Beruflich Dokumente
Kultur Dokumente
by
Devendra Ray
First and foremost, I would like to thank my Lord Shiva for giving me the
strength and passion to arrive at this point and for putting the right people in my life at
the right time to help guide me. I dedicate this thesis to my parents, Ramkailash Ray,
father, and late mother, Ramshakhi Devi Ray, for their love, encouragement and the
many sacrifices they made that have enabled me to arrive at this point of my life. I would
also like to dedicate this thesis to my wife, Sima Ray. Finally, I would like to thank all
my family and friends who have encouraged and supported me to reach this milestone in
my life.
iii
BIOGRAPHICAL SKETCH
Devendra Ray was born in Nepal and immigrated to the United States of America
Technical State University. He has performed research on solid oxide fuel cells (SOFC)
electrolyte materials.
iv
ACKNOWLEDGEMENTS
I would like to begin by thanking my major advisor, Dr. Jag Sankar, and senior
research scientist/co-advisor, Dr. Sergey Yarmolenko, for their consistent guidance and
support over the past two years. I significantly benefited from their knowledge and
valuable experience in the field of material science and its characterization. I would also
like to thank all my friends and the staff of the Center for Advanced Materials and Smart
Structures (CAMSS). I would also like to express my thanks to Dr. Z. Xu and Dr. M. T.
Saad for their suggestions. I am very grateful to my family for their support, sacrifice,
v
TABLE OF CONTENTS
ABSTRACT..................................................................................................................... xiii
vi
3.2.1 Powder Compaction.....................................................................................18
3.2.2 Sintering.......................................................................................................19
3.4.2.1 Nanoindentation..................................................................................24
3.4.2.2 Microindentation.................................................................................31
vii
4.4 Coefficients of Thermal Expansion (CTE) ...........................................................71
4.5.2 Nanoindentation...........................................................................................87
REFERENCES ..................................................................................................................91
viii
LIST OF FIGURES
FIGURE PAGE
3.4 Typical images collected at 500x on optical microscope for hardness study ............33
3.5 (a) Schematic figure having crack and indent length; (b) Image collected at 1000x
on optical microscope (Indentation load is 200g)..................................................35
3.8 (a) Typical AFM images for grain size study; (b) Grain size and cavitations
measurements............................................................................................................42
3.10 Constant, linear and quadratic fitting with temperature vs. log (LT) (P1600-
powder) .....................................................................................................................47
ix
4.2 XRD pattern at different temperatures.......................................................................54
4.7 Grain structure of samples produced from DKKK powder sintered at temperatures
1400ºC (left), 1500ºC (middle) and 1600ºC (right) ...................................................60
4.8 SEM images of polished and low angle ion-milled surface of sample P1600 (a) and
thermally etched surfaces of samples P1600 (b) and J1600(c).................................61
4.11 SEM images of DKKK samples from left to right sinter correspondingly
1200-1600°C.............................................................................................................64
4.12 SEM images of Praxair samples from left to right sinter correspondingly
1500-1600°C.............................................................................................................64
4.14 SEM images collected after 1hr ion milling at 30 deg (left - J1600 right - P1600,
right - P1600) ............................................................................................................66
4.15 SEM images collected after ion milling (left-J1600 milled at 9-deg for 18hrs,
right - P1600 milled at 5 deg for 20hrs)....................................................................67
4.16 SEM images of P1600 collected after ion milled at 5 deg for 20hrs (right -
high porous area, left - less porous) ..........................................................................67
4.18 Changes in XRD pattern of powdered sample J1600 heated from 300 to 500°C
with step 10°C...........................................................................................................70
x
4.19 Phase transition in DKKK ceramic sintered at 1600°C............................................70
4.21 Constant, linear and quadratic fitting with temperature vs. log (LT) (Praxair
powder) .....................................................................................................................74
4.23 Constant, linear and quadratic fitting with temperature vs. log (LT)
(DKKK powder) .......................................................................................................75
4.25 Constant, linear and quadratic fitting with temperature vs. log (LT) (J1500
Powder) .....................................................................................................................76
4.27 Constant, linear and quadratic fitting with temperature vs. log (LT) (J1500
ceramic).....................................................................................................................77
4.28 Hardness vs. sintering temperature plot for etched and non-etched sample.............78
4.30 (a) Effect of sintering temperature on microhardness and fracture toughness at the
load 1000g; (b) Micrograph of typical Vickers indents and cracks at
different loads (thermally etched J1600) ..................................................................81
4.31 ISE analysis using power law (a) and Nix-Gao (b) models......................................86
xi
LIST OF TABLES
TABLE PAGE
4.5 Grain size and in-grain cavities at different sintering temperatures ..........................65
4.6 Coefficients of thermal expansion for Praxair and DKKK samples in ceramic
and powder forms ......................................................................................................73
4.7 Hardness (H) and fracture toughness (K1C) values at different sintering
temperatures and loads for non-etched samples .......................................................82
4.8 Hardness (H) and fracture toughness (K1C) values at different sintering
temperatures and loads for thermally etched samples ...............................................83
xii
ABSTRACT
coefficients of thermal expansion are very important parameters for stability and
reliability of solid oxide fuel cell (SOFC) operation. Ceria doped Sc2O3-CeO2- ZrO2
ceramics are also very promising materials as electrolyte in solid oxide fuel cells (SOFC)
In this experimental work, sintering behavior of two commercial powders with the
surface technologies, USA and DKKK, Japan was studied. Ceramics were made by
principle. Porosity level in ceramics was estimated using actual and theoretical densities
calculated using lattice parameter a=5.09 Ǻ of FCC unit cell obtained from XRD data and
actual composition of powder. Hardness of most dense ceramics reaches 15GPa and
fracture toughness is in the range of 1.8-2.5MPa×m 0.5. These results are in good
effect (ISE). Indentation size effect parameters were analyzed using Power law model.
xiii
CTEs were studied. These results can be further used for the optimal design of SOFC
layered structures as well as for determination of their reliability and durability under
operational conditions.
xiv
CHAPTER 1
INTRODUCTION
It is well known that production and distribution of energy affects all sectors of
the global economy. The increasing industrialization of the world requires sustainable,
production will impact the quality of human life on Earth. Dependency on fossil fuel will
not help to maintain the global economy in the long term due to the growing consumption
and limitation of fossil fuel. For this reason, the application of fuel cell technologies may
alternative whereby electrical energy can be made available with smaller losses.
hydrogen and an oxidant such as oxygen. If the fuel is clean, the effluents are in principle
only water, heat and CO2. Fuel cell plants can be modular in design, and the energy
production can be adjusted to meet the actual demand, which is a convenient feature for a
The efficiency of a fuel cell is dictated by factors regarding the kinetics and
which is to achieve optimum efficiency. Only a few materials have survived as practical
solid oxide fuel cell (SOFC) electrolytes because many electrical, chemical, and
1
improved through the introduction of superior electrolyte materials and enhanced
The objective of this research was to further the development of better electrolyte
material. This includes the study of mechanical and microstructure properties of two
powders received from Praxair surface technology, USA, and DKKK, Japan. Knowledge
(K1c), thermal properties and densities are very important for the design of electrolyte in
SOFC. It is equally important to know the microstructure, such as the grain size of two
the material.
2
CHAPTER 2
LITERATURE REVIEW
2.1 Background
This chapter provides a brief review of solid oxide fuel cells (SOFC), what they
are and how they work. It also details their history, the types of fuel cells, working
principles and materials employed for SOFC production. This research work emphases
the scandia stabilized zirconia (ScSZ) electrolyte material which was selected for the
A fuel cell is a “factory” that takes fuel as input and produces electricity as output.
chemical energy from a reaction between a fuel and oxidant into electrical energy as long
as fuel is supplied. This is the key difference between a fuel cell and a battery. While
both rely on electrochemistry, a fuel cell is not consumed when it produces electricity. It
is really a factory, a shell, which transforms the chemical energy stored in a fuel into
electrical energy. The basic elements of a typical fuel cell, as depicted in Figure 2.1,
consists of electrolyte in intimate contact with a porous anode (negative electrode) and
porous cathode (positive electrode). The fuel and oxidant gases flow along the surface of
the anode and cathode, respectively, and then react electrochemically in the three-phase-
boundary region established at the gas-electrolyte electrode interface. A fuel cell can
3
theoretically produce electrical energy for as long as fuel and oxidant are fed to the
porous electrodes, but the degradation or malfunction of some of its components limits
Anode e-
Fuel
Direct current
Electrolyte External load Exhaust gases
and heat
Cathode e-
Oxidant
The history of the fuel cell dates back to 1839 when Sir William Grove [3] first
described its principle and demonstrated a fuel cell at room temperature using a liquid
electrolyte. In 1899, Nernst discovered the solid oxide electrolyte when using stabilized
zirconia in making filaments for electric glowers. In the middle of the 20th century the
development accelerated. Several types of fuel cells were developed in the race for
conquering space. In the eighties, focus on pollution and the demand for higher efficiency
in the exploitation of fossil resources initiated a new wave of fuel cell development. At
present, several types of fuel cells are approaching the consumer market within a limited
4
number of years. The primary challenges are cost and durability, to be solved by
Currently, there are five major types of fuel cells, differentiated from one another
While all five fuel types are based upon the same underlying electrochemical
and often differ in their fuel tolerance and performance characteristics, some of which are
listed in Table 2.1[1, 2, 3, and 52]. Among these, operation temperature, electrical
efficiency and demands for fuel composition are the defining prospects of each type.
applications because they provide a continuous electrical energy supply from fuel at high
levels of efficiency and power density. They also offer the the advantage of minimal
maintenance because there are no moving parts in the power generating stacks of the fuel
cell system.
5
Table 2.1. The description of major fuel cell types
6
Applications2 >Electric utility >Electric utility >Military >Electric utility >Electric utility
>Transportation >Space >Transportation >Portable power
>Military >Transportation
Advantages2 >High efficiency >High efficiency >Cathode reaction >Can use impure >Solid electrolyte
>Fuel flexibility >Fuel flexibility faster in alkaline H2 as fuel reduces corrosion
>Can use a variety >Can use a electrolyte that and management
of catalysts variety of catalysts leads to high problems
>Reduces performance >Low temperature
corrosion problems
Disadvantages2 >High temperature >High temperature >Requires removal >Requires >Low operating
accelerates enhance corrosion of CO2 from fuel platinum catalyst temperature
breakdown of cell and breakdown of and air streams, >Low current requires expensive
components cell components which is expensive and power catalysts
Electrical 60% 60% 40% 40% 40%
Efficiency3
The low temperature fuel cells (AFC, PAFC) also have a potential for propulsion
of cars, where a short heating time is needed and the efficiency has to be compared with
that of a combustion engine (~20%) given by the limitations of the Carnot cycle. In
miniature applications such as cellular phones, laptops and digital cameras, methanol is
reformed to hydrogen in the fuel cell itself (reformation reaction). For applications that
require higher output, such as cars, methanol can be reformed in a device that is located
outside the fuel cell. This reformation can be accomplished through steam reforming or
partial oxidation [2]. Recently invented micro fuel cells are smaller, lighter, cleaner,
There are different types of electrolytes that are commercially available; however,
more research and development is needed to characterize and explore new and promising
electrolytes for an intermediate temperature (IT) SOFC. The basic components of SOFC
are:
1. Cathode
2. Electrolyte
3. Anode
4. Interconnect
7
2.5.1 Electrolyte
The most effective electrolytes employed in fuel cells have ionic conductivities of
around 0.10 Ώ-1cm-1. The search for better electrolytes has led to the development of
three major candidate material classes for fuel cells: aqueous, polymer, and ceramic
electrolytes. Regardless of the class, however, any fuel cell electrolyte must meet the
• Ease of manufacturability.
Except for the high conductivity requirement, the electrolyte stability requirement
is often the hardest to fulfill. It is difficult to find an electrolyte that is stable in both the
highly reducing environment of the anode and the highly oxidizing environment of the
The electrolyte is the central part of an SOFC. Within the electrolyte the oxygen
ions (O2-), which are reduced on the air electrode side (cathode), are transported and
react with, for example, hydrogen to form water on the fuel electrode side (anode).
Conversely, electrons (e-) are formed and while moving in the opposite direction, are
available for an outside current use. Nowadays, the most frequently used electrolyte
8
material is zirconia (ZrO2). Zirconia is at ambient conditions a poor ionic conductor. If
(ScSZ) the cubic structure is stabilized even at ambient conditions. Because of this
proceeds through oxygen vacancies due to insertion of a di- or trivalent element (Ca2+,
Y3+, Sc3+) instead of the tetravalent Zr4+. The lack of positive charge is balanced by free
oxygen lattice sites. Through these free sites oxygen can move through the cubic
structure.
Besides good ionic conductivity, gas tightness of the electrolyte is the most
important characteristic the material should have. If the gas tightness of the electrolyte is
insufficient, a reaction between oxygen (cathode side) and hydrogen (anode side) may
approximately 1400°C. Lower temperatures lead to inadequate gas tightness, but pure
contrast to the high sintering temperature, which is necessary for the electrolyte, the
higher values, the amount of triple phase boundaries reduces drastically due to enhanced
sintering. These two facts are the reason for one area of major research and development
9
using a nanosized starting material or b) the use of sintering additives. Additional
research and development is focused on the coating technologies for SOFC electrolytes
[4].
Zirconia has become one of the most industrially important ceramic materials of
the present time. The traditional applications of ZrO2 and ZrO2-containing materials are
foundry sands and flours, refractory ceramic, and abrasives. Because of its high oxygen
ion conduction and high refractive index, it is also used in a wide range of newer
applications which include fuel cells, catalysts, oxygen sensors, and jewelry. Zirconias
have also been utilized in many mechanical applications. Along with high strength and
toughness, zirconia also possesses good hardness, wear resistance, and thermal shock
number of engineering applications such as automobile engine parts, wire drawing dies
and cutting tools. The low thermal conductivity together with relatively high coefficiency
of thermal expansion makes zirconia a suitable material for thermal barrier coating on
with monoclinic, tetragonal and cubic structures. The monoclinic from is stable up to
2370°C; from 237°C to the melting point (2680°C), the cubic form is stable. The reverse
10
transformations take place on cooling. The phase transformations are martensitic in
the ceramic. The phase transformation can be avoided and the high temperature cubic
phase can be stabilized at low temperature by substituting low-valency cations for the
Cubic ZrO2 has the fluorite structure with the O2- ions arranged in simple cubic
packing and half the interstices in this lattice occupied by Zr4+ ions. The substitution of
11
lower-valency cations leads to O2- ion vacancies. The vacancies that stabilized the
structure also lead to high mobility in the oxygen sublattice and to behavior as a high-ion
conductor.
The elements that stabilized the cubic fluorite structure in zirconia include
has been found to give a material with a higher conductivity, which is particularly
valuable at a lower temperature. Scandia stabilized zirconia (ScSZ) is one of the solutions
Fully stabilized zirconia based ceramic materials have several areas of application
such as an electrolyte in solid oxide fuel cells (SOFC), oxygen sensors and structural
applications due to their high oxygen conductivity in the 800-1000oC temperature range,
electrical and mechanical properties. Scandia stabilized zirconia (ScSZ) ceramics have
the highest oxygen conductivity among zirconia based materials [7] and, with the trend in
of 600-800ºC, ScSZ ceramics are very attractive materials for IT SOFCs and catalytic
the electrical and mechanical properties, high density and a low level of internal flaws.
temperatures above 1400ºC in order to form dense ceramics. High sintering temperatures
create a variety of problems such as temperature induced degradation and shape change
12
excessive grain growth as well as energy production costs. Lower processing
temperatures can limit grain growth to nanocrystalline size which has beneficial effect on
the improvement of electrical properties of solid electrolytes [8] due to specific grain
films prepared by sol-gel using a polymer precursor solution showed that the electrical
transport of this material can also be enhanced [9]. Therefore the reduction of the
and electrical properties and make possible co-firing zirconia-based electrolytes with
other SOFC components and significantly lower production costs. Properties such as
hardness, fracture toughness and coefficients of thermal expansion (CTEs) are very
Extensive research has shown that a phase transition from cubic to rhombohedral
600°C [10, 11]. This phase transformation prevents the direct use of this ceramic in
SOFC applications. The aging process of scandia doped zirconia is also accompanied by
conductivity of Sc2O3-ZrO2 electrolytes [12, 13]. Usually the cubic phase can be
stabilizer such as Yb2O3 [13], Y2O3 [10], Bi2O3 [14], Al2O3 [15], and so on. The addition
of a low amount of other oxides usually does not greatly deteriorate the sintering activity
and other properties of the resulting powder. It was reported recently that replacement of
1mol% Sc2O3 by CeO2 stabilized cubic phase led to only 50ºC increase in sintering
13
temperature [14]. Also it was found recently that scandia-ceria doped zirconia ceramics
(ScCeSZ) that can be used to fabricate dense and stable ceramic electrolytes was carried
out. The crystal structure, the morphology and phase transitions of powders and ceramics
were also investigated. Mechanical properties, such as hardness and fracture toughness,
of the Sc2O3-CeO2-ZrO2 ceramics were reported. These results can be further used for the
optimal design of SOFC layered structures as well as for determination of their reliability
Table 2.2 briefly describes all types of commercially available SOFC electrolyte
materials currently available in different forms for developing solid oxide fuel (SOFC)
[18].
14
Table 2.2. (Continued)
15
Table 2.2. (Continued)
16
Table 2.2. (Continued)
17
CHAPTER 3
3.1 Materials
Powder compaction was performed using the uniaxial pressing method. Uniaxial
pressing involves the compaction of powder into a rigid die by applying pressure in a
single axial direction through a rigid punch or piston. The presses were hydraulic and
needed manual operation. A 31mm steel die was used for the preparation of all samples.
The setup consisted of a steel die, a steel punch and two plugs. The powder was poured in
between the two plugs. The steel punch was placed on top of the second plug, and the
pressure was applied. The applied force was 15000N. The calculation is as follows:
Applied Force( F )
Pr essure, P =
Area( A)
∏d2
Area, A = = 1.1697 in2
4
18
Applies Force, F = 15000N = 3372.134 lb
3372.134
Pr essure, P =
1.1697
= 2882.92 Psi
3.2.2 Sintering
joined to one another by grain boundaries. Grain boundaries vary in thickness from about
100µm to over 1µm. Samples were numbered based on the company name for all
different temperatures as listed in Table 3.1. They may consist of crystalline or vitreous
second phases, or may simply be a disordered form of the major phase because of the
differing lattice orientations in neighboring grains. They impart strength to the body,
19
typically in the range 70-350 MPa (cross-breaking). Grain boundaries are generally not as
The sintering process of two uniaxially pressed powders was performed in a high
sintering temperature was from 1100-1600°C in air. The sintering time was 2 hours and
Sample preparation is one of the most important steps during the various
experimental testing of ceramic materials. This process involves several steps to produce
a smooth surface for analysis. The main objective is to obtain an absolutely flat, highly
polished, and clean sample. This process was completed basically in two different steps
as follows:
• Polishing
• Etching.
Polishing
This is an integral part of most types of testing. The best and the fastest polishing
to produce flat and smooth surfaces was performed with the lapping tool typically used
for SEM, AFM and TEM sample preparation and polishing by diamond grinding paper
with a grain size from 30 µm to 0.5 µm. The following steps were taken during the
20
• The sample was mounted on a lapping stage using wax and heat.
• Diamond paper was placed on a clean wet glass plate and air bubbles were
• The lapping tool with the sample was slid under the self-weight of the lapping
tool on 30 micron diamond paper using water for lubrication to obtain a fully flat
surface.
• The sample was flipped over to begin final polishing on the other side
Figure 3.1 represents the polished surface obtained by different grit sized diamond
paper.
Etching
Etching is performed to make visible the grain boundaries for study of grain size
diamond paper and were etched thermally in air for 30 minutes at 1300°C regardless of
21
the sintering temperature. A Thermcraft Incorporated Furnace model#TSL-2-0-6-1S-
Density is the quotient of the mass and volume. The unit is g/cm3 which is
Archimedes’ principle which states that every solid body immersed in a fluid apparently
Archimedes’ method in acetone. For each sample 12 readings were taken. The apparatus
used for this purpose was Mettler Toledo of Model: AX105 Delta Range(R). The
• A balance with all necessary setup (draft shield elements, bracket, equalizing
• The sample was immersed in acetone and the reading was taken after stabilization
22
The density ρ was calculated from the two weighings as follows
A
Density : ρ = (ρ 0− ρ L ) + ρ L
A−B
A − B
Volume :V = α
ρ o − ρ L
where,
ρ = Density of sample
A = Weight of sample
Theoretical Density
Theoretical density is based on crystal structure and molar mass of material, the
Atomic Weight ( M * N )
Theoretical Density : ρ =
N A * a sin α * b sin β * c sin γ
lattice parameters, obtained from XRD experiments, Angle α, β, and γ = 90°, The
23
3.4.2 Indentation Techniques for Hardness and Fracture Toughness Analysis
3.4.2.1 Nanonindentation
known mechanical properties such as hardness and elastic modulus into a material whose
through the material and the images of the indent impressions are captured. Hardness of
the material is defined as the mean pressure the material can support under load and is
calculated as H = P/A, where P is the applied load and A is the projected contact area of
the impression. Elastic modulus is a measure of the recovery of the material during
techniques involve optical imaging of the indentations for calculation of the contact area.
Microindentation testing carried out using Rockwell and Vickers’ hardness tests involve
high loads and are applicable for bulk materials. Nanoindentation, a recently developed
technique, involves indentation depths in the order of nanometers and loads of few
millinewtons.
control and monitor the loads and displacements of an indenter that is driven into and
withdrawn from the material being tested [22-28, 29]. Depending on the specifications of
24
measured [30]. The first major contribution in indentation experiments was made by I.N.
Sneddon (Sneddon, 1965). Sneddon studied the relationships between the load,
displacement and contact area for different indenter geometries and derived with an
equation, P = αhm, where P is the indenter load, h is the elastic displacement of the
indenter and α and m are constants [31]. Values of the exponent m for some common
punch geometries are m=1 for flat cylinders, m=2 for paraboloids of revolution [22].
variety of mechanical properties such as hardness, stiffness and elastic modulus. Methods
have also been developed for evaluating the yield stress and strain-hardening
storage and loss modulus, activation energy and stress exponent for creep are also being
evaluated from the nanoindentation data [35-37]. Fracture toughness is also being
estimated from the optical measurements of the cracks that have formed at the corners of
[21]. Stillwell and Tabor noticed that at least in metals, the impression formed by the
spherical indenter is still spherical with a slightly larger radius than the indenter, and the
impression formed by a conical indenter is still conical with a large included tip angle.
Tabor showed that the shape of the entire unloading curve and the total amount of
recovered displacement could be accurately related to the elastic modulus and the size of
25
the impression for both spherical and conical indenters [40]. It is also notable that the
diameter of the contact impression on the surface formed by the conical indenter does not
recover during unloading and only the depth recovers. The indenter is loaded and
unloaded a few times before the load displacement behavior becomes perfectly
reversible. The plasticity can be dealt with by taking into account the shape of the
perturbed surface in the analysis of the elastic unloading curve. Therefore, the shape of
the unloading curve and the recovered displacement characterizes the elastic modulus of
the material.
Hardness and elastic modulus are the most frequently measured mechanical
pressed into the material, both elastic and plastic deformation processes occur, creating
an impression that conforms to the shape of the indenter of some contact depth, hc. The
26
radius of the contact circle is a. As the indenter is withdrawn, only the elastic portion of
the displacement is recovered and it distinguishes the elastic properties of the material
from its plastic properties [30]. Here hs is the displacement of the surface at the perimeter
of the contact; h is the total indentation depth; hf is the final depth of the residual
The critical parameters used in the calculation of hardness and elastic modulus are
the peak load (Pmax), the maximum indentation depth (hmax), the residual or final
indentation depth after unloading (hf) and the initial unloading stiffness (S = dP/dh). S is
also known as the elastic contact stiffness, or simply contact stiffness. Hardness, the
P
H = (1)
A
27
where P is the load and A is the projected contact area at that load. The hardness
determined by dividing the applied load by the projected contact area under the load
should not be confused with the traditional definition of hardness, the load divided by the
projected area of contact of the residual hardness impression. These two definitions yield
similar values for hardness when the deformation process is mainly dominated by a
plastic region and a fully plastic permanent hardness impression is formed. When the
contact is predominantly elastic, they exhibit different values for hardness since the
residual contact area is very small for a purely elastic contact, which results in infinite
hardness based on the traditional hardness definition [30]. This phenomenon plays a
crucial role while indenting materials with very sharp indenters at small indentation
depths. In these conditions, traditional definition of hardness yields a higher value than
1 1 − ν 2 1 − ν i2
= + (2)
Er E Ei
where Er is the reduced elastic modulus, E and ν are the elastic modulus and Poisson’s
ratio of the test material, and Ei and νi are the elastic modulus and Poisson’s ratio of the
indenter, respectively, and for a diamond indenter these values are 1141 GPa and 0.07
respectively. Reduced modulus is used to account for the elastic displacements that occur
in both the indenter and the sample. Er is related to the contact stiffness by the equation,
π S
Er = (3)
2β A
28
where S is the contact stiffness, A is the projected contact area and β is a constant that
depends on the geometry of the indenter. This equation is derived from elastic contact
theory [41-42] and holds good for any indenter that can be described as a body of
revolution of smooth function [43]. Though the equation formally applies only to circular
contacts, it has been shown that it works well for different geometries provided a
different value of β is used [44-45]. For indenters with square cross sections β = 1.012;
for triangular cross sections like the Berkovich and cube-corner indenters, β = 1.034. It
has been shown recently that yet another correction factor has to be added to the equation
[42, 46-49].
the nanoindentation technique is the manner in which the contact area is determined.
Instead of imaging of the indent impressions, the contact area is estimated from the
indentation load-displacement data. Oliver and Pharr [22] developed a method by fitting
P = A(h − h f ) m (4)
where A and m are constants determined by a least squares fitting procedure; hf is the
final displacement after complete unloading determined from curve fit. The initial
unloading slope is found by differentiating the equation and evaluating the derivative at
the peak load and displacement. Contact stiffness is obtained by differentiating Equation
29
P
hc = h − ε (5)
S
where ε is a constant that depends on the indenter geometry. For spherical indenters ε =
0.75, for conical indenters ε = 0.72 and for Berkovich indenters ε = 0.75. Equation (5)
does not account for pile-up formed during indentation because it is assumed that the
contact is perfectly elastic and only sink-in occurs continuously [30]. The contact area is
for a perfect Berkovich indenter. Taking into consideration the tip blunting effects, the
A(hc ) = 24.5hc2 + C1hc1 + C 2 hc1 / 2 + C 3 hc1 / 4 + C 4 hc1 / 8 + C 5 hc1 / 16 + C 6 hc1 / 32 + C 7 hc1 / 64 + C8 hc1 / 128
(7)
where C1 through C8 are constants. The area function is determined by making a series of
The basic assumption in this process is that the elastic modulus is independent of the
indentation depth. It is also imperative that there is no pile-up. Fused silica is the widely
accepted calibration sample. The machine compliance is also calculated during the
data are reduced and used to obtain the contact stiffness (S) and the contact depths (hc).
For fused silica material, E = 72 GPa; ν = 0.17 and for diamond indenter, E = 1141 GPa;
ν = 0.07. The contact area, determined using Equation (6), is plotted against contact
depth, hc, and is fitted according to a polynomial order. A weighted fitting procedure
30
ensures that data from all depths are given equal importance. The area function
3.4.2.2 Microindentation
It is one of the most frequently measured properties of a ceramic. There are many
densification, and fracture. In fact, many ceramic specifications list minimum hardness
requirements. For example, a new ASTM zirconia’s specification for surgical implants, F
1873-98, stipulates that Vickers hardness (HV) shall be no less than 11.8 GPa (1200 kg/
mm2) at a load of 9.8 N (1 kg). Although measuring and interpreting ceramic hardness
The Vickers hardness test was used for hardness and fracture toughness
Vickers, with loads typically in the range of a few newtons to 9.8 N (1 kg) and occasional
data for soft or high-toughness ceramics as high as 98 N (10 kg). Knoop hardness is
more frequently used for glass, glass ceramics, and ceramic white wares. About 35% are
Knoop with loads from as low as 0.98 N (100 g) to 19.6 N (2 kg). Rockwell hardness are
about 5% [21]. For research purposes, Vickers, Knoop, and Berkovich (triangular
pyramid) indenters are customary; Rockwell and Brinell indenters are rarely suitable for
31
A LECO Microhardness testing Machine Model: M-400-H1/H2/H3 was used for
indentations. The indentions loads were 100, 200, 300, 500, 1000g. At each load 20
indents were created. The following procedures were undertaken for indentations:
• The sample was focused and adjusted to create 20 indents at each load from 100,
• A loaded indentor was lowered against a polished flat surface of the sample for a
• During loading, the indentor penetrated into the ceramic and, on retracting, left a
adjusting the indent on the center of the lens to avoid possible errors of focusing
on Zeiss Axiovert 10 at magnification 500x (1000, 500g) and 1000x (100, 200,
300g).
32
Figure 3.4: Typical images collected at 500x on optical microscope for hardness
study
F
VHN = 1.8544 2 GPa
D
9.81 MPa.)
starting from a pre-existing crack. It is one of the most important properties of material
for virtually all design applications. If the material has a large value of fracture
with a low fracture toughness value. There are four different types of fracture toughness,
33
KC, K1C, K11C, and K111C. KC is the fracture toughness of the sample, which is in a state
plane stress. K1C, K11C, and K111C are the fracture toughness of material under three
Ceramic materials have a much lower K1C value than the metals. The low K1C
value shows that ceramic materials are very susceptible to cracks and undergo brittle
fracture, whereas the metals undergo ductile fracture. Thus for this study, the fracture
nanoindentation data 230GPa and 210GPa for DKKK, Japan and Praxair samples
respectively.
1/ 2
E P
K 1C = 0.016
H C 3/ 2
Indentation load (mN), C = Crack length (µm). Figure 3.5 shows diagonals and cracks
The Young’s modulus value for the DKKK based ceramic sample sintered at 1600°C was
230GPa. The same value for the Praxair based ceramic sintered at 1600°C was 210GPa.
Both values are in good agreement with the literature review of this experimental work.
34
D
C
(a)
(b)
Figure 3.5: (a) Schematic figure having crack and indent length; (b) Image collected
at 1000x on optical microscope (Indentation load is 200g)
35
3.5 Microstructure Characterization
performed on the thermally etched samples of DKKK, Japan and Praxair materials using
a scanning electron microscope (SEM) Model: Hitachi S3000N and atomic force
The SEM uses electrons rather than light to form an image. There are many
advantages to using the SEM instead of a light microscope. The SEM has a large depth of
field, which allows a large amount of the sample to be in focus at one time. The SEM
also produces images of high resolution, which means that closely spaced features can be
most SEMs only require the sample to be conductive. The combination of higher
magnification, larger depth of focus, greater resolution, and ease of sample observation
makes the SEM one of the most heavily used instruments in materials research today.
The scanning electron microscope (SEM) has a very fine “probe” of electrons
with energies of up to 40 keV focused at the surface of the specimen in the microscope
occur at the surface under electron impact: most important for scanning microscopy is the
36
Backscattered Electron Imaging
and backscattered electrons is very sensitive to the angle at which the electron beam
strikes the surface, that is, to topographical features on the specimen. When the electron
beam strikes the sample some of the electrons will interact with the nucleus of the atom.
The negatively-charged electron will be attracted to the positive nucleus but if the angle
is just right instead of being captured by the "gravitational pull" of the nucleus, it will
circle the nucleus and come back out of the sample without slowing down. These
electrons are called backscattered electrons because they come back out of the sample.
Because they are moving so fast, they travel in straight lines. In order to form an image
with BSE (backscattered electrons), a detector is placed in their path. When they hit the
Different elements have different size of nuclei. As the size of the atom nucleus
increases, the number of BSE increases. Thus, BSE can be used to derive an image that
shows the different elements present in a sample. A typical image collected at SEM is
The magnification of this microscope is the ratio between the dimensions of the final
image display and the field scanned on the specimen. Usually, a magnification range of
37
Figure 3.6: Typical SEM images for grain size study
There are several types of SEMs designed for specific purposes ranging from
compositional analyses.
The AFM is also called the scanning force microscope (SFM) and was invented in
1986 by Binnig, Quate and Gerber. The AFM utilizes a sharp probe moving over the
surface of a sample in a raster scan. The probe is a tip on the end of a cantilever which
bends in response to the force between the tip and the sample. Figure 3.7 illustrates the
working of the AFM. As the cantilever flexes, the light from the laser is reflected onto the
38
split photo-diode. By measuring the difference signal (A-B), changes in the bending of
The interaction force between the tip and the sample can be found as the
cantilever obeys Hooke's Law for small displacements. The movement of the tip or
ceramics, most often in the form of a tube scanner. The scanner is capable of sub-
to the sample.
39
The AFM operates by measuring attractive or repulsive forces between a tip and
the sample. In its repulsive contact mode, the instrument lightly touches a tip at the end of
a cantilever to the sample. As a raster-scan drags the tip over the sample, some sort of
detection apparatus measures the vertical deflection of the cantilever, which indicates the
local sample height. Thus, in contact mode, the AFM measures hard-sphere repulsion
In non-contact mode, the tip does not touch the sample and the AFM derives
resolution of 10 pm, and unlike electron microscopes, can image samples in air and under
liquids.
Feedback operation
Once the electronic feedback is switched on, then the positioning piezo which is
moving the sample or tip up and down can respond to any changes in force which are
detected, and alter the tip-sample separation to restore the force to a predetermined value.
This mode of operation is known as constant force mode, and usually enables a fairly
When the feedback electronics are switched off, then the microscope is said to be
operating in constant height or deflection mode. This is particularly useful for imaging
40
very flat samples at high resolution. The following procedures were considered to obtain
• The controller and the main computer were turned on and the SPM cockpit
• The close-contact EZ-Mode was selected and the corresponding configuration file
was loaded.
• The cantilever close-contact mode operation was installed in the AFM scanner
head.
• The cantilever tip was aligned so that the laser light was projected from the back
• The AFM scanner was moved down and the surface was focused to obtain the
• The tip approach was carried out and a specific area on the sample was selected to
• Grain size measurements were performed by the intersect method of Image Pro-
Plus software using Scanning Electron Microscope (SEM) and Atomic Force
41
(a)
(b)
Figure 3.8: (a) Typical AFM images for grain size study; (b) Grain size and
cavitations measurements
42
The measurements were performed, collecting data by moving the red lines from
top to bottom at 10 locations in the same images. Almost 20 grain intersects per line were
collected (see Figure 3.8 (b)). For each sample, the data were collected from 4/5 images
which gave 800-1000 measurements. Cavities density was collected using the threshold
crystallographic structure and chemical composition of the materials. In the present work,
the XRD experiment was used for phase identification of ceramic sintered at different
parameters.
diffractometer with Cu Kα X-ray radiation source and Eurlean cradle. Parallel beam
optics was used to minimize errors associated with sample displacement and surface
roughness. The Anton Paar H-900 high temperature stage mounted on the XYZ-
positioning stage was used for measurements at temperatures up to 900ºC. The zero
point of detector was <0.004º, and sample height was controlled with accuracy of 0.005
mm. Ceramic samples were polished down to 0.5 mm thick tablet form with final grit
substrate in the slurry form using colloidal silver paste. Silver matrix has been used for
43
temperature calibration. Diffraction patterns were collected at 25ºC and then between 100
and 800ºC in 100ºC intervals for CTE measurements and 10°C for phase transition
analysis. The heating rate was 10 ºC/min and the waiting time was 30 minutes before
XRD scans at every temperature for temperature stabilization. High temperature XRD
scans were performed in the interval 2θ of 28-66º with integration time 5s/point and step
0.01º. The waiting time for temperature stabilization before the XRD scan at each
expansion is the relative change in a given dimension when the body is heated. Thermal
expansion of the crystalline phase means a change in the cell lattice parameters with the
temperature variation. Increase of the lattice parameter in the cubic phase with
example of this behavior is shown in Figure 3.9 for the P1600 powdered sample. To
were obtained by Retvield refinement of at least 5 peaks in XRD data using Full-Proof
1 dL
CTE =
L dT
where, L is original length, dL is change in length and dT is temperature difference. This
demonstrates the linear expansion of solid material. The cubical or volume expansivities
44
of solids are three times the linear expansivity. The unit of linear CTE is /°C or
mm/mm/°C.
1) Model-I:CTE- constant
CTE = b o
45
ln L T = b o * T + const .
CTE = b o + b1T
b1
ln L T = b o * T + T 2
+ const .
2
3) Model-III: CTE - varying qudratically
CTE = b o + b1T + b 2 T 2
b1 b2
ln L T = b o * T + T 2
+ T 3
+ const .
2 3
For all three models, coefficients bo, b1, b2 were obtained through curve fitting log
(LT) by linear, quadratic, and cubic functions respectively. An example of this fitting is
presented in Figure 3.10. Analysis of fitting residuals can help to select the appropriate
model for CTE (see Figure 3.11). By substituting value of coefficients and given
temperature, the CTE was calculated. In the case of the P1600 powdered sample
dependence Ln (LT) is not linear and the best fit is achieved by linear and quadratic
models. In cases like this, selection between linear and quadratic models is difficult
because residuals are similar. Despite the constant and the linear models, the quadratic
model is the better fitting (less residuals) model. The accuracy of coefficients of fitting
was used for selection of the appropriate model. In the case of the P1600 powdered
46
sample (see Figures 3.9, 3.10, 3.11) for the linear model b0 = (7.14±0.39)×10-6 and
coefficients for the linear model is significantly better than for the quadratic model.
Therefore, the best model was decided based on analysis of residuals, overall standard
deviations and standard errors of parameters. To compare our data with the known value
of CTE of cubic zirconia (10.5×10-6 deg-1), we have used the results of a constant model.
1.636
Constant
Linear
1.635
Qudratic
1.634
1.633
1.632
Ln(LT)
1.631
1.630
1.629
1.628
Figure 3.10: Constant, linear and quadratic fitting with temperature vs. log (LT)
(P1600-powder)
47
Residuals Constant
Linear
0.00015 Quadratic
0.00010
0.00005
Delta
0.00000
-0.00005
-0.00010
-0.00015
parameters. The actual value of temperature is essential for the accurate calculation of
temperature of silver was taken as reference value for temperature calibration. By back
substitution and polynomial fitting using Microsoft excel and the Origin program, the
coefficient of polynomial equation was obtained for known temperature, and for each set
temperature the actual temperature was calculated as shown in Figure 3.12 and Table 3.3.
48
1100
1000
900
Actual Temperature (K)
800
700
600
500
400
300
200
200 400 600 800 1000 1200
Set Tem perature (K)
49
CHAPTER 4
Mass Spectroscopy at the University of St. Andrews, UK by J. Irvine and C. Savanie) and
theoretical densities were calculated using the lattice parameter obtained by XRD
DKKK Powder
Praxair Powder
The actual composition of Praxair Powder is 15% Sc2O3 +0.05% HfO2 + 0.4%
50
Table 4.1. Theoretical density from the actual composition of DKKK powder
51
Oxygen 1.91670 122.67
Sum 456.34
23
Avogadro’s Number 6.032*10
Lattice Parameters 5.09*10-8cm
Theoretical Density 5.74 gm/cm2
Table 4.2. Theoretical density from the actual composition of Praxair powder
52
CeO2 140.12 140.12 1 16 2 0.4
TiO2 47.9 47.9 1 16 2 2 0.0250 0.0273 0.054 5.241
ZrO2 91.22 91.22 1 16 2 82.55 0.6699 0.732 1.464 267.14
Total Moles 0.9150 1
Oxygen 1.881 120.39
Sum 437.13
Avogadro’s Number 23
6.032*10
Lattice Parameters 5.09*10-8cm
Theoretical Density 5.49gm/cm2
Table 4.3. Powder characterization
It was found that the DKKK powder has a composition close to the nominal
composition while the Praxair powder has an excessive amount of Sc2O3 and a lower
amount of CeO2. The HfO2 is considerably less in the Praxair than in the DKKK and
there is no big difference in composition of TiO2 in both powders. The c-phases at room
temperature has almost identical lattice parameters as shown in Table 4.3. The result
shows that the DKKK powder has a composition close to the nominal while the Praxair
powder has an excessive amount of Sc2O3 and a lower amount of CeO2. From Figure 4.1,
Figure 4.1: SEM micrographs of powders (a-Praxair, USA) and (b-DKKK, Japan)
53
it is very clear that Praxair powder consists of relatively large 1-2 µm pre-sintered
agglomerates of smaller ~300 nm particles and DKKK powder has significantly higher
surface area due to relatively uniform nanograins with size 60-100 nm.
ranges RT-600°C as shown in Figure 4.2. It was found that both powders had a mixture
I
I X X I
I
600°C
O X I
I X I
O
OI OO O OO
500°C
X
O X I
I I
O O OO
O OO RT
30 35 40 45 50 55 60
2Theta (degree)
54
The result showed that with the increase of temperature the rhombohedra phase
disappeared in the XRD pattern and only remained in a cubic phase at temperatures
higher than 500°C. All samples sintered at temperatures higher than 1200°C existed in
XRD of Praxair
The following XRD patterns were obtained for the as-received Praxair powder
R e tv ie ld r e fin e m e n t o f a s -re c e iv e d P r a x a ir p o w d e r
5000 M u lt i- p h a s e a n a ly s is r e s u lt s :
Ag 8 .5 7 (0 .1 4 )%
C u b ic 8 1 .3 7 (0 .6 2 )%
R h o m b o h e d r a l 1 0 .0 7 (0 .2 5 )%
4000
C /R h = 8 .0 8
3000
Intensity (counts)
2000
4 9 .5 5 0 .0 5 0 .5 5 1 .0 5 1 .5 5 2 .0
1000 2 T h e ta (d e g r e e )
0 S ilv e r
C u b ic
R h o m b o h e d ra l
The XRD pattern showed that at room temperature, the powder has both
rhombohedra and cubic phases. The cubic to rhombohedral phase ratio for Praxair
powder is larger than for DKKK powder. The whole pattern was successfully fitted by
three known phases: cubic ScSZ, rhombohedral phase (b-phase) of ScSZ and FCC silver
55
(used as internal standard). This powder has small amounts (~12%) of secondary b-
phase. The insert above shows details of multi-phase fitting used in the Retvield
refinement procedure: small amounts of secondary phase can be traced but quantitative
analysis cannot be done correctly. The cubic phase had lattice parameter
the exact content of powder. The apparent crystallite size of c-phase is 68.1 nm, which
grains.
The following XRD patterns were obtained for the as-received DKKK powder
10000 M u lt i- p h a s e a n a ly s is r e s u lt s :
Ag 4 .2 4 (0 .0 2 ) %
C u b ic 7 5 .6 2 (0 .4 3 ) %
R h o m b o h e d ra l 2 0 .1 4 (0 .3 4 ) %
8000
C /R h = 3 .7 5
Intensity (counts)
2 9 .0 2 9 .5 3 0 .0 3 0 .5 3 1 .0 3 1 .5
6000
2 T h e ta (d e g re e )
4000
2000
0 S ilv e r
C u b ic
R h o m b o h e d ra l
30 40 50 60
2 T h e ta (d e g r e e )
56
DKKK has also both rhombohedral and cubic phases at room temperature. A
phase transition was found at a temperature higher than 500°C and cubic phases recorded
at 600°C. Whole patterns can be successfully fitted by three known phases: cubic ScSZ,
rhombohedral phase (b-phase) of ScSZ and FCC silver. This powder has also significant
amounts (~21%) of secondary β-phase. The insert above shows details of multi-phase
fitting used in the Retvield refinement procedure: the amount of secondary phase can be
quantitatively analysed. The cubic phase has a lattice parameter a=b=c=5.09119(7) Å and
its corresponding theoretical density is 5.74 g/cm3, based on the exact content of powder.
The apparent crystallite size of the c-phase is 36.8 nm, which is in good agreement with
SEM data. The grain size of DKKK powder is two times smaller than the Praxair powder
microstructure of powder. It was found that the DKKK powder had significantly higher
sintering activity and required ~200ºC less sintering temperature than the Praxair powder.
The difference in the sintering activity of powders was a result of differences in the
powder morphology, chemical composition and levels of impurities present in the DKKK
and Praxair powder. It was also observed that the agglomeration had a significant
57
4.2.1 Density Measurements of ScCeSZ and its Porosity
The sintering temperature and density data of ScCeSZ are listed in Table 4.4.
The results show that the porosity level in ceramics sintered from Praxair and
DKKK powders has significant sinteribility difference. It was found that the DKKK
powder was significantly more active in the sintering process and required ~200ºC lower
sintered from Praxair powder (Figure 4.1). Amounts of porosity had been calculated
using theoretical densities 5.49 and 5.74 g/cm3 of ScCeZrO2 ceramics Praxair and DKKK
powder respectively which were estimated using XRD data and the actual chemical
composition of the powders. Density data showed that for ceramics produced from
58
DKKK powder, density increases rapidly from 1100 to 1400ºC and at higher
temperatures become constant (Table 4.4). Similarly, in the same temperature range,
grain size in DKKK ceramics increases significantly from 200 nm to 3 µm and stabilizes
at higher temperatures. Even at 1200°C samples produced from DKKK powder are dense
enough for polishing, grain size measurements and microhardness testing while samples
sintered from Praxair powder at 1400°C are still porous for microindentation. Analysis of
grain size with increase of sintering temperature allowed for the conclusion that DKKK
powder produces dense ceramics at temperatures higher than 1400ºC while Praxair
powder is still not fully sintered at 1600ºC. This conclusion was supported by analysis of
fracture surfaces. Figure 4.5 shows typical fracture surfaces of the samples sintered at
1600ºC from Praxair powders and DKKK powder. It is evident that DKKK powder
created almost
59
fully dense ceramic while the sample produced from the Praxair powder had noticeable
surfaces of samples produced from Praxair powders and DKKK are different (see Figures
4.6-4.8).
Figure 4.6: Grain structure of samples produced from Praxair powder sintered at
temperatures 1500ºC (left) and 1600ºC (right)
Figure 4.7: Grain structure of samples produced from DKKK powder sintered at
temperatures 1400ºC (left), 1500ºC (middle) and 1600ºC (right)
60
Figure 4.8: SEM images of polished and low angle ion-milled surface of sample
P1600 (a) and thermally etched surfaces of samples P1600 (b) and J1600
(c)
due to thermal etching with sintering temperature while the in-grain cavitations level in
DKKK-based samples remain constant. SEM analysis reveals that all Praxair-based
samples have significant non-uniformity. Even the densest sample, P1600, has highly
dense and very porous areas with well distinguished borders between them (see Figure
4.8(a)). The difference in overall density between samples P1400, P1500 and P1600
(Table 4.4) results predominantly from different ratios between dense and porous areas in
the samples. Grain structures in Figure 4.8(b) demonstrate a higher density area of
sample P1600 while average porosity of the sample is 4.4%. Contrary to samples
(Praxair), DKKK-based samples have high uniformity, and the density of in-grain
cavities remains constant for samples J1400, J1500 and J1600 (Table 4.4). One can
conclude that DKKK powder produces dense ceramics at temperatures higher than
1400ºC while ceramics produced from Praxair powder are still not fully sintered at
61
6.0
5.5
5.0
Density (g/ccm)
4.5
4.0
3.5
Praxair
3.0
Japan
2.5
2.0
45
40
P ra xair (I)
35
D KK K (II)
30
Porosity Level (%)
25
20
15
P orous
10
5
0
1100 1200 1300 1400 1500 1600
0
Sintering T em perature ( C )
62
The difference in sintering activity of Praxair and DKKK powders is a result of
powder has a significantly higher surface area due to relatively uniform nanograins (see
Figure 4.1) which favors lower sintering temperature for DKKK powder. Also Praxair
and DKKK powders have a slightly different chemical composition that could have an
and AFM of sintered ceramics. The high quality images were obtained with distinct grain
boundary and grain size measurement was performed as described in Chapter 4 using
Image Pro-Plus. It is observed from the experimental data that the sintering temperature
has a significant effect on the grain size of the ceramic and its microstructure is shown in
Figures 4.11, 4.12, and 4.13. Results of grain size and cavities are listed in Table 4.5.
63
(a)
J1500°C J1600°C
Grain size: 3.61µ Grain Size: 3.81µ
(b)
Figure 4.11: SEM images of DKKK samples from left to right sinter
correspondingly 1200-1600°C
P1500°C P1600°C
Grain size: 2.48µm Grain Size: 4.42µm
Figure 4.12: SEM images of Praxair samples from left to right sinter
correspondingly 1500-1600°C
64
0 2 4 6 8 10
10 10
6.0
5.5 DKKK, Japan
5.0
Praxair, USA
8 8
4.5
4.0
6
3.5 6
Grain size (µ)
3.0
2.5
4 4
2.0
1.5
1.0
2 2
0.5
0.0
0
-0.5 0
0 1100 2 1200 1300
4 1400
6 1500 8 1600 10
Table 4.5: Grain size and in-grain cavities at different sintering temperatures
65
The ceramic sintered at higher temperature has a large grain size while ceramic
sintered at low temperature has a smaller size. The comparative study of grain size of
ceramics J1600 and P1600 showed that the P1600 has a larger grain size. While Praxair
sintered at 1500ºC has significant difference in grain size (Table 4.5) in comparison to the
Japan powder (DKKK). It was found that below 1500ºC, ceramics from Praxair powders
is not fully sintered; on the other hand, pallets of Japan (DKKK) powders are fully
sintered even at 1400ºC. The grain size versus the sintering temperature plot reveals that
the Japan powder can produce the same dense ceramic even at 200 lower temperatures
than the Praxair powder (see Figure 4.11). The Praxair powder cannot produce fully
dense material even at 1600°C (see Table 4.4). This result was also supported by the
images collected after ion milling (see Figures 4.14, 4.15, 4.16).
Figure 4.14: SEM images collected after 1hr ion milling at 30 deg (left-J1600, right-
P1600)
66
Figure 4.15: SEM images collected after ion milling (left-J1600 milled at 9-deg for
18hrs, right-P1600 milled at 5 deg for 20hrs)
Figure 4.16: SEM images of P1600 collected after ion milled at 5 deg for 20hrs
(right-high porous area, left-less porous)
Transition of phase has a greater impact on the different mechanical, thermal, and
67
SOFC electrolyte design. This is because of the significantly higher ionic conductivity of
the cubic phase as compared to the rhombohedral phase. To analyze the DKKK and
Praxair materials, several XRD experiments were performed. The cubic and
rhombohedral phases were identified along with their phase stability temperature and
two phases at room temperature. Ceramics sintered at temperatures 1300ºC and higher
exist in cubic phase at room temperature. Phase transition from rhombohedral to cubic
1.0 0.0
0.8 0.2
0.6 0.4
0.4 0.6
0.2 0.8
0.0 1.0
100 200 300 400 500 600 700
Temperature (°C)
(a)
68
400°C
440°C
460°C
480°C
500°C
540°C
30 31 50 51
2 Theta (°)
(b)
High temperature XRD experiments were performed to study the phase transition
behavior of J1600 and P1600 samples (see Figures 4.18-4.19). Silver peaks were used for
internal temperature calibration. It was found that the DKKK based samples underwent
420°C. Results show that all Praxair-based samples exist in c-phase in the temperature
range100-850°C.
69
Figure 4.18: Changes in XRD pattern of powdered sample J1600 heated from 300 to
500°C with step 10°C.
1.0
-0.005
Fraction of Cubic Phase from XRD
0.8
-0.010
Heat Flow (W/g)
0.6
-0.015
0.4
-0.020
449.4 °C 0.2
-0.025
3.462 J/g
1 hr
0.0
-0.030
300 350 400 450 500
Temperature (°C)
70
The differential scanning calorimetry (DSC) results show that when this material
was heated at temperatures higher than 300°C DKKK samples underwent the slow cubic
to rhombohedral phase transformation. This process was not registered as a peak in slow
DSC scan. As the temperatures increased above 420°C the reverse rhombohedral to cubic
phase transformation occurred at significantly faster rate and this endothermic process
was registered by the DSC (see Figure 4.19). There was no such phase transition in the
Praxair samples and only c-phase was found in the range of RT-850°C. It is my endeavor
to study this behavior by running different samples at different heating and cooling rates
for further investigation of material properties. Further experiments are under way at the
The CTE calculation was based on the procedure described in Chapter 4. The
results obtained for coefficient of thermal expansion of as-received powders Praxair and
DKKK ceramic samples P1600 and J1400 and powders produced from them were
obtained by high temperature XRD method in the temperature range 100-850ºC. Powders
were used in the form of slurry coating on Si (100) substrate with small amounts of silver
paint used as a binder for powders and as internal reference material for temperature
calibration. Ceramic samples with thickness equal to Si substrate were polished down to
0.5µm grit size and annealed at 850ºC for 20 minutes to release stresses on sample
surface.
71
DKKK-based powders have an almost constant CTE in the temperature range
100-850ºC while CTE of all Praxair-based samples and ceramic sample J1400 almost
linearly increase with temperature. CTE of ceramics have the highest increase in this
range (almost 2 times for the sample P1600 and 1.7 times for J1400). This difference
temperature. This also can explain formation of crater-like cavities on fracture surfaces at
The three models of fittings are listed in Table 4.6, and are recognized by bold,
The best fitting for as-received Praxair powder, as-received DKKK powder and
J1500-powder and J1500-ceramic is shown in Figures 4.20-4.25. The best fitting model
was chosen based on an analysis of residuals, overall standard deviations and standard
errors of parameters.
72
Table 4.6. Coefficients of thermal expansion for Praxair and DKKK samples in ceramic and powder forms
# of
Sample Form CTE×106 (deg-1) SD×106
Temp.
Praxair Powder 10.02±0.21 155.1
as-received (6.51±0.24) + (5.16±0.35)×10-3T 27.1 9
(7.43±1.19) + (2.25±3.69)×10-3T + (2.12±2.68)×10-6T2 28.0
P1600 Powder 10.45±0.19 118.4
(7.14±0.39) + (4.61±0.53)×10-3T 32.5 8
(5.83±2.33) + (8.47±6.75)×10-3T - (2.68±4.67)×10-6T2 35.0
P1600 Ceramic 10.51±0.23 244.3
(4.85±0.43) + (8.27±0.62)×10-3T 69.9 18
(5.24±2.05) + (7.05±6.37)×10-3T + (0.89±4.62)×10-6T2 72.2
DKKK Powder 9.74±0.16 89.2
as-received (7.61±0.91) + (3.17±1.35)×10-3T 60.0 7
73
(4.32±6.06) + (13.4±18.7)×10-3T - (7.6±13.8)×10-6T2 66.0
J1400 Powder 10.35±0.20 124.6
(7.67±1.13) + (3.80±1.60)×10-3T 92.6 8
(11.8±6.8) - (8.7±20.3)×10-3T + (8.9±14.4)×10-6T2 99.0
J1400 Ceramic 10.60±0.29 212.8
(5.68±0.36) + (7.36±0.53)×10-3T 39.9 9
(7.37±1.75) + (1.92±5.52)×10-3T + (4.04±4.09)×10-6T2 40.0
J1500 Powder 10.56±0.23 115.1
(4.89±1.51) + (7.27±1.92)×10-3T 64.2 8
(22.8±10.2) - (39.8±26.8)×10-3T + (30.2±17.1)×10-6T2 53.9
J1600 Powder 10.45±0.20 121.6
(8.49±2.29) + (2.47±2.87)×10-3T 123.0 13
(24.0±19.0) - (37.6±48.8)×10-3T + (25.2±30.6)×10-6T2 125.0
Note: Bold – best fit, Italic – acceptable fit, Normal – poor fit.
Praxair as received powder
The following figures represent the best fitting model for the Praxair as received
10
0.00025
0.00020
8 Constant
0.00015 Linear
0.00010
Qudratic
6
0.00005
Delt
4
0.00000
-0.00005
2
-0.00010
-0.00015
0
0
200 300 2
400 500 4600 700 6800 900 8 1000 1100 10
Temperature, K
1.636
Constant
1.634 Linear
Q uadratic
1.632
Ln(a)
1.630
1.628
Figure 4.21: Constant, linear and quadratic fitting with temperature vs. log (LT)
(Praxair powder)
74
DKKK as received powder
The following figures represent the best fitting model for the DKKK as received
Constant
0.00010
Linear
Qudratic
0.00005
Delta
0.00000
-0.00005
-0.00010
100 200 300 400 500 600 700
Temperature,K
Constant
1.634
Linear
Quadratic
1.632
Ln(a)
1.630
1.628
Figure 4.23: Constant, linear and quadratic fitting with temperature vs. log
(LT) (DKKK powder)
75
J1500 Powder
The following figures represent the best fitting model for the J1500 powder (see
Figures 4.24-4.25).
0.00020
Constant
0.00015
Linear
Quadratic
0.00010
Residuals
0.00005
0.00000
-0.00005
-0.00010
-0.00015
-0.00020
300 400 500 600 700 800 900 1000 1100
Temperature (K)
1.636
Constant
1.634 Linear
Quadratic
1.632
Ln(a)
1.630
1.628
Figure 4.25: Constant, linear and quadratic fitting with temperature vs. log (LT)
(J1500 Powder)
76
J1500 Ceramic
The following figures represent the best fitting model for the J1500 ceramic (see
Figures 4.26-4.27).
0.0004
0.0003 Constant
Linear
Quadratic
0.0002
Ln(a)Calc - Ln(a)
0.0001
0.0000
-0.0001
-0.0002
200 300 400 500 600 700 800 900 1000 1100
Temperature (K)
1.636
Constant
1.634 Linear
Quadratic
1.632
Ln(a)
1.630
1.628
Figure 4.27: Constant, linear and quadratic fitting with temperature vs. log (LT)
(J1500 ceramic)
77
4.5 Hardness and Fracture Toughness
high quality indent images on Optical Microscope. The measurements were performed as
14
Non-etched(DKKK)
Non-etched(Praxair)
13 Etched(DKKK)
Etched(Praxair)
12
Hardness(GPa)
11
10
8
1200 1300 1400 1500 1600
Sintering temperature(°C)
Figure 4.28: Hardness vs. sintering temperature plot for etched and non-etched
sample
78
3.8
3.6
Non-etched(DKKK)
3.4 Non-etched(Praxair)
Fracture toughness(MPa·m0.5)
Etched(DKKK)
3.2
Etched(Praxair)
3.0
2.8
2.6
2.4
2.2
2.0
1.8
phenomenon, namely, indentation size effect (ISE), is typical for single crystals and
developed are based on the dislocation nature of the ISE. ISE also can be considered as
nanoroughness range did not exceed 50 nm). To avoid the effect of mechanical polishing
we selected loads which create indents at least 1 µm or 3 times deeper than the estimated
79
threshold of damaged surface layer due to polishing. Simple evaluation of ISE can be
done by linear log H – log h relationship based on empirical power-law relation H = HI*
h-m [50], where h is the depth of indentation, m is power-law exponent of the ISE index,
80
5
13 Praxair Hardness
)
-0.5
DKKK Hardness
11
Hardness (GPa)
3
10
9 2
8
Praxair K1C 1
7
DKKK K1C
6 0
1200 1300 1400 1500 1600
Sintering Temperature (°C)
(b)
Tables 4.7-4.8 contain the hardness and fracture toughness values for each load from
1000g-100g. The weighted value for hardness and fracture toughness was also calculated.
The results obtained for DKKK based and Praxair-based ceramics are listed in Table 4.7
(for non-etched samples), and Table 4.8 (for thermally etched samples). Table 4.9
contains hardness and fracture toughness values for both etched and non-etched samples.
81
Table 4.7. Hardness (H) and fracture (K1C) toughness values at different sintering temperatures and loads for non-
etched samples
82
J1500 H(StDev) 12.70(0.57) 13.23(0.5) 13.81(0.5) 14.63(0.51) 15.28(1.25) 13.93(1.03) 13.70(0.25)
K1C(StDev) 2.36(0.19) 2.35(0.32) 2.68(0.28) 2.6(0.32) 2.49(0.16) 2.46(0.13)
J1600 H(StDev) 12.78(0.41) 12.98(0.5) 13.07(0.4) 13.20(0.53) 13.32(0.61) 13.07(0.48) 13.03(0.2)
K1C(StDev) 2.06(0.17) 1.92(0.14) 2.08(0.25) 1.93(0.22) 1.91(0.23) 1.98(0.08) 1.97(0.08)
P1500 H(StDev) 7.88(1.5) 8.69(1.85) 8.81(1.69) 8.84(1.49) 8.93(2.31) 8.63(0.43) 8.57(0.78)
K1C(StDev) 3.53(0.48) 3.23(0.4) 2.52(0.24) 2.28(0.18) 2.89(0.6) 2.55(0.13)
P1600 H(StDev) 10.20(1.21) 1.03(0.96) 10.6(0.74) 10.86(0.81) 11.70(1.06) 10.74(0.69) 10.74(0.41)
K1C(StDev) 2.01(0.31) 2.33(0.33) 2.31(0.35) 2.37(0.27) 2.07(0.33) 2.24(0.17) 2.24(0.14)
Table 4.8. Hardness (H) and fracture toughness (K1C) values at different sintering temperatures and loads for
thermally etched samples
Avg. Weighted
Sample Property 1000g 500g 300g 200g 100g
(GPa) Avg.(GPa)
H(StDev) 9.05(1.10) 8.55(0.87) 9.74(0.88) 9.94(1.02) 10.56(0.97) 9.57(0.78) 9.55(0.43)
J1200
K1C(StDev) 3.09(0.29) 2.70(1.07) 1.86(0.79) 2.55(0.63) 2.93(0.26)
H(StDev) 10.7(0.68) 11.07(0.54) 11.0(0.64) 11.91(0.5) 11.31(0.87) 11.21(0.45) 11.01(0.20)
J1300
K1C(StDev) 2.75(0.27) 3.08(0.08) 2.88(0.21) 2.50(0.30) 2.80(0.24) 3.00(0.07)
H(StDev) 11.9(0.54) 11.94(0.26) 12.15(0.8) 12.2(0.76) 12.15(0.82) 12.0(0.12) 11.99(0.21)
J1400
K1C(StDev) 2.05(0.14) 2.02(0.29) 2.29(0.37) 2.16(0.30) 1.76(0.42) 2.06(0.18) 2.06(0.11)
H(StDev) 13.3(0.28) 13.09(0.54) 13.1(0.64) 13.6(0.55) 13.25(0.85) 13.3(0.21) 13.28(0.28)
J1500
83
K1C(StDev) 2.00(0.12) 2.02(0.16) 2.23(0.32) 2.02(.19) 1.74(0.02) 2.00(018) 1.75(0.02)
H(StDev) 12.70.41) 12.92(0.55) 13.2(0.39) 13.3(0.45) 13.63(0.23) 13.15(0.3) 13.32(0.16)
J1600
K1C(StDev) 1.91(0.21) 2.12(0.15) 2.11(0.28) 1.79(015) 1.65(0.40) 1.92(0.20) 1.95(0.09)
H(StDev) 8.81(1.58) 8.44(1.19) 8.88(1.29) 9.49(1.56) 8.85(0.97) 8.89(0.38) 8.84(0.56)
P1500
K1C(StDev) 2.68(0.53) 2.81(0.16) 2.55(0.32) 2.29(0.53) 1.69(0.5) 2.41(0.44) 2.66(0.13)
H(StDev) 10.471.0) 9.93(0.99) 10.24(0.3) 10.99(1.0) 10.26(0.54) 10.38(0.4) 10.28(0.27)
P1600
K1C(StDev) 2.11(0.21) 2.18(0.24) 2.30(0.34) 1.70(0.16) 1.05(0.52) 1.87(0.51) 1.93(0.11)
Table 4.9. Comparison of hardness and fracture toughness of non-etched and
etched samples
depths was predicted using the empirical power-law as described in Page 80. The results
of this analysis are presented in Table 4.5 and in Figure 4.10(a). Most of the samples
exhibited relatively weak ISE compared to single crystals of metals and semiconductors
(typically m from 0.12 to 0.32). ISE for polycrystals, was found to depend on the grain
size. Generally, no ISE is observed if impression covers several grains, and for large-
Nanocrystalline sample J1200 does not exhibit ISE. In the case of other samples,
indentation size for most loads are comparable or exceed grain size. We have noticed that
at higher loads (500 or 1000 g) hardness values are scattered significantly for samples
with large grains. These samples have less porosity. They are dense and are expected to
create more regular plastic deformation zones for every indent. This unexpected behavior
84
of large-grained samples can be explained by significant variation of ISE effect from
indent to indent due to significant scattering of the volume of stored dislocations under
indentation.
The Nix and Gao [51] model relates the ISE to geometrically necessary
inverse indentation depth. The additional hardening of GNDs is mainly based on total
line length of dislocations, necessary to form the permanent geometry of the impression,
and on the volume, in which the dislocations are stored. Using strain-gradient plasticity
theory, Nix and Gao created a model describing the contribution of GNDs to the hardness
(strain gradient effect) properties of material. Based on this model the strain gradient law
H h*
= 1+ 4.1
H0 h
depth, and h* is characteristic length (depth) at which the effect of strain gradient
From the linear dependences of H2 from the reversed indentation depth (see
Values of h* indicate that geometrically necessary dislocations are limited by grain size.
85
14
13 180
12 160
I-1500
Hardness (GPa)
11 140 I-1600
II-1200
H (GPa )
2
10 120 II-1300
II-1400
2
100 II-1500
9
II-1600
80
8
60
(a) (b)
Figure 4.31: ISE analysis using power law (a) and Nix-Gao (b) models
86
4.5.2 Nanoindentation
Nanoindentation tests were performed on the ceramic samples J1600 and P1600
sided Berkovich indenter was used with a 60nm tip radius and 65.3° included half-angle.
The indentation was carried out at a depth of 1000nm for ceramic. Sixteen indents for
P1600 and sixty four indents for J1600 samples were created. The data was averaged for
It was observed that the hardness and modulus of DKKK ceramic is slightly
higher than the Praxair ceramic sintered at 1600°C shown in Figures 4.32-4.33. It was
also found that the density of J1600 has a slightly larger density than the P1600. The
DKKK samples were fully sintered and had a fewer number of pores and cavities in
comparison to the Praxair ceramic (see Figures 4.8-4.11).There was not much difference
in hardness and modulus data depending on the number of grains covered by the indentor
87
300
Praxair, P1600
280 DKKK , J1600
240
220
200
180
160
24 Praxair P1600
DKKK J1600
22
20
Hardness (GPa)
18
16
14
12
10
0 100 200 300 400 500
Indentation Depth (nm)
88
CHAPTER 5
5.1 Conclusion
The following conclusions can be made based on the results of the research work
outlined:
• Phase and chemical composition, morphology, and grain size of the 10 mol%
• The sintering rate of the ceramic powder manufactured by DKKK is much higher
than the powder manufactured by Praxair. The DKKK powder was able to reach
its lowest porosity level at lower temperatures and shorter sintering times than the
Praxair powder.
required to use high heating and cooling rates to avoid formation of significant
amounts of b-phase.
89
• Coefficients of thermal expansion in the range RT-800ºC of both types of
ceramics are in the range 10.5 – 10.6×10-6 grad-1 and very close to CTE of cubic
zirconia. CTE of DKKK samples have less linear dependency on temperature than
Praxair samples.
• Young’s modulus and hardness of DKKK samples are higher than Praxair
samples.
are very brittle, especially at elevated temperatures. The highest K1C was 2.5
samples J1300-J1600
90
REFERENCES
[1] Rayne O’Hayre, Suk-Won Cha, Whitney Colella, Fritz B. Prinz, Fuel Cell
Fundamentals, Chapter 1, Published by John Wiley & Sons, New York, (2006).
[2] María Mercedes González Cuenca, Novel anode materials For Solid Oxide Fuel
Cells, Chapter1, Twente University Press, P.O. Box 217, 7500 AE Enschede, the
Netherlands, www.tup.utwente.nl.
[3] Sharon Thomas and Marcia Zalbowitz at Los Alamos National Laboratory in Los
Alamos, New Mexico, Fuel Cells Green Powder
http://www.scied.science.doe.gov/nmsb/hydrogen/Guide%20to%20Fuel%20Cells
. pdf.
91
[12] C. Haering, A. Roosen, H. Schichl, M. Schnoeller, "Degradation of the electrical
conductivity in stabilized zirconia system. Part II: Scandia-stabilized zirconia",
Solid State Ionics, 176(3-4), 2005, 261-268.
[16] I. Kosacki, "Nanoscaled oxide thin films for energy conversion.", NATO Science
Series, II: Mathematics, Physics and Chemistry, 202 (Fuel Cell Technologies),
2005, 395-416.
[19] Mettler Toledo GmbH, Laboratory & weighing Technologies, Ch- 8606
Greifensee, Switzerland, Phone +41-1-944 22 11, Model: AX105 Delta Range(R).
Density Determination of Solid.
92
[24] G.M.Pharr, Measurement of Mechanical Properties by ultra-low Load
Indentation, Mater. Sci. Eng. A, Vol 253, 1998, 151-159.
[25] W.D.Nix, Mechanical Properties of Thin Films, Metall. Trans. A, Vol 20, 1989,
2217-2245.
[26] M.F.Doerner and W.D.Nix, A Method for Interpreting the Data from Depth-
Sensing Indentation Instruments, J.Mater. Res., Vol 1, 1986, 601-609.
[29] Yeon-Gil Jung and Brian R Lawn, Evaluation of Elastic Modulus and Hardness of
Thin Films by Nanoindentation, J. Mater. Res., Vol 19 (No. 10), 3076-3080
[30] J.L.Hay and G.M.Pharr, Instrumented Indentation Testing, ASM Handbook, Vol
8, Mechanical Testing and Evaluation (2000).
[31] J.W.Harding and I.N.Sneddon, Proc. Cambridge Philo. Soc., Vol 41, 1945, 12.
[32] J.S.Field and M.V.Swain, A Simple Predictive Model for Spherical Indentation,
J.Mater. Res., Vol 8 (No. 2), 1993, 297-306.
93
[37] B.N.Lucas, W.C.Oliver, J.L.Loubet, and G.M.Pharr, Understanding Time
Dependent Deformation During Indentation Testing, in Thin Films-Stresses and
Mechanical Properties VI, MRS Symposium Proc., Vol 436, Materials Research
Society, 1997, 233-238.
[41] I.N.Sneddon, The Relation Between Load and Penetration in the Axisymmetric
Boussinesq Problem for a Punch of Arbitrary Profile, Intl. J. Eng. Sci., Vol 3,
1965, 47-56.
[44] R.B.King, Elastic Analysis of Some Punch Problems for a Layered Medium, Int.,
J. Solids Struct., Vol 23, 1987, 1657-1664.
[45] B.C.Hendrix, The Use of Shape Correction Factors for Elastic Indentation
Measurements, J. Mater. Res., Vol 10 (No. 2), 1995, 255-257.
94
[48] J.C.Hay, A.Bolshakov and G.M.Pharr, Applicability of Sneddon Relationships to
the Real Case of a Rigid Cone Penetrating an Infinite Half Space, in
Fundamentals of Nanoindentation and Nanotribology, MRS Symposium Proc.,
Vol 522, Materials Research Society, 1998, 263-368.
[49] J.C.Hay and G.M.Pharr, Experimental Investigations of the Sneddon Solution and
an Improved Solution for the Analysis of Nanoindentation Data, in Fundamentals
of Nanoindentation and Nanotribology, MRS Symposium Proc., Vol 522,
Materials Research Society, 1998, 39-44.
[50] I .Manika, J. Maniks, "Size effects in micro- and nanoscale indentation", Acta
Materialia, 54(8), 2006, 2049-2056
[51] W.D. Nix, H. Gao, "Indentation size effects in crystalline materials: a law for
strain gradient plasticity", Journal of the Mechanics and Physics of Solids, 46(3),
1998, 411-425
[52] MRS Bulletin, Publication of the materials research society, Volume 30, Number-
8, ISSN: 0883-7694 CODEN: MRSBEA, (2005).
95