle rates, the capacity decreases by negligible
amounts over thousands of cycles[16]. According to literature reports, prototype thin- lm
batteries fabricated with crystalline LiCoO2 cathode consistently provide the power lev-
els up to 30mW=cm2 . They exhibit long cycle life, negligible self-discharge, and rapid
charge rates [17]. Nevertheless, none of the commercially available thin- lm cells possess
such high energy densities; lithium thin- lm batteries typically provide substantially less
than 1mAh=cm2 capacity [18]. Promising results have also been obtained with crystalline
LiMn2O4 cathodes, although no good reproducibility has been achieved with this system
[17]. A viable approach toward  exible power systems consisted in utilizing as the electrodes
nonporous cellulose paper embedded with aligned carbon nanotubes. Such building blocks
have been used for devising  exible supercapacitors, batteries, and hybrid power systems,
as well as dual-storage battery-in-supercapacitor devices. It was demonstrated that thin
freestanding nano-composite paper devices o er complete mechanical  exibility during op-
eration. Hence, these easy-to-assemble integrated nanocomposite energy-storage systems
could provide unprecedented design freedom for a variety of devices operating over a wide
range of temperature and environmental conditions [15]. Other very promising electrode
supports are polypyrrole  lms. They can be prepared by electrochemical polymerization
3
methods, and are obtained as highly  exible and bendable free-standing  lms. These bat-
tery components look paper-like, are soft, lightweight, mechanically robust, and exhibit
high electrical conductivity. Such free-standing  lms show promise as cathodes for  exible
and bendable batteries [19].
1.2 Goals and Accomplishment of This Thesis
In this thesis I present the  st systematic evaluation of a cathode electrode containing
ruthenium oxide used in a thin- lm  exible galvanic cell. I describe the impact of varying
di erent component parameters for this battery, such as: electrolyte pH, separator material,
and the electrode current collector materials.
I have found that each of the parameters studied proved optimal in the sense disclosed:
 Electrolyte pHof 2.48 gives the highest cell capacity.
 The  ber-glass separator gives the highest cell capacity and provide lowest internal
resistance. However, this separator has the highest thickness as compare to other
separators studied.
 Di erent materials have been tested as cathode current collectors for optimizing the
cell. The 0.13 mm thick graphite sheet has been chosen as the current collector for
its low cost, low internal resistance and high continuous current. Depending on our
applications and limitations di erent cathode current collector can be used.
The  nished, optimized product, provides a capacity > 20mAHr=cm2 . This is the
highest thin  lm battery storage capacity reported in open literature.
1.3 Background
1.3.1 Thin-Film Lithium and Lithium-Ion Batteries
These galvanic cells belong to several types, according to their cathode materials. Cath-
odes utilized in these batteries include crystalline or nano-crystalline oxide-based lithium
intercalation compounds. Anodes are made of lithium metal, inorganic compounds, e.g.,
4
silicon-tin oxynitrides, Sn3N4, and Zn3N2, or metal  lms, such as Cu, in which the anode
is formed by lithium plating on the initial charging. Typically, the electrolyte is a glassy
lithium phosphorus oxynitride (\Lipon") [13].
The construction and testing of batteries that operate with cathodes made of (i) LiCoO2,
(ii) LiMnO2, and (iii) Li8V2O5, or (iv) hydrated RuO2 appear in literature[20, 21, 22]. The
latter being utilized mainly as electrode materials for supercapacitors. We will brie y review
the manufacturing, testing, and performance of such galvanic cells.
So-called \Li-free" thin- lm batteries have the following cell con guration: Lidi usion
blocking overlayer/Cu/solid lithium electrolyte (Lipon)/LiCoO2. These cells are activated
during the initial charge by in situ plating of metallic Lionto the Cuanode current collector.
Electrochemical cycling between 4.2 and 3.0 V was demonstrated for a number of over
1000 cycles, at the current density of 1mA=cmsq or for 500 cycles at 5mA=cm2 . The
overlayer confers high cycle stability; otherwise the plated Lirapidly develops a detrimental
morphology, and the battery loses most of its capacity within a few cycles. Li-free thin- lm
batteries retain their high cell voltage, while permitting a simple fabrication process in air,
thus avoiding problems related to the presence of a metallic Lianode [20].
The performance of lithium-ion thin- lm batteries could be improved by heat treat-
ment (at 250 C in air for 1h). Cells consisted of an amorphous silicon tin oxynitride anode
(?SiTON?), an amorphous lithium phosphorus oxynitride electrolyte (?Lipon?), and a crys-
talline LiCoO2 cathode. They sustained the capacity of 5mA=cm2 in the voltage range
from 4.2 to 2.7 V. Long-term cycling stability test of a SiTON/Lipon/LiCoO2 batteries
(between 3.93 and 2.7 V) demonstrated that their capacity faded only by 0:001% per cycle
[21].
Thin- lm rechargeable lithium batteries with amorphous and crystalline LiCoO2 cath-
odes were investigated. The lithium cobalt oxide  lms were deposited by radio-frequency
(RF) magnetron sputtering of a LiCoO2 target in a 3:1 Ar/O2 mixture gas. From proton-
induced gamma-ray emission analysis (PIGE) and Rutherford backscattering spectrometry
(RBS), the average composition of these  lms was determined to be Li1:15CoO2[15] or,
within experimental uncertainty, LiCoO2 + 0:08Li2O. The x-ray powder di raction pat-
terns of  lms annealed in air at 500 to 700 C were consistent with the regular hexagonal
5
structure observed for crystalline LiCoO2. The discharge curves of the cells with amor-
phous LiCoO2 cathodes showed no obvious structural transition between 4:2 and 2:0V,
while the discharge curves of the cells with polycrystalline cathodes were consistent with a
two-phase potential plateau at 3.9 V with a relatively large capacity. Two lower capacity
plateaus were observed at  4:2 and 4:1V with the 600 and 700 C annealed cathodes; the
 dq=dV peaks were broader and weaker for the 600 C annealed cathodes and were not
present at all with the 500 C annealed  lms. The chemical di usion coe cients of Li+ in
the cathodes obtained from ac impedance measurements at cell potentials of  4V ranged
from  10 12cm2=s for the as-deposited amorphous cathodes to  10 9cm2=s for the
 lms annealed at 700 C. The capacity loss on extended cycling of the thin- lm cells varied
with the crystallinity and thickness of the cathodes and with temperature. With the highly
crystalline, 700 C annealed material, losses on cycling between 4:2 and 3:8V at 25 C ranged
from 0:0001%/cycle (> 104 cycles) to 0:002%/cycle for cells with cathodes from 0:05 m to
0:5 m thick [22].
Thin- lm cathodes consisting of crystalline lithium manganese oxide with the general
composition Li1+xMn2 yO4enabled for extended cycling of lithium cells, at the tempera-
tures of 25 and 100 C [16]. Innovative electrolyte materials, such as polymers or inorganic
glasses, allow for designing  at lithium primary or secondary batteries. The so-called "hy-
brid plastic electrolytes" enable for devising thick  lm cells (with the thickness in the range
of 1mm-3mm) that possess a speci c surface capacity of several mAh cm 2. Secondary
Li-ion batteries typically last less than 500 cycles. By contrast, all solid state thin- lm
batteries o er better cycleability and exhibit liquid leakage. Their speci c surface capacity
is, however, about one order of magnitude lower than for the polymer electrolyte batteries.
Solid cells can be manufactured via sputtering and vacuum evaporation techniques [13].
All-solid-state thin- lm rocking chair lithium batteries have been fabricated by means of
sequential thin- lm deposition techniques on  exible 80  m thick aluminum foil substrates.
Their cell con guration was V2O5/Lipon/Li8V2O5. The total thickness of the device did
not exceed 1.41  m. Such galvanic cells comprised thin- lm electrodes of highly lithiated
vanadium (V) oxide as the anode and vanadium (V) oxide as the cathode. By controlling
the lithium concentration in the anode, one obtained an operating voltage in the range from
6
1 to 3 V. These cells had an active area of 15 cm2, and delivered a discharge capacity of
50  Ah, at room temperature. Additionally, they showed negligible capacity fade from the
second cycle to over 5800 cycles, at the current density of 10  A/cm2 [23].
Bates and co-workers fabricated and characterized thin- lm rechargeable batteries equipped
with an anode of metallic lithium and a cathode, which consisted of a mixture of amorphous
V2O5, crystalline and amorphous LixMn2O4 [24]. These thin- lm batteries operate with an
amorphous inorganic electrolyte, which consists of amorphous lithium phosphorus oxyni-
tride, known as Lipon. This electrolyte is typically deposited by RF magnetron sputtering
[14]. The performance of the thin- lm cells was evaluated at di erent current densities,
at several temperatures. Electrical measurements showed that the current density of the
thin- lm cells was limited by the mobility of the Li+ ions in the cathode materials. The
internal resistance of the Li/LixMn2O4 cells with crystalline cathodes was about two orders
of magnitude lower than that of Li/V2O5 cells with amorphous cathodes [24].
Sequential thin- lm deposition techniques enabled for the manufacturing of all-solid-
state rechargeable thin- lm batteries with the Li/Lipon/LiMn2O4 cell con guration. When
cycled at room-temperature, these cells delivered the nearly constant voltage of 4.0 V at
open circuit, good Coulombic e ciency, and the capability of carrying high current density.
Planar microbatteries have been connected in series via metallization. For example, eight
unit cells in series can deliver high voltage values up to ca. 32 V [25].
1.3.2 Flexible Zinc-Carbon Batteries
Circuit interconnects and various passive components can be fabricated on a variety of
 exible substrates, by using the o set lithographic process. This approach enables for the
manufacturing of Leclanch-type galvanic cells. In order to accomplish this task, it was
necessary to develop zinc and carbon-based inks with appropriate rheological properties for
lithographic printing. Zinc and carbon electrodes were fabricated with silver-based current
collectors, and then, a separator material soaked in electrolyte has been in between zinc-
carbon electrode structures and MnO2 paste to form a galvanic cell [26].
Use of a combination of functional nanostructured materials allowed for the fabrication
of optimum performance solid-state and  exible zinc-carbon batteries. Flexible carbon
7
nano ber mats were obtained by electrospinning and utilized as the current collector and
the cathode support. The cathode layer consisted of Mn(IV) oxide particles, which were
mixed with single-walled carbon nanotubes (SWCN) for improved conductivity. A thin zinc
foil served as the anode, while the electrolyte layer and a polyethylene oxide layer containing
titanium oxide nanoparticles completed the system. This battery retained its performance
even under mechanically stressed conditions, and its shelf-life was good [27].
A novel zinc ion conducting polymer gel electrolytes (PGEs) was made with non-volatile
room temperature ionic liquids. PGEs consisted of an ionic liquid with a zinc salt dissolved
in it, blended with a polymer matrix, poly (vinylidene  uoride-co-hexa uoropropylene),
PVDF-HFP. Thus, one obtained electrolyte membranes, which were freestanding, translu-
cent,  exible, and elastic; also, they exhibited excellent mechanical integrity and strength,
possessed exceptional thermal stability, and did not show weight loss under dynamic vac-
uum or upon heating to 200  C. The gel phase was stable in a wide temperature range, with
ionic conductivities on the order of 10 3S=cm at room temperature, 10 4S=cm at  20 C,
and (3 5) 10 3S=cm at 80 C.
Electrochemical tests demonstrated that zinc ions possess mobility in the membranes,
while zinc metal is capable of dissolution into the membrane, and can redeposit from the
membranes. The membranes are stable in a wide Electrochemical potential window. PGEs
based on ionic liquids show promise for potential application in next-generation non-aqueous
zinc battery systems [28]. Recently, binary room temperature molten electrolytes based on
acetamide and zinc perchlorate have been prepared and characterized. These electrolytes
showed high conductivity at ambient conditions, and had favorable physicochemical and
Electrochemical characteristics. Raman and infrared spectroscopic studies revealed the
presence of large free-ion concentration in the molten liquid. Rechargeable zinc batteries
assembled with a gamma-MnO2 cathode, a metallic Znanode, and operated with the molten
electrolyte showed high discharge capacity over several cycles. It is expected that this class of
acetamide-based, room temperature molten liquids should become a viable green alternative
for electrolytes used in rechargeable zinc-based secondary batteries [29].
Room-temperature zinc ion-conducting molten electrolytes based on acetamide, urea,
and zinc perchlorate or zinc tri ate have been prepared and characterized by various physic-
8
ochemical, spectroscopic, and Electrochemical techniques. The ternary molten electrolytes
are easy to make and can be handled under ambient conditions. They show excellent stabil-
ity, high ionic conductivity, relatively low viscosity, and other favorable physicochemical and
Electrochemical properties that make them good electrolytes for rechargeable zinc batteries.
Speci c conductivities of 3.4 and 0.5 mS cm 1 at 25  C are obtained for zinc perchlorate
and zinc-tri ate-containing melts, respectively. IR spectroscopy reveals that the free ion
concentration is high in the optimized composition. Rechargeable Znbatteries have been
assembled using the molten electrolytes, with gamma-MnO2 as the positive electrode and
Znas the negative electrode. They show excellent Electrochemical characteristics with high
discharge capacities. This study opens up the possibility of using acetamide-based molten
electrolytes as alternate electrolytes in rechargeable zinc batteries [30].
A novel type of storage battery, which can be devised in various geometries, including
nanoscopic and  exible systems, was built with microfabricated Zn/MnO2 electrodes by
using very special membranes, based on so-called "nanograss" tubes. These tubes were
developed in 2002 by Bell Labs and Lucent Technologies to provide a "super-hydrophobic
nanostructured surface" atop of which a droplet of water or aqueous solution can be placed.
Such a droplet would sit above the tubes, showing negligible interaction with the tubes.
Inspiration came from smart surfaces found in nature, e.g., the surface of lotus leaves, which
is not wetted by water. A honeycomb type silicon structure was manufactured by top down
processing techniques, and this structure served as support for the battery, keeping the
electrolyte physically separated from the electrodes until the battery was activated. [31].
It was demonstrated that upon applying an electrical voltage and current to the system, a
droplet of liquid can be reversibly switched between the super-hydrophobic state or "rolling
ball" state and the hydrophilic or immobile state [32, 33].
As a function of the applied voltage (as low as 22 V), liquid droplets on a nanostructured
surface undergo sharp transitions between three possible wetting states, and liquid surface
tension [34]. Under the action of an external voltage, the surface transitioned from super-
lyophobic to hydrophilic state. This phenomenon provides a novel means for manipulating
liquids on the reduced micro scale [35]. By this the permeability of the entire structure
can be tuned and such "smart" energy storage devices can be activated on demand. Upon
9
rigorously maintaining the spacing between the electrodes and the electrolyte, miniaturized
reserve batteries can be obtained, which do not lose their power when not in use. This
behavior is favorable, as over storage an average battery would typically lose as much as
7-10% of its power [31, 36].
One should emphasize here that the density of this cell that stays dormant over extended
time periods enables the nanograss-based battery to serve as a so-called reserve battery or
backup battery, traditionally used for special purposes, such as in emergency or military
applications. [37]. \Nanograss" Zn/MnO2 batteries provide an open circuit voltage of 1.55
V and a capacity of 200  Ah cm 2 [38].
A  exible galvanic cell and zinc-air battery were obtained by screen-printing on paper[39].
The galvanic cell consisted of a zinc/carbon/polymer composite anode, on one side of the
sheet, and poly (3,4-ethylenedioxythiophene) (PEDOT) cathode, on the other side of the
sheet. Lithium chloride and LiOH-based electrolyte was applied onto paper by inkjet print-
ing, in between the two electrodes. This paper-based battery had an open-circuit voltage
of ca. 1.2 V and a speci c areal capacity of 0.5 mAh cm 2 [39].
1.3.3 Flexible Hybrid Power Sources
Comprising a thin- lm organic or hybrid solar cell connected to a lithium-polymer battery,
the so-called EURO-PSB device possesses attractive characteristics like low mass (below
10g), small thickness (less than 1mm), and mechanical  exibility. Owing to an e cient
interconnection between the two devices, the battery is always charged with optimized
voltage, independently of the illumination conditions. This product is compatible with roll-
to-roll production processes[40] As pointed out in the Introduction to this chapter, various
power storage devices, including hybrid systems and dual-storage battery-in-supercapacitor
devices can be made with nonporous cellulose paper embedded with aligned carbon nan-
otubes [15].
1.3.4 Thin-Film Hydrated RuO2 Electrodes
Electrodes based on hydrated ruthenium (IV) oxide are particularly relevant to the work
disclosed in this Thesis. In all previously reported cases, however, RuO2 has been utilized
10
as electrodes in supercapacitors, rather than in galvanic cells. Highlighted below are a few
examples of such electrodes, including their preparation, analysis, and the system in which
they are being employed.
Thin- lm ruthenium oxide electrodes can be prepared by cathodic electrodeposition on
a titanium substrate. Film thickness is being controlled by the deposition time. Such
electrodes are used in supercapacitor applications, with 0:5M H2SO4 aqueous electrolyte.
Morphological changes with increasing  lm thickness a ect the speci c capacitance and
charge-discharge times of these supercapacitors. Maximum speci c capacitance of 788F=g
was achieved with an electrode with 1:4mg=cm2 RuO2. These results are explained by
considering [41].
An alternative technique employed for depositing ruthenium (IV) oxide thin- lms for
solid-state thin- lm supercapacitors consisted in direct current reactive sputtering, per-
formed at 400 C. (TFSC). The supercapacitor has a cell structure of RuO2/Li2:94PO2:37N0:75
(Lipon)/RuO2/Pt. Radio frequency, reactive sputtering deposition of a Li2:94PO2:37N0:75
electrolyte  lm is performed on the bottom RuO2  lm, at room temperature, for electrically
separating the bottom and the top RuO2 electrodes. Scanning electron microscopy (SEM)
revealed that the RuO2/Lipon/RuO2 hetero-interfaces had no inter-di usion problems.
Since the electrolyte thin- lm has low ionic mobility, the capacity and cycle performance of
the cell are inferior to those of a bulk type of supercapacitor [42].
A SnO2-RuO, composite (SnRuO) thin- lm possesses unique Electrochemical prop-
erties for a battery-supercapacitor hybrid due to the fact that the SnRuOsystem shows
battery and supercapacitor characteristics simultaneously. SnRuOthin- lms were prepared
via magnetron co-sputtering method. The SnRuOthin- lm as the anode in a secondary
battery demonstrated a  rst discharge capacity of 1:557mAh=cm2; the second discharge ca-
pacity was, however, only half (52%) of the  rst discharge capacity. The SnRuOthin- lm
exhibited a speci c capacitance of 14mF=cm2 over 1000 cycles. These results suggest that
SnRuOthin- lms can be utilized as both a thin- lm supercapacitor and a thin- lm battery
[43].
Electrostatic spray deposition (ESD) enabled for the manufacturing of ruthenium (IV)
oxide thin- lm electrodes with an average speci c capacitance of 650F=g and high rate
11
capability. This technique is a one-step process as compared to multistep sol-gel processes.
In addition, ESD o ers the features of low-temperature synthesis and allows for an easy
control of surface morphology. While as-prepared hydrous ruthenium (IV) oxide thin- lm
was amorphous, the hydrous ruthenium oxide, RuO2 H2O, thin- lm became crystalline
after annealing at temperatures > 200 C [44].
12
Chapter 2
THE ELECTROCHEMISTRY OF GALVANIC
CELLS
2.1 Brief description of battery operation
As stated in the previous chapter, one half of a battery is the electrolyte solution in contact
with an electrode, called the anode, where the oxidation takes place. The other half-
cell contains electrolyte in contact with another electrode, called the cathode, where the
reduction proceeds [45]. Other essential components of a battery are the separator, current
collectors, and packaging. In order to minimize the internal resistance of a battery the
electrodes should be as close to each other as possible. The separator should be porous
and thin. The separator is an insulating material, which prevents electrodes from touching
each other, and, as a result, shorting the battery [46]. For open assembly of a battery in
a container there is no need for a separator; the electrodes are immersed in the \sea" of
electrolyte, and they will not touch each other. In case of a packaged battery, the pores
of the separator will be impregnated with electrolyte, and the ions transfer through these
pores.
At both electrodes chemical reactions take place, and generate electricity. Each electrode
contains chemicals that participate in the reaction, and they are called active materials. The
driving force for the external current derived from a cell is the di erence in the electrode
potentials of the two half-cell reactions [46]. Let us examine the chemistry that drives the
galvanic cells.
The chemical reaction at the negative side releases electrons, and these electrons are
13
accepted by the current-collector of the positive electrode via the external circuit, when the
battery is connected to a load. This electron transfer is complemented by the migration of
ions in the internal circuit. Simpli ed reactions at the negative electrode and the positive
electrode are illustrated bellow by equations (2.1) and (2.2), respectively:
Anode (-) : X !Xn+ +ne (2.1)
Cathode (+) : nY +ne  !nY (2.2)
During discharge of a battery, negative ions move toward the negative electrode, and
positive ions move toward the positive electrode. The  ow of these ions is reversed during
the charging process [46].
2.2 Thermodynamic Considerations
Galvanic cells generally operate close to thermodynamic equilibrium; hence, the \ideal" cell
potential (standard cell voltages) can be predicted from thermodynamics. But most cells
are not ideal. Thermodynamic analysis alone cannot tell us anything about the details of
charge transfer (a non-equilibrium process), the over-potential (de ned as the departure of
cell potentials from thermodynamically determined standard potentials), or indicate what
reaction(s) will proceed, and at what rate [47]. In the sub-sections below, I address some
of these \real-world" chemistry issues.
2.2.1 Oxidation-Reduction Reactions
Oxidation-reduction reactions are chemical processes that involve transfer of electrons from
a reducing agent to an oxidizing agent. In these reactions the reducing agent loses electrons,
while the oxidizing agent gains those electrons. Below is an example of an oxidation-
reduction reaction between Fe2+ and MnO 4 , in acid solution, which proceeds according to
equation (2.3):
14
5Fe2+(aq) + MnO 4 (aq) + 8H+(aq) !5Fe3+(aq) + Mn2+(aq) + 4H2O(l) (2.3)
Where Fe2+ is oxidized and MnO 4 is getting reduced. The above reaction can be broken
down into two half reactions, described by Equations (2.4) and (2.5):
MnO 4 (aq) + 8H+(aq) + 5e reduction      !Mn2+(aq) + 4H2O(l) (2.4)
5Fe2+(aq) oxidation      !5Fe3+(aq) + 5e (2.5)
As can be seen from the (2.3), electrons transfer directly from the reducing agent Fe+2 to
the oxidizing agent MnO 4 . This is, however, a chemical reaction, which does not generate
current. When both the reducing agent and the oxidizing agent are in the same solution,
electron transfer will happen directly, from one reactant to the other, in the solution, and
no useful work is obtained.
In order for current to  ow, the reducing agent and the oxidizing agent should be physi-
cally separated; this means that each half reaction should take place at a di erent electrode,
and the electrons should transfer through a load connected to the electrodes by external
wires. Hence, the key requirements for converting a redox system into an electrochemi-
cal system are (i) the physical separation of the reducing agent from the oxidizing agent,
(ii) each electron exchange should occur at a di erent electrode, and (iii) the transfer of
electrons should happen through an external loop, i.e., over an in nitely long pathway, as
compared to the atomic scale, characteristic to redox reactions [48]. The result current can
then power a load.
2.2.2 Electrochemical Potential
As discussed in the previous section, in a galvanic cell reduction and oxidation processes
occur simultaneously. Each half-reaction has a potential associated with it. When we design
a galvanic cell from a given pair of half reactions, the sum of electrochemical potentials of the
15
half-cell reactions determines the voltage of the battery. Standard reduction potentials for
half-reactions are listed relative to the standard hydrogen electrode (SHE). By convention
the reduction potential of a SHE is 0:000V. For example, in a galvanic cell with zinc as
the anode and SHE as the cathode, the potential of Znis +0:76V, which is the standard
reduction potential for zinc. In this cell, the anode is zinc metal in contact with Zn2+
solution. The oxidation half-reaction is, according to Equation (2.6):
Zn(s) !Zn2+(aq) + 2e (2.6)
On the cathode side, i.e., at the SHE, the reduction of protons takes place, as shown in
equation (2.7):
2H+(aq) + 2e  !H2(g) (2.7)
The overall cell reaction can be written as equation (2.8):
2H+(aq) + Zn(s) !Zn2+(aq) + H2(g) (2.8)
By de nition, the voltage of this cell is equal to the standard reduction potential of zinc,
as shown in equation (2.9):
E cell = E (Zn2+=Zn) E (H+=H2) = 0:76V + 0:00V = 0:76V(vs. SHE) (2.9)
Another representative example, known as Daniell?s cell, consists of a combination of
a zinc anode and a copper cathode, each in contact with its own ions. The overall cell
reaction can be written as equation (2.10):
16
Zn(s) + Cu2+(aq) !Zn2+(aq) + Cu(s) (2.10)
The standard voltage of a Daniell?s cell is calculated from equation (2.11):
E cell = E (Zn2+=Zn) E (Cu2+=Cu) = 0:76V  ( 0:34)V = 1:10V (2.11)
Similarly, for any given cathode and anode pair the standard cell voltage can be calcu-
lated from tabulated standard potential values [49].
Thermodynamics provides an answer to the fundamental question of how one assigns a
magnitude to the ideal cell voltage. But there are other questions that must be answered.
We would like to know what combination of anode and cathode will or will not be able to
perform work (i.e., to generate electrical energy.) We would also like to know the magnitude
of currents that can be drawn from a particular cell. These quantities are limited by the
chemical kinetics of the system. The rate of Electrochemical reactions is related to and
measured in terms of current densities. In the next section we will examine the parameters,
which a ect and determine the magnitude of the current that can be delivered by a galvanic
cell.
2.3 Kinetics of a Battery
We should de ne a series of electrochemical parameters involved in our discussion. First,
the overpotential will be addressed, and next the voltage of a cell will be examined, as a sum
of standard potentials, overpotentials, the Ohmic potential drop, and polarization losses.
2.3.1 Overpotential
The overpotential is an electrochemical parameter, which refers to the potential (voltage)
di erence between a half-reaction?s thermodynamically determined reduction potential and
the potential at which the redox event is experimentally observed [45]. The term is directly
17
related to a cell?s voltage e ciency. Driving a reaction in an electrolytic cell requires more
energy than thermodynamically expected. This excess of energy is quantitatively measured
by the overpotential. In other words, no reactions can proceed in a system, while it is at
equilibrium. In order to make the charge transfer reactions happen, one has to move out
the system from its equilibrium state.
The overpotential is the extent to which the system is being pushed out from equilibrium.
When a galvanic cell is discharged, less energy is recovered than thermodynamics predicts.
This can be explained by asserting the existence of overpotential. Upon re charging the cell,
more energy is needed than predicted by thermodynamics, which means that the cathode
and anode overpotentials will add to the charging voltage. In each case the extra or missing
energy is lost as heat. Overpotential is speci c to each cell design, and will vary between
cells and operational conditions even for the same reaction. But when the cell is connected
to a load, i.e., current is drawn from the system, and then the cell voltage will not be at its
maximum value anymore [45, 50].
Equation (2.12) is the Tafel relationship between the overpotential,  , and the log of
current density, logi,  owing through the system; a and b are empirical constants [51].
 = a+blogi (2.12)
This empirical equation tells us how much one must change the potential (i.e., the
magnitude of overpotential) in order to accomplish a certain current density [51]. This can
also be written as equation (2.13):
i = A eB RT (2.13)
Where A and B are di erent from a and b, as one switches from 10-base logarithm to
natural logarithm, and generalizes the relationship to any temperature (rather than 298K
room temperature in equation (2.12)).
18
While Tafel discovered his relationship empirically, later it was derived on a theoretical
basis, as a limiting form of the Butler-Volmer equation, equation (2.14), which predicts the
net current  owing through the system
i = i0
h
e nF =RT e(1  )nF =RT
i
(2.14)
Where i0 is the exchange current density,  is the electron transfer coe cient, n is
the number of electrons involved in the Faradaic charge transfer process, T is absolute
temperature, and F is Faraday?s constant [52]. At extreme overpotentials the rate of one of
the reactions, and, consequently, one term of the Butler-Volmer equation becomes negligible.
Thus, when the positive polarization of the electrode exceeds +120mV=n relative to the
equilibrium only the oxidation reaction will proceed (according to (2.4)), the reduction being
insigni cant. By contrast, at negative polarization values of  < 120mV=n, reduction will
take place, and no oxidation allowed. equation (2.15), which applies for anodic polarization,
can be linearized as (2.16).
i = i0 e1  nF =RT (2.15)
lni = lni0 + (1  )nF =RT (2.16)
When converted to 10-base logarithm, and rearranged, (2.16) yields to (2.17), which is
equivalent to the empirical Tafel relationship (2.12):
 = [2:3RT=(1  )nF] logi0 + [2:3RT=(1  )nF] logi (2.17)
Hence, the physical meaning of the empirical constantsaandbin (2.1) can be understood
from the theoretical treatment, equations (2.16), as equations.(2.18) and (2.19) [50]:
19
a = [2:2:3RT=(1  )nF] logi0 (2.18)
b = 2:3RT=(1  )nF (2.19)
Constant b, termed as the Tafel slope, provides information about the reaction mecha-
nism, while constant a mainly re ects the rate constant of the reaction. Rigorously, Equa-
tion 1 should be written by using the absolute value of the current density, as shown in
equation (2.20):
 = a blogjij (2.20)
Where the plus sign stands for the anodic overpotential, while the minus sign is used
for the cathodic overpotential [50].
It should be noted here that the Ohmic resistance of the solution may cause deviations
from the ideal electrode response described by equation (2.14) Therefore, in the real case
one should amend the overpotential in equation (2.21) with a term, which accounts for
solution resistance, Rs, the density of current  owing through the cell, i, and the exposed
surface area of the working electrode, S, as shown in equation (2.21) [53, 54, 55]:
 = E E  RsIS (2.21)
Use of Tafel plots, as an analytical tool for evaluating fuel cell performance has become
popular over the past years; however, most studies are restricted to chemical power sources,
rather than being extended to microbial fuel cells. Most likely, authors choose Tafel plots
as their electroanalytical technique for uncomplicated electroactive processes, which have
a well-de ned stoichiometry. For such simple charge transfer reactions one can exploit
the linear portion of the Tafel plot for deriving the number of electrons transferred in the
process and the exchange current density, i0 (by extrapolating the linear portion of the plot
20
to  = 0). The latter reveals the ability of the system to produce current, when moved out
from the state of equilibrium.
Steady-state Tafel plots enable assessment of the electrocatalytic activity of various
electrodes used in methanol or hydrogen based fuel cells. Hence, Tafel plots provide insight
into the reaction mechanism of methanol oxidation on carbon-supported binary and ternary
alloys of Pt, Ru, and Sncatalysts. Furthermore, these plots reveal changes in the rate-
determining step, and when recorded at di erent temperatures, they allow for comparing
the activation energy of di erent catalyst [56]. Also, they can identify the composition of
catalysts with the greatest activity in methanol oxidation [57, 58, 59, 60]. Also, this analysis
tool can be used for the performance assessment of the anode and cathode during direct
methanol fuel cell operating conditions [61].
Another widespread application of Tafel plots in chemical fuel cells is the monitoring of
reaction kinetics of oxygen reduction. Recent examples of this evergreen topic addressed by
means of Tafel plots are the investigation of O2 reduction in subfreezing polymer electrolyte
membrane (PEM) fuel cells [62], the study of oxygen discharge mechanism in concentrated
alkaline solution on hydrophobic cathodes [63], and the long-term durability testing of H2-air
proton exchange membrane fuel cells [64].
The voltage of a battery, measured while connected to a load, is lower than its open
circuit voltage. This can be the result of the internal impedance of the battery, which is
made of up of [46]:
 Polarization losses at the electrode.
 Ohmic losses at the cathode, anode, in the electrolyte, and across the separator.
 Contact resistance.
All these phenomena can cause an unwanted voltage drop. Creating a limitation in
charge transfer between the two electrodes. This is an intrinsic property of the electrodes
in contact with the electrolyte. Another reason for the drop is due to the slow di usion of
the ions across the separator. This phenomenon depends on the thickness and porosity of
the electrodes as well as of the separator. Ideally, we want the voltage drop across the cell
21
should be minimal, when connected to the load. In order to achieve this goal, studies need
to be done for reducing the polarization losses at the electrodes. Various compounds, such
as manganese (IV) oxide, commonly referred to as manganese dioxide or ruthenium (IV)
oxide can be used as depolarizer.
Overpotential losses are present both at the cathode side and the anode side, and man-
ifest as both charging potential and discharging potential. The increased and decreased
cell potential during charging and discharging can be calculated form equations (2.22) and
(2.23) [46].
Vd = Vr  +    IR (2.22)
Vch = Vr + + +  +IR (2.23)
Where Vd, Vch, and Vr are discharge voltage, charge voltage, and reversible voltage or
open-circuit voltage, respectively.
Ideally, we want to minimize the overpotentials (make them zero), so that the equation
above can be written as:
Vd = Vr IR (2.24)
Vch = Vr +IR (2.25)
Since the ideal case is not achievable, we will not be able to reduce the over potential
to zero. As mentioned earlier, a system with overpotential Zero is in equilibrium, hence
it cannot perform work, and hence, it cannot generate electricity. We need, however, to
reduce as much as possible all unnecessary overpotentials in our cell, e.g., by using di erent
kinds of depolarizer or employing more porous and thinner separators.
22
2.4 Chemical and Physical Considerations of a Galvanic
Cell
2.4.1 Electrolyte E ects
The electrons exchanged in the chemical processes that drive a galvanic cell are carried
across the internal circuit, i.e., by the electrolyte, via the directional stream of oppositely
charged ions [51]. Hence, ions act as charge carriers in the internal circuit; once they
reach the electrodes, they may undergo electrochemical processes or, alternatively, other
species may discharge at the electrodes, and account for the charge transfer in the internal
circuit of the cell. In order to achieve highest e ciency, the electrolyte must contain high
concentrations of the ions. On their turn, ions should exhibit high mobility. Owing to its
high polarity (with a dipole moment of 1.85 debye), water is a favorable solvent of inorganic
and organic salts, contained in the electrolytes [51]. Pure water undergoes ionization to a
very limited extent (2.26):
H2O  !H+ + OH (2.26)
Where the concentrations of [H+] = [OH ] = 10 7mol=cm3 (at 20 C). Such low ion
concentrations do not allow the current to be transported through pure water, which is an
insulator (speci c resistivity of de-ionized water is 18:2M cm, at 25 C). Consequently,
water can be used as an electrolyte only after a \supply of ions" has been added in it, i.e.,
an acid, a base, or a salt has been dissolved in it.
Electrolytes for Zn  RuOx cells can be made in an alternative manner. They are not
necessarily aqueous solutions, but hydrogels with incorporated ions can also be used as the
electrolyte. Such solid electrolytes exhibit advantages over liquid ones, owing to their ability
of preventing battery dry out and leakage.
23
Fundamentals
As stated previously, batteries consist of two electrodes, which are electronic conductors,
separated by an electrolyte solution or gel, which is an ionic conductor. The latter plays
the role of closing the internal circuit of the cell. Several features need to be considered for
choosing an electrolyte for a system. Below is a list of such parameters:
 The electrolyte should be compatible with the system. It should not be too acidic or
excessively alkaline as to dissolve the electrode (or separator) materials. On the other
hand, close to neutral electrolytes have proven Electro chemically inactive. Depending
on the type of the electrodes, the electrolyte should be chosen to contain su cient
concentration of ions for transporting an ion current equivalent to the electron current
streaming through the external circuit.
 Electrolyte should be chemically stable, i.e., it should not decompose over time and
its pHshould be constant.
 Electrolyte should be electro chemically stable, which means that it should exhibit a
potential window compatible with the operation voltage of the galvanic cell.
 High conductivity represents a major requirement of a good electrolyte, so the ions
can transfer without being hindered. High conductivity of the electrolyte will reduce
the overall internal resistance of the cell and, consequently, the Ohmic potential drop,
which may cause signi cant losses in e ciency of the galvanic cell (the cell voltage
of the battery would be decreased). We attempt to lower the internal resistance of a
battery as much as possible.
 As we are concerned about the safety, the electrolyte should not to be  ammable,
which applies for most electrolytes based on organic solvents. Also, it should be
environmentally friendly and nontoxic.
 The e ect of the pHof the electrolyte appears to be a complex phenomenon. Dif-
ferent pHvalues may cause di erent results in battery performance and cyclability.
Depending on the type of electrodes and their purity, the separator material and its
24
thickness, as well as the form factor of the galvanic cell as a, electrolytes of di erent
pHvalue will perform di erently.
Concepts of Electrolyte Conductivity
The ability of an electrolyte solution to sustain the channel of electrical current depends
on the mobility of its ions, driven by the electric  eld between electrodes [65]. If ions of
charge Ze0 are accelerated by the electric  eld strength, E, and are subject to the Stokes
frictional (or viscous) force, F, which increases with the drift velocity vd, and is given for
simple spherical ions of radius ri, then the frictional force can be calculated by equation
(2.27), where  is the viscosity of the medium [65].
F = 6  rivd (2.27)
The absolute mobility of ions, uabs, is de ned for di using particles by dividing the drift
velocity by either the di usional driving force or the equal and opposite Stokes viscous force,
equation (2.28) [65]:
uabs = vd=6  rivd = 1=6  ri (2.28)
In the electrochemical literature ion mobility is typically expressed as conventional
(electro chemical) mobility, uconv, which represents the drift velocity in unit electric  eld
(1V=cm). The relationship between absolute and conventional mobility is given by equation
(2.29):
uconv = uabsZe0 (2.29)
When the velocity attains a limiting value Vmax corresponding to the exact balance
between the electrical and frictional forces, equation (2.30) applies:
25
Ze0jEj= 6  rivd (2.30)
For a given values of  and jEj, each type of ion will exhibit a transport velocity, vd,
dependent on the charge and the radius of the solvated ion, and a direction of migration,
which is determined by the sign of the charge [65].
uconv = Ze0=6  ri (2.31)
The ion of greatest mobility is H+, therefore, its presence in the supporting electrolyte
will enhance the overall conductivity. In our system, acidic electrolytes have proven e cient.
However, the pHvalue of the solution (or gel) should represent a trade-o between high
electrochemical activity in the galvanic cell and moderate chemical activity, when the cell
is not in use, so that the unwanted chemical corrosion of zinc anodes can be avoided, and
the shelf life of the battery increased.
Conductivity
Conductivity, just like the resistivity, is a characteristic property of the material, and does
not depend on the geometric factors. We seek to de ne the conductivity for a given elec-
trolyte solution, which will be characteristic of the solution, and independent of geometrical
factors. For a cell with a uniform cross-section A, while the two electrodes are separated by
a distance d, and has the conductance L, then the conductivity K1 can be calculated from
equation (2.32) [65]:
K1 = d L=A (2.32)
Conductivity is measured in units of   1m 1 or Sm 1.
26
2.4.2 Separators
Battery separators provide a barrier between the anode and the cathode. The separator
physically separates and electrically insulates the anode from the cathode but it allows
electrolytic or ionic conduction to occur through the electrolyte [66]. Hence, the battery
separator must be a good electronic insulator, Electro chemically stable in the presence of
the electrolyte and electrode materials, and it must be ionically conductive [67].
The major parameters that a ect the performance of a separator in a battery are:
 Electrical resistance which can be calculated with Ohm?s law
 Porosity
 Water absorption
 Electrolyte absorption and retention as needed by the ion transport.
 Dimensional stability, meaning that the separator should stay  at and not curl at the
edges.
 The separator should not swell or shrink during electrolyte absorption.
 Chemical stability over a long time period meaning that the separator should be
able to tolerate the strongly oxidizing cathode and the corrosive electrolyte at high
temperatures.
 Thickness and uniformity of properties. For  exible thin- lm batteries, a thin sepa-
rator is needed to maintain the form factor.
The electrical resistance of the separator is the most important factor. The goal is to
pick a separator that has the smallest internal resistance while having the smallest thickness.
There is a trade-o between the thickness and the capacity of the cell when it comes to the
separator. Analysis in section 5.3 shows the e ect of di erent separators and their thickness
on the total capacity and performance of the battery. Most likely, having a thicker and more
porous separator will give a better result because more electrolyte can be absorbed by the
separator. This will help the battery to have longer life time. The internal resistance of
27
the cell can be calculated by Ohm?s law. subsection 4.2.5 explains in more detail how the
internal resistance of a cell can be calculated. To summarize, the internal resistance is the
cause of the extra voltage drop of the cell, when connected to the load. Table 5.4 shows the
internal resistance of the cells with di erent separator materials.
28
Chapter 3
FABRICATION AND CHARACTERIZATION
OF NOVEL ZINC-RUTHENIUM OXIDE CELLS
Let us now turn to the details of the ruthenium oxide/zinc cells that form the basis of this
Thesis. We address each of the separate components of the battery sequentially.
3.1 Chemical Considerations for Novel Zinc-Ruthenium
Oxide Cell
This section of this Thesis deals with the chemical consideration of Zn- ruthenium oxide cell.
It looks closely at the cathode side and the anode side and their e ects on the performance
of a battery. It also looks into the e ects of electrolyte on the battery operation.
3.1.1 Close Look at the Cathode Side
Ruthenium oxide has been used in recent research to achieve high value capacitances. It
has the highest capacitive storage of any known material in the world ( 1400F=g), and has
been the subject of many papers [68]. Ruthenium oxide is non-toxic and environmentally
friendly. It can also be found abundantly in the planet. Initially, ruthenium dioxide was used
for super capacitors, but due to the materials cost and internal leakage current, its market
is very limited now. However, ruthenium dioxide does not have these disadvantages, when
used in batteries. The proposed battery in this study shows that none of these shortcomings
have been observed when using RuO2 as the cathode material. In the proposed battery,
oxidation-reduction takes place at a metallic anode, such as a zinc or aluminum. The choice
29
depends on the desired output voltage.
RuOx  lms can be produced Electro chemically or thermochemically. Electro chemically
formed RuOx  lms are often hydrous, and thus, are open to much easier proton transfer
through the lattice. These hydrous  lms exhibit a great component of pseudocapacitance,
as compared to a much smaller component of double layer capacitance. The conductivity
of RuO2 allows for easy electron transfer from the surroundings, and is facilitated by H+
transfer from an electrolyte. Therefore, RuO2 electrodes can be charged/discharged by a
coupled electron/proton transfer process.
Ruthenium has many oxidation states, range from 0 to +8, and 2. The +2 and +3 and
+4 states are the most common stages [69]. Most reversible redox reactions are observed
between the states +4 and +2. In surface regions of RuOx  lms, oxidation states between
+2 and +6 can be assessed. There is some support for the belief that reaction, yielding
for Ruoxidation states greater than +6, especially at high potentials, may not be reversible
[69].
For hydrated ruthenium oxide, the following type of redox reaction occurs between states
+2 and +4 equation (3.1):
RuOx+  (y  )H2O + 2 H+ + 2 e  !RuOx yH2O (3.1)
In this reaction, the left-to-right direction involves the reduction of ruthenium oxide.
For instance, take x = 2 and  = 1. On the left-hand side, the hydrated RuOx specimen is
Ru2O3, with the ruthenium at the oxidation state of +3. In order to be converted to RuO2
on the right-hand side, the ruthenium gives up one oxygen to form one more molecule of
water of hydration with two extra hydrogen ions, which should be available in the medium
(e.g. from an acidic electrolyte). In the process, the oxygen atom is being released from
the ruthenium oxide, and requires two additional electrons and two protons to form the
covalent bonds H O H, i.e., the \new" hydration water molecule. These electrons should
be supplied from an external circuit, while the supporting electrolyte contained by the cell
supplies the necessary H+ ions.
30
It has been suggested that in some fraction of available surface states, the reaction shown
in equation (3.2) takes place:
RuOx(OH)y + H+ + e  !RuOx  (OH)y+ (3.2)
With x = 2 and y = 0, this reaction becomes, equation (3.3):
RuO2 + H+ + e  !RuO2  (OH) ; where 0   2. (3.3)
The above proton-assisted reduction processes of ruthenium oxides were investigated by
cyclic voltammetry, abbreviated as CV. section 5.5 looks into cyclic voltammetry experi-
ments in more detail.
The discharge of a RuO2 cell shows a number of plateaus, indicating the participation
of a number of valence states. Let us assume that the entire mass of the RuOx nanosphere
was hydrated in the discharge reaction. According to Conway, the hydration reaction may
extend into the bulk. Our particles may be small enough to assure complete hydration of
the material mass. We can derive an expression (based on density, molecular weight, and
an assumed number of participating electrons) for the mass of the oxide (M) required for
achieving the observed capacity, equation (3.4):
M = 2:5Cp (3.4)
Where M is in mg and Cp is in mAh=cm2. We have assumed that each Ru(IV) oxide unit
contributes 2 electrons. Thus, to achieve a capacity of 5mAh=cm2, we would need 12:5mg
of the oxide. As the density of the oxide is 7:14g=cm3, we would need a thickness of 17:8 m.
The amount of the ruthenium used in one cell in normally more than above, therefore we
expect to get more capacity per area. Some of the cells provide up to 10mAh=cm2 before
dry out.
31
3.1.2 Close Look at the Anode Side
The reaction on the zinc side is relatively simple. Below is the oxidation reaction at the zinc
side, equation (3.5),(3.6). Considering the electron con guration of zinc, it can be noted
that zinc has two free electrons in the outer shell and during the oxidation it loses these
two electrons becoming Zn2+:
Zn : [Ar]4s23d10 !Zn2+ : [Ar]3d10 (3.5)
The reaction on the zinc side is however pretty simple.
Zn(s) !Zn2+(aq) + 2e (3.6)
As described in chapter 2, the standard electrochemical potential of zinc is 0:76V and
the electrochemical potential of the ruthenium is 0:45V. As a result, the overall potential
of the cell is around 1:2V. Depending on the electrolyte the potential may vary by  0:2V.
One  nds the following relationship for capacity (in mAh=cm2) as a function of zinc foil
thickness and surface area of the cell:
d = 1:74Cp=A (3.7)
Where d is the thickness in microns, Cp is the capacity (mAh=cm2), and A is the
projected geometric surface area of the cell. Using equation (3.7), we can calculate the
thickness needed for a cell. For example, for getting 5mAh=cm2 of capacity, we would need
thickness of the zinc equal to 8:7 m for a 1cm2 cell. A commercially available 25 m zinc
sheet will more than su ce for each 10mAh of capacity, with su cient safety margin to
prevent etch-through and the development of a non-recharging surface.
Di erent zinc thicknesses are used to make our novel battery. The goal is to reduce the
32
thickness of the zinc electrode as much as possible. Zinc foils as thin as 0:001" or 0:003"
can be used in battery manufacturing.
3.1.3 Close Look at the Cell Capacity
The general method used to experimentally determine the charge capacity of a cell is to run
it across a constant load by using potentiostat. Many attempts have been made to increase
the capacity of a cell. One fundamental technique of changing the capacity is to use di erent
electrolytes. In an e ort to obtain the greatest capacity value, we have tested a variety of
electrolyte compositions. Figure 3.1 is an example of constant load discharge of a 4cm2 cell
through 1k load. Close look at the curves shows a very fast drop of the potential, which
indicates the magnitude of internal resistance. Then, in 6h into the discharge, we see a very
rapid drop-o in voltage. Next, we observe a plateau lasting out to 14h, where the voltage
drops to 0:5V. A secondary plateau along with a long tail follows this.
We interpreted these results as follows. Through a reduction process at the cathode
side, the cathode assembly starts hydrating and going through its possible valence states.
As the RuO2 is reaching a saturating level of hydration, we theorize that electrons (derived
from the anode) continue to be injected into the electrolyte, discharging hydronium ions,
i.e. H3O+, and electrolyzing water, as well, and that this creates the secondary plateau.
Further, atomic hydrogen, created by the previous processes and stored as an adsorbate on
the surface of activated carbon (henceforth: AC) particles, may yields molecular hydrogen,
subsequently coalescing to form bubbles, which can increase the series resistance of the
battery, and cause the decline of the output voltage across the load.
3.1.4 Initial Electrolyte Utilized in Zn  RuOx Galvanic Cells
The initial cell was implemented with an electrolyte composed of a weak acid and ethylene
glycol, with ammonium hydroxide blended in the ethylene glycol. This mixture had a close
to neutral pHvalue, and prevented rapid dissolution of both the electrode materials and the
package sealing materials, while providing the needed conductivity. However, the capacity
drawn with this electrolyte was only 0:0325mAh=cm2. As part of the improvement of
battery performance, various types of electrolytes with di erent compositions and di erent
33
Load: 1000 , Cell area: 4cm2
Y Axis: Voltage (V), X Axis: Time (h)
Fig. 3.1: Constant load discharge of a 4cm2 cell through 1k load.
pHvalue were tested.
More recent research revealed that there are tradeo s in increasing the conductivity of
our electrolyte, as follows: (i) Although, high ion concentration will augment the conduc-
tivity, it will also cause faster dry out in the cell. This degrades the shelf life of the galvanic
cell. It appears that higher ion concentration implies more water molecules and are tided
up hydrating the ions. As the hydration water is tightly attached to the cations and anions
through dipole charge interaction, less free water is available at the RuOx-based anode
for balancing the Faradaic process at the cathode; (ii) While high ion mobility of H+, in
acidic electrolytes, or OH ions, in alkaline electrolytes confer favorable conductivity to
the electrolyte, they also promote unwanted corrosion phenomena in the cell. Corrosion is
detrimental to shelf life.
The use of a gel electrolyte can yield important advantages for ease-of-manufacturing
and battery operational life, as well as in preventing the increase of internal resistance due
to hydrogen evolution. Also, it may improve cycle life by reducing dendrite formation.
Replacing the electrolyte solution by a thin gel, such that it maintains the performance of
batteries equipped with a regular separator is a challenging task. Our goal is to make a gel
electrolyte that provides the same charge capacity as aqueous electrolytes, while it prevents
the dry-out of the cell. Some explorations have been conducted in the space of electrolyte
34
viscosity, and we have demonstrated the operation of batteries with an alkaline gel-based
electrolyte. Further explorations and experiments are needed for a better understanding
of the material composition of gels and thinning techniques for rendering their thickness
appropriate for use in  exible batteries. The further study of gel electrolytes and their
performance was beyond the scope of this Thesis.
3.2 Physical Consideration and Fabrication of a Novel RuO2
Cell
In this section we overview the fabrication process for the battery. We implemented the
cells with an electrolyte made of an acid and ethylene glycol. Di erent types of electrolyte
have been used that will explain in subsection 5.1.2. The systems of the materials and
the packaging have been chosen to yield a light, small volume,  exible and thin cells.
RuO2  xH2O (hydrated ruthenium oxide) was compounded with another material, and
used in the battery as the cathode material. The compound forms a conducting paste that
is added as a thin layer atop of a  exible substrate. Typically, we use graphite sheet as the
backing for the cathode side. However, di erent types of current collectors have been tested
to give the highest  exibility and conductivity. The electrolyte is added over the powder
to make a slurry-like mixture. This layer is covered with a separator, and then an anode
electrode will go over the separator. The cell will be packed in a thin plastic package, and
sealed using its adhesive side. The edges of the plastic packaging will be sealed with heat
or tape. Figure 3.2 shows the cross sectional view and the top view of the battery.
Fig. 3.2: (a)Top view of the battery. (b) Cross sectional view of battery.
35
The battery is rechargeable and  exible, so that  exible electronics may be integrated
on it, which may be used for power distributed networks node devices. In summary, the
cell includes: an anode, a cathode, an electrolyte, separator, and  exible packaging.
Varieties of these parts have been used to make the battery as thin as possible, while
keeping its  exibility and high current capacity.
In fabricating a ruthenium oxide battery we should consider following factors to get the
highest performance. Acetone and aqueous ammonia solution (3wt:%) are used to clean
the surface of the anode, which is typically zinc foil. While acetone removes the organic
impurities, such as grease or oil, ammonia solution etches slightly the surface of Zn. It
dissolves and removes a very thin Znlayer from the electrode surface, and thus, exposes a
fresh, active Znlayer to the electrolyte.
The separator needs to be cut slightly bigger than the active area of the cell so it
prevents any accidental short-circuiting internal to the battery. Since drops of electrolyte
will be added to the separator, caution needs to be taken not to add excessive amounts
of electrolyte, which would leak through the battery. The added aqueous solution serves
as the ionic conductor in the internal circuit of the cell. The entire volume of electrolyte
should be accommodated by the galvanic cell, so that no leakage occurs.
The contacts on the electrodes should be large enough to prevent any current crowding.
In other words, the contacts should be large enough, so that ions have enough space to
migrate. Small contact will cause ions to cluster at the surface of the contact and increase the
internal resistance of the battery. Internal resistance is an important to the performance of
a battery. We attempt to make internal resistance of the battery as small as possible. So the
voltage drop across the resistance will be smaller and this yields a better performance of the
battery. Contacts are made to the electrodes with the aid of conductive epoxy (if necessary).
Details about the internal resistance of a battery are explained in subsection 4.2.5. Finally,
the battery needs to be sealed properly, for preventing any liquid evaporation that results
in the premature death of the battery. Another important factor to keep in mind is that
the existence of the contacts will not interfere with the voltage of a cell. The contacts are
not in direct contact with the electrolyte so they don?t interfere with the chemical reaction
that takes place inside the cell.
36
3.2.1 Initial Separator Utilized in Zn  RuOx Galvanic Cell
In the novel  exible battery we have used glass  ber separators designed for air sampling and
other applications. These  lter papers, such as Model FP810 by F&J Specialty Products,
are made of 100%high quality borosilicate glass micro  bers, and contain less than 5%
acrylic binder. Their thickness is 0:38mm and their dry tensile strength is 5lb per inch
width. They are chemically insert in acidic or alkaline solutions, and are thermally stable
at temperatures up to 500 C. These separators have high load capacity, meaning that they
have the ability of collecting particles without clogging or cessation of  ow.
3.2.2 Thin Separators
Our ultimate goal is to reduce the thickness of the cell by thinning di erent components of
the cell. Given the relatively high thickness of the glass  ber separator, di erent commercial
membranes available for Li-ion batteries were tested for thinning the overall cell. Battery
separators manufactured by Celgard, LLC are highly-engineered and o er many options of
varying thickness, porosity and Gurley number. Also, American Membrane Corporation
provides a variety of separators with di erent pore size. The thickness of the separator is
around 90 m, with a variety of pore sizes spanning from 0:03 m to 10 m. The results of
battery performance of with di erent types of separator are summarized in section 5.3.
3.2.3 Cathode Current Collectors
A variety of cathode materials have been explored for their ability to collect electron charge,
and provide a low resistance material, able to transfer electron current into the electrolytic
medium [68]. Throughout this Thesis, the standards Zn RuOx batteries feature a 0:13mm
thick graphite foil as the cathode current collector. The cathode side of the battery consists
of ruthenium hydroxide mixed with activated carbon, in a paste. Two problems are encoun-
tered when using graphite foil in the system. The thickness of the graphite has a major
impact on the overall thickness of the battery. Since the goal is to reduce the thickness
as much as possible, one of our main concerns is to replace the graphite foil with thinner
materials. Another weak point of using graphite foil is the experimental observation that
37
the material absorbs electrolyte, reducing the amount of available liquid inside the battery;
this phenomenon contributes to the mechanical problem of dry-out noticed for our cells.
To overcome this problem, we have investigated alternative cathode materials, such as
carbon-based mesh, carbon-based cloth, carbon- ber veil, as well as carbon nanotube  lms,
known as \buckeye". We achieved relatively high capacity from these alternatives; however,
the graphite foil is still our primary current collector, because of its low cost and easy
handling. section 4.2 illustrates the performance of batteries and their internal resistance,
when using each of these materials.
Since the current collector does not exert any e ect on the chemistry of the cell, it
should be as thin as possible, and should have very small internal resistance. One approach
to identify such materials is to evaluate ultrathin  bers or similar materials, as the substrate
coated with a thin layer of chemically insulating, electrically conductive polymers [68].
3.2.4 Anode Current Collectors
In a galvanic cell, the anode is the negative electrode from which electrons  ow out to the
external circuit. Since electrons move from the negative electrode to the positive electrode
externally the current  ows form the cathode towards the anode via a load.
Anodes are metals, which are corrodible in electrolyte that leads to reduction reaction
inside the cell [70]. In the general case, the reaction on the anode side can be summarized
as:
M(s) = M2+(aq) + 2e (3.8)
Where M represents the pure metal, while M2+ are its ions.
Metallic anodes have the advantage of a high electrical conductivity. This reduces the
internal resistance of the cell on the anode side. There are two important criteria in choosing
the anode materials. Weight and volume are the main factors for  exible battery systems.
The parameters under consideration are the density of the materials, its energy density,
and its equivalent weight relative to the released charge. Table 3.1 shows some examples of
38
these parameters for di erent anodic materials used in batteries [70].
Aluminum and zinc are the common anode materials, which are used throughout this
Thesis. The anode material is important in releasing a hydrogen atom in reduction reactions.
Zinc is one good example of anode electrode that has been used in di erent types of batteries
such as Ni  Zn, Zn-carbon and alkaline batteries.
Active Material Mass Density Equivalent weight Energy density
(g=mL) (g/charge) (Ah=kg)
Al 2.7 9 2980
Cd 8.7 56 480
Li 0.53 7 3830
Mg 1.7 12 2230
Na 0.97 23 1170
Zn 7.1 33 812
Tab. 3.1: The anodic material used in batteries.
3.2.5 Anode in Zn  RuOx Galvanic Cell
Zinc is a main anode material used in the novel ruthenium cell. As described before, the
reaction on the zinc side is pretty simple. Below is the oxidation reaction at the zinc
side. Considering the electron con guration of zinc, it can be noted that zinc has two free
electrons in the outer shell, and during the oxidation it loses them to become Zn2+ ion.
Zn : [Ar]4s23d10 (3.9)
and when it loses the valence electrons it becomes:
Zn2+ : [Ar]3d10 (3.10)
The reaction on the zinc side is pretty simple, according to equation (3.11):
39
Zn(s) !Zn2+(aq) + 2e (3.11)
As described in chapter 2, the electrochemical standard potential for the Zn2+=0 redox
couple is 0:76V, and the electrochemical potential of the ruthenium is 0:45V (at pHzero).
As a result, the overall standard cell voltage is around 1:21V. Depending on di erent
pHvalues of the electrolytes, the potential may vary by  0:2V.
Based on Faraday?s law of electrolysis, one can derive the relationship in equation (3.12)
for capacity (in mAh=cm2) as a function of zinc foil thickness and projected geometric
surface area of the anode in the cell:
d = 1:74Cp=A (3.12)
Where d is the thickness of the Znfoil in microns, Cp is the capacity (mAh=cm2), and
A is the projected geometric surface area of the cell. Using the above equation we can
calculate the thickness needed for a cell. For example, to get 5mAh=cm2 of capacity, we
would need thickness of the zinc equal to 8:7 m for a 1cm2 cell. A commercially available
25 m zinc sheet will more than su ce for each 10mAh of capacity, with reasonable safety
margin to prevent etch-through and the development of a non-recharging surface.
Di erent zinc foil thicknesses are used to make our novel battery. The goal is to reduce
the thickness of the zinc electrode as much as possible. Zinc foils as thin as 0:025mm or
0:075mm are used in battery manufacturing. Since the cost of thin zinc foils is high for thin
sheets, commercially available thicker zinc material will be used in battery testing; their
thickness is 0:6mm, and it is one order of magnitude cheaper.
There are some disadvantages associated with zinc-based systems, as listed below [68]:
 The materials form dendrites that cause the separator to become rigid or penetrate
through the separator and short out the cells.
 Zinc surface changes shape during the charging and discharging processes. Zinc clumps
40
in one area and disappears in other areas. This causes current crowding and local
heating and, as a result, malfunctions of the electrode.
 Formation of insoluble zinc oxides.
Adding ethylene glycol to the electrolyte provides help in solving the dendrite problem.
However, we have not experienced any dendrite shorting, as the glass- ber separator, which
is very thick. We have observed a piece of  lter paper under microscope, after the battery
went through a discharge. Figure 5.15 shows this image. The black areas correspond to the
ruthenium remaining on the  lter paper, and the white and green areas represent the  lter
paper. As can be seen, there is no sign of dendrite formation on the  lter.
Fig. 3.3: Piece of a separator under the microscope.
Since using thick zinc allows for a considerable safety margin, the size e ect issue is under
control. However, utilizing thinner zinc with our current electrolyte sometimes results in
eating through the zinc anode, and creating holes. Part of the ongoing e ort is to determine
the minimum zinc thickness required for successful cell operation. section 2.4.1 looks into
di erent methods used to obtain thin zinc. It will also examine the performance of the
battery obtained via these Zn\thinning" methods.
41
Adjusting the pHof the electrolyte eliminated the formation of insoluble zinc oxides.
Initially the battery was made by using pH5 electrolyte. Changing the pHto more acidic
values helped in solving this problem. subsection 5.1.2 analyzes the e ect of electrolytes
with di erent pHand their chemistry on the battery.
Other Anode Materials
In addition to zinc, which has been so far the default anode material for our novel ruthenium
oxide batteries, it is possible to explore other anode materials. A possible candidate is
aluminum, a main group metal. Aluminum has been used as an anode material owing to
its low price, light weight, and also its availability in di erent thicknesses. Considering the
electron con guration of aluminum, shown below, equation (3.13), it can be observed that
it has three electrons in its outer shell; these are the valence electrons that are involved in
the oxidation reaction of Al.
Al : [Ne]3s23p1 (3.13)
Consequently, according to equation (3.14) the reaction on the aluminum side will be:
Al(s) !Al3+(aq) + 3e (3.14)
The standard reduction potential of aluminum is  1:662V [71], and since the electro-
chemical standard potential of ruthenium (IV) oxide is +1:120V [71], the total cell voltage
is expected to be 2:782V. The practical value of the cell voltage of the Al RuOx battery
will, however, not exceed 1:4V, as the battery is not operated under standard pressure and
temperature (1atm and 298:15K), and the electrolyte in contact with the electrodes is of
di erent composition than that corresponding to standard conditions (H+ concentration of
1mol=L, i.e., pH = 0). Given that our electrolyte is mildly acidic (typical pHvalues in the
range from pH2:5 to 5:5), it may not remove the native Al2O3 layer from the aluminum
42
surface. This thin, highly adherent aluminum oxide layer prevents Aldissolution in many
environments [72], which a ects on one hand, the measured cell voltage (i. e., the ther-
modynamics of the process), on the other hand, it decreases the rate at which metallic
aluminum can be oxidized (i. e., the kinetics of the process).
The advantage of aluminum over other metals is that cells equipped with Alanodes can
be recharged at very low voltage. Envisioned applications are in devices that need to be
recharged by harvesting various low power energy sources, such as RF energy or microwaves.
It is mostly used in sensor application, where the battery should be charged remotely.
To summarize, Alis a good candidate for anode material, owing to its low cost and its
light weight. However, aluminum is widely known to oxidize very fast and for this reason
it is not our main option for the anode material.
Anode Optimization
As already described, our battery uses metallic zinc as the oxidizing half cell (anode). Zinc
o ers several advantages over other candidate metals, such as (i) exchange of two electrons
per atom, (ii) Znis available in chunks or plates at low cost. In ingot form, bulk zinc costs
only about 0.3 cents per gram. Thin zinc foils are, however, expensive. Hence, anode cost
is a signi cant component of the overall galvanic cell. In order to reduce the cost of the
anode, we initiated avenues toward electroplating zinc onto thin metallic supports, including
copper, aluminum, and stainless steel mesh.
Our thinnest prototypes to date use very thin foils of zinc (1 to 3 mils in thickness)
for demonstrating energy density and  exibility. These foils are manufactured with much
higher tolerances than our galvanic cells require, and involve labor intensive processing,
which makes them far more expensive than is necessary for use in a battery. At present,
they are the highest cost battery component, despite being the lowest cost raw material in
the battery. Thus, one of the objectives is to develop a low cost method for producing a
50 m to 100 m thick zinc anode.
This process involves three stages: (i) degreasing and cleaning of the conductive surface
(e.g., steel), (ii) electroplating (with sacri cial or non-sacri cial anode), and (iii) post-
treatment (rinsing and drying, phospahting, chromating, or oiling the surface)[68].
43
Although the primary result for the replacement of electroplated zinc by zinc foil showed
promising results, further analysis needs to be done in this area to get the optimum result.
The result from the battery made with the electroplated zinc shows that the battery does
not deliver the same current capacity as the regular batteries. This might be due to the
shortage of zinc particles in the battery. Since a zinc layer is coated on di erent substrates,
drawing current from the battery will use up the zinc particles, and there will not be enough
zinc on the substrate to run the battery. Cycle life with electroplated Znanodes appears to
pose additional problems. So far, we could not exceed 10 charge/discharge cycles.
44
Chapter 4
EXPERIMENTAL TECHNIQUES
In this section I will present the working principle of the main electrochemical techniques
utilized in the framework of this Thesis. First, I will provide the basics of the instrumen-
tation, and next will describe the methods. There are multiple techniques for measuring
the performance of a battery. This section focuses on the measurement techniques used
to characterize the charge capacity, cycle life, and internal of the galvanic cells fabricated
throughout this Thesis work.
4.1 ELECTROCHEMICAL Instrumentation
Electrochemical instrumentation generally consists of a potentiostat, which imposes a con-
trolled potential or current at one of the electrodes, termed the working electrode (abbre-
viated as WE). The potentiostat supplies power to enforce charge to the electrode, so that
it can recharge secondary batteries.
Electrochemical workstations or potentiostats are instruments, which apply a known
potential and monitor the current response to the potentiostat. In present, the most widely
used potentiostats are assembled from discrete integrated-circuit operational ampli ers and
other digital modules. In some larger instruments the potentiostat package also includes
electrometer circuits, A/D and D/A converters, and dedicated microprocessors with mem-
ory. In other words, the potentiostat is a device that controls the potential applied across
the working electrode-reference electrode pair, and it adjusts this voltage to maintain the
potential di erence between the working and reference electrodes [73].
The schematic of a potentiostat is shown in Figure 4.1. The system maintains the
45
potential of the working electrode (WE) at a constant level with respect to the reference
electrode (Ref), by adjusting the current at a counter electrode (CE).
Fig. 4.1: Schematic of the potentiostat: the equivalent circuit of a 3-electrode cell (Ei is
the input voltage, Rm - current-measuring resistor, and CA - control ampli er).
The potentiostat used through out this Thesis for battery testing is a Model VMP3 Elec-
trochemical workstation, manufactured by BioLogic Scienti c Instruments (Claix, France),
it is a 16-channel multiplex instrument. The equipment uses a program function generator
to run various experiments. The manufacturer?s software for this device is called EC-Lab,
and it performs a variety of Electrochemical techniques from basic electrochemistry to cor-
rosion/impedance measurements and battery test. Each of the techniques provides  exible
user input, technique speci c plotting formats, and data analysis tools [74].
EC-Lab o ers more than 45 techniques that enable the user to address all Electro-
chemical applications. This software uses advanced graphic utilities. It provides two-axes
representation on Y1 and Y2. It also allows the user to observe more than one graph in one
window, and can monitor several parameters at same time. The simultaneous display of
several graphs allows the user to follow an active experiment, while analyzing data acquired
previously.
The software saves the setting for each channel and saves the data (in a so-called .mpr
 le format, which is the raw data  le). These  les can only be used by the BioLogic
potentiostat software. We can also save data as a text  le (in .mpt format), which can be
imported into any text editor software.
46
4.1.1 Electrochemical Techniques Performed in 3-Electrode Geometry
Three electrodes are being utilized in these Electrochemical measurements, with the fol-
lowing role: the electrode of interest, referred to as the working electrode (WE), has a
controlled potential set and monitored relative to the reference electrode (RE), while the
current is passed between the WE and the counter or auxiliary electrode (both terms are
being used in the scienti c literature, and are abbreviated as CTR and AUX, respectively).
The CTR (or AUX) electrode waives the RE from passing/handling large magnitudes of
currents, which would be damaging to the RE.
In 3-electrode geometry the electrodes can be connected to the BioLogic potentiostat,
as shown below:
 the electrode under investigation should be designated as the working electrode, and
attached to the potentiostat by means of the red alligator clip;
 the counter electrode, typically a Ptwire electrode, is attached by means of the blue
alligator clip, while
 the Ag=AgCl;3mol=LKCl reference electrode, commonly abbreviated as Ag(RE), is
clamped by means of the white alligator clip
4.1.2 Electrochemical Techniques Performed in 2-Electrode Geometry
For selected type of measurements, such as recording open circuit potential vs. time (ocP
vs. time) or performing bulk electrolysis without potential control of the WE, this simpler
2-electrode geometry can be utilized. Most battery testing experiments are conducted in
the 2-electrode geometry.
In this geometry, the WE will be utilized relative to CTR (or RE), where the counter
electrode is connected to the RE. For zinc-RuO2 cells the counter electrode is connected to
the zinc, while the working electrode is attached to the graphite  lm, which serves as the
current collector for the ruthenium side.
47
4.2 Electrochemical Techniques and Testing Procedures
This section summarizes di erent methods and techniques for battery testing. It also delves
into di erent experiments used to determine various characteristics of a zinc-RuO2 battery.
4.2.1 Cyclic Voltammetry (CV)
Beyond doubt, this is the most versatile, the most popular, and the best known Electro-
chemical method, with a wide range of applications from testing battery components to
evaluating biological electron transfer, from determining corrosion rates to examining elec-
trocatalytic properties of electrode materials. Also, CV is feasible for exploring various
electroactive materials, from the gram-scale to the nano-scale. Given the wealth of its ap-
plications and the easiness of performing the analysis, one can consider cyclic voltammetry
as the queen of electrochemical techniques.
The working principle of CV is presented in Figure 4.2, where the potential change vs.
time, referred to as the scan rate, is imposed by the researcher, and the switching potentials
are chosen according to the features of the system under investigation.
One important setting in CV is the so-called sign convention. The European sign con-
vention displays the anodic peaks up and the increasingly oxidizing potentials from the
left to the right, while the American or polarographic sign convention is exactly the oppo-
site. For our workstation the setting of the instrument is according to the European sign
convention, as shown in Figure 4.3.
4.2.2 Open Circuit Voltage
The open circuit voltage (ocV) measurements refer to a time span during which no potential
or current is applied to the working electrode. In ocP mode the cell is not connected to any
power ampli er. The evolution of the rest potential for the cell can be observed over time.
This data can be used as preconditioning time or for equilibration of the Electrochemical
cell [75]. Utilizing the zinc-RuO2 cell with electrolyte of pH2.5 and a glass- ber separator
as an example, the nominal cell voltage is 1:4V. Therefore, the operating voltage of the
cell will be from 1:4V to 0:6V, and the end voltage is typically 0:5V, for moderate and low-
48
Fig. 4.2: Principle of cyclic voltammetry. In the left bottom panel the applied potential
is displayed vs. time, and the current response vs. time is shown above it, while
the right panel shows the output, the voltammogram in which current is plotted
relative to the applied potential.
drain discharges, and 0:3V, for engine cranking. The open circuit voltage of the Zn RuO2
battery is derived from the potential of the active anode and cathode materials, i.e., zinc and
ruthenium (IV) oxide. The open circuit voltage depends also on the type of the ruthenium
oxide (hydrated or anhydrous), and on the composition and pH of the electrolyte. As a
result, the open circuit potential for a Zn RuO2 battery can be as low as 1:1V to as high
as 1:5V.
4.2.3 Charge /Discharge Studies
Di erent methods have been used to charge the battery, namely, constant current regime
and constant voltage regimes. In these applications, charge and discharge sequences with
possible open circuit periods can be measured. We can also look at the constant load
discharge of a cell with di erent external loads. This is the most common test on a cell
to determine its capacity. Furthermore, the life cycle of a cell can be explored. One can
examine how the amount of charge going through a number of charge/discharge cycles can
49
?
? 62
 The working principle of CV is presen ted in \autore f{fig:4.2}, where the pote ntial 
change vs. ti me, r eferred  to as the \ emph{ scan rate} , is i mpos ed by  the resea rcher, and the  
switching pot entials are ch osen according to the featur es of the sy stem  under investigation. 
Cyc lic  vo lta mme try (CV)lic mm e
tim e
E
time
i
E
i
}
Re d ? ne
-
? Ox
Ox  + ne
-
? Re d
 
 Fi gure 4. 2. Prin cip le of cyclic v oltammetry.  In  th e left  botto m p anel th e ap plied  po ten tial is d isp layed  
vs. tim e, an d the curren t respo nse v s. tim e is sho wn ab ov e it, wh ile th e rig ht pan el sh ows th e ou tpu t, 
th e vo ltamm ogram  in  wh ich  cu rren t is p lotted  relativ e to  the app lied  potential.   
 
 One i mportant setting in CV is the so-called \em ph{ sign conventio n} .  The European  
sign conventi on displa ys the anodic peaks up and the increasingl y oxidizing potentials from  
the left to the right, while the American or polarographic sign convention is  exactly  the 
opposite.  For our wor kstation t he setting of t he instr ument is acco rding t o the E uropean sign 
convention , as shown in \a utoref{fig:4. 3}. 
 
 
 
Figure 4.3.  E uropean sign convention  
+?E?
i
a
Fig. 4.3: European sign convention.
in uence the performance of the cell. The battery will also exhibit self discharge over time
that can be measured for a long time period.
Constant Current Charging
The title technique uses a constant current to charge the device under test, while monitoring
the voltage across the cell. In order to charge the battery, a proper value for the current
should be chosen. The battery undergoes the charging process, the voltage of the cell being
monitored until it reaches a certain value. The voltage it stops changing over time. If
the current is too low, then the overall charging process is very slow [76]. The described
methodology is customarily applied for charging Li-ion batteries. If the current is too
high, then it will generate excessive gas in selected types of batteries, due to the fast rate
electrolysis of the supporting electrolyte inside the cell. One possible way of charging a
battery using constant current regime is to apply to the battery a high current for a short
time, and then charge the battery at lower current values over a longer time. This technique
is called two-step constant current charging, which uses medium-to-high rate charge in the
 rst period, followed by a very low current call trickle current, until the battery reaches full
charge [76]. The advantage of using very small currents for charging a cell is evident, when
a series of batteries are connected for charging. This allows the batteries to be rebalanced
at the top-of-charge, i.e., at the nominal voltage of a cell, without causing any damage to
the batteries that are already fully charged [76]. Overall, the constant current charging is
a very slow process that can be considered for overnight charging.
Two sets of currents were used to charge the novel Zn RuO2 battery. Figure 4.4 shows
50
the charging curve for the current set to 240 A. The charging process takes 3h and 30min,
until the battery reaches the nominal voltage of 1:15V, and the voltage  ats out at that
voltage.
By comparison to the previous set up, a second test uses 50 A constant current to
charge a battery, which leads to a longer charging time. In this process, it takes for 8h and
30min to charge a battery to the constant value of its nominal voltage. Figure 4.5 displays
the charging curve for this method.
Since the batteries had been discharged several times the initial nominal voltage of
1:4V, could not be reached. The maximum current provided by the potentiostat was only
240 A, therefore, excessive time was needed to charge up the battery, and, also, the highest
potential could not be reached.
Comparing the two techniques, one can see that when the higher current is used to
charge a battery, the faster the charging voltage will achieve its highest point. The smaller
current can also be useful when time is not the main concern.
Fig. 4.4: Constant current charging with 240 A current.
Constant Voltage Charging
As the name implies, this regime involves the use of a constant voltage throughout the
charging process. The current starts o at a very high value, and then decreases almost
exponentially to a low value. When the battery is deeply discharged, the starting current
51
Fig. 4.5: Constant current charging with 50 A current.
can be very high; this may result in internal heating of the cell. Figure 4.6 is an example
of current-time curve, when the Zn  RuO2 battery is charged with 1:4V for 1h. As the
plot implies, the charging current starts as high as 74mA, and drops to 0:062mA in 1h.
Depending on how deep the battery was discharged previously to the charge, the initial
current and the time needed for the current to drop to very low values may vary. Further
studies are needed on the charge time of a battery for reaching an accurately set potential
value.
Fig. 4.6: Constant voltage charge current of a Zn  RuO2 cell, when charged at 1:4V for
1h
Constant voltage charging can be used as a fast technique for returning maximum charge
in a minimum time [76]. The starting potential, when a battery is charged at a certain
potential, will be the same as the charging voltage under normal conditions. For example,
52
if a battery is charged at 1:4V, then the starting voltage after the charging will be 1:4V, as
well.
Constant Load Discharge
Constant load discharge (CLD) is the technique to measure the total capacity of a cell. As
explained in section 4.1, the BioLogic potentiostat allows the user to set the load. Constant
load discharge implies that the load resistance stays constant throughout the discharge for
speci c duration of time. The magnitude of the external load used for testing the battery
mainly depends on the application of the battery. If the battery needs to provide a large
burst current in subsequent a short given time, then smaller load resistance is used. By
contrast, when it is needed to source small constant current over a longer time period, then
a greater load resistance will be used to test the performance of the battery. The charge
capacity of a galvanic cell can be expressed in Ah or mAh units. To calculate the capacity
of a cell constant load discharge technique is used.Figure 4.7 shows an example of constant
load discharge of a cell for measuring its capacity.
Fig. 4.7: Constant load discharge of a battery through 1k external load, providing a
charge capacity of over 10mAh=cm2.
Constant Current Discharge or Chronopotentiometry
Chronopotentiometry (CP) is a controlled current technique. The current is controlled, and
the potential is the variable determined as a function of time. This technique is mostly used
53
when one is interested in the performance of a cell as a response to a pulse. CP technique
enables for the monitoring of the potential of a cell, when speci c amount of current is drawn
over time. Figure 4.8 shows an example of using this technique, when the potentiostat is
set up to draw speci c current for a particular time period, while monitoring the voltage.
The magnitude of 0:1mA current was drawn from a cell for 30s, and the potential of the cell
was monitored. As will be explained in subsection 4.2.5, the CP can be used for internal
resistance testing or for the experiments or for protocols where an exact magnitude of peak
current is required.
Fig. 4.8: Voltage vs. time discharge recorded by CP for calculating the internal resistance
of a cell.
Cycling
Rechargeable batteries, also known as secondary cells, are batteries with reversible electro-
chemical reactions, which allow for reversible electrical operation of the cell. Ruthenium
(IV) oxide is an example of a cathode that can be used in rechargeable batteries. The
number of cycles that a battery is able to perform is referred to either as cycleability or
cycle life. To test this feature, the battery should be discharge through a constant load for
a certain duration, and then recharged. The charge time depends on both the load and
the depth of the discharge. In this Thesis, the typical cycleability test is 1h discharge of a
cell through 1k load, followed by 1h charging at 1:4V potential. Figure 4.9 is an example
54
of approximately 100 full 1h cycles achieved with a 4cm2 Zn  RuO2 cell. As the plot
reveals, the voltage slowly decays as the number of cycles increases. What is desired from
a secondary battery is to be able to undergo a number of discharges, while maintaining the
dropping potential constant.
Figure 4.10 shows the amount of charge that was drawn from the battery in each cycle.
On the average, the battery is able to source 0:9mAh. Further investigation shows that 95
cycles have the current capacity above 80% of the capacity of the  rst four cycles.
Fig. 4.9: 100 cycles accomplished from a 4cm2 Zinc-RuO2 battery with 1k load and 1:4V
charging intervals.
Fig. 4.10: The capacity plot drawn from a 4cm2 Zn-RuO2 battery
The battery is not able to accomplish cyclability exceeding 100 cycles for discharges 1h
duration through 1k load. More cycles can be accomplished with smaller rate load and
55
faster discharges. Figure 4.11 shows the completion of 800 cycles that involve very short
discharge intervals of 1min discharges with 3min charges.
Fig. 4.11: 800 cycles accomplished by the Zn  RuO2 battery.
Listed below are some considerations with regards to cyclability for zinc batteries. It is
di cult to accomplish a large number of cycles with zinc anodes [77]. There are three main
reasons for this inability:
1. The material tends to form dendrites that pierce the separator, and short out the
galvanic cell;
2. There is a \size e ect," whereby the zinc surface changes shape during discharge charge
cycles. Zinc clumps in one area, and disappears from others. This leads to current
crowding, local heating, and the eventual destruction of the electrode; and
3. Formation of insoluble zinc oxides.
Further studies have been done on the above features for solving these problems. The
zinc particles need to be kept in place, and chemical reactions should proceed on the zinc
surface. The goal is to protect zinc from surface corrosion. Di erent additives have been
tested to keep the zinc particles in place, which is beyond the scope of this Thesis.
56
4.2.4 Shelf Life/Self-Discharge
Shelf life of a cell is de ned as the length of time over which the battery maintains its
potential and capacity, when not in use. The shelf life of a cell is directly tied to the extent
of its self discharge. In other words, the time period that battery is being stored and still
is suitable for operation there after that depends on how much the battery is going to
discharge without being connected to a load, i.e. in open circuit mode. Our experimental
 ndings show that self-discharge of a cell in the Zn  RuO2 battery depends primarily on
electrolyte pH. Near neutral pH, the cell shows very low self-discharge. Figure 4.12 reveals
the low self-discharge for a 4 4cm2 battery discharged through 50k load, over three
months. A negligible drop in cell voltage occurred.
Fig. 4.12: Three-month discharge of a cell with 16cm2 of active area, operating with elec-
trolyte of near neutral pH.
As shown in subsection 5.1.2 the acidic pHenables for higher current capacity, owing
to the lower internal resistance of the cell. There is a trade-o that we should keep in
mind, while choosing the electrolyte, although the lower pHhas smaller internal resistance,
it increases self-discharge of the cell. In other words, whenever an electrolyte is Electro
chemically active, it typically is active in the purely chemical process, as well, i.e., it speeds
up the corrosion of the zinc electrode, and by this it decreases shelf life. We will need to
57
optimize the cell system with this trade-o in mind.
A series of batteries have been tested for shelf life and self-discharge. These batteries
were made with electrolyte having the pHof 2:50 0:05 and the active area of 2cm 2cm.
Four such batteries were connected in series as part of a battery cloth. The potential of
the battery was monitored over time, and plotted as shown in Figure 4.13. A comparison
between the performance of these two batteries reveals that the cell with the close to neutral
pHhas a 27% drop in potential over 90 days, while the battery with the more acidic pHhas
dropped by 38% in 64 days. For the same time period of 64 days the battery with the
neutral pHhas dropped only by 9%.
Fig. 4.13: Two months discharge of a 5V cell with 4cm2 of active area and acidic pH.
4.2.5 Internal Resistance
The internal resistance of a battery, also calls equivalent series resistance (ESR), is de ned
as the \force" opposing to the  ow of current within the battery [78]. Any electronic device
that generates voltage can be represented by an ideal voltage source with impedance in
series with it. Considering the time dependency of the cell behavior, the ideal voltage
source can be in parallel with a capacitor. The impedance is called the internal resistance
of a battery. Figure 4.14 illustrates the model for a battery.
58
Fig. 4.14: The equivalent circuit of a battery.
There are two factors that impact the internal resistance of a cell. They are the elec-
tronic resistance and the ionic resistance [78]. The summation of these two resistances is
called the total e ective resistance of a battery. The ionic resistances are due to various
Electrochemical parameters, such as electrolyte conductivity, ion mobility, and the surface
area of electrodes [78]. These factors a ect the performance of the cell to a lesser extent
than the electronic resistance. These factors typically act when the battery is connected
to a load. Electronic resistance also in uences the performance of a battery. Rather than
being internal, the electronic resistance is mostly due to the connection of a battery to other
devices.
Figure 4.15 shows the experiment done on the novel ruthenium oxide cell to calculate
its internal resistance. For 5s a battery was in open circuit, hence, no current was drawn
form the battery. In the meanwhile, the voltage over time was monitored. Then, over the
next 30s larger magnitude of current was drawn from the cell. In this step, the battery
was set up to source 100 A current. Using Ohm?s law the internal resistance of the battery
can be calculated by dividing the drop in voltage by the change in current. This yields the
internal resistance of 5 .
In order to get more accurate results on the internal resistance of the cell, this experiment
needs to be repeated for di erent current rages. chapter 5 looks into the internal resistance
of the batteries, when di erent parameters of the battery (such as electrolyte of separators)
were changed.
As explained above, di erent parameters will in uence the internal resistance. There-
fore, the material used in the cell, e.g.,  lter paper separators, will largely a ect the internal
59
resistance. Same technique has been used to calculate the internal resistance of a cell in
chapter 5.
Fig. 4.15: Calculating the internal resistance of a cell by using CP technique; it yields the
value of 5 .
As described in chapter 3, a simpler form of the cathode reaction is as shown in equation
(4.1):
RuO2 + 2H+ + 2e  !Ru(OH)2 (4.1)
It is possible for a cell with an acidic electrolyte to exhibit the reaction according to
equation (4.2):
2H+(aq) + 2e  !H2(g) (4.2)
The formation of hydrogen gas also adds to the internal resistance of a cell. Creating a
hydrogen bubble enforces ions to avoid the bubbles, by traveling around the gas bubbles,
as opposed to going in straight directions. This phenomenon will lower the rate of ion
migration, and result in the increase of the internal resistance.
60
Chapter 5
MEASUREMENTS AND TESTING RESULTS
As described throughout this Thesis di erent parameters will a ect the performance of a
battery. This section looks into the performance of a Zn  RuO2 battery when di erent
parameters are being changed in the battery. The battery performance has been analyzed
when di erent pHelectrolytes as well as di erent separators and anode materials are used.
5.1 Electrolyte E ects Measurements
The e ect of electrolyte on the battery performance has been described previously. This
section demonstrates the batteries? capacity when electrolytes of di erent pHvalues are
used. Di erent materials are used to make an electrolyte. Some of the electrolytes consist
of same materials with di erent amounts of acid added to adjust the pH. The battery
with highest current capacity of 33:35mAh=cm2 was made with the mixture of zinc chloride
(ZnCl2) and ammonium chloride (NH4Cl) and uses hydrochloric acid (HCl) to adjust the
pH. Basic electrolytes with higher pHwere made with the zinc acetate family, which did
not provide outstanding results.
5.1.1 Original Electrolyte
The original cell was implemented with an electrolyte composed of a weak acid and ethylene
glycol that is close to neutral pH, to prevent rapid dissolution of the electrode materials
or package sealing materials, while providing the needed redox reactions. For the original
standard electrolyte a mixture of boric acid, citric acid and ammonium hydroxide mixed
with ethylene glycol was used. The charge capacity accomplished with this electrolyte was
61
0:0325mAh=cm2. As part of the improvements for the performance of the battery, di erent
types of electrolytes with di erent compositions and di erent pHwere used.
5.1.2 Overview of Electrolytes of Di erent pHValue
As shown above, the composition and pHof the electrolyte will a ect battery performance
signi cantly. Listed in Table 5.1 are the compositions and pHvalues of various electrolytes
tested in Zn  RuOx batteries. The second column displays the internal denomination of
each electrolyte, while column 4 shows the best charge capacity accomplished with each
electrolyte.
Fig. 5.1: Plot of battery capacity versus pH for di erent electrolytes. Note that connecting
point is just for visual clarity.
As the plot in Figure 5.1 illustrates, at pH2:48 of the electrolyte, the galvanic cell reached
the maximum capacity of 33:35mAh=cm2. This electrolyte (E10A4) was made with a
mixture of zinc chloride and ammonium chloride, both salts used at high concentrations, the
pHof the solution being adjusted with hydrochloric acid. Another battery was made with
a di erent salt mixture, at the same pHof 2:48. The latter electrolyte (E10A5) contained a
ternary salt mixture: zinc chloride, ammonium chloride, and zinc acetate. Batteries made
with the latter electrolyte delivered high current capacity, but not as high as with the E10A4
electrolyte. Other high capacity points were recorded at pHranges from pH1:7 to 3:1. Also,
at the pHof 4 and 4:3, the capacity values were slightly less than half of the maximum
accomplished point.
62
Salt Composition Electrolyte pH Capacity
(mAh=cm2)
Boric Acid, Citric Acid, NaCl E05 0.32 1.11
ZnCl2, NH4Cl E10A2 1.71 13.75
ZnCl2, NH4Cl E12 2.36 17.5
ZnCl2, NH4Cl E10A4 2.48 33.35
ZnCl2,Zn(C2H3O2), NH4Cl E10A5 2.48 21
ZnCl2,Zn(C2H3O2), NH4Cl E12A 2.52 15
Zn(C2H3O2)2, NH4C2H3O2 E16B 2.6 11.04
ZnCl2, NH4Cl E12B 3.1 15.67
ZnCl2, NH4Cl E10A1 3.18 6.43
ZnCl2, NH4Cl E10A 3.5 1.03
Zn(C2H3O2)2, NH4Cl E014A 4 3.67
ZnCl2, NH4Cl E12C 4 13.37
Zn(C2H3O2)2 E013A 4.28 1.11
ZnCl2, NH4Cl E10A3 4.31 13.7
Zn(C2H3O2)2, (NH4)2SO4 E015A 4.6 2.41
Zn(C2H3O2)2,NH4C2H3O2 E16A 4.8 1.04
ZnCl2, NH4Cl E10B 5 1.12
Boric Acid, Citric Acid E01 5 0.032
ZnCl2, NH4Cl E11 5.5 2.12
Zn(C2H3O2)2, NH4Cl E014 5.85 1.48
Zn(C2H3O2)2 E013 6 0.78
Zn(C2H3O2)2, (NH4)2SO4 E015 6.25 4.27
Zn(C2H3O2)2, NH4C2H3O2 E16 6.58 0.8
Tab. 5.1: Denomination of di erent electrolytes with di erent pHvalue and the charge
capacities enabled by them. All batteries were made under similar conditions
and testing environment.
As column one of the table implies, the electrolytes were made with di erent salt com-
positions, at di erent molarities of the salts, which are not mentioned in this Thesis due
to the con dentiality of the material. Figure 5.2 provides a relevant correlation of charge
capacity and pH, as it shows electrolytes that belong to the same family of salts. The
electrolytes were made with di erent molarities of zinc chloride and ammonium chloride.
As shown in the plot, the electrolytes made with zinc chloride and ammonium chloride
have high current capacity except for pHgreater than 5.
In an attempt of explaining the outstanding performance of the electrolytes, which
contain zinc salts, we hypothesize an equilibrium process between the metallic zinc, which
is the anode material, and the hydrated zinc ions, present in the electrolyte, according to
equation (5.1):
63
Fig. 5.2: Capacity versus di erent pHvalues of the electrolytes that belong to the same
family; they all contain di erent concentrations of zinc chloride and ammonium
chloride. Note that connecting point is just for visual clarity.
Zn(s) + Zn2+(aq) !Zn2+(aq) + Zn(s) (5.1)
Where  marks the metallic zinc, which constitutes the anode. This equilibrium may
control the dissolution of metallic zinc from the anode, and the zinc ions added to the
electrolyte is likely to limit the extent and rate of metallic zinc dissolution. Therefore, the
equilibrium shown in equation (5.1) improves the charge capacity and the cycle life, while
it also extends the shelf life of the galvanic cell.
In addition, the high concentration of salts present in the zinc salt based electrolytes
confers good ionic conductivity to the system, which limits the magnitude of unwanted
ohmic drops in the cell. Our most recent experiments revealed that not the highest con-
centrations of zinc and ammonium salts are the most e cient. Upon reducing the Zn2+
salt concentration to by 1=3 and the NH+4 to approximately 1=2 of that used in E010A4,
the charge capacity of the battery has increased. We believe that this  nding may be a
consequence of a good compromise between the conductivity of the electrolyte and its lesser
salt concentration, which prevents the dry out of the cell.
64
5.2 Cathode Current Collector
We represent here the performance of the Zn RuOx battery, when di erent current collec-
tors are utilized. Table 5.2 shows the thicknesses of di erent current collectors, their Ohmic
resistance, and their performance in battery operation.
Cathode Current Collector Thickness (mm) Internal Resistance (Ohm) Capacity (mAh=cm2)
Carbon Mesh 0.26 82 58
Carbon Veil 0.10 57 30
Carbon Cloth 0.30 66 35
Buckeye 0.10 140 50
Graphite sheet 0.13 40 34
Tab. 5.2: Di erent current collectors.
It can be observed that the carbon mesh is delivering the highest capacity of 58mAh=cm2.
However, it also has the greatest thickness among the current collectors, and this prevents
accomplishing our goal. The buckeye paper has higher capacity than the graphite sheet.
However, purchasing these materials is currently not cost e cient. Buckeye paper is 7 times
more expensive than that of graphite.
Graphite sheet possesses two advantages over other materials. First, it is easier to work
with it and is uncomplicated to handle, as compared to carbon veil or buckeye paper. Those
materials are fragile and often break. Therefore, on the lab scale, where batteries are hand-
made, the handling of graphite sheet is easier than of other materials. Second, graphite
sheets exhibit the lowest internal resistance of 34 Ohm.
Table 5.3 illustrates the nominal continuous current for each of the tested current col-
lectors. As the table reveals, although graphite mesh has the highest capacity, its nominal
continuous current is 35% less than that obtained with graphite sheet. It can also be ob-
served that the battery with graphite sheet has high nominal current, same as the case with
buckeye and carbon cloth, although the capacity is lower.
5.3 Separator Analysis
This section looks into the e ect of separators on the performance of the batteries. The
capacity of a cell has been analyzed, and its internal resistance and cyclability, as well.
65
Cathode Collector Nominal Continuous Current (mA)
Carbon Mesh 0.74
Carbon Veil 0.54
Carbon Cloth 0.86
Buckeye 0.84
Graphite sheet 0.84
Tab. 5.3: Nominal continuous current for di erent cathode current collectors
Various separators with di erent thickness were used in making the batteries.
Table 5.4 lists some separators obtained from commercial suppliers, each having a di er-
ent thickness and a characteristic performance. All batteries contained the same electrolyte,
with the acidity pHset to 2:43, and were made under similar conditions. The thickness of
a cell versus its capacity as well as its internal resistance and the number of cycles were
measured.
Name Thickness Capacity # of Cycles Internal Resistance
( m) (mAh=cm2) ( )
F&J Specialty 350 33.35 40 32
Fisher Scienti c 240 20.30 61 73
Whatman 216 12.5 68 65
AMC 100 10.5 33 61
KIMTECH Wipes 60 7.87 20 29
Celgard 20 9.87 2 72
Tab. 5.4: Di erent separators with di erent thickness and their performance in capacity,
number of cycles, and internal resistance.
Figure 5.3 shows the thickness of the utilized separator versus the capacity of the cell.
As it can be seen, the thicker the separator, the more capacity can be drawn. This may be
a consequence of more electrolyte absorbed by the thicker glass- ber separator.
The batteries with di erent separators also underwent cycling test. The cycling tests
for these batteries were recorded as described in chapter 4. Figure 5.4 is the plot of the
separator thickness versus the number of cycles. The cycleability of a battery was based on
the number of 1-hour cycles it underwent over a 1k external load. As the result implies,
unlike the capacity test the separator thickness is not proportional to the number of cycles
and the separator with 216 m.
The internal resistances of cells related to di erent separators were also measured ac-
66
Fig. 5.3: Capacity versus the thickness of a separator obtained from Zinc-RuO2 battery
system. Note that connecting point is just for visual clarity.
Fig. 5.4: Number of cycles versus the thickness of a separator obtained from the Zn RuO2
battery system. Note that connecting point is just for visual clarity.
cording to the procedure discussed in subsection 4.2.5.
The internal resistance is not necessarily related to the separator thickness. As the plot
in Figure 5.5 shows, the KIMTECH wipes have a very small internal resistance, while the
capacity of the cell and its number of cycles are not the best.
One important factor that we need to keep in mind while comparing the separators is
that not all separators tested by us belonged to the same family. The F&J and the Fisher
Scienti c separators are made of glass- bers, while Celgard is based on polypropylene layers.
The membrane material for AMC is modi ed hydrophilic polyethersulfone.
The above considerations are of important qualitative value. Their quantitative infor-
67
Fig. 5.5: Internal resistance of a cell versus the di erent separators thickness. Note that
connecting point is just for visual clarity.
mation is hindered by the fact that we were not able to make the same type of battery
at di erent thicknesses. The unavailability of di erent thicknesses of separators having the
same composition prevented a more accurate comparison.
5.4 Anode Materials
While ruthenium is a novel material used as a cathode in this study, zinc and aluminum
batteries have been known for decades. As described in subsection 3.1.2 the reaction on the
anode side is quiet simple. However, the reduction potential of the anode will determine
the cell voltage. In this section, I present my analysis of the performance of the Zinc-RuOx
and Al  RuOx batteries.
5.4.1 Zinc-RuOx Batteries
The capacity and cycle life for the zinc battery is demonstrated here. Typically, batteries
were made with 0:6mm thick zinc as the anode and a mildly acidic electrolyte, having
pH2:48. The electrolyte was the combination of a zinc salt and an ammonium salt, in most
cases both were chlorides. The pHof the electrolyte was adjusted by adding rigorously
controlled amounts of an acid (HCl, H2SO4, or acetic acid). The separator was glass- ber
from F&J Company; it had the thickness of 350 m. The battery provided 33mAh=cm2 ,
68
when connected to a 1k external load.
A battery with the same characteristics was made for testing the cycling ability of the
cell. The battery was discharges through 1k load for 1h, and then charged at the applied
voltage of 1:4V for another hour. The goal was to assess the number of cycles the battery
is able to provide charge, i.e., how long it can cycle consistently. The battery went through
98 cycles before it hit the pre-set end potential of 0:7V. Figure 5.7 shows the cycling of the
battery. Currently, the desired cycle life for the secondary battery is not yet achieved. The
battery dies after a number of cycles because of (i) the formation of gas bubbles that a ect
the battery membrane, and increase the internal ohmic resistance of the galvanic cell, and
(ii) the dry out of the cell; a very small volume of electrolyte is added to the cell, which is
eventually lost by electrolytic decomposition and evaporation.
Fig. 5.6: Discharge curve for Zn  RuO2 battery with pH2:48 electrolyte over 1k load
Fig. 5.7: Approximately 100 cycles accomplished with Zn  RuO2 battery through 1k 
load with charging cycle in between two discharges, at an applied voltage of 1:4V.
69
5.4.2 Al  RuOx Battery
As explained in section 2.4.1, Al can be used as an anode in the RuOx cells. We employed
two di erent electrolytes to make the Alcells. One with pH6:2, containing boric acid,
citric acid, and ethylene glycol. The battery was not able to deliver more than 102 Ah
before it dropped to a very low voltage and died. Figure , shows the voltage curve of
the battery. The battery has the initial potential of ca. 1V. Another strongly acidic
electrolyte, with pH 0:2 was used to test the Al RuO2 battery. The electrolyte has the
same composition as previously, except that it also contains hydrochloric acid. Figure 5.8(b)
shows the discharge curve for this battery. The battery had higher initial voltage than the
battery with greater pH.
Fig. 5.8: (a). Discharge curve for the Al  RuO2 battery with pH6:2 electrolyte. (b).
Discharge curve for the Al  RuO2 battery with pH0:2.
5.5 Monitoring Hydrogen Discharge and Water
Decomposition by Cyclic Voltammetry in Open Cell
Experiments
The goal of this investigation was to determine the potential window in which each elec-
trolyte is stable, and to get a preliminary diagnostics on gas evolution, in both the cathode
and anode reaction. The technique described here enables for examining in a systematic
manner the e ect of pHon gas formation.
We tested several electrolytes that we customarily utilize in the galvanic cells, as follow:
E010A4 (pH-0.15), E010A5 (pH1.64), E010A7 (pH2.40), E013 (pH5.68), E014 (pH5.85),
and E030A (pH2.26).
70
Tests were performed by cyclic voltammetry (CV), in 3-electrode geometry. The working
electrode was graphite  lm. Metallic zinc plate was the counter electrode, and Ag=AgCl,
3MKCl, abbreviated as Ag(RE), served as the reference. This setup can be extended to
the use of working electrodes identical to those employed in sealed galvanic cells, namely,
hydrated Ru(IV) oxide con ned to graphite support
Cyclic voltammetry was recorded in a 10mL glass cell, which contained 5mL of elec-
trolyte, and the above electrodes were immersed into the liquid. Measurements were per-
formed at room temperature, without degassing the electrolyte. CV scans were recorded in
an increasingly wide voltage window, e.g., from the initial 0:6V to +0:8V to the more and
more positive upper value concomitant with the more and more negative limit. The scan
rate was kept at moderately fast value (at or around 100mV=s). CV scans will be recorded
at identical scan rate, at room temperature, and without degassing the electrolyte solutions.
These conditions are comparable to the ones in the FlexEl galvanic cells. Therefore, the ex
situ experiments are expected to be relevant to our real life batteries.
Scans recorded so far showed that the negative limit of the voltage range typically
cannot be pushed below  1:15V, while the upper limit can be as high as +1:3V or even
more positive. It was also demonstrated that metallic zinc cannot serve as the working
electrode (given that it behaved as an ideally non-polarizable electrode), but it can be used
as the counter electrode. This setting corresponds to the setting of the two electrodes in
the galvanic cell.
Figure 5.9 displays an overlay of two CV scans recorded for electrolyte E10A5, in a
potential range, which has the same upper limit, but a slightly di erent negative potential
limit. One can notice a striking di erence, when the negative end of the scan is extended
by only 100mV; while at  1:05V vs.Ag(RE) no hydrogen evolution occurs, ar  1:15V vs.
Ag(RE), strong hydrogen discharge proceeds.
In E010A4 electrolyte (pH < 0), strong hydrogen evolution occurred, as soon as the
zinc plate was immersed in the solution. It looked like boiling water. Hence, open cell
experiments reveal the same behavior of zinc in electrolytes, as evidenced by the ex situ
zinc corrosion experiments. Potential windows for various electrolytes are listed in Table 5.5.
The pHvalue of the electrolyte was measured with a Model SevenEasy pHMeter (Toledo-
71
Fig. 5.9: CV scans recorded for electrolyte E10A5, at 100mV=s: inner (blue) scan: from
 1:05V to +1:22V vs. Ag(RE), while the external curve was obtained in the
potential range from  1:05V to +1:22V vs. Ag (RE).
Mettler) equipped with an InLab c Routine Pro pHprobe.
The described methodology will be extended for all electrolytes in our stock. Primarily,
we will focus on electrolytes with a pHvalue relatively to the neutral pH. This means
the range 2:5 < pH < 5:5. We are, however, interested in the E020 series of strongly
basic electrolytes, as well. In addition to E022, which has already been tested in open cell
experiments, with graphite  lm WE (see table above), we will perform similar experiments
on E020 and E022 listed in Table below, and some other chlorate based electrolyte from
the E030 series in Table 5.2, which have not been tested, so far.
Once we down-select the electrolytes o ering the largest voltage window, we utilize
them in testing the real cathode material, with incorporated RuO2 nH2O. For this, we
use carbon paste electrode, which served well for the preliminary ex situ investigation of
various dendrimers.
We will add RuO2 nH2O powder in 1:1 volume ratio to the carbon paste, will homoge-
nize the mixture by stirring it with a thin spatula, and then  ll with this mixture the cavity
of the carbon paste electrode. The same paste could be made with RuO2 nH2O and our
activated carbon (used routinely in preparing the cathode paste of our galvanic cells) by
adding the powder mixture the minimum volume of silicon oil, which still allows for the
72
formation of the paste, and keeping the powder together. Then, this paste would be  lled
in the cavity of the carbon paste electrode, and tested by CV according to the previously
established procedures for carbon paste electrodes.
Electrolyte pH H2 discharge H2O Decomposition.
(new probe) [V vs. Ag(RE)] [V vs. Ag (RE)]
C C + RuO2 C C + RuO2
A10A4 -0.15 -0.850 +1.220
A10A5 1.64 -1.050 -1.000 +1.220 +1.080
E10A7 PEG6 1.66 -1.200 -1.135 +1.130 +1.080
E032A 2.05 -1.250 -1.200 +1.410 +1.410
E030A 2.26 -1.150 -1.120 +1.300 +1.400
E10A7 2.40 -1.100 +1.250
E10A3 3.93 -1.060 -1.030 +1.120 +1.075
E031A 4.16 -1.040 +1.250
E10A 4.54 -1.050 -1.030 +1.150 +1.100
E10A13 5.68 -1.120 +1.650
E032 5.72 -1.200 -1.100 +1.330 +1.340
E015 5.75 -1.150 -1.125 +1.300 +1.280
NaHCO3 8.25 -1.350 -0.500 +1.320 +1.220
E022 14.45 -0.520 +0.450
Tab. 5.5: Potential windows for various electrolytes.
5.6 XPS Analysis
We have analyzed a freshly made cell and a completely discharged copy of the same cell
using XPS (X-Ray Photon Spectroscopy), at the Shared Instrumentation Facilities of the
Dept. of Chemistry, University of Maryland in College Park. The XPS facilities are directed
and used by Dr. Karen Gaskell.
XPS works by sending monochromatic x-rays to the sample and measuring the kinetic
energy and number of emitted electrons. The results are given in a binding energy-vs-count
of electrons detected per second. The visible peaks in a typical spectrum correspond to
orbital binding energies and characteristic Auger emission energies of materials. Figure 5.10
below is an example graph taken from Wikipedia, showing the 2p-level-split of Silicon.
Some features of XPS analysis that is relevant to our particular case are listed below:
1. Some characteristic peaks of certain materials overlap. Signi cant in our case is the
73
Fig. 5.10: The Si(2p) peak obtained with XPS.
overlap between C(1s) and Ru(3d). In such cases, other peaks of similar materials
can be consulted for veri cation.
2. The place and shape of peaks can be used to deduce oxidation states of a given
material, as well as to distinguish, for instance, between metallic Znand ZnO. Metallic
material peaks tend to be narrower and asymmetric.
3. It is possible to get a rough idea of the sample composition in terms of number-of-
atoms ratios between di erent species in the sample.
4. XPS is ran in a high vacuum. Therefore, during the measurement the ruthenium oxide
in the samples is probably losing its water of hydration. This is a possible phenomena
to be aware of during the analysis.
5.6.1 Analysis on the Cathode Side
Figure 5.11 shows the full survey results for the \black powder" (mainly the compounded mix
of ruthenium oxide and activated carbon, plus leftover solute materials from the electrolyte
after the solvent has dried in the vacuum) for the new and discharged batteries.
Examining this result, several points can be immediately made:
1. Metallic zinc (Zn) is present on the cathode side even in the non-discharged (new)
battery. This indicates that some Znis immediately dissolved when the battery is
74
made, and that Zn2+ ions can travel through the standard separator we have been
using ( lter paper). After the discharge, the amount of Znincreases.
2. The C: Ruatomic ratio is about 12:1. From the compounding ratio of AC:RuOx
we use, we would have expected 6.7:1 if the ruthenium oxide had not been hydrated,
and a higher amount since it is hydrated (but we are uncertain of the number of
water of hydration). Since there is further carbon present from the elements of the
electrolyte, such as ethylene glycol, the measured result there is a deviation in the
expected direction.
3. Zooming on the peak shared between C1s and Ru3d shows an interesting e ect of
the discharge, which is also veri able by checking the Ru3p peaks (see Figure 5.12).
In the discharged battery, the peaks caused by ruthenium have almost disappeared.
Since no ruthenium can be detected from samples taken from anywhere else in the
battery, particularly not on the other side of the separator, one plausible conclusion is
that ruthenium is being covered/coated with a material during discharge. See ahead
for what can be done to verify this theory.
4. The analysis on oxidation states of ruthenium is not yet complete. Some Ru4+,
consistent with the presence of RuO2, is detectable in the new battery. However, the
discharged battery does not show any relevant peak due to possible coating-over.
5.6.2 Analysis on the Anode Side
On the anode side, which is constructed by using a zinc sheet in these batteries, a white
powder precipitate was available to brush o and use as the analytical sample. From our
previous work, we can theorize that this was either ZnO or Zn(OH)2. It is a relevant
observation that the precipitate is present not only in the discharged cell, but also in the
fresh cell, although it must be noted that the \fresh" cell, while never discharged, had been
made several days prior to being cut and opened for sample collection. This is consistent
with the observation that zinc ions have made their way over to the cathode side already
in the new cell, indicating zinc dissolution.
75
Preliminary examination of this result leads to several conclusions:
1. There is no ruthenium and very little carbon present on this side. This is consistent
with the conclusions that neither carbon nor ruthenium dioxide migrates inside the
cell during use. The small amount of carbon present is very likely remains from the
organic electrolyte molecules, including citric acid and ethylene glycol.
2. The only form of zinc that can be discerned is oxidized, non-metallic zinc. This can be
ZnO or Zn(OH)2. See Figure 5.14 below for a zoom-in on two characteristic Znpeaks.
The exact oxidation state, and, therefore, the probable compounding, is the subject
of further analysis, to be discussed ahead.
5.6.3 Further XPS Measurements and Analysis
For investigating the problem of ruthenium \disappearing" on the cathode side in the dis-
charged battery, we use the depth-pro ling capability of the XPS system. This capability
allows the user to sputter the sample with argon ions for pitting away the top few nm
of the sample, and looking at what is underneath. This way, it may be possible to both
identify the oxidation states of ruthenium in the discharged sample, as well as, by noticing
which materials decrease percentage-wise as the depth increases, to determine what may be
coating the ruthenium paste.
We perform further detailed analysis on samples on either side to determine the exact
oxidation states and more precise ratio information on the species.
In particular, with the anode side sample, deciding between ZnO and Zn(OH)2 (or, pos-
sibly, another Zncompound) requires the performance of detailed peak-shifting and Auger
parameter analyses, which are underway. These analyses will be carried out comparatively
between the new and discharged samples to determine how oxidation progresses during
discharge.
Finally, the discharged battery sample provided to Dr. Gaskell for analysis had actually
been put through a hard discharge, consisting of several days? worth of both terminals
being shorted. We intend to carry out similar analyses on both samples at several levels
(\depths") of discharge, as well as on samples that have recently undergone recharging.
76
Fig. 5.11: Full XPS spectrum for cathode-side samples from undischarged (top) and dis-
charged (bottom) batteries.
77
Fig. 5.12: Top: Zoom in on the C1s/Ru3d peak; left: new battery; right: discharged
battery. Bottom: Zoom in on the Ru3p split peak; left: new battery; right:
discharged battery. Of the four peaks visible in the bottom left graph, from
left to right (higher-to-lower binding energy) the peaks are ZnAuger, Ru1p,
ZnAuger, the other Ru1p. In the discharged battery graph, only the Znpeaks
remain clearly visible.
78
Fig. 5.13: Full XPS spectrum for anode-side (Zn) samples from undischarged (top) and
discharged (bottom) batteries.
Fig. 5.14: Zoom in on the Zn2p peaks from the new (left) and discharged (right) batteries.
79
Fig. 5.15: Piece of a separator under the microscope.
80
Chapter 6
MANUFACTURING AND COST ANALYSIS OF
Zn RuO2 BATTERIES
This section addresses the current state of the Zn RuOx battery manufacturing in terms
of thickness,  exibility, cost, cycleability, and shelf life. As a result of this examination,
the chapter suggests techniques for improving battery performance by controlling and opti-
mizing these parameters. In the current state of the performance the novel ruthenium(IV)
oxide cell hold the \world record" in terms of capacity of a  exible thin cell.
6.1 Current Stage of the Novel Ruthenium(IV) Oxide Cells
6.1.1 Thickness
The thickness of the batteries depends on the material used for electrodes, separators,
current collectors, and on the thickness of the packaging material, as well. Our intent was
to optimize these parameters, such that highest possible capacity of the cell is reached.
Currently, all the batteries manufactured on the lab scale are made with 0:65mm thick
zinc plate, given its low cost. These materials are used for manufacturing kitchen counter
tops, and the material has proven suitable for devising battery anodes, as it consists of
99:99% pure zinc. A graphite sheet of 0:13mm thickness is used as the cathode current
collector. These materials are utilized in most of the batteries, since they o er the best
performance. For a regular battery, the separator thickness is of 0:350mm, which enables
for the highest capacity, as described in section 5.3. Using the three main components above
plus the packaging, the nominal thickness of the battery becomes 1:5mm. One should note
81
here that these batteries can also be made with thinner zinc, as thin as 0:0762mm or even
0:0254mm. The batteries equipped with thin zinc foils delivers, however, charge capacity,
as a consequence of faster corrosion of the zinc. section 5.3 of this thesis discussed di erent
cathode current collector that can possibly be used instead of graphite sheet. As explained
in section 5.3, the tested materials provide less charge capacity; therefore, to the present
they have not been used as a replacement of graphite sheets.
6.1.2 Capacity
We have shown that when RuOx is used with appropriate electrolyte, it continues operating
as a cathode material even after it has been reduced from Ru(IV) to lower oxidation states.
Its extended job is most likely the result of catalyzing hydrogen ion (or water) reduction at
the RuOx/electrolyte interface [79].
We have demonstrated cell capacities as high as 33:35mAh=cm2 for Zn-RuOx batteries.
This capacity has been achieved by using thick zinc (0:65mm), graphite sheet, and  ber-
glass  lter paper with the pHset to the slightly acidic value of pH2:43. Battery made with
same material but thinner zinc (0:0762mm) delivered the lower capacity of 16mAh=cm2,
because of zinc corrosion. Table 6.1 below compares the capacity of the Zn RuOx cell with
several competitor batteries. As revealed by the listed values, the FlexEl battery exhibits the
highest charge capacity relative to other non-rechargeable  exible batteries[79, 80, 81, 82].
RuO2=Zn RuO2=Zn Enfucell Enfucell Power Paper Blue Spark
Nominal voltage [V] 1.4 1.4 1.5 3 1.5 1.5
Nominal capacity 16 33.35 2.1 0.4 1.2 1.3
(mAh=cm2)
Tab. 6.1: The Capacity Comparison of Non-Rechargeable Flexible Batteries.
It can be observed that the Zn  RuOx battery has one order of magnitude greater
speci c surface capacity (in mAh=cm2) as compared to other thin- lm, non-rechargeable
 exible batteries. The record for rechargeable,  exible, thin- lm batteries is clearly held by
the FlexEl galvanic cell.
82
6.1.3 Costs
Ruthenium and zinc are the most expensive components of the proposed battery. Therefore,
we explored possibilities for reducing these costs. We expect these costs to be signi cantly
lower for large scale production.
Ruthenium Costs
Hydrated Ru(IV) oxide is the main component of the proposed battery; this is a very
expensive compound. Due to its high cost, it has largely been ignored as a viable candidate
for battery cathode material. As explained in section 3.2, the paste used in the battery is
a mixture of RuOx and activated carbon (AC); by \diluting" Ru(IV) oxide with AC, we
reduce the cost of the ruthenium signi cantly. Also, utilizing nano-coating technologies,
we can reduce to a signi cant extent the amount of ruthenium incorporated in our cells.
This technique has been examined, but it represents an approach beyond the tasks of this
Thesis.
Zinc Costs
Zinc is the second most expensive component of the novel Zn  RuOx batteries. What
makes this metal expensive is the need for shaping it into a very thin zinc foil. The cost of
zinc increases noticeably as the metal sheets become the thinner and thinner. As explained
previously, in present we are using 0:65mm thick zinc plates, which enable low expense
laboratory scale battery manufacturing. Znplate is noticeably cheaper than thin zinc foil.
The materials cost $0.005 per square centimeter. Comparably, the cost of thin zinc foil is
almost 50 times more expensive, of approximately $0.23 per square centimeter. Therefore,
there is a need for identifying a thinning technology, which can reduce the thickness of the
0:75mm zinc plate, while keeping its cost low.
6.1.4 Cycle Life
Studies show that ruthenium is the best candidate for the non-rechargeable  exible batteries.
It can also be used in rechargeable batteries. We have demonstrated more than 800 cycles
83
with 1min discharges into a 10 load followed by 3min charging as shown in section 5.4.
Over the  rst 30s these discharges source approximately 100mA for a 4cm2 cell, when the
discharge occurs at the average voltage of 1V. Cycle life at deep discharges is currently
limited by the hydrogen evolution, and the discussion of the latter phenomenon does not
represent the scope of this thesis. High internal resistance of a cell and cell dry-out are two
important factors, which limit the cycleability of the cell. Formation of zinc oxide can also
a ect the cycleability of a cell. Details about cycling of Zn  RuOx batteries have been
provided in subsection 5.4.1.
6.1.5 Safety and Sustainability
Most of the batteries currently commercialized are lithium-ion based. Lithium like all alkali
metals, is highly reactive and  ammable. Therefore, these batteries are not environmentally
friendly, and can be extremely harmful. The RuOx cell chemistry is inherently non-toxic,
and environmentally benign. Zinc is nontoxic, being used in anti- re medicine. Acidity of
the electrolytes used to make batteries is comparable to that of a salad dressing, so it will
not cause any harm to the human body.
84
Chapter 7
ACCOMPLISHMENTS AND FUTURE
CONSIDERATIONS
7.1 Summary of Main Accomplishments
The main focus of this Thesis was to design and manufacture ultrathin galvanic cells that
can comply with a variety of form factors and electronic system packages. Such cells are in
great demand, as they show promise for  exible electronic systems. Described is a  exible,
high energy density galvanic cell, based on hydrated ruthenium (IV) oxide-zinc electrodes,
which utilizes RuO2 nH2O nanoparticles, in amounts that are not cost-prohibitive. So far,
Ru(IV) oxide was utilized almost exclusively in supercapacitors, their possible application as
a cathode material in batteries being almost completely neglected. Therefore, our approach
of building battery systems the topic of this Thesis represents a breakthrough scienti c
novelty. The obtained battery is suitable for use in a number of environments, is non-toxic,
environmentally friendly, and, being based on aqueous electrolytes, it is safe in operation.
It can be optimized for volume manufacture at low cost. As it functions at much lower cell
voltage than Li-ion batteries, it can be recharged remotely, at conveniently low voltage, by
harvesting, for example, radio-frequency (RF) energy.
The  rst chapter of the Thesis is a literature review, which summarizes the data from
public domain on thin- lm Li-ion batteries, the use of hydrated ruthenium (IV) oxide
in supercapacitors, describes the relatively few commercially available thin- lm battery
systems, and provides insight in hybrid power systems, consisting of a galvanic cell and a
supercapacitor.
85
Next, battery operation is explained, with sections dedicated to thermodynamic con-
siderations and the kinetics of the galvanic cell. Attention is paid to the e ect exerted by
various electrolytes and to the cell separators.
chapter 3 takes a close look into the fabrication and characterization of the novel Zn 
RuO2 cells. Described are problems and technical solutions for the anode and the cathode
half-cell, the proper choice of electrolyte, of the cell separator, and of current collectors.
Speci c investigation techniques and the needed instrumentation are addressed in chap-
ter 4, where the electrochemical techniques are grouped into the ones performed in 3-
electrode geometry and others, which can be carried out in 2-electrode geometry. Discussed
are cyclic voltammetry, open circuit voltage vs. time measurements, charge/discharge stud-
ies (including charging at constant current or at constant voltage, and discharge at constant
external load or at constant current, i.e., by chronopotentiometry). Additional tests dis-
closed in this chapter include cycling experiments, shelf life/self-discharge examination, and
internal resistance measurements.
The most signi cant experimental results are described in a systematic manner in chap-
ter 5. In order to learn more about the mechanism of the electroactive processes that drive
the galvanic cells, the investigation went beyond the Electrochemical testing, the cathode
and anode materials being analyzed by XPS (subsection 5.6.3).
Current stage of the novel zinc-ruthenium (IV) oxide cell is evaluated in chapter 6,
where the most important parameters are discussed in separate sections, namely: thickness
of the device, capacity of the galvanic cell, and the fabrication costs. In present this cell
demonstrates a speci c capacity of over 35mAh=cm2, which is, so far, the largest reported
for thin- lm cells. Also, its cycle life of up to 200 charge/discharge cycles is very promising
for use as a secondary galvanic cell.
My last section takes a look on the prospective of the described galvanic cell. It describes
brie y research strategies and research plans toward improving the existing system, mainly
via increasing its charge capacity and its cycle life.
86
7.2 Future Considerations
Optimization of each battery component is needed. Below are the suggested technologies
that can help the performance of the battery.
7.2.1 Anode Optimization
To date, the bulk of our research has focused on Znas an anode material, but other met-
als may be considered as candidates. Although, primary experiments have demonstrated
favorable performance of zinc anodes, I plan performing a broader exploration of anode
materials, to con rm that zinc is the most suitable choice for pursuing further develop-
ments. Assuming that we will able to con rm the choice of zinc as an anode material, we
are going to explore techniques for low-cost thin zinc manufacturing. Additionally, we will
want to explore a variety of compounds added to zinc for lowering its internal resistance
and improving its cycling ability through reduction in shape change and corrosion.
7.2.2 Cathode Optimization
A variety of cathode materials have been explored for their ability to collect electrons, and
to provide low resistance current collectors to the electrolytic medium. In addition, there
is a need for new approaches for creating ultrathin cathode current collectors. Ultrathin
carbon structures have been already used in other applications [83, 84]. One common
method is to use a  uid  ltration and drying procedure, similarly to the one utilized in
paper manufacturing. The result is commonly referred to as \buckeye". Several companies
o er this as a commercial product, made of carbon nanotube, where the thickness of the
paper determines its cost [84]. section 5.2 of this Thesis shows the result obtained with
buckeye paper having the thickness of 0:1 m. Our goal is, however, to be able to reduce
the thickness even further. Experiments need to be done for investigating the performance
of the battery, when thinner buckeye paper is used. We plan on exploring commercially
available thin- lm carbon conductors for our initial experiments and, if successful, we will
create and incorporate in our cells eve thinner conductors.
Other considered techniques are various coating technologies. This approach not only
87
helps for thinning the battery structure, but also enables for reducing the amount of hy-
drated ruthenium oxide, which needs to be used per square centimeter of the battery. The
goal is to coat ruthenium and activated carbon compounds on thin current collectors. Fur-
ther investigation needs to be done for identifying an appropriate technology, which can
coat powders on a conductive substrate, such that the resulting structure should have a low
Ohmic resistance.
7.2.3 Electrolyte Optimization
There is a need for exploring the full range of pHvalues, which may allow for maximizing
the cell capacity and cycleability, as well. We are going to explore the possible recipes for
electrolytes that provide the highest capacity, while increasing the shelf life.
The next very important step in optimizing the electrolyte is switching to gel-type elec-
trolytes. Use of such electrolytes may o er important advantages for ease-of-manufacturing
and battery lifetimes, as well as preventing the increase of internal resistance, as the un-
wanted result of hydrogen evolution. Also, cycle life may be improved, by reducing dendrite
formation. Lastly, utilizing gel type electrolytes that can be cast directly onto the surface of
one of the battery electrodes eliminates the need for a cell separator. This renders the bat-
tery design simpler and more e cient. Although, we were able to prepare several alkaline
gel electrolytes, the results obtained with these batteries were not as good as with liquid
electrolytes. Therefore, we are going to further investigate the technique for preparing gel
electrolytes, such that battery performance is maintained at its highest level.
88
BIBLIOGRAPHY
[1] Selvapathy, P., Madhavan, S.D., \Risk Assessment of used Household Batteries in the
Municipal Solid Waste - A Case Study," 2003. Workshop on Sustainable Land ll Man-
agement,December, 3-5, 2003, Chennai, India, pp. 219-224.
[2] Scrosati, B. \Battery technology - Challenge of Portable Power," 1995. Nature, 373
(6515), pp. 557-558.
[3] \Basic Research Needs for Electrical Energy Storage," 2007. Report of the Basic Energy
Sciences Workshop for Electrical Energy Storage, April 2-4, 2007. O ce of Basic Energy
Sciences, Department of Energy, July 2007.
[4] \World Batteries to 2012 - Market Research, Market Share, Market Size, Sales, Demand
Forecast, Market Leaders, Company Pro les, Industry Trends," Freedonia Industry Study
#: 2375, http://www.freedoniagroup.com/> (Sept. 21, 2010).
[5] Lowy, D.A., \An Electrochemical Journey at the Nano Scale," 2009. Ed. Casa Stiintei,
Cluj, pp. 4-5.
[6] Bates, J.B., Dudney, N.J., Lubben, D.C., et al. \Thin-Film Rechargeable Lithium Bat-
teries," 1995. J. Power Sources, 54 (1), pp. 58-62.
[7] West, W.C., Whitacre, J.F., White, V., Ratnakumar, B.V. \Fabrication and Testing of
All Solid-State Microscale Lithium Batteries for Microspacecraft Applications," 2002. J.
Micromech. Microeng., 12 (1), pp. 58-62.
[8] Buchmann, I. \Batteries in Portable World. Handbook on Rechargeable Batteries for
Non-Engineers," 2001. Cadex Electronics Inc.
[9] \Smart Nanobattery Makes History," 2008. Am. Ceramic Soc. Bull., 87 (5), pp. 34-34.
89
[10] \Smart Nanobattery Designed to Be >Green.<", 2007. Am. Ceramic Soc. Bull., 86 (3),
pp. A12-A12.
[11] Barak, M. \Primary Batteries - Introduction," 1981. In: Bockris, J.O?M., Conway,
B.E., Yeager, E., White, R.E. (Editors), \Comprehensive Treatise of Electrochemistry,"
Plenum Press, New York, London, p. 193.
[12] Sacrosati, B. \Paper Powers Battery Breakthrough," 2007. Nature Nanotech., 2 (10),
pp. 598-599.
[13] Souquet, J.L., Duclot, M., \Thin- lm lithium batteries," 2002. Solid State Ionics, 148
(3-4), pp. 375-379.
[14] Dudney, N. J., Neudecker B.J. \Solid state thin- lm lithium battery systems," 1999.
Current Opinion in Solid State & Materials Science, 4, pp. 479-482.
[15] Pushparaj, V.L., Shaijumon, M.M., Kumar, A., Murugesan, S., Ci, L., Vajtai, R.,
Linhardt, R.J., Nalamasu, O., Ajayan, P.M., \Flexible energy storage devices based on
nanocomposite paper," 2007. Proc. National Acad. Science USA, 104 (34), pp. 13574-
13577.
[16] Bates, J.B., Dudney, N.J., Neudecker, B., Ueda, A., Evans, C.D. 2000, \Thin- lm
lithium and lithium-ion batteries," Solid State Ionics, 135, pp. 33-36.
[17] Dudney, N.J. \Solid-state thin- lm rechargeable batteries," 2005. Materials Science &
Engineering B, 116, pp. 245-249.
[18] Dunn, B., Long, J.W., Rolison, D.R., \Rethinking Multifunction in Three Dimensions
for Miniaturizing Electrical Energy Storage," 2008. The Electrochemical Society Inter-
face, Fall issue, pp.49-53.
[19] Wang, J.-Z., Chou, S.-L., Liu, H., Wang, G.X., Zhong, C., Chew, S.Y., Liu, H.K.,
\Highly  exible and bendable free-standing thin- lm polymer for battery application,"
2009. Materials. Letters, 63 (27), 2352-2354.
90
[20] Neudecker, B.J., Dudney, N.J., Bates, J.B., \>>Lithium-Free<< Thin-Film Battery with
In Situ Plated Li Anode," J. Electrochem. Soc., 2000. 147 (2), pp. 517-523.
[21] Neudecker, B.J., Zuhr, R.A., Bates, J.B., \Lithium silicon tin oxynitride (LiySiTON):
high-performance anode in thin- lm lithium-ion batteries for microelectronics," 1999. J.
Power Sources, 81-82, pp. 27-.
[22] Wang, B., Bates, J.B., Hart, F.X., Sales, B.C., Zuhr, R.A., Robertson, J.D., \Charac-
terization of Thin-Film Rechargeable Lithium Batteries with Lithium Cobalt Oxide Cath-
odes," 1996. J. Electrochem. Soc., 143 (10), pp. 3203-3213.
[23] Lee, S.H., Liu, P., Tracy, C. E., Benson, D.K., \All-Solid-State Rocking Chair Lithium
Battery on a Flexible Al Substrate," 1999. Electrochem. Solid-State Lett., 2 (9), pp.
425-427.
[24] Bates, J.B., Dudney, N.J., Lubben, D.C., Gruzalski, G.R., Kwak, B.S., Yu, X., Zuhr,
R.A. \Thin- lm rechargeable lithium batteries," 1995. J. Power Sources, 54, pp. 58-62.
[25] Young-Shin Park, Se-Hee Lee, Byung-Il Lee, and Seung-Ki Joo, 1999. \All-Solid-State
Lithium Thin-Film Rechargeable Battery with Lithium Manganese Oxide," Electrochem.
Solid-State Lett., 2 (2), pp. 58-59.
[26] Southee, D., Hay, G.I., Evans, P.S.A., Harrison, D.J., \Lithographically Printed Voltaic
Cells - A Feasibility Study," 2007. Circuit World, 33 (1), pp. 31-35.
[27] Hiralal, P., Imaizumi, S., Unalan, H.E., Matsumoto, H., Minagawa, M., Rouvala, M.,
Tanioka, A., Amaratunga, G.A. \Nanomaterial-Enhanced All-Solid Flexible Zinc-Carbon
Batteries," 2010. ACS NANO, 4 (5), pp. 2730-2734.
[28] Xu, J.J., Ye, H., Huang, J. \Novel Zinc Ion Conducting Polymer Gel Electrolytes Based
on Ionic Liquids," 2005. Electrochem. Comm. 7 (12), pp. 1309-1317.
[29] Narayanan, N.S.V., Ashokraj, B.V., Sampath, S. \Ambient Temperature, Zinc Ion-
Conducting, Binary Molten Electrolyte Based on Acetamide and Zinc Perchlorate: Ap-
plication in Rechargeable Zinc Batteries," 2010. J. Colloid & Interface Science, 342 (2),
505-512.
91
[30] Narayanan, N.S.V., Ashokraj, B.V., Sampath, S. \Physicochemical, Electro chemical,
and Spectroscopic Characterization of Zinc-Based Room-Temperature Molten Electrolytes
and Their Application in Rechargeable Batteries," 2009. J. Electrochem. Soc., 156 (11),
A863-A872.
[31] Simon, S., Lifton, V., Allen, F. \Extending the Life of Wireless Networks Using the
Smart NanoBattery. White Paper," 2008. mPhase Technologies, Inc. / AlwaysReady,
Inc., Little Falls, New Jersey, September 2008.
[32] Krupenkin, T.N., Taylor, J.A., Kolodner, P., Hodes, M. \Electrically Tunable Super-
hydrophobic Nanostructured Surfaces," 2005. Bell Labs Tech. J. 10 (3), pp. 161-170.
[33] Krupenkin, T.N., Taylor, J.A., Wang, E.N., Kolodner. P., Hodes. M., Salamon,
T.R. \Reversible Wetting-Dewetting Transitions on Electrically Tunable Superhydropho-
bic Nanostructured Surfaces," 2007. Langmuir, 23 (18), pp. 9128-9133.
[34] Krupenkin, T.N., Taylor, J.A., Schneider, T.M., Yang, S. \From Rolling Ball to Com-
plete Wetting: The Dynamic Tuning of Liquids on Nanostructured Surfaces," 2004. Lang-
muir, 20 (10), pp. 3824-3827.
[35] Ahuja, A., Taylor, J.A., Lifton, V., Sidorenko, A.A., Salamon, T.R., Lobaton, E. J.,
Kolodner, P., Krupenkin, T.N. \Nanonails: A Simple Geometrical Approach to Electri-
cally Tunable Superlyophobic Surfaces," 2008. Langmuir, 24 (1), pp. 9-14.
[36] Krupenkin, T.N., Mandich, M.L., Taylor, J.A. 2005. EP 1591415A2, 2005. US
2005269743A1, 2005. JP 05349558A, 2005. DE 6005000397E, 2005. DE 6005000397T2,
2005. US 20057323033B2.
[37] Brazier, A., Dupont, L., Dantras-La ont, L., Kuwata, N., Kawamura, J., Tarascon,
J.-M. \First Cross-Section Observation of an All Solid State Lithium-Ion "Nanobattery"
by Transmission Electron Microscopy," 2008. Chem. Mater., 20 (6), pp. 2352-2359
[38] Lifton, V.A., Taylor, J.A., Vyas, B., Kolodner, P., Cirelli, R., Basavanhally, N., Pa-
pazian, A., Frahm, R., Simon, S., Krupenkin, T. \Superhydrophobic Membranes with
92
Electrically Controllable Permeabilityand Their Application to "Smart" Microbatteries,"
2008. Appl. Phys. Lett., 93, Paper #043112.
[39] Hilder, M., Winther-Jensen, B., Clark, N. B. \Paper-Based, Printed Zinc-Air Battery,"
2009. J. Power Sources, 194 (2, Sp. Iss.), pp. 1135-1141.1.40
[40] Dennler G., Bereznev, S., Fichou, D., Holl, K., Ilic, D., Koeppe, R., Krebs, M.,
Labouret, A., Lungenschmied, C., Marchenko, A., Meissner, D., Mellikov, E., Mot, J.,
Meyer, A., Meyer, T., Neugebauer, H., pik A., Sariciftci, N.S., Taillemite, S., Whrle T.
\A self-rechargeable and  exible polymer solar battery," 2007. Solar Energy, 81 (8), pp.
947-957.
[41] Park, B.-O., Lokhande, C.D., Park, H.-S., Jung, K.-D., Joo, O.-S. \Performance of su-
percapacitor with electrodeposited ruthenium oxide  lm electrodes-e ect of  lm thickness,"
2004. J. Power Sources, 134, pp. 148-152.
[42] Yoon, Y.S., Cho, W.I., Lim, J.H., Choi, D.J. \Solid-state thin- lm supercapacitor with
ruthenium oxide and solid electrolyte thin- lms," 2001. J. Power Sources, 101 (1), pp.
126-129.
[43] Choi, S.H., Kim. J., Yoon, Y.S. \Fabrication and characterization of a SnO2-RuO2
composite anode for a hybrid thin- lm battery," 2005. Metals & Materials International,
11 (1), 77-83.
[44] Kim, I.H., Kim, K.B. \Ruthenium oxide thin- lm electrodes for supercapacitors," 2001.
Electrochem. Solid State Lett., 4 (5), A62-A64.
[45] "Battery" in wikipedia: http://en.wikipedia.org/wiki/Battery%28electricity%29#Primarybatteries
[46] Dell, R.M. Rand D.A.J. "Understanding Batteries," Royal Society of Chemistry, 2001,
PP. 9-25
[47] Barak, M. "Primary batteries - Introduction," in: Bockris, J. Conway, B. Yeager, E.
White, R. (Eds.) " Comprehensive Treaties of Electrochemistry,"
93
[48] Zumdahl S.S. Zumdahl S.A. "Chemistry," 2007. 7th Edition, Houghton Mi in Com-
pany, Boston Newyork, pp. 891-821.
[49] "Standard Electrode Potential" in wikipedia: http://en.wikipedia.org/wiki/Tableofstandardelectrodepotentials
[50] Bard, A.J., Faulkner, L.R. "Electrochemical Methods. Fundamentals and Applica-
tions,". 2001. 2nd edition, John Wiley & Sons, Inc., New York, Chichester, Weinheim,
Brisbane, Singapore, Toronto. pp. 92. 100-106.
[51] Kordesch, K. (1981) Electrochemical energy storage. In Comprehensive Treatise of
Electrochemistry. Vol. 3: Electrochemical Energy Conversion and Storage (eds. Bockris,
J.O?M., Conway, B.E., Yeager, E. and White, R.E.) pp.202.139-141., Plenum Press, New
York, London.
[52] Kissinger, P.T., Preddy, C.R., Shoup, R.E., Heineman, W.R. "Fundamental Concepts
of Analytical Chemistry. In Laboratory Techniques in Electroanalytical Chemistry," 1996
(Eds. Kissinger, P.T., Heineman, W.R.) 2nd edition, Marcel Dekker, New York, Basel,
pp.37-38.
[53] Rocchini, G. "Corrosion rate monitoring by the Linear-polarization method," 1993.
Corrosion Science, 34 (12), pp. 2031-2044.
[54] Rocchini, G. " The in uence of the potential sweep rate on the computation of the
polarization resistance," 1996. Corrosion Science, 38 (12), pp. 2095-2109.
[55] Rocchini, G. "Experimental veri cation of the validity of the linear polarization
method," 1997. Corrosion Science, 39 (5), pp. 877-891.
[56] Zhu, J.; Cheng, F.Y.; Tao, Z.; Chen, J. (2008) Electrocatalytic methanol oxidation of
Pt0.5Ru0.5-xSnx/C (x=0-0.5). J. Phys. Chem. C 112(16), 6337-6345.
[57] Eguiluz, K.I.B., Salazar-Banda, G.R., Miwa, D., Machado, S.A.S., Avaca, L.A. "E ect
of the catalyst composition in the Pt-x(Ru-Ir)(1-x)/C system on the electro-oxidation of
methanol in acid media," 2008. J. Power Sources 179 (1), pp. 42-49.
94
[58] Innocente, A.F., Angelo, A.C.D. "Hydrogen oxidation on ordered intermetallic elec-
trodes covered with CO," 2008. J. Power Sources 175(2), pp. 779-783.
[59] Sieben, J.M., Duarte, M.M.E., Mayer, C.E. "Supported Pt and Pt-Ru catalysts pre-
pared by potentiostatic electrodeposition for methanol electrooxidation," 2008. J. Appl.
Electrochem. 38(4), pp. 483-490.
[60] Salazar-Banda, G.R., Su redini H.B., Calegaro, M.L., Tanimoto, S.T., Avaca, L.A.
"Sol-gel-modi ed boron-doped diamond surfaces for methanol and ethanol electro-
oxidation in acid medium," 2006. J. Power Sources 162(1), pp. 9-20.
[61] Kim, Y.J., Hong, W.H., Woo, S.I., Lee, H.K. "Analysis of the polarization of a direct
methanol fuel cell using a pseudo-reversible hydrogen reference electrode," 2006. J. Power
Sources 159(1), pp. 491-500.
[62] Thompson, E.L., Jorne, J., Gasteiger, H.A. "Oxygen reduction reaction kinetics in
subfreezing PEM fuel cells," 2007. J. Electrochem. Electrochem. Soc. 154 (8), pp. B783-
B792.
[63] Chirkov, Y.G., Rostokin V.I. "Hydrophobized oxygen cathode of a fuel cell with a liquid
electrolyte: Calculating overall currents and thicknesses," 2007. Russian J. Electrochem.
43 (2), pp. 146-156.
[64] Liu, D., Case, S. "Durability study of proton exchange membrane fuel cells under
dynamic testing conditions with cyclic current pro le," 2006. J. Power Sources, 179 (1),
pp. 42-49.
[65] Bockris, J. O?M., Reddy, A.K.N. \Modern Electrochemistry. Ionics," 1998. Second
edition, Plenum Press, New York and London, Vol. 1, pp. 13,454-456.
[66] Linden, D., Reddy, T.B. (Eds.) \Handbook of Batteries," 2002. Third edition, McGraw-
Hill, New York, Chicago, , Toronto, pp. 8.15.
[67] FlexEl LLC. "Novel Ultra-High Capacity PHEV battery," Internal Technical Docu-
mentation, College Park, MD. 2010
95
[68] FlexEl LLC. "Novel Ultra-High Capacity PHEV battery," Internal Technical Docu-
mentation, College Park, MD. 2010
[69] "Battery" in wikipedia: http://en.wikipedia.org/wiki/Ruthenium
[70] Barak, M. "Primary batteries - Introduction," in: Bockris, J. Conway, B. Yeager, E.
White, R. (Eds.) " Comprehensive Treaties of Electrochemistry," pp.202-205.
[71] Linden, D., Reddy, T.B. (Eds.) \Handbook of Batteries," 2002. Third edition, McGraw-
Hill, New York, Chicago, , Toronto, pp.8-23.
[72] S.K. Toh, D.G. McCulloch, J. Du Plessis, P.J.K. Paterson, A.E. Hughes, D. Jamieson,
B. Rout, J.M. Long and A. Stonham, Surface Rev. Lett. 10 (2003), 365.
[73] Bard, A.J., Faulkner, L.R. "Electrochemical Methods. Fundamentals and Applica-
tions,". 2001. 2nd edition, John Wiley & Sons, Inc., New York, Chichester, Weinheim,
Brisbane, Singapore, Toronto. pp. 92. 100-106.
[74] "Potentiostats" in BioLogic: http://www.bio-logic.info/potentiostat/brochures/20071123%20-
%20MRes%20-%20VMP3.pdf
[75] BioLogic Sience Instruments. " EC-Lab Software User?s Manual," Version 9.9, France,
2010
[76] Dell, R.M. Rand D.A.J. "Understanding Batteries," Royal Society of Chemistry, 2001,
P. 38
[77] Zhang, X.G. "Corrosion and electrochemistry of zinc," 1996. Plenum Press, New York,
pp. 385-389.
[78] Internal Resistance(IR) in Energizer: http://data.energizer.com/PDFs/BatteryIR.pdf
[79] FlexEl LLC. "Novel Ultra-High Capacity PHEV battery," Internal Technical Docu-
mentation, College Park, MD. 2010
[80] Power Paper Batteries in: http://www.powerpaper.com/
[81] Blue Spark battery products in: http://www.bluesparktechnologies.com/2.1.0.cfm
96
[82] Enfucell batteries in: http://www.enfucell.com/
[83] Long, J.W. Rolison, D.R." Carbon nanoarchitectures with ultrathin, conformal poly-
mer coating for Electrochemical capacitors," US patent 7672114
[84] "Buckeye Composites," 2010 in: http://www.buckeyecomposites.com/
97