Citation
Influences of composition, melt viscosity and crystallization on the color strength and stability of multi-oxide glass frit/zircon-vanadium pigment systems for ceramic whitewares coatings applications

Material Information

Title:
Influences of composition, melt viscosity and crystallization on the color strength and stability of multi-oxide glass frit/zircon-vanadium pigment systems for ceramic whitewares coatings applications
Creator:
Earl, David A., 1961-
Publication Date:
Language:
English
Physical Description:
xxi, 367 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Alkalies ( jstor )
Ceramic glazes ( jstor )
Ceramic materials ( jstor )
Crystallization ( jstor )
Frit ( jstor )
Oxides ( jstor )
Pigments ( jstor )
Reflectance ( jstor )
Viscosity ( jstor )
Wavelengths ( jstor )
Dissertations, Academic -- Materials Science and Engineering -- UF ( lcsh )
Materials Science and Engineering thesis, Ph.D ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Includes bibliographical references (leaves 359-366).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by David A. Earl.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
030019851 ( ALEPH )
40878780 ( OCLC )

Downloads

This item has the following downloads:


Full Text













INFLUENCES OF COMPOSITION, MELT VISCOSITY AND CRYSTALLIZATION
ON THE COLOR STRENGTH AND STABILITY OF MULTI-OXIDE GLASS
FRIT/ZIRCON-VANADIUM PIGMENT SYSTEMS FOR CERAMIC
WHITEWARES COATINGS APPLICATIONS











By

DAVID A. EARL


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1998
















ACKNOWLEDGMENTS


I am grateful for the guidance and inspiration provided

by Dr. David E. Clark, chairman of my supervisory committee.

My association with Dr. Clark has greatly enhanced my

academic, professional and personal growth over the past five

years.

I would like to thank Dr. Joseph Simmons, Dr. E. Dow

Whitney, Dr. Jack Mecholsky and Dr. Dinesh Shah for

participating on my supervisory committee. Thanks also go to

Kristie Leiser, Mark Moore, Robert DiFiori, Diane Folz and

Rebecca Schulz of Dr. Clark's research group for their

advice.

In addition I would like to acknowledge the industrial

support of this research. I would like to thank Florida Tile

Industries; Bob Blonski, Klaus Meinssen, Bruno Burzacchini

and Marzia Barrattini of Ferro Corporation; and Dan Swiler,

Hong Chen and Pam Lucas of Ceredec Corporation.

Finally, and most importantly, I am grateful for the

patience and encouragement of my wife, Jacquie. This










research effort would not have been possible without her

support.


iii
















TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ........................................... ii

LIST OF TABLES ........................................... viii

LIST OF FIGURES............................................. x

ABSTRACT..................................................... xix

CHAPTERS

1. INTRODUCTION ........................................ 1

1.1 Color in the Ceramics Industry .................... 1
1.2 Glaze Colorants ................................ 5
1.3 Potential Influence of Frit ......................... 7
1.4 Overview of Dissertation Goals .................... 12
1.5 Guide for Using This Dissertation ................. 16

2. BACKGROUND .......................................... 21

2.1 Color Theory ........................................ 21
2.1.1 Light and the Visible Spectrum ......... 22
2.1.2 Materials Interactions with Light ...... 29
2.1.2a Refraction ........................... 30
2.1.2b Surface Reflection and Gloss ...... 36
2.1.2c Opacity and Translucency ............ 42
2.1.2d Absorption, Transmission and
Color ................................... 51
2.1.3 Color Perception by the Human Eye ...... 60
2.1.4 Color Measurement ......................... 65
2.1.4a Spectrophotometry .................... 65
2.1.4b Basis for Color Quantification .... 72
2.1.4c CIE L*a*b* Measurement Scale ...... 78










2.2 Color in Ceramic Glazes ........................... 86
2.2.1 Silicate Glass Structures and
Properties .................................. 87
2.2.2 Glaze Base Materials and Formulas ...... 99
2.2.3 Fast-Fire Whiteware Glazes and Frits... 105
2.2.4 Ceramic Colorants ......................... 113
2.2.4a Zircon Triaxial Pigments ............ 115
2.2.4b Kubelka-Munk Analysis of
Colorant Layers ............................. 127
2.2.5 Frit Influence on Color Development .... 132
2.2.5a Crystallization Mechanisms ........ 132
2.2.5b Zircon Crystallization and
Dissolution ............................. 145
2.2.5c Liquid-Liquid Phase Separation .... 152
2.2.5d Viscosity Relationships ............. 155

3. EXPERIMENTAL PROCEDURES ............................ 161

3.1 Materials and Methods ............................. 161
3.1.1 Glass Frits and Zr-V Pigment ............. 161
3.1.2 Coatings Preparation and Application... 165
3.1.3 Firing Curves .......................... 167
3.2 Materials Characterization and Analytical
Techniques .................................. 169
3.2.1 AAS and XRF ............................ 169
3.2.2 Frit Density Determination ............... 169
3.2.3 Laser Diffraction Particle Size
Analysis .................................... 170
3.2.4 Spectrophotometry and Color
Calculations ................................ 170
3.2.5 Gloss Measurements ........................ 172
3.2.6 Heating Microscopy ........................ 173
3.2.7 Dilatometry ............................ 174
3.2.8 X-Ray Diffraction (XRD) ................... 174
3.2.9 Scanning Electron Microscopy (SEM) and
Energy Dispersive X-Ray Spectroscopy (EDS).. 176
3.3 Statistical Methods for Deriving Equations .... 177

4. RESULTS ............................................ 180

4.1 Frit and Pigment Properties ...................... 180
4.2 Color of Fired Coatings ........................... 181
4.2.1 Spectral Reflectance Curves .............. 182
4.2.2 Pigment Absorption Factors (K/S) ......... 188










4.2.3 Color in L*, a* and b* Values ............ 193
4.2.4 Color Stability ........................... 206
4.2.5 Specular Gloss ............................. 208
4.3 Viscosity of Coatings During Heating ............ 210
4.3.1 Heating Microscope Images ................ 210
4.3.2 Dilatometric Tg and T2 ................... 213
4.3.3 Viscosity vs. Temperature ................ 216
4.4 Derived Statistical Models ....................... 219
4.4.1 K/S vs. Coating Composition and
Temperature ................................. 220
4.4.2 AE* vs. Coating Composition .............. 235
4.4.3 Log Viscosity vs. Coating Composition
and Temperature ............................. 244
4.5 Evolved Crystalline Species ...................... 246
4.5.1 XRD, SEM and EDS Evaluations ............. 246
4.5.1a Frits with ZrO2 ...................... 246
4.5.1b Frits without ZrO2 ................... 260
4.5.2 Zircon Quantitative Analysis ............. 277
4.5.2a Frits with ZrO2 and ZnO ............. 280
4.5.2b Frits with ZrO2 and SrO ............. 280
4.5.2c Frits without ZrO2 ................... 281

5. DISCUSSION ......................................... 282

5.1 Color Strength and Stability Dependency ....... 282
5.1.1 Coating Composition ....................... 284
5.1.1a Zr-V Loading ......................... 284
5.1.1b ZrO2 .............................. 286
5.1.1c SrO vs. ZnO .......................... 290
5.1.!d A1203/Alkalis ....................... 293
5.1.2 Crystalline Species ....................... 296
5.1.2a Zircon ............................ 296
5.1.2b Diopside ............................. 305
5.1.2c Hardystonite ......................... 307
5.1.2d Strontium Calcium Silicate ........ 308
5.2 Melt Viscosity ................................ 312
5.2.1 Influence on Crystallization and
Zr-V Dissolution ........................ 312
5.2.2 Value as a Predictor of K/S and E* .... 315










6. SUMMARY AND CONCLUSIONS ............................

6.1 Zr-V Pigment and Color Values .................
6.2 Frit Oxide Composition ........................
6.3 Viscosity, Crystallization and Zr-V
Dissolution ......................................

7. FUTURE WORK ........................................

APPENDICES

A UNITS FOR DESCRIBING LIGHT AND COLOR ...............

B THE 15 CAUSES OF COLOR .............................

C DENSITY, PARTICLE SIZE AND APPLICATION WEIGHT DATA.

D DATA FROM COATINGS BATCHED WITH FRIT, 2.5%
BENTONITE AND Zr-V PIGMENT, AND FIRED USING A
45-MINUTE CERAMIC TILE CYCLE .....................

E FRIT SPECTRAL REFLECTANCE DATA AND CURVES AT EACH
TEMPERATURE AND PIGMENT LOADING ....................

REFERENCES ..............................................

BIOGRAPHICAL SKETCH .....................................


vii


320

324
326

328

333




337

339

341




343


347

359

367















LIST OF TABLES


Table Pace

1.1 Example of Ceramic Coatings Applications on a
Decorated Floor Tile ........................... 8

1.2 Variables That Influence Ceramic Glaze Color ...... 9

2.1 Index of Refraction of Selected Materials at
589 nm Wavelength in Air ....................... 32

2.2 Properties of Materials Used for Opacifying
Ceramic Glazes ................................. 45

2.3 Transition Elements and Their Properties .......... 58

2.4 Factors for Uniform Color Scales for Normalizing
to a Standard Reference White .................. 81

2.5 Glass Formers, Intermediates and Modifiers
Materials Commonly Employed in Whiteware
Glazes .......................................... 92

2.6 Properties Associated with the Presence of
Various Oxides in Glass ........................ 98

2.7 Common Ceramic Tile Glaze Base Materials .......... 100

2.8 Examples of Compositions (Weight %) of Commercial
Glazes ......................................... 1 01

2.9 Summary of Important Glaze Properties and
Characteristics ................................ 102

2.10 Seger's Formula for Classifying Glazes ............ 103

2.11 Typical Empirical Formulas, in Molar Equivalents,
for Fast-Fire Gloss and Matte Glazes ............ ill


viii











2.12 Glaze Pigments and Their Requirements ............. 117

3.1 Frits Investigated ................................ 162

E.1 Engobe and Frits A and B Reflectance Data ......... 347

E.2 Frits C and D Reflectance Data .................... 348

E.3 Frits E and F Reflectance Data .................... 349

E.4 Frits G and H Reflectance Data .................... 350















LIST OF FIGURES


Figure Pace

1.1 Flow chart summary of main research variables ..... 17

2.1 Electromagnetic spectrum .......................... 25

2.2 Wavelength vs. energy distribution of daylight .... 26

2.3 Refractive index vs. wavelength of incident
light for three glasses ......................... 34

2.4 Reflection and transmission of light by a glassy
material containing suspended particles ......... 37

2.5 Fresnel reflection for a) an air/glass boundary
and b) total reflectivity for different index
of refraction ................................... 40

2.6 Reflectance vs. wavelength of light for a TiO2
opacified white glaze fired to 10000C, at
various glaze application weights in g/ft2 ...... 50

2.7 The 5 d orbitals .................................. 57

2.8 Human optical system .............................. 61

2.9 Luminosity functions of the rods (nighttime
scotopic vision) and cones (daytime photopic
vision) of the human eye ........................ 63

2.10 Reflectance versus wavelength for opaque coatings
colored with pigments that absorb a portion of
incident light .................................. 67

2.11 Basic components of spectrophotometers ............ 69










2.12 Schematic of the Hardy spectrophotometer .......... 70

2.13 Weighting functions used for the standard
observer at a 20 field of view .................. 74

2.14 Luminosity or lightness (Y) and chromaticity
(x, y) MacAdam limits for colors viewed in
daylight ........................................ 79

2.15 Schematic of L*a*b* color space ................... 83

2.16 Comparison of structures and XRD patterns of
crystalline and vitreous silica ................. 90

2.17 Two-dimensional representation of modifiers
(a) Na 1 and (b) Ca+2 in the silicate glass
structure ....................................... 95

2.18 Inorganic pigment family .......................... 114

2.19 CIE a* and b* chroma of ceramic pigments .......... 116

2.20 Typical forms of zircon crystals. (a-c): a{ 1001
m{110}, p I01}, x{211} ; and zircon lattice
structure (d,e) ................................. 120

2.21 Splitting of the d orbital in V+4 by tetrahedral
(Td) and tetragonal (D2d) crystal fields ......... 124

2.22 Schematic of basis for Kubelka-Munk analysis of
colorant layers ................................. 128

2.23 Relationship between viscosity and temperature
favoring nucleation and growth in glazes ........ 137

2.24 Crystal growth rate as a function of temperature
in Na20-CaO-Al20l-SiO2 glass ..................... 144

2.25 Binary phase diagram of ZrO2 and Si02 system ...... 149

2.26 Viscosities of some commercial silicate glasses... 157











3.1 Time-temperature profiles used to fire the tiles.. 168

4.1 Spectral reflectance of unfired raw materials and
the engobe substrate backing .................... 183

4.2 Spectral reflectance of coatings batched with
frit C, fired to 1000'C ......................... 186

4.3 Spectral reflectance of coatings batched with
frit H, fired to 1100C ......................... 187

4.4 Spectral reflectance of coatings batched with
2.0% Zr-V, fired to 1000*C ...................... 189

4.5 Spectral reflectance of coatings batched with
2.0% Zr-V, fired to 1050'C ...................... 190

4.6 Spectral reflectance of coatings batched with
2.0% Zr-V, fired to 1100'C ...................... 191

4.7 Pigment absorption factors versus weight percent
Zr-V batched in coatings fired to 1000C, 1050C
and 1100*C peak temperature ..................... 192

4.8 Color values of coatings batched with frits
(A-H) and Zr-V pigment, and fired to 1050C ..... 194

4.9 Color values of coatings batched with frits
(A-H) and Zr-V pigment, and fired to I100C ..... 195

4.10 Color values of coatings batched with frits
!A-H) and no Zr-V, and fired to 1000'C, 1050C
or 1100'C ....................................... 198

4.11 Color values of coatings batched with frits
(A-H) and 0.5% Zr-V, and fired to 1000'C, 1050'C
or 1100C ....................................... 199

4.12 Color values of coatings batched with frits
(A-H) and 2.0% Zr-V and fired to 1000'C, 1050C
or 1100C ....................................... 200


xii











4.13 Color values of coatings batched with frits
(A-H) and 5.0% Zr-V, and fired to 1000'C, 1050'C
or 1100 C ....................................... 201

4.14 Relationship between lightness (L*) and blueness
(-b*) of coatings batched with Zr-V pigment and
fired to 10000C, 10500C and 1100'C peak
temperature ..................................... 203

4.15 Zr-V pigment absorption factor relationships
with L*, a* and b* color values of fired
coatings ........................................ 205

4.16 Color changes (AE*) due to a variation in peak
firing temperature from 10500C to 11000C for
coatings batched with frits (A-H) and Zr-V
pigment ......................................... 207

4.17 Specular gloss of fired coatings at a 60' angle
of incidence .................................... 209

4.18 Heating microscope images of coatings batched
with frits (A-H) and 2.0% Zr-V, at 900'C, 1000'C
and 11000C ...................................... 211

4.19 Heating microscope images of characteristic
stages of flow of coatings batched with frits
(A-H) and 2.0% Zr-V ............................. 212

4.20 Thermal dilatometric analyses of coatings A-D
batched with 2.0% Zr-V .......................... 214

4.21 Thermal dilatometric analyses of coatings E-H
batched with 2.0% Zr-V .......................... 215

4.22 Log il versus temperature of coatings batched with
2.0% Zr-V. Frits A-D incorporate 8% ZrO2 ....... 217

4.23 Log 'q versus temperature of coatings batched with
2.0% Zr-V. Frits E-H contain no ZrO2 ........... 218


xiii











4.24 Pigment absorption factor actual and equation
(4.2) predicted results for fired coatings
batched with 2.0% Zr-V ............................. 224

4.25 Individual effects of frit oxides on K/S based on
statistical model (4.2) for 2.0% Zr-V and a
peak firing temperature of 10500C................. 225

4.26 Variables weight of influence on K/S, based on
statistical models 4.1 (0.5% Zr-V), 4.2
(2% Zr-V) and 4.3 (5% Zr-V) ........................ 228

4.27 Pigment absorption factor trends with frit
A1203:alkalis ratio of coatings batched with
2.0% Zr-V and fired to 10000C, 1050'C or I100'C. 229

4.28 Pigment absorption factor versus frit ZnO and
SrO molar equivalents of fired coatings batched
with 2.0% Zr-V and fired to 10000C, 1050'C or
1100C peak temperature ............................ 230

4.29 influence of frit ZrO2, in the presence of SrO or
ZnO, on K/S of fired coatings batched with 2.0%
Zr-V and fired to 10000C, 10500C or 11000C ...... 231

4.30 Variations in pigment absorption factor (K/S) due
to changes in peak firing temperature, for
coatings watched with 2.0% Zr-V and frits
containing 8% ZrO2 versus 0% ZrO2 .................. 232

4.31 Delta E* actual and equation (4.7) predicted
results for fired coatings batched with 2.0%
Zr-V ............................................ 239

4.32 Individual effects of frit oxides on Delta E*,
based on statistical model 4.7 for 2.0% Zr-V .... 240

4.33 Variables weight of influence on Delta E*, based
on statistical models 4.6 (0.5% Zr-V), 4.7
(2.0% Zr-V) and 4.8 (5.0% Zr-V) .................... 242


xiv











4.34 Color changes due to variations in peak firing
temperature and frit A1203:alkali ratio of
coatings batched with 2.0% Zr-V ................... 243

4.35 XRD patterns from coatings batched with frit A
and 2.0% Zr-V, and fired to 1000'C, 1050'C and
11000C. Frit A includes 8% ZrO2, 12% ZnO,
0% SrO, 5% alkalis .............................. 247

4.36 SEM micrographs of a coating batched with frit A
and 2.0% Zr-V, and fired to 11000C.
Magnification is (a) X 1,000 and (b) X 6,000.
Shown are large Zr-V particles surrounded by
fine zircon precipitates ........................... 248

4.37 XRD patterns from coatings batched with frit B
and 2.0% Zr-V, and fired to 10000C, 1050'C and
11000C. Frit B includes 8% ZrO2, 12% ZnO,
0% SrO, 10% alkalis ............................. 249

4.38 SEM micrographs of a coating batched with frit B
and 2.0% Zr-V, and fired to 11000C.
Magnification is (a) X 1,000 and (b) X 6,000.
Shown are large Zr-V particles, fine zircon
precipitates and large zircon fibers .............. 250

4.39 XRD patterns from coatings batched with frit C
and 2.0% Zr-V, and fired to 10000C, 10500C and
I100'C. Frit C includes 8% ZrO2, 0% ZnO,
12% SrO, 5% alkalis ............................. 251

4.40 XRD patterns from coatings batched with frit D
and 2.0% Zr-V, and fired to 10000C, 10500C and
11000C. Frit D includes 8% ZrO2, 0% ZnO,
12% SrO, 10% alkalis ............................ 252

4.41 SEM micrographs of coatings batched with 2.0%
Zr-V, (a) frit C and (b) frit D, and fired to
11000C. (magnification X 1,000). Particles
shown are Zr-V pigment .............................. 253

4.42 X-ray diffraction profile for zircon, ZrSiO4 ...... 255


xv











4.43 SEM micrograph of zircon-vanadium (Zr-V) pigment,
Ceredec 41715A, X 1,000 magnification ............. 256

4.44 XRD patterns from coatings batched with frit E
and 2.0% Zr-V, and fired to 1000'C, 10500C and
1100'C. Frit E includes 0% ZrO2, 12% ZnO,
0% SrO and 5% alkalis ........................... 261

4.45 SEM micrograph of a coating batched with frit
E and 2.0% Zr-V, and fired to 1050'C.
Magnification is X 1,000. Shown are large
Zr-V particles surrounded by dispersed diopside. 262

4.46 XRD patterns from coatings batched with frit
F and 2.0% Zr-V, and fired to 10000C, 10500C
and 11000C. Frit F includes 0% ZrO2, 12% ZnO,
0% SrO and 10% alkalis .......................... 263

4.47 SEM micrographs of a coating batched with frit
F and 2.0% Zr-V, and fired to 11000C.
Magnification is (a) X 1,000 and (b) 6,000.
All particles shown are hardystonite .............. 264

4.48 XRD patterns from coatings batched with frit G
and 2.0% Zr-V, and fired to 10000C, 10500C and
11000C. Frit G includes 0% ZrO2, 0% ZnO, 12%
SrO and 5% alkalis .............................. 265

4.49 SEM micrograph of a coating batched with frit
G and 2.0% Zr-V, fired to 11000C (magnification
X 1,000). Shown is crystallized SrCa2Si3O9 ...... 266

4.50 XRD patterns from coatings batched with frit
H and 2.0% Zr-V, and fired to 10000C, 10500C
and 11000C. Frit H includes 0% ZrO2, 0% ZnO,
12% SrO and 10% alkalis ............................ 267

4.51 SEM micrographs of a coating batched with frit
H and 2.0% Zr-V, and fired to 11000C.
Magnification is (a) X 1,000 and (b) X 6,000.
Shown is crystallized SrCa2Si3O9 ................. .. 268


xvi










4.52 Relationship between XRD [312] integrated
intensity and weight percent zircon in unfired
coatings ........................................ 278

4.53 Weight percent zircon in fired coatings batched
with 2.0% Zr-V .................................. 279

5.1 Changes in reflectance distributions at 400 nm
and 640 nm wavelengths due to increases in peak
firing temperature for coatings batched with
0.5%, 2.0% and 5.0% Zr-V ....................... 285

5.2 Influence of frit density on the color stability
of coatings batched with 2.0% Zr-V pigment ...... 289

5.3 Pigment absorption factors (K/S) for coatings
batched with Zr-V pigment and frits containing
8% ZrO2, and fired to 10000C, 10500C or 11000C
peak temperature ................................ 291

5.4 Visual lightness (L*), greenness (-a*) and
blueness (-b*) as a function of weight percent
zircon in the fired coatings batched with 2.0%
Zr-V ............................................ 299

5.5 Pigment absorption factor (K/S) as a function of
Zr-V pigment and zircon contents in coatings
batched with 2.0% Zr-V .......................... 302

5.6 Changes in color stability denoted by K/S and
Delta E*, as a function of weight percent
zircon in coatings batched with 2.0% Zr-V ....... 304

5.7 Colcr lightness (L*) and blueness (-b*)
progression with diopside crystallization
and pigment dissolution in coatings batched
with frit E and 2.0% Zr-V ....................... 306

5.8 Color lightness (L*) and blueness (-b*)
progression with hardystonite crystallization
and pigment dissolution in coatings batched
with frit F and 2.0% Zr-V ....................... 309


xvii










5.9 Color lightness (L*) and blueness (-b*)
progression with SrCa2Si3O9 crystallization
and pigment dissolution in coatings batched
with 2.0% Zr-V .................................. 311

5.10 Integrated log viscosity from 700'C to 11000C
versus fired color strength (K/S) and stability
(Delta E*) in coatings batched with 2.0% Zr-V... 316

5.11 Slope in the log viscosity versus temperature
near the softening point versus fired color
strength (K/S) and stability (Delta E*) in
coatings batched with 2.0% Zr-V ................. 318

E.1 Frit A spectral reflectance curves ................ 351

E.2 Frit B spectral reflectance curves ................ 352

E.3 Frit C spectral reflectance curves ................ 353

E.4 Frit D spectral reflectance curves ................ 354

E.5 Frit E spectral reflectance curves ................ 355

E.6 Frit F spectral reflectance curves ................ 356

E.7 Frit G spectral reflectance curves ................ 357

E.8 Frit H spectral reflectance curves ................ 358


xviii
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

INFLUENCES OF COMPOSITION, MELT VISCOSITY AND CRYSTALLIZATION
ON THE COLOR STRENGTH AND STABILITY OF MULTI-OXIDE GLASS
FRIT/,ZIRCON-VANADIUM PIGMENT SYSTEMS FOR CERAMIC
WHITEWARES COATINGS APPLICATIONS

By

David A. Earl

December 1998

Chairman: David E. Clark
Major Department: Materials Science and Engineering


Color control is becoming increasingly important in the

industrial processing of ceramics coatings. Multi-oxide

glass frits are the predominant materials in ceramic

whitewares coatings, and zircon doped pigments are the most

commonly used colorants.

The primary objective of this research was to determine

if glass frits could be formulated to improve the fired color

strength and high-temperature stability of ceramic coatings

colored with zircon-vanadium (Zr-V) blue pigments. The

results would also be applicable to other ceramic pigments

that utilize the same zircon structure to incorporate

xix










colorant metal ions. A secondary goal was to relate the frit

oxide composition, pigment content, firing temperature, melt

viscosity and microstructural development to the fired color.

A ceramic tile process was applied to fabricate sample

coatings for the study. A coating's color was quantified and

related to human perception with CIE L*, a* and b* values and

pigment absorption factors (K/S), calculated based on

spectral reflectance data.

The research was successful in quantifying the influence

of individual glass frit oxides on the fired color strength

and high-temperature stability of the coatings. Opaque and

transparent glossy frit compositions which yield excellent

color strength and stability were formulated. Mathematical

models for predicting a coating's color strength and

stability given the frit oxide composition, Zr-V pigment

loading and peak firing temperature were derived. Frit

oxides of ZrO2, SrO, ZnO, A1203, Na20 and K20 were found to

have a significant influence on crystallization, pigment

dissolution and color development. The properties, sizes,

morphologies and quantities of crystalline phases that

precipitated in the coatings during firing were related to

the color. A technique for producing uniquely light yet high


xx










chroma colors through control of zircon precipitate particle

size was demonstrated.

In addition, a statistical model was developed for

calculating the coating melt viscosity as a function of the

frit oxide composition and temperature. Melt viscosity

versus temperature curves were applied to predict a frit's

potential for producing strong and stable color.


xxi














CHAPTER 1
INTRODUCTION


I.1 Color in the Ceramics Industry


The color of a product often determines its aesthetic

and monetary value. Selling prices of ceramics such as tile,

brick, artware, dinnerware, cookware, porcelain enamel,

concrete, bathroom fixtures and decorative glass are directly

related to their appearance.

Customers often select ceramic products based on viewinq

sample chips or prototypes at stores or retail distributors.

Sometimes buyers wish to color coordinate with the fixtures,

appliances or paint of an existing building. Occasionally,

customers purchase more of the same product in the future,

for example, adding the same color floor tile to an enlarged

room. In all of these cases, it is imperative that

manufacturers strictly conform to tight color tolerances,

year after year, in order for goods to comply visually with

customer expectations.

Inconsistencies in manufactured colors lead to

significant increases in industrial costs. Noticeable color

1











variations in fired coatings due to normal fluctuations in

manufacturing processing conditions result in nonstandard,

scrap products and lower productivity. For example, a major

portion of the $120 million worth of products scrapped in

1997 by U.S. ceramic tile manufacturers resulted from out-of-

tolerance colors. Since the U.S. ceramic tile market only

comprises about 0.6% of the world ceramic whitewares sector,

problems with color variations result in significant

industrial waste worldwide [Sez98] In addition, produce

inventory levels must be increased in order to accommodate

muiriole color shades per item.

Color consistency in manufacturing has become

increasingly difficult due to the rapid evolution of

processing technology to improve productivity and

profitability. Often during the transition, accessories for

the same product line must match in color but are made with

wo different processes. For example, in the ceramic tile

industry, flat "field" tile is predominately manufactured

with completely automated fast-fire roller kiln technology,

while difficult to handle trim and angle shapes are

constrained to high-labor slow-fire tunnel kiln processes.

Firina differences between the processes (30-minute fast-fire

vs. 14-hcur slow-f-ire cycles) complicate the formation of











matching glaze colors, but customers demand no noticeable

difference between coordinating trim and field glazes. Color

incompatibility between these products is currently the

largest customer complaint in the ceramic tile industry.

Other whitewares companies, including some manufacturers of

sanitaryware and dinnerware, are planning to convert to the

same fast-fire roller kiln technology and will encounter

similar problems with color variations.

Undesirable color differences between coatings batched

with the same formula occur even in facilities that utilize

the same type of firing technology throughout the factory.

Firing variations result from different kiln models, sizes

and shapes, the debugging of new kiln equipment, kiln fan and

burner wear over time, and changes in ambient conditions

which cause the kiln intake air density, humidity and

temperature to vary. Gaps in the product load entering a

kiln, normally due to breakdowns in machines upstream in the

production line, also cause firing temperature versus time

profiles to fluctuate. Shifts in firing conditions result in

variations in crystallization and pigment dissolution in

ceramic coatings, which alter their fired gloss, opacity and

color. This burdens industry with low yields, customer

complaints and potential loss of business.











Another source of color variation can be attributed to

the continued effort to reformulate ceramic coatings in order

to lower raw material costs. Unfortunately, some of the

cheaper systems have also yielded lower color strength and

stability during high-temperature processing. Due to

increasing foreign competition, domestic companies have been

compelled to lower manufacturing costs to enable selling

price reductions and gain a competitive edge. For example,

in the ceramic tile industry, import market share rose from

about 20% in the mid-1970s to over 60% in the 1990s [ Ear94].

Glaze raw material costs comprise roughly 10% of all tile

manufacturing expenses or approximately $100 million per year

[Fer96, Sez98] In order to stay in business, it is crucial

for whitewares companies to optimize color consistency while

minimizing glaze costs.

There is an increasing interest in the ceramics industry

to develop low cost color systems which are reproducible over

a range of processing conditions. High-temperature

interactions between multi-oxide glass frits and pigments in

ceramic glazes greatly affect the fired color. The influence

of frit is most significant since it is a relatively

expensive, carefully manufactured material usually added at

the highest weight percent of ingredients in fast-fire glaze











batches. Compared to other glaze components, frit normally

has the lowest melting temperature and is the most corrosive

to ceramic colorants. Most frit in the U.S. is used by

ceramic tile manufacturers, but frit consumption by other

whitewares industries will increase as they convert to fast-

fire roller kiln technology. Zircon doped pigments are the

most commonly used colorants for whitewares coatings because

they have the best high-temperature stability.

The overall objective of this research was to determine

the influence of various oxides in glass frits on color

development with zircon doped pigments during firing. The

results could be used to improve the color strength and

stability of industrial ceramic glaze systems.

This chapter very briefly introduces the reader to color

in ceramic glazes, the potential influence of frit on color

and an overview of the dissertation goals. More detailed

discussions are included in succeeding chapters.



1.2 Glaze Colorants


Ceramic glazes provide an impervious glassy decorative

coatinq for whitewares. Colors are produced with selective

scattering or absorption of incident light by colloidal-sized

particles suspended in the glassy matrix.











Solution colorants are sometimes introduced into the

glaze batch as oxides, then dissolved and precipitated as

metal ions during the firing process. Oxides such as Cr2O,

(green), CoO (blue), CuO (green to red), Fe203 (yellow to

brown) and MnO2 (purple to brown) were common sources of metal

ion solution colorants.

They are now rarely used in high volume whitewares

processes such as ceramic tile manufacturing because their

solubility and reprecipitation, and thus the color produced,

are extremely sensitive to the glaze composition, particle

size distribution, firing time and kiln atmosphere.

More typically, manufactured inert pigment particles are

added to glazes to obtain color. The most widely used for

industrial glazes are zircon crystal lattices doped with

metal ions. They provide a lower chroma than other pigments,

but the zircon structure is less soluble at high temperature.

Besides, most of the demand in the whitewares market is for

relatively weak, light colors.

The most common zircon-based pigments are zircon-

vanadium blue, zircon-iron coral and zircon-praseodymium

yellow. They are referred to as the triaxial pigments and

can be blended to achieve most glaze colors desired by

customers. Even though zircon crystals are more stable at











high temperature than other colorants, color variations still

occur. During firing, zircon may remain stable and protect

the metal ions, partially dissolve with that portion

reprecipitating or remaining in solution, or totally

dissolve.

Color control is further complicated because most

decorated ceramic whitewares contain multiple layers of

coatings. For example, a typical "stone look" floor tile

produced by Florida Tile Industries requires nine different

coatings over the pressed body (Table 1.1). Some

applications overlap, while others are distinctly separate on

the surface to create a more natural appearance and depth of

design. Both physical and chemical interactions between

layers influence pigment dissolution and the nucleation and

growth of new phases which affect opacity, gloss and color.


1.3 Potential Influence of Frit


Many variables influence color development in ceramic

glazes, as outlined in Table 1.2. The strength and high

temperature stability of ceramic pigments are highly

dependent upon the base glaze composition. Limited

preliminary studies [ Dec93, Byr94, Blo93] indicate that














Table 1.1. Example of Ceramic Coatings Applications on a
Decorated Floor Tile.



Application Application
Glaze Coating Color/Opacity Method Weight
(g/cm2)


Engobe
(primer coat)


Fume





Middle coat



Fume


Ink 1



Ink 2


Ink 3



Ink 4


Topcoat


White/Opaque



Taupe/Opague





Off-white/
Translucent


Grey/Opaque


Light Grey/
Opaque


Gold/Opaque


Light Beige/
Opaque


White/Opaque


Transparent


Rotating disk



Spray gun


Brushing machine


Rotating disk



Spray gun


Screen printer



Screen printer


Screen printer



Screen printer


Spray gun


0.05






O.OO7




u.09
0.007








<0.001





<0 .001




0.02

0. 02












Table 1.2. Variables That Influence Ceramic Glaze Color.




I. Batch Composition

a. Base ingredients (oxide composition and phases
present)
b. Pigments (composition, structure and loading)
c. Chemical additives (influence application drying
rate and smoothness)

II. Glaze Preparation with Ball Milling

a. Particle size distribution (influences melting
point)

III. Application

a. Thickness (hiding power)
b. Drying rate (can affect composition gradient caused
by differences in particle settling rates)
c. Smoothness (influences gloss or degree of specular
reflectance)

IV Firing

a. Time vs. temperature (phase dissolution and
precipitation)
b. Kiln atmosphere (oxidation/reduction reactions,
sulfur "scumming," etc.)











improved slow-fire glaze compositions may reduce color

changes resulting from variations in glaze preparation,

application and firing conditions. However, there is no

published comprehensive or quantitative research on the

subject. There is also a great lack of research on fast-fire

glaze systems.

Frits are ceramic compositions that have been fused,

quenched to form a glass and granulated [ Dod94] They are

the primary ingredients in fast-fire ceramic glazes and in

most cases are the most reactive and corrosive part of the

formula.

Frits for whitewares coatings are classified as either

opaque (opacified; containing ZrO2) or transparent

(unopacified; no ZrO2), and glossy or matte. Besides ZrO2,

frits also normally employ SiO2 as the primary glass former,

alkalis (K20 and Na2C), B203 and ZnO or SrO as the main

fluxes, and other oxides such as CaO, Al203 and MgO. These

oxides are cost effective, environmentally safe and provide

the desired glaze properties. Compositions with BaO or PbO

are avoided because these elements are deemed hazardous by

the EPA. There is also an increasing interest in replacing

ZnO with SrO because ZnO is classified as a regulated

chemical by EPA. Sections 2.2.1-2.2.3 and 2.2.5 detail the











theoretical effects of individual frit oxides on the

properties of glasses and ceramic glazes.

High temperature properties of frits influence crystal

growth and dissolution rates in glazes. It was observed

[Jam85, Dor94] that dissolution and crystallization

velocities in multi-oxide glasses are diffusion transport

related and inversely proportional to the glass melt

viscosity, although no accurate models have been developed.

Glass viscosity, in turn, varies with composition and overall

has an Arrhenius-type inverse exponential relationship to

temperature. Phase changes in the glass resulting from these

phenomena alter the optical properties and color.

During fast-fire ceramic processing, a glaze is

typically in the molten stage at the peak temperature for

only 3 to 5 minutes. The rapid changes in heating and

cooling rates create a complex thermodynamic system where

phases are often not brought to equilibrium at high

temperature. Frit compositions which reduce the sensitivity

of crystallization and pigment dissolution to processing

variations would be beneficial for color control. The

"ideal" frit would preserve the pigment and precipitate the

same quantity and morphology of desired crystalline phases

over a wide range of firing conditions. The frit should also











produce a coating with enough opacity to hide the substrate

without significantly concealing the pigment and achieve the

desired surface gloss without defects. Currently, it is

not known if zircon pigments dissolve during fast-fire

cycles, and there is uncertainty regarding what phases

precipitate.

1.4 Overview of Dissertation Goals


The main goals of this investigation were to

1. Determine if glass frit oxide compositions could be

formulated to improve the fired color strength and high-

temperature stability of industrial whitewares coatings

colored with zircon pigments.

2. Relate the optical properties resulting from zircon-

vanadium pigment in a glass matrix to the color

perceived.

3. Quantify the influence of individual frit oxides,

pigment loading and peak firing temperature on a

coating's color.

4. Correlate the evolution of the coating's structure and

properties to the original frit oxide composition and

the fired color.











5. Ascertain whether frit melt viscosity data can be

applied as an industrial quality control tool for

predicting a frits potential for producing strong and/or

stable color with zircon pigments.

"Fast-fire" ceramic tile manufacturing constitutes a

major portion of the whitewares industry and consumes most of

the frit produced in the U.S. This was the chosen processing

method for preparing and firing coating samples. Materials

selected for the study were eight laboratory-smelted frits

and a zircon-vanadium (Zr-V) blue pigment. Oxide

compositions of the frits were designed to provide cost

effective, environmentally safe formulas and comply with

Seger's rules (Section 2.2.2) for ensuring insolubility of

the frit and fired coating, and ready fusion at high

temperature. The range of oxide contents tested encompassed

and exceeded the range normally employed for glossy ceramic

tile glazes. Special emphasis was placed on comparing frit

compositions with ZrO2 (opacified) versus no ZrO2

(unopacified), SrO versus ZnO as the secondary flux, and

alkali/silica ratios. The B203 contents were kept low in

order to avoid liquid-liquid phase separation. The Zr-V

pigment tested was the blue colorant most commonly used in

the ceramic tile industry. The results will also be











applicable to other ceramic pigments which utilize the same

zircon structure to incorporate colorant metal ions.

Research goals were achieved by performing the following

tasks:

1. Each of the eight experimental frits were loaded with

four different pigment concentrations of 0%, 0.5%, 2.0%

and 5.0% by weight. They were blended with water and a

suspending agent to produce 32 different glaze coatings.

2. Coatings were applied to opaque 2"X6" wall tile body

substrates using a wet spray method. Samples of each

formula were fired to 10000C, 10500C and 11000C peak

temperatures using a standard "fast-fire" ceramic tile

industrial heating profile. A total of 96 different

fired coatings were produced.

3. The spectral reflectance versus wavelength and the gloss

of each fired coating were measured. The CIE L* a* b*

color values, pigment absorption factors (K/S), and

color differences between tiles fired to 10500C and

1100'C (AE*) were calculated. Relationships between

light absorption by the pigment and color values based

on human perception were quantified.








15


4. Dilatometry and heating microscopy methods were employed

to measure the Tg, Ts and melt viscosity versus

temperature of coatings batched with 2.0% Zr-V.

5. In coatings batched with 2.0% Zr-V, phase changes and

resulting microstructures that formed during firing were

identified using x-ray diffraction, scanning electron

microscopy and energy dispersive x-ray spectroscopy

techniques. Fired coatings were quantitatively analyzed

for contents of Zr-V pigment and zircon which

precipitated from ZrO, and SiO2 in the frit.

6. Statistical models were derived to predict K/S and LE*

given the original frit oxide composition, pigment

loading and peak firing temperature. An equation was

also developed for calculating log 1] of the frit with

2.0% Zr-V given the frit oxide composition and

temperature.

Color strength and stability were correlated to melt

viscosity, crystallization and Zr-V pigment

dissolution.

8. The mathematical models and experimental observations

were related to current scientific literature in order

to collate hypotheses which explain the results.











The foundation of materials science and engineering

research is to gain a better understanding of relationships

between processing, structure and properties of materials.

Figure 1.1 summarizes critical steps taken during this

research to define the processing-structure-properties

relationships of interest. This investigation focused mainly

on variables that influence a ceramic coating's color

strength and stability.



1.5 Guide for Using This Dissertation


The present document is greater in length than most

dissertations. The primary objective was not only to unveil

valuable information for basic science interests but also to

compile a text that could be used as a reference by engineers

working in industry. Thus, some sections may be bypassed if

only very specific information is desired.

In the Background chapter, Section 2.1 overviews current

scientific principles behind (a) materials interactions with

light with a focus on ceramics, (b) color perception by the

human eye and (c) the most common industrial method for

quantifying color and correlating it to human vision.

Section 2.2 summarizes current knowledge of the materials,

processing, structures and properties of ceramic whitewares









MATERIALS

8 Glass Frits (oxide composition, density)

Zircon / Vanadium Blue Pigment




PROCESSING

- Wet Mix 32 Formulas: 8 Frits X 4 Pigment Concentrations

Wet Applications of Coatings to Tile Substrates

Firing Temperatures: 1OQOC, 1050C, 11 OOC

96 Fired Coatings Produced


Figure 1.1. Flow chart summary of main research variables.


COMPOSITION

AAS, XRF, EDS


STRUCTURE

XRD, SEM


PROPERTIES

- Melt Viscosity: Heating Microscopy and Dilatometry

- Color Strength & Stability: Spectrophotometry &
L*a*b* Color System

Gloss: Gloss Meter











coatings and their influences on color. A special emphasis

is placed on ceramic tile glazes. A review of common frit

compositions, ceramic colorants, crystallization, phase

separation and melt viscosity relationships is given.

Throughout the Procedure, Results and Discussion

chapters, references are made to specific principles and

equations outlined in Chapter 2.

Chapter 3 details the experimental procedures applied

for the research. This includes descriptions of typical

industrial "fast-fire" ceramic glaze frit compositions, wet

coating application methods and firing profiles. The two

main categories of frits investigated were with and without

ZrO2. Materials characterization and analytical techniques

and procedures typically applied by industry to evaluate

whitewares coatings are also reviewed.

Chapter 4 shows the results of the research performed

for this dissertation. The particle size distributions,

densities and chemical analyses of the starting frits and

Zr-V blue pigment are given in Section 4.1. Section 4.2

details the optical properties of each of the 96 fired

coatings samples, as indicated by spectral reflectance

curves, gloss measurements and calculated color values of L*,

a*, b*, K/S and AE*0i1C50 I0oc* Section 4.3 reveals the











viscosity versus temperature profiles, heating microscope

images and dilatometric data for coatings loaded with 2.0%

Zr-V. Statistical models for predicting color strength

(K/S), color stability (AE*1050 1 00oC) and melt viscosity as a

function of the frit oxide composition and firing temperature

are given in Section 4.4. These equations provide a method

for engineers in the whitewares industry to estimate the

potential color strength, color stability and melt viscosity

resulting from various frit compositions when utilizing a

typical "fast-fire" heating profile. If the reader is only

interested in specifying frit compositions to obtain certain

color or viscosity results but is not concerned with the

crystallization or pigment dissolution processes responsible

for the optical properties, then it is not necessary to read

Sections 4.5 and 5.1.

Section 4.5 details the structures, compositions and

morphologies of crystalline species that precipitated in the

coatings during firing. An analysis for zircon present in

the coatings quantifies the amount of zircon precipitation

from fritted SiO2 and ZrO, and Zr-V pigment dissolution that

occurred during firing. Results from XRD, SEM and EDS

analyses are shown.











In Chapter 5, the Discussion, relationships are

established between color strength and stability, specific

oxides in the glass frit, Zr-V loading, crystallization, Zr-V

dissolution and melt viscosity. Scientific explanations for

observed phenomena are proposed. The Discussion focuses on

basic science interests, although Section 5.2.2 is also

noteworthy for engineers in industry. It correlates

important characteristics of a coating's log rj versus

temperature plot to the color strength and stability.














CHAPTER 2
BACKGROUND


2.1 Color Theory


The color of an object is its most apparent attribute.

Other properties such as gloss and opacity also contribute to

appearance.

Color is a term that can be used to describe the

reflection or transmission of light in visible wavelengths,

the properties of an object and the perception of the eye.

As one pioneer of color science, Dean B. Judd, stated

Nas83]

Color is that aspect of the appearance of objects
and lights which depends upon the spectral
composition of the radiant energy reaching the
retina of the eye and upon its temporal and spatial
distribution thereon. (p. 3)

Thus, in order to understand color, we must consider the

three elements involved with the production of color: the

light source,the physical modifications of light by matter

and the human eye as a color sensor. Various units typically

applied to quantify light and color are listed in Appendix A

along with essential conversion factors.

21










2.1.1 Liqht and the Visible Spectrum


Light is the visible radiant energy which interacts with

matter to produce what our eyes detect as color. It is an

electromagnetic wave that propagates as electric and magnetic

fields. Maxwell's theory [Hal86] predicted the existence of

a spectrum of electromagnetic waves differing only in

wavelength and traveling through space in a vacuum with a

speed of c = 3 x 108 m/s. Electromagnetic waves other than

light include radio waves, microwaves and x-rays.

The energy of a light wave is quantized into small

bundles called photons. According to Einstein, the energy

(E) of a photon is [Tip8O]
c
E = hf = h- (2.1)
x

where h = 6.63 x 30-34 J-sec is Planck's constant, f is its

frequency and the wavelength.

Other photonic relationships frequently applied are

Hum93]

E mc (2.2)
2m
p : mc (2.3)

kp h (2.4)

where m is the mass of a particle and p is its momentum.

These equatio-ns allow *.s to contemplate a photon or light as








23


either a particle of energy E or a wave with a characteristic

frequency and wavelength.

The wave-particle duality of light can be described

mathematically by considering two harmonic waves with

slightly different frequencies which contain time and space

dependent components [Hum93]

TP = sin [ kx-(ot] (2.5)

and

TP2 = sin [ (k+Ak)x -(o)+Ao))t] (2.6)

2 t
where k the wave number (2.7
x
0=27f = angular frequency (2.8

Superposition of XV and V2 and considering sin a + sin
1
= 2cos(-P).sinx+p) yield a new wave [Hum93]:


T T, + 2 =


2 cos Ot xi sin (k + x L0 + lt (2.9)
2 2 2 2


In (2.9), if Awo= 0 and Ak = 0, a monochromatic wave results

of the form

= 2 sin (kx-ot) (2.10)

Equation (2.10) illustrates the wave characteristics of

photons.











If Aw andAk are very large, the cosine part of (2.9)

modulates the amplitude of the wave, resulting in a string of

wave packets. If many waves are considered with frequencies

ranging between c and Aw, one wave packet results and the

photon can be depicted as a particle. A better intuitive

understanding of materials interactions with light can be

achieved by noting light's wave and particle characteristics.

The color of light is related to its energy and thus its

frequency or wavelength. As shown on the electromagnetic

spectrum in Figure 2.1 [Hun87] visible light falls in a

range of 380 nm to 760 nm in wavelength. White light is

comprised of all the visible wavelengths.

Light can be produced by heating objects to

incandescence, or by exciting atoms or molecules with other

forms of energy.

Incandescent sources are applied to produce light with a

wavelength energy distribution similar to daylight (Figure

2.2) [Bil67] When a material is heated to incandescence,

the increased vibration of its atoms results in kinetic

energy that is sufficient to excite electrons to higher

energy levels. Photons are released when the electrons drop

back to their normal energy levels. As atom vibrations

become more energetic, the frequency and energy of emitted


















FREQUENCY IN CYCLES PER SECOND
tn16 In14 1n12 in10


-, ,u ... -o


10a to6 10~


GAMMA RAYS
- X-RAYS
HARD SOFT
VACUUM U.V.


300 4


HERTZIAN WAVES


ULTRAVIOLET INFRARED
S NR*FAR
_NEAR, FAR__ DIRECTIONAL
RADIO IRADARI
L -FM
....... ~LIGHT ".-..
TELEVISION
LET BLUE GREEN YELLOW RED SHORT WAVE
I -I I II-
00 500 600 700 760 BROADCAST


WAVELENGTH IN NANOMETERS


00


,,, 1I I


0



,10 4 I 4
,o' o"1,,2 10'


WAVELENGTH IN METERS


Figure 2.1. Electromagnetic spectrum. [Adapted from Hun87]


tU


I I I i li i I


POWER
TRANSMISSION






i 0 I
i06 l08


-T


b- -


il


I I I .


24 .22 .20 is,1

















150














C.



50









400 500 600 700
Wavelength, nm








Figure 2.2. Wavelength vs. energy distribution of daylight.
[Adapted from Bi167] (Note: Relative energy distribution
plots set the energy at 555 nm equal to 100, and the rest of
the curve is relative to the distribution radiated from the
source.)











light increases. The color of light produced changes from

red at low temperatures to nearly white at higher

temperatures.

The correlated color temperature of an incandescent

source is defined by the temperature at which a black body

would operate to produce a visual color match with the

incandescent source [Hun87]. The color of a real black body

depends only on its temperature, not its composition.

According to Wien's displacement law [Wea79] the product of

the absolute temperature of a radiating black body and the

wavelength corresponding to the maximum energy is a constant:

X: T = W (2.

where W Wien's displacement constant.

A common incandescent light source which operates at a

high enough temperature to emit a spectrum close to daylight

involves heating a tungsten filament enclosed in an evacuated

fused silica buln. Tungsten filaments are close

approximations to black bodies. Some tungsten filament lamps

used for light sources in color measurement devices

incorporate glass filters to provide a more accurate match to

the daylight spectral energy distribution.

Light can also be generated from luminescence or outer

electron shell interactions in fluorescent and phosphorescent











materials [Ask94] In luminescence, kinetic heat energy is

not essential for the mechanism of excitation. Luminescence

occurs when light has sufficient energy to excite valence

band electrons through the energy gap and into the conduction

band. When the electrons eventually fall back to the valence

band, photons are emitted. If the photon energy corresponds

to wavelengths between 380 nm-760 nm, visible light is

produced. In fluorescent materials, photon wavelengths can

be calculated with
hc
E (2.12)
E0


where Eg is the energy gap. Fluorescent lamps are

electrically excited to produce light. Normally with such

light, the spectral energy distribution is not as continuous

as with incandescent sources. Spectral lines are typically

narrow at specific wavelengths. In order to broaden the

spectral curve, sometimes more than one source is used in

combination. For example, in common household fluorescent

lamps, a spectrum from mercury vapor, electrically stimulated

inside the bulb, interacts with fluorescent powder on the

inside of the glass tube to generate a "cool white" light.

It is common to us- from one to several light sources acting











in combination as illuminates in color measurement devices

such as spectrophotometers [Bil67].


2.1.2 Materials Interactions with LiQht


Properties of a material that influence its appearance

include index of refraction, gloss, translucency,

reflectivity, transparency, absorption and color. These

properties result primarily from the interaction between

light and a material's electronic structure and

microstructure. When photons interact with a material, they

are either attenuated (reflected or absorbed) or transmitted.

As photons enter a material, their speed will also change,

resulting in refraction. Light may be partially reflected,

absorbed, or transmitted as related to the incoming beam

intensity [ io] by [ Ask94]

I = Ir + Ia + 1, (2.13

where Ir, IC and I, are portions of the incident beam that is

reflected, absorbed and transmitted, respectively. The

combination of wavelengths of light reflected from an

opaque material or reflected and/or transmitted from a

nonopaque material produce its color.










2.1.2a Refraction


The refraction of light by a material is estimated by

Snell's law [Che83]:
sin 0 n,
n i (2 .1 4 )
sin 0 n


where

0i = angle of incidence from the line normal to the

irradiated surface.

0- = angle of transmission from the same plane.

n, = index of refraction of material.

n. = index of refraction of medium.

n. relative index of Yefraction of material to

medium.

Normally, n is used as the notation for index of refraction

if the medium is air or a vacuum. Snell's law can be derived

from Fermat's principle [ Tip80] :

The paTri taken by light in traveling from one point
to another is such that the time of travel is a
minimum compared with nearby paths. (p. 612)

When light passes from air into a denser material, its

velocity decreases. Following Fermat's principle, liaht will

minimize its travel time by increasing its optical path

length in air relative to the path length in the denser

material. This results in a change in Qi, Ot and n. Thus, n











is related to the velocity and wavelength of light, which is

given by

n =' 2.5
1)mat-eriai

where u and Xmaterial are the velocity and wavelength of light


passing through a material and Xair is the wavelength of

incident light in air.

At a constant wavelength of incoming light, n tends to

increase for denser materials. When comparing values for

ceramic vs. polymeric (teflon and polystyrene) materials in

Table 2.1 [Wea79] it becomes evident that the structure of a

material also influences n.

If a material is easily polarized, there are increased

interactions of its electronic structure with incident

photons. In dielectrics such as ceramics, the index of

refraction is related to the relative dielectric constant

(K') by [ Kin76]

n = K' + K (2.16)

where Ki is the index of absorption.

Polarization (P) of the electron cloud around an atomic

nucleus is proportional to the electric field strength (E) of

the incoming light:


P = N 2 E


(2.17)











Table 2.1. Index of Refraction of Selected Materials
at 589 nm Wavelength in Air [Wea79].




Material Density Mean Refractive Index
(g/cm3) (n)



Water 1.0 1.33
Polystyrene 1.06 1.60
SiO2 (glass) 1.41-1.46 1.46

Teflon 2.17 1.35

Silicon 2.33 3.49

SiO2 (quartz) 2.64-2.66 1.55

CaCO 2.93 1.60

Diamond 3.51 2.41

(X A!903 3.97 1.77

Fe203 (hematite) 5.24 2.95











where a is the average dipole moment per unit field strength

or the polarizability and N is the number of material

particles per unit volume.

By the Lorentz-Lorentz equation, electronic polarization

is linked to the refractive index of a monatomic gas:
3 F_
n (2.18)
No n

where Eo is the dielectric constant in a vacuum, N. is

avagadro's number, and n_ is the molar refractivity. The

molar refractivity n_ is determined by measuring n at various

wavelengths of light as shown in Figure 2.3 and then

extrapolating to infinite wavelength.

Since the electron density is uniform within an atomic

radius (ro), polarizability is also related to the atomic

volume of a material:

X r (2.19)

Equation (2.19) shows that larger atoms, which contain more

electrons, exhibit greater polarizability and thus tend to

have a higher refractive index.

Ionic charge also plays an important role in influencing

the index of refraction. As the ionic charge becomes

increasingly negative, outer electrons are more loosely bound

and can increase& !olarizability. In addition, the refractive

index is dependent upon crystal structure symmetry, except in


















1.70



1.65
Dense flint glass


1.60 Light flint glass
.6
.-
~Light flint glass


ab-1.55
Booiiae glass


1.50


1.45 !!!
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Wavelength (microns)








Figure 2.3. Refractive index vs. wavelength of incident
light for three glasses. [Adapted from Kin76]











glasses and cubic crystals which are isotropic. Crystals

have a higher index of refraction in denser, close-packed

directions. For anisotropic substances, the mean index of

refraction is estimated as [Par73]


X+ + y(2.20)
n =
3
where X, P and y are refractive indices of the principle

crystallographic directions.

Multiphase crystalline and glassy substances have

specific refractive energies that are the sum of the

re.racrtve energies of their components [ Par73]:

n-l
Kr (2.21)


and


K, = k I00+ k p I00 +
10 0 100


where


K_ = specific refractive energy of a

substance.

n mean index of refraction of a substance.

p = density of a substance.

ki, k2 : specific refractive energies of the

components.

PI, P2 weight percentages of the components.











Reference values for specific refractive energies of oxides

can be used to estimate the refractive indices of glasses,

glazes and crystalline species.



2.1.2b Surface Reflection and Gloss


When a beam of light strikes a material, a portion of

the photons may be reflected. The light reflected at an

angle equal to the incident angle is referred to as specular

reflection (Figure 2.4). This "mirror-like" scattering

occurs at one angle from the point of reflection on a smooth,

nonmetallic surface. If the surface has some roughness, some

light may be scattered at all angles from the point of

reflection. This diffuse reflection is common in ceramics

where surfaces are not normally perfectly smooth. The total

reflectivity is the fraction of incident light specularly and

diffusely reflected. The gloss of a surface is related to

the relative amounts of specular and diffuse reflections. It

can be defined as the degree of approach to a mirror surface.

A perfect mirror surface has a maximum gloss and reflects all

visible light. This surface itself is invisible and has no

apparent microstructure.

Within the visible spectrum, the reflectivity (R) from a

perfectly smooth surface in an air medium is the fraction of























Incident beam

Diffuse
reflection


Diffuse
transmission


Specular
transmission


Figure 2.4. Reflection and transmission of light by a glassy
material containing suspended particles. [Adapted from
Kin76]











incident light reflected at an angle equal to the incident

angle. This fraction of specular reflectance from normal

incidence (Oi=0) is calculated for the optical region of the

spectrum with Fresnel's formula [ Kin76:


{n-i l: (2.23
R (n+1j

Equation (2.22) shows that materials with a high index of

refraction have a high reflectivity. Since the index of

refraction varies with the wavelength of light, so does R.

From Fresnel's law, equations can be written to compare

the reflectJivity of plane polarized light so oriented to a

plane mirror surface that its reflection (ratio of reflected

to incident flux) is most facilitated (R ) or most hindered

(R,) in an air medium [ Jud65]:


cosO, n sin (2.24)

cos01+ n2 si


and


n cos 01 n sin 0i (2.25)

n cos i+ n sin 0,

where 0i is the angle of incidence.

For unpolarized incident light, the total reflectivity

(RT) is the average of R and











RT= (R 1 + R) / 2 (2.26)

Thus, the reflectivity, RT, for unpolarized light is the

average of reflectivities for plane polarized light in (RP

and perpendicular to (R.) the plane of the incident beam. If

0i = O', the incident beam is perpendicular to the surface and

perfect mirror specular reflection results. In this case,

equations (2.24) and (2.25) reduce to equation (2.23) (R -

R). If 0i = 900, only grazing incidence occurs and R R1

RT = 1. Between perpendicular and grazing incidence, R ,R

and RT vary.

For a perfect mirror surface R = R = RT = 1.0

regardless of Oi. For real colorant layers with a glossy

medium, mirror reflectance approaches unity with grazing

incidence as 0i -- 90'. In this case, the incident beam

contacts only that part of the colorant layer which is just

below the surface. This is the high-gloss transparent medium

carrying suspended pigment particles.

Figure 2.5(a) shows the relationship between R R and

RT for 0i from 50 to 90' for an air/glass boundary where the

refractive index of the glass is 1.5. Figure 2.5(b) plots

total reflectivity as a function of 0i and index of

refraction. Figure. 5 can be used to estimate the gloss or










0.11

0.10 1.0

0.09 n2zn1=l.5 0.9

0.08 0.8

0.07 0.7

0.06- 0.6

0.05 0.5
,,0.4
~0.040.
0.03 0.3

0.02 0.2

0.0 -0.1

10 20 30* 40 50' 60" 70 80* 90.
ANGLE OF INCIDENCE (O0
a)



0.11

0.10 1.0
1k.9
0.09 0.9

0.08 -0.8

0.07 -0.7

0.06 0.6

0.050.5

ko 0.04 -0.4

0.03 0.3

0.02 0.2

0.010.1

10' 20' 30' 40' 50" 60' 70' 80 90*
ANGLE OF INCIDENCE (0k)

b)


Figure 2.5. Fresnel reflection for a) an air/glass boundary
and b) total reflectivity for different index of refraction
glasses.











degree of approach to a mirror surface for mediums such as

plastics, glass, textile fibers and paint vehicles [ Jud65].

Note in Figure 2.5(a) that for a smooth glass with n =

1.5, R -- 0 at 0, = 56'. This follows Brewster's law, which

states that mirror reflectance is most facilitated (R = 0)

when [ Jud65]

tan 0i = n (2.27)

Thus, for a glass of n = 1.5 at incident and viewing angles

of approximately 560, there is a lack of polarized diffusely

reflected iaght. Parallel polarized light in the reflected

beam is the only component present from the unpolarized

incident light. This principle has been applied to measure

gloss through the use of a polarizing element which subtracts

the specular component at Brewster's angle where R, = 0.

Brewster's law is also employed to measure the index of

refraction of smooth glass [ Fle93] A light polarizer and

detector are used to measure the polarization effectiveness

of light reflecting from the glass surface. The angle of

incidence (0k) where reflected polarization is most efficient

is applied to calculate n with equation (2.27).

As the surface roughness of an opaque coating or a glass

matrix with suspended colDrant particles a greater

portion of the reflection becomes diffuse. This broadening











of the reflection band and lowering of the specular intensity

tends to lower the gloss. Gloss is greater for smoother,

higher index of refraction surfaces. Surface roughness in

ceramic glazes caused by crystal structures, defects,

interfaces, or a uneven application usually lowers the gloss

by increasing the amount of diffuse reflection.

The relative amounts of specular and diffuse reflection,

and the gloss of the material, have an effect on the color

revealed. For example, glossy paint will appear to lose its

color in daylight glare, but "flat" paint will look nearly

the same. Most color measurement devices take reflection and

gloss into account when generating values for color.



2.1.2c Opacity and Translucency


Subsurface reflections can occur in materials that are

not completely opaque. The reflectivity of opaque metals is

typically 0.9 to 0.95, where most clear glasses are closer to

0.05 [Ask94] .

Optical characteristics of ceramic glazes, glasses and

enamels can be developed by modifying their internal

reflective and light-scattering properties. Characteristics

such as the portion of light specularly reflected (gloss),

the portion of light diffusely reflected (opacity), and the











portions of light directly and diffusely transmitted are

important (Figure 2.4). These attributes are influenced by

the light-scattering properties of phases or small opacifying

particles suspended in the glassy matrix. Good opacification

is obtained with the high light scattering and reflectivity

this mechanism provides.

Common opacifiers used in ceramics coatings include

zircon (ZrO .SiO zirconia (ZrO), SnO2, TiO, and Al20

They are selected based on their optical properties, the

desired coating composition, processing limitations and

properties that affect compatibility with the matrix phase

such as thermal expansion, solubility, hardness and melting

temperature.

Opacifiers can be inert with respect to the host matrix

phase, formed during melting, or crystallized from the glass

melt. Very fine particles can be obtained from materials

such as TiO, or at higher firing temperatures ZrO2, which can

be melted, nucleated and re-crystallized during the heating

cycle.

The degree of opacity of ceramics which contain

particles suspended in a glassy matrix depends on five main

factors [ Par73]:











1. The difference in refractive index between the matrix

and dispersed particles. This effect is described by

Kerstan:
(n n )
S= Io (2.28)
(n + nJ2

where

Ir = intensity of diffuse reflected light or intensity

of opacification.

I0 = incident beam intensity.

ni = index of refraction of the glassy matrix.

n2 = index of refraction of the opacifier.


In ceramic coatings, this mechanism provides the

greatest influence on opacity. As the difference between the

refractive indices of the matrix and particle phases

increases, more light scattering and higher opacification is

achieved. Gencrayv, glazes and enamels have refractive

indices ranging from 1.50 to 1.70 while opacifiers range from

2.0 to 2.8 (Table 2.2) [1 Ree83] .



2. The size of dispersed particles which scatter light.

The light-scattering ability of an optically

heterogeneous material can be estimated with the scattering

or turbidity coefricient (S). This coefficient is a













Table 2.2. Properties of Materials Used
Ceramic Glazes [Par73].


Opacifier Mean Index ot Difference in n:
Material Refraction Opacifier-Glaze
(n)


TiO2 2.50 0.95


ZrO2 2.40 0.85


ZrO 2SiO2 2.05 0.50


SnO2 2.04 0.49


Air i.00 0.55


Glaze


for Opacifying


Melting
Temp.
(C)


1,830


2,715


2,430


1, 625


Coefwcien-
of Expansion






70



4 1


41


1.50-
1.70











measurement of the attenuation due to scattering of light as

it traverses a medium containing scattering particles

[ Kin76]
3
S S V r (2.29)
4 fp

and

Io
exp -IV 4r exp(-Sx) (2.30)

where

Sf = scattering factor that varies between 0 and 4.

S= volume fraction of scattering particles.

r = radius of the scattering particle.

I iIc =ratio of light intensity scattered/initial

nteisity

x = optical path length.


Equation (2.29, shows that scattering, and thus opacity,

tends to increase up to a point with decreasing particle

size. The scattering constant (Sf) increases with particle

size (r) and is inversely proportional to the fourth power of

wavelength for particle sizes much smaller than the

wavelength of incident light. When the particle size is

approximately equal to the wavelength of light, Sf reaches its

maximum value, then decreases with increasing particle size.

Thus, maximum scattering occurs when the opacifier has a











particle size similar to the wavelength of light used, which

is in the range of 0.38 to 0.76 microns in diameter for

visible light.



3. The number of reflecting particles per unit volume.

Equations (2.29) and (2.30) indicate that opacity

increases with the number of particles. As the concentration

of opacifier increases, the rate of increase of opacity

decreases. Glazes and enamels typically utilize a maximum of

17% zircon by weight for opacification. Approximately 3% to

4% becomes an intermediate part of the amorphous glass

structure, while the rest serves as light-scattering

particles.



4. Higher opacification is obtained when there is a

distinct boundary and steep concentration gradient

between the matrix and particle phases.

Diffusion and the consequent reduction in concentration

gradient between the matrix and particles results in lower

opacification. Opacifier/glass systems with relatively low

diffusion coefficients at the required processing

temperatures are most effective. Crystals precipitated in a











glass during the heating or cooling cycle tend to have sharp

interface boundaries.



5. The thickness of a coating applied to a substrate.

Coating thickness effects on covering power can be

related with the Kubelka and Munk equations [ Kin76]:


(1/R_)(R'-R_) R RR-I-exp Sx --R_.

(R' -RJ- RI exp Sx(1 RJ



and
K K 2K
R =1+ + (2. 32
SS S- S


where

RR, coating reflectance.

R'= substrate reflectance.

R-= total reflectivity of a colorant layer so thick

that further increases in layer thickness do not

change the reflectivity.

S = scattering coefficient

K = absorption coefficient.

x = coating thickness.











Equation (2.31) shows that coating reflectance increases

up to a point with thicker applications and higher reflection

substrates. Figure 2.6 demonstrates the increase in

reflectance with application thickness of a typical TiO,

opacified white glaze fired to 1000'C. The main

disadvantages of a thicker coating is the increased material

requirements and cost.

Equations (2.31) and (2.32) indicate that opacifiers

with a high scattering coefficient (S) and low absorption

coefficient (K) are most powerful. The quantity Sx is often

applied to estimate the scattering power of a coating.

Another method gauges opacity with the ratio of reflectance

obtained from a coating over a black (R' = 0) versus a white

(R' = 0.89' backing. This is the Tappi Opacity Method for

determining the contrast ratio C0.9 = R'0/RI0.89.

Liquid-liquid phase separation during glass formation is

another method by which glaze opacification can occur. For

example, the Li20 SiO, Ti02 system separates at high

temperature into a low refractive index silicate glass and a

high index titania rich glass [Kim59] One disadvantage of

this mechanism is that it is very sensitive to processing

temperature. Therefore, in current manufacturing settings

where heating variations during firing are common, the
















100 1

90 30 ,,

80Z N

70 1

60

Ot50

0
u40

0
30
S
20


I0

400 450 500 550 600 650 700
Wavelength, millimicrons









Figure 2.6. Reflectance vs. wavelength of light for a Ti02
opacified white glaze fired to 1000C, at various glaze
application weights in g/ft2. [Adapted from 2ar731











relative proportions of glassy phases and opacification would

be inconsistent and cause product appearance variations.

If translucent rather than opacified appearances are

desired, particles in tne glassy phase must create diffuse

transmission. Translucency is important for products such as

opal glass, where opaque substrates are not utilized. It is

most common to achieve translucency by dispersing a

particulate phase with a slightly different index of

refraction than the glassy matrix. Translucency is also

often controlled with porosity, where lower pore

concentrations (higher material density) or higher pore size

at a given concentration increase translucency.



2.!.2d Absorption, Transmission and Color


Light that is not reflected or transmitted by a material

is absorbed. The linear absorption coefficient (K) indicates

the portion of normally incident radiant energy absorbed

through a unit distance (x) in a single phase material by

[Wea79]
I
T t exp(-Kx) (2.33)


where

T : fraction of light transmitted as it passes through

a material.











It/Ii = transmitted intensity of light/initial intensity

of beam after reflection.


In the Raleigh scattering mechanism of absorption,

photons are deflected from electrons orbiting an atom without

any change in energy. This mechanism is more common for high

atomic number atoms and low photon energies [Ask94]. If an

electron is ejected from an atom, consuming some of the

photon energy, this is referred to as Compton scattering.

Resonance occurs when the frequency of material oscillations

is close to the frequency of incident radiation, which

results in the absorption of radiation.

If incident light stimulates electrons to change their

energy level, the photons are absorbed and the material is

opaque to this particular wavelength of light. Because there

is no energy gap in metals, electron movement into higher

energy levels occurs at almost any photon energy. Therefore,

metals have a high absorption coefficient and are opaque to

most wavelengths of electromagnetic radiation.

The energy gap in semiconductors is greater than metals

and smaller than insulators. Semiconductors with small band

gaps can transmit photons with energies below the energy gap

Eg or become opaque and absorb photons of higher energy. For











example, at 300 K, Si has a gap energy of 1.12 eV, while

diamond has a gap energy of 5.47 eV [Hum93]. Therefore,

silicon requires less energy for electron transitions and

appears opaque in daylight, while pure diamond is

transparent. In the visible spectrum, Eg ? 3.1 eV materials

do not absorb any photons, where Eg 1.8 eV materials absorb

all visible light [Ask94]. For intermediate energy gaps, a

fraction of the incident visible light is absorbed.

In ionic ceramics, filled shells of tightly bound

electrons do not have energy levels available for electron

movement [Ric92], and most single crystals are transparent.

Covalent ceramics, however, vary in the level of absorption.

For example, diamond and graphite both have covalently bonded

carbon atoms, but their optical properties are significantly

different. Diamond is transparent while graphite appears

black. Although there is a strong covalent bonding within

the graphite hexagonal network, weak Vander Waal's bonding

between the layers allows for electron movement. This

results in electron transitions and absorption of visible

light. Good insulators with a large Eg such as diamond tend

to transmit light.

Absorption due to electron transitions and resonance is

intrinsic, while extrinsic effects in ceramics can also cause











absorption and color. Extrinsic effects include grain

boundaries, pores, inclusions, anisotropy and atom vacancies.

In ceramics, the absorption coefficient (K) is related

to the index of absorption (ki) (also referred to as the

attenuation index or extinction coefficient) by [Kin76]

K = 4nki/ (2.34)

From equations (2.15), (2.16) and (2.34), K increases with ki

and the index of refraction of a ceramic material and

decreases with higher wavelengths of incident light.

The overall fraction of light transmitted (T') after

both reflectance and absorption losses is
r

T (I- R) exp (-Kx)) (2.35)


where R is the reflectivity and It/IO is the ratio of

transmitted to incident light intensities. By equations

(2.13), (2.23), (2.28) and (2.35), all of the incoming light

can be accounted for by reflection, absorption and

transmission.

Total light interaction with a material can thus be

written [Ask94]
Irf IR (2.36)

I a = (iIIf)- [I(l-R) exp(-Kx} ] (2.37)

In, = I R(I-R) exp(-Kx) (2.38)








55

I, = Io(l-R) exp(-Kx) (2.39)

= Ir + Ia + Ir, + I, (2.40)

where

Irf = intensity of light reflected at the incident

surface.

Ia = intensity of light absorbed by the material.

Irb = intensity of light reflected at the back face.


Color is produced in many materials through selective

scattering and absorption of incident light. This

selectivity often results from variations in the absorption

coefficient with wavelength. Four electron transitions

concurrent with this type of absorption are common causes of

color [Ric92]:

1. Internal transitions with rare-earth or transition

metals or other ions with incomplete inner electron

shells.

2. Charge transfer, where electrons are transferred

from one ion to another.

3. Electronic transition caused by crystal

imperfections.

4. Bad gap transitions found in many semiconducting

compounds, as discussed earlier in this section.











Transitions (1), (2) and (3) usually are caused by impurities

or defects in a material, while (4) is a bulk property.

Often, particles are suspended in a matrix such as a glass to

create electronic transitions and color. The absorption

coefficient is proportional to the concentration of absorbing

ion (c), according to Beer's law [Ree83]:

T = exp (-Ecx) (2.41)
and
K = cc (2.42)

where E is the extinction coefficient observed per unit

concentration. This is the fundamental law of simple

subtractive colorant mixing.

The most commonly used colorant ions are from transition

metal compounds or impurities such as V, Cr, Mn, Fe, Co, Ni

and Cu shown in Table 2.3 [Ree83]. They provide color in

many ceramic bodies, glazes, glasses, minerals, gems,

pigments and paints.

Crystal or ligand field theories describe how these

elements produce color [ Pet72] Transition metals have

unfilled inner orbitals available for the creation of split

energy levels for electronic transitions. In "free" ions,

orbitals have equal energies but different spatial

orientations, as shown in Figure 2.7 for the five d orbitals.

But the coordination of negatively charged anions about the







































A .f -


~.. ;.~


dx2-y2


Figure 2.7. The 5 d orbitals. [Adapted from Kin76]













Table 2.3. Transition Elements and Their Properties [ Pet72].


V Cr Mn Fe Co Ni Cu
Atomic number 23 24 25 26 27 28 29
Atomic Radius, 1.31 1.25 1.37 1.24 1.25 1.25 1.28
Angstroms
Electronic 3dS4s2 3d54s' 3d54s2 3d64s2 3,74s2 3d84s2 3d104.3
configuration1
Ionization
energies2
First 155 156 171 182 181 176
Second 338 334 361 373 393 418
Third 676 713 777 706 772 810
Oxidation +1.2 +0.91 +1.18 +0.44 +0.28 +0.25 -0.34
potential3
Oxidation 2,3,4,5 2,3,6 2,3,4,7 2,3 2,3 2 1,2
States4
Melting point, 1710 1930 1220 1535 1495 1455 1083
C

Density, g/cc 5.96 7.20 7.20 7.86 8.9 8.90 8.92

Hardness5 9.0 5.0 4.5 2.5-3.0

Electrical --- 62 32 16 17 24 96
conductivity6
lEach atom has an argon inner core configuration.
2Values are in kcal/mole.
3For the oxidation process: M(s) = M2+(aq) + 2e-.
4Common oxidation states; the most stable one is italic.
5Hardness values are on the Mohs scale.
6Compared to an arbitrarily assigned value of 100 for silver.











metal cation produces an electrostatic field that raises

inner orbital energies nonuniformly. This electrostatic

interaction between anions and metal ion's electron clouds

splits inner d or f oroitals into different energy levels.

The energy and corresponding wavelength of light absorbed by

the metal which produces color is equal to the difference in

the split energy levels.

For example, in a tetrahedral structure surrounding a

metal ion with unfilled d orbitals, the df dx and d...

orbitals have more energy than the d-_y- and d,- orbitals.

Color-producing transitions are allowed between these two

split groups, and the wavelengths of light absorbed depend

upon the magnitude of the splitting. Thus, only a limited

range of colors can be produced by any given ion.

The oxidation state of the metal also has an influence

on the magnitude of splitting and resulting spectral

properties. For example, Cu is colorless in solution while

Cu+2 has a strong blue color [ Pot67] When the valence of a

given element increases (e.g., smaller d occupancy), so does

the strength of the ligand field.

This section reviewed the most common causes of color in

materials and the method by which color will be derived

during the subsequent research involved with this











dissertation. A comprehensive list of all of the possible

causes of color are listed in Appendix B.



2.1.3 Color Perception by the Human Eye


The eye is the human optical system (Figure 2.8) [ Tip80]

that allows us to perceive the color, gloss, opacity and

dimensions of an object. The eye is sensitive to light

between wavelengths of 400 and 700 nanometers.

Light enters the eye through the pupil and is focused by

the cornea-lens system on the retina. As the distance of an

object from the eye varies, the ciliary muscle changes the

lens shape to improve focus of the image on the retina.

The photosensitive parts of the eye are the rods and

cones of the retina. These tiny structures receive images

and transmit information along the optic nerve to the brain.

The size of an image on the retina increases with the number

of rods and cones activated, which is proportional to the

apparent size cf the object being viewed.

Rods respond to very small amounts of radiant energy

and, thus, serve for night vision. Rods do not detect hue or

chromatic colors but only perceive neutral colors such as

white, gray and black.




















Anterior


Central retinal
artery and vein


Figure 2.8. Human optical system. [Adapted from Tip80]











On the other hand, cones sense chromatic as well as

neutral colors. Cones, which are responsible for day vision,

can detect a much higher density of radiant flux than rods

but are less sensitive at very low levels of light. Rods

respond to minute quantities of light as low as 10-6 candelas

per square meter (cd/m2), while cones require at least 10-3

cd/m2 [ Jud65] .

At illumination commonly referred to as twilight, both

rods and cones are active. The approximate range of

luminances which correspond to twilight or the mesopic region

is from l0- cd/in2 to 10 cd/M2 [ Jud65] In this range, color

judgments are extremely unreliable because the relative

degree of rod and cone vision continually changes.

Therefore, color inspections in manufacturing should not be

carried out in this condition, although many factories and

industrial inspection areas are dimly lit. Luminance at

approximately 10' cd!m" is the maximum level where the human

eye can perceive color with cone vision.

The eye is not equally sensitive to all wavelengths of

light. It has been demonstrated that 555 nm light is viewed

more easily than other wavelengths. Figure 2.9 graphically

shows the sensitivity or relative response of the human eye

at daytime (cone or photopic curve) and nighttime (rod or









63




















1.0


I (NIGHTIME SCOTOPIC j (PHOTOPIC LUMINOSITY
LUMINOSITY CURVE) / CURVE)
/
/\
Ii .
U)
z
0


.5
,-J.. /\

I-. /
-j
,,.,, //
w/
/\
/i
Ix



0 .
400 500 600 700
WAVELENGTH nm














Figure 2.9. Luminosity functions of the rods (nighttime
scotopic vision) and cones (daytime photopic vision) of the
human eye. (Adapted from Hun87]











scotopic curve) for the same amount of energy at different

wavelengths of light in the visible spectrum. The curves

were developed from experiments where 52 human observers

adjusted the intensity of light at different wavelengths

until they appeared equally luminous or bright [Hun87] The

property of light by which we define how easily we can see it

is referred to as luminosity.

Both the Young-Helmholtz and Hering experiments of the

mid-1800s confirmed that human observers see colors with

three spectrally unique receptors which detect black-white,

red-green and yellow-blue. Subsequent numeric scales

developed for quantifying color contain three values; one for

each opponent-color pair.

Color sensory responses of the eye are not linear with

the amount of stimulus. For example, there is a logarithmic

relationship between the actual light level reflected from an

object and its perceived lightness. Color perception is also

a function of the light source and directions of illumination

and view. For these reasons standardized observation

conditions are required for industrial inspection. Even so,

differences between individuals are great enough to affect

visual color quality control. Most often, human observations

are coupled with color measurements performed by machines











such as spectrophotometers in order to make final

determinations.



2.1.4 Color Measurement


Color measurement, like human eye perception, depends on

the light source, the sample being viewed and the observing

apparatus. Properly operated color measurement equipment,

however, can provide more repeatable results than subjective

human perception. Through a series of calculations, measured

colorimetric data can be converted into values that better

relate to human vision.



2.1.4a Spectrophotometry


The spectral characteristics of an object determine its

perceived color. Spectral characteristics are defined by the

reflectance or transmittance of light from a material as a

function of its wavelength. Spectrophotometers are used to

measure reflectance or transmittance from a sample as a

percentage of incident light at each wavelength in the

visible spectrum, normally in 0.5 nm increments [Mac9la].

Typically, reflectance is measured for opaque materials and

transmittance for transparent materials where the color of

light after it passes through a material is important.











Figure 2.10 reveals reflectance curves for opaque

coatings colored with pigments which absorb a portion of the

incident light [Hun87] The white coating reflects a high

portion of the incident light across the whole visible

spectrum while the black coating absorbs most of the light

flux over the wavelength range. Colors of blue, green,

yellow and red are created through selective absorption and

reflection of different light wavelengths by the pigment.

For example, the blue coating is shown to absorb primarily

yellow to red light (550-700 nm) while reflecting blue (450-

550 nm). The color of a ceramic whiteware glaze coating

results from this mechanism. Glazes are applied over an

opaque white substrate, and the reflectance curve produced is

the sum of reflectances from the pigment particles, other

crystalline species present, and the white substrate minus

specific wavelengths of light absorbed by the pigment

colorant. Normally, the background substrate and undoped

pigment crystal strongly reflect all visible wavelengths and

appear white or light yellow without the light absorbing

metal ion incorporated in the pigment. In contrast, the

color of light transmitted through a nonopaque glass results

from the incident beam minus both light absorbed by the

structure and reflected from the irradiated side.
















100%


PERCENT ULUL
REFLECTANCE
















BLACK
0%
WAVELENGTH, NANOMETERS




Figure 2.10. Reflectance versus wavelength for opaque
coatings colored with pigments that absorb a portion of
incident light. [Adapted from Hun87].











The basic components of spectrophotometers are outlined

in Figure 2.11. Reflectance factors are measured one

wavelength at a time, normally at 0.5 nm increments, by

isolating wavelengths with gratings, prisms or interference

filters and slits. This is the monochromator device in

Figure 2.11 [Hun87] Current spectrophotometers are similar

to original mechanism developed by A. C. Hardy in 1928, shown

in Figure 2.12. The position of mirror slit #2 is adjusted

for wavelength isolation.

Typical light sources include a tungsten filament lamp

or a pulsed xenon bulb, which, in conjunction with a prism,

produce white light. Illumination is normally near 10' cd/mr

where only cone vision occurs. For accurate color

comparisons, the relative energy versus wavelength

distribution from the source must exactly match the desired

standard, but the total energy or illuminance from the source

can range from 10 cd,/n2 to 106 cd/i2 where the rods are

inactive.

Real liQht sources are difficult to standardize, and it

is often useful to compare the color of objects viewed under

various wavelength energy distributions. Normally

reflectance values from the real light source are






















SPECTRALLY
CONTINUOUS
SOURCE


WAVELENGTH
SELECTOR
(MONOCHRO-
MATOR)


-~ \-
I")


SPECTRAL CURVE


Figure 2.11. Basic components of spectrophotometers.
[ Adapted from Hun87]























PHOTOMETER
Anti-hunt
generator
W atee n g t S p e c u l a r
Apertre po.ailfng rau lSa nd
Lens Aperturt Slit #e r filter t










Collimator PDom#1codeset
enI N Lamm
Photometer #2m Sev motor I s ...mp ..le





Photometer caNOTEgls o
sc~~D aleh motor srophttue


~ du amplifier aelctdud


















Figure 2.12. Schematic of the Hardy spectrophotometer.
Adapted from Bi167]











mathematically converted to represent theoretical sources or

illuminants before color values are derived. Outputs from

spectrophotometric measurements include color values derived

from relative energy distributions of standard light sources

such as D65 (average noon daylight from the total sky),

illuminant A (incandescent lamp), illuminant B (near

sunlight) and illuminant C (average daylight from the total

sky).

The standard full visual field of view utilized to

detect light reflecting from an object is 2' angular

subtense. Occasionally, a 100 observer is used to provide a

larger field of view.

Light flux reflected from a sample is collected for

measurement by a white-lined integrating sphere. Elimination

of surface gloss from the color measurement provides results

which better correlate to visual inspection. This can be

accomplished by replacing the white plug on the sphere's

specular cup with a black plug. Since the specular cup is

offset to be illuminated at an angle of reflection equal to

the angle of the incident beam, the black plug absorbs

primarily specularly reflected light. Diffusely reflected

flux is diverted up the sphere to the photodetector. The

collected photon energy distribution is converted into an











electrical signal and sent to a computer. The computer

program converts measured spectral data into numbers that can

be more easily interpreted and correlated to the response of

the human eye.

Measurements of gloss can be performed separately with

goniophotometers or gloss meters, which measure the spectral

reflectance or quantity of light emitted in directions

related to the surface characteristics of the object. The

gloss of ceramic coatings is normally measured at a 600 angle

of incidence, where mirror reflectance is most facilitated,

according to Brewster's law (equation 2.27).



2.1.4b Basis for Color -uantification


The average sensitivity of the human eye to each

wavelength of light has been determined through extensive

experimentation [ Bil67, Hun87, Jud65, MacA35, MacA42] Human

observers were asked to visually match the colors of light

from individual wavelengths by mixing together lights from

three colored primaries. Three primaries were applied

because the eye contains three spectrally unique receptors

for detecting colors. The amounts of energy of each of the

three lights required to match single wavelength colors were

used to develop standard observer functions for the basis of











all color measurement. These weighting functions are applied

to transform spectrophotometric data into numbers that better

correlate to the way the human eye perceives colors.

The weighting functions ( x y and z ) are plotted in

Figure 2.13. Mathematical functions for describing colors

obtained by mixing different sets of primary colors have been

shown to always be related by a set of linear transformations

[Hun87] Therefore, there was some flexibility for selecting

the three standard primaries which providex y and z

the most user-friendly set of weighting functions. The

curves in Figure 2.13 were derived with the following useful

properties:

1. One of the functions, the y curve, was made to equal

the photopic plot shown previously in Figure 2.9, which

indicates the eye's response to luminosity or color

brightness.

2. The areas underneath the three curves were made equal

for light of equal energy at all wavelengths.

3. One function was selected to be as near to zero as

possible for as much of the spectrum as possible.

In this form x y and z do not represent any real

colored primaries but can be converted to values that are









74











200








150


En
I-I

z
Do
0

w T
> 100

-J





50 _








0
400 500 600 700
WAVELENGTH. NM








Figure 2.13. Weighting functions used for the standard
observer at a 20 field of view. [Adapted from Bi167]











easier to apply. The weighting functions are used to

transform spectral reflectance curves into three numbers

referred to as tristimulus values, X, Y and Z. These values

specify color in terms of the mixture of red (X), green (Y)

and blue (Z) primary light that would produce the same color.

The Y value also still includes brightness detected. If two

materials are found to have the same measured X, Y and Z,

they will appear to have the same color under that specific

viewing condition.

At any one wavelength, X = x, Y = y and Z = z. For

example, at a wavelength of 450 nm in Figure 2.13, the light

detected would consist of proportions 32 : 5 : 175 of red ( x

or X) : green (y or Y) : blue (z or Z) light. For all

wavelengths in the visible spectrum, the contribution of each

tristimulus value can be calculated with [ Hun87]

X = Sk x + S XX + S) X + ... + S? X, (2.43)

or
7CO
X f f S. x dA (2.44)
4CC
where

x. = weighting function (x) value at X wavelength.

SX= spectral energy at ? wavelength.











Y (y) and Z ( z ) can be calculated in the same manner as X

x) with (2.43) and 2.44). The spectral energy, in turn, is

a function of the properties of the light source for an

illuminate or aperture color (SXsource) or the reflectance from

a reflecting object (S~materiaI) :

S source :E (2.45)

and

SmateriaI EX Rx (2.46)

where

EX = energy of the light source at ). wavelength.

RX = percent reflectance of light of ? wavelength from

the material.


Since objects are viewed in relation to their surroundings,

X, Y and Z are normally expressed relative to the luminosity

of a perfect white opaque material where R = 100, as

fE R x d%.
X = 100 f X (2.47)
f E y Ldk





f ER z ydX
Z 100 f (2.49'
E y dk











Thus, for a perfectly white material, Y is 100 and X and Z

vary depending upon the light source.

Trichromatic coefficients (x, y, z) are often calculated

from tristimulus values:

x (2.50)
X
x + Y + z

Y (2.51)
y x + Y + z

z (2. 52
Z
x + Y + z


where x + y + z = 1.0. The x and y are coefficients used to

indicate chromaticity or color, while tristimulus Y is

normally kept to represent luminosity. MacAdam in 1935

[MacA35] proposed the first color measurement space with Y,

x, y cartesian coordinates.

Most current color measurement systems apply tristimulus

values rather than trichromatic coefficients. Tristimulus

values are further converted to allow for easier

interpretation of color in three-dimensional black-white,

red-green and yellow-blue space. Since the tristimulus

system was sanctioned by the International Commission on

Illumination or CIE in 1931, there have been over 30 three-

dimensional color spaces developed through transformations of

X, Y and Z values. The best scales provide an approximately











uniform color space where equal distances within the space

represent nearly equal visual color differences. The

current most commonly used color space in the world is the

CIE L*a*b* scale which was developed in 1976 (Mac96].



2.1.4c CIE L*a*b* Color Measurement Scale


Tristimulus data are converted into scales which, based

on visual discrimination experiments, correlate to perceived

color differences. Approximately uniform scales have been

developed where differences between measured colors

throughout the color space are proportional to visual

distinction.

Human eye sensitivity for detecting color differences

varies across the visible spectrum of wavelengths observed.

Visual color discrimination is greatest near 485 nm and 590

nrm and least around 425 nm and 650 nm. This is represented

in the MacAdam color limits (Y, x, y) diagram in Figure

2.14). The greater the distance between two wavelengths on

the perimeter of the plot, the greater the range of colors

that can be perceived in the interval. The third dimension

indicated in the graph is the Y-value from 0 through 95.

Note as the Y-value or lightness increases, the potential







































0.4 600



"- 770 nm


0.2





470) 380

0 450 ____ ________
0 0.2 0.4 0.6 0.8
x





Figure 2.14. Luminosity or lightness (Y) and chromaticity
(x, y) MacAdam limits for colors viewed in daylight.
[ Adapted from Bi167]




Full Text

PAGE 1

,1)/8(1&(6 2) &20326,7,21 0(/7 9,6&26,7< $1' &5<67$//,=$7,21 21 7+( &2/25 675(1*7+ $1' 67$%,/,7< 2) 08/7,2;,'( */$66 )5,7=,5&219$1$',80 3,*0(17 6<67(06 )25 &(5$0,& :+,7(:$5(6 &2$7,1*6 $33/,&$7,216 %\ '$9,' $ ($5/ $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

$&.12:/('*0(176 DP JUDWHIXO IRU WKH JXLGDQFH DQG LQVSLUDWLRQ SURYLGHG E\ 'U 'DYLG ( &ODUN FKDLUPDQ RI P\ VXSHUYLVRU\ FRPPLWWHH 0\ DVVRFLDWLRQ ZLWK 'U &ODUN KDV JUHDWO\ HQKDQFHG P\ DFDGHPLF SURIHVVLRQDO DQG SHUVRQDO JURZWK RYHU WKH SDVW ILYH \HDUV ZRXOG OLNH WR WKDQN 'U -RVHSK 6LPPRQV 'U ( 'RZ :KLWQH\ 'U -DFN 0HFKROVN\ DQG 'U 'LQHVK 6KDK IRU SDUWLFLSDWLQJ RQ P\ VXSHUYLVRU\ FRPPLWWHH 7KDQNV DOVR JR WR .ULVWLH /HLVHU 0DUN 0RRUH 5REHUW 'L)LRUL 'LDQH )RO] DQG 5HEHFFD 6FKXO] RI 'U &ODUNnV UHVHDUFK JURXS IRU WKHLU DGYLFH ,Q DGGLWLRQ ZRXOG OLNH WR DFNQRZOHGJH WKH LQGXVWULDO VXSSRUW RI WKLV UHVHDUFK ZRXOG OLNH WR WKDQN )ORULGD 7LOH ,QGXVWULHV %RE %ORQVNL .ODXV 0HLQVVHQ %UXQR %XU]DFFKLQL DQG 0DU]LD %DUUDWWLQL RI )HUUR &RUSRUDWLRQ DQG 'DQ 6ZLOHU +RQJ &KHQ DQG 3DP /XFDV RI &HUHGHF &RUSRUDWLRQ )LQDOO\ DQG PRVW LPSRUWDQWO\ DP JUDWHIXO IRU WKH SDWLHQFH DQG HQFRXUDJHPHQW RI P\ ZLIH -DFTXLH 7KLV Q

PAGE 3

UHVHDUFK HIIRUW ZRXOG QRW KDYH EHHQ SRVVLEOH ZLWKRXW KHU VXSSRUW LQ

PAGE 4

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rDrEr 0HDVXUHPHQW 6FDOH ,9

PAGE 5

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f 6FDQQLQJ (OHFWURQ 0LFURVFRS\ 6(0f DQG (QHUJ\ 'LVSHUVLYH ;5D\ 6SHFWURVFRS\ ('6f 6WDWLVWLFDO 0HWKRGV IRU 'HULYLQJ (TXDWLRQV 5(68/76 )ULW DQG 3LJPHQW 3URSHUWLHV &RORU RI )LUHG &RDWLQJV 6SHFWUDO 5HIOHFWDQFH &XUYHV 3LJPHQW $EVRUSWLRQ )DFWRUV .6f Y

PAGE 6

&RORU LQ /r Dr DQG Er 9DOXHV &RORU 6WDELOLW\ 6SHFXODU *ORVV 9LVFRVLW\ RI &RDWLQJV 'XULQJ +HDWLQJ +HDWLQJ 0LFURVFRSH ,PDJHV 'LODWRPHWULF 7J DQG 7 9LVFRVLW\ YV 7HPSHUDWXUH 'HULYHG 6WDWLVWLFDO 0RGHOV .6 YV &RDWLQJ &RPSRVLWLRQ DQG 7HPSHUDWXUH $(r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r YL

PAGE 7

6800$5< $1' &21&/86,216 =U9 3LJPHQW DQG &RORU 9DOXHV )ULW 2[LGH &RPSRVLWLRQ 9LVFRVLW\ &U\VWDOOL]DWLRQ DQG =U9 'LVVROXWLRQ )8785( :25. $33(1',&(6 $ 81,76 )25 '(6&5,%,1* /,*+7 $1' &2/25 % 7+( &$86(6 2) &2/25 & '(16,7< 3$57,&/( 6,=( $1' $33/,&$7,21 :(,*+7 '$7$ '$7$ )520 &2$7,1*6 %$7&+(' :,7+ )5,7 b %(1721,7( $1' =U9 3,*0(17 $1' ),5(' 86,1* $ 0,187( &(5$0,& 7,/( &<&/( ( )5,7 63(&75$/ 5()/(&7$1&( '$7$ $1' &859(6 $7 ($&+ 7(03(5$785( $1' 3,*0(17 /2$',1* 5()(5(1&(6 %,2*5$3+,&$/ 6.(7&+ YL

PAGE 8

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bf RI &RPPHUFLDO *OD]HV 6XPPDU\ RI ,PSRUWDQW *OD]H 3URSHUWLHV DQG &KDUDFWHULVWLFV 6HJHUnV )RUPXOD IRU &ODVVLI\LQJ *OD]HV 7\SLFDO (PSLULFDO )RUPXODV LQ 0RODU (TXLYDOHQWV IRU )DVW)LUH *ORVV DQG 0DWWH *OD]HV ,OO YLLL

PAGE 9

*OD]H 3LJPHQWV DQG 7KHLU 5HTXLUHPHQWV )ULWV ,QYHVWLJDWHG (O (QJREH DQG )ULWV $ DQG % 5HIOHFWDQFH 'DWD ( )ULWV & DQG 5HIOHFWDQFH 'DWD ( )ULWV ( DQG ) 5HIOHFWDQFH 'DWD ( )ULWV DQG + 5HIOHFWDQFH 'DWD ,;

PAGE 10

/,67 2) ),*85(6 )LJXUH 3DJH )ORZ FKDUW VXPPDU\ RI PDLQ UHVHDUFK YDULDEOHV (OHFWURPDJQHWLF VSHFWUXP :DYHOHQJWK YV HQHUJ\ GLVWULEXWLRQ RI GD\OLJKW 5HIUDFWLYH LQGH[ YV ZDYHOHQJWK RI LQFLGHQW OLJKW IRU WKUHH JODVVHV 5HIOHFWLRQ DQG WUDQVPLVVLRQ RI OLJKW E\ D JODVV\ PDWHULDO FRQWDLQLQJ VXVSHQGHG SDUWLFOHV )UHVQHO UHIOHFWLRQ IRU Df DQ DLUJODVV ERXQGDU\ DQG Ef WRWDO UHIOHFWLYLW\ IRU GLIIHUHQW LQGH[ RI UHIUDFWLRQ 5HIOHFWDQFH YV ZDYHOHQJWK RI OLJKW IRU D 72 RSDFLILHG ZKLWH JOD]H ILUHG WR r& DW YDULRXV JOD]H DSSOLFDWLRQ ZHLJKWV LQ JIW 7KH G RUELWDOV +XPDQ RSWLFDO V\VWHP /XPLQRVLW\ IXQFWLRQV RI WKH URGV QLJKWWLPH VFRWRSLF YLVLRQf DQG FRQHV GD\WLPH SKRWRSLF YLVLRQf RI WKH KXPDQ H\H 5HIOHFWDQFH YHUVXV ZDYHOHQJWK IRU RSDTXH FRDWLQJV FRORUHG ZLWK SLJPHQWV WKDW DEVRUE D SRUWLRQ RI LQFLGHQW OLJKW %DVLF FRPSRQHQWV RI VSHFWURSKRWRPHWHUV [

PAGE 11

6FKHPDWLF RI WKH +DUG\ VSHFWURSKRWRPHWHU :HLJKWLQJ IXQFWLRQV XVHG IRU WKH VWDQGDUG REVHUYHU DW D r ILHOG RI YLHZ /XPLQRVLW\ RU OLJKWQHVV
PAGE 12

7LPHWHPSHUDWXUH SURILOHV XVHG WR ILUH WKH WLOHV 6SHFWUDO UHIOHFWDQFH RI XQILUHG UDZ PDWHULDOV DQG WKH HQJREH VXEVWUDWH EDFNLQJ 6SHFWUDO UHIOHFWDQFH RI FRDWLQJV EDWFKHG ZLWK IULW & ILUHG WR &7& 6SHFWUDO UHIOHFWDQFH RI FRDWLQJV EDWFKHG ZLWK IULW + ILUHG WR r& 6SHFWUDO UHIOHFWDQFH RI FRDWLQJV EDWFKHG ZLWK b =U9 ILUHG WR r& 6SHFWUDO UHIOHFWDQFH RI FRDWLQJV EDWFKHG ZLWK b =U9 ILUHG WR r& 6SHFWUDO UHIOHFWDQFH RI FRDWLQJV EDWFKHG ZLWK b =U9 ILUHG WR R& 3LJPHQW DEVRUSWLRQ IDFWRUV YHUVXV ZHLJKW SHUFHQW =U9 EDWFKHG LQ FRDWLQJV ILUHG WR r& & DQG r& SHDN WHPSHUDWXUH &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG =U9 SLJPHQW DQG ILUHG WR r& &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG =U9 SLJPHQW DQG ILUHG WR n& &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG QR =U9 DQG ILUHG WR r& r& RU r& &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG b =U9 DQG ILUHG WR r& r& RU f& &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG b =U9 DQG ILUHG WR r& r& RU 8r& [L L

PAGE 13

&RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG b =U9 DQG ILUHG WR r& r& RU r& 5HODWLRQVKLS EHWZHHQ OLJKWQHVV /rf DQG EOXHQHVV Erf RI FRDWLQJV EDWFKHG ZLWK =U9 SLJPHQW DQG ILUHG WR r& r& DQG r& SHDN WHPSHUDWXUH =U9 SLJPHQW DEVRUSWLRQ IDFWRU UHODWLRQVKLSV ZLWK /r Dr DQG Er FRORU YDOXHV RI ILUHG FRDWLQJV &RORU FKDQJHV $(rf GXH WR D YDULDWLRQ LQ SHDN ILULQJ WHPSHUDWXUH IURP r& WR r& IRU FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG =U9 SLJPHQW 6SHFXODU JORVV RI ILUHG FRDWLQJV DW D r DQJOH RI LQFLGHQFH +HDWLQJ PLFURVFRSH LPDJHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG b =U9 DW r& r& DQG r& +HDWLQJ PLFURVFRSH LPDJHV RI FKDUDFWHULVWLF VWDJHV RI IORZ RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG b =U9 7KHUPDO GLODWRPHWULF DQDO\VHV RI FRDWLQJV $' EDWFKHG ZLWK b =U9 7KHUPDO GLODWRPHWULF DQDO\VHV RI FRDWLQJV (+ EDWFKHG ZLWK b =U9 /RJ U_ YHUVXV WHPSHUDWXUH RI FRDWLQJV EDWFKHG ZLWK b =U9 )ULWV $' LQFRUSRUDWH b =U /RJ U_ YHUVXV WHPSHUDWXUH RI FRDWLQJV EDWFKHG ZLWK b =U9 )ULWV (+ FRQWDLQ QR =U [LLL

PAGE 14

3LJPHQW DEVRUSWLRQ IDFWRU DFWXDO DQG HTXDWLRQ f SUHGLFWHG UHVXOWV IRU ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 ,QGLYLGXDO HIIHFWV RI IULW R[LGHV RQ .6 EDVHG RQ VWDWLVWLFDO PRGHO f IRU b =U9 DQG D SHDN ILULQJ WHPSHUDWXUH RI r& 9DULDEOHV ZHLJKW RI LQIOXHQFH RQ .6 EDVHG RQ VWDWLVWLFDO PRGHOV b =U9f b =U9f DQG b =U9f 3LJPHQW DEVRUSWLRQ IDFWRU WUHQGV ZLWK IULW $,2DONDOLV UDWLR RI FRDWLQJV EDWFKHG ZLWK b =U9 DQG ILUHG WR r& r& RU r& 3LJPHQW DEVRUSWLRQ IDFWRU YHUVXV IULW =Q2 DQG 6U2 PRODU HTXLYDOHQWV RI ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 DQG ILUHG WR r& r& RU r& SHDN WHPSHUDWXUH ,QIOXHQFH RI IULW =U&! LQ WKH SUHVHQFH RI 6U2 RU =Q2 RQ .6 RI ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 DQG ILUHG WR r& r& RU r& 9DULDWLRQV LQ SLJPHQW DEVRUSWLRQ IDFWRU .6f GXH WR FKDQJHV LQ SHDN ILULQJ WHPSHUDWXUH IRU FRDWLQJV EDWFKHG ZLWK b =U9 DQG IULWV FRQWDLQLQJ b =U YHUVXV b =U 'HOWD (r DFWXDO DQG HTXDWLRQ f SUHGLFWHG UHVXOWV IRU ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 ,QGLYLGXDO HIIHFWV RI IULW R[LGHV RQ 'HOWD (r EDVHG RQ VWDWLVWLFDO PRGHO IRU b =U9 9DULDEOHV ZHLJKW RI LQIOXHQFH RQ 'HOWD (r EDVHG RQ VWDWLVWLFDO PRGHOV b =U9f b =U9f DQG b =U9f [LY

PAGE 15

&RORU FKDQJHV GXH WR YDULDWLRQV LQ SHDN ILULQJ WHPSHUDWXUH DQG IULW $,2 DONDOL UDWLR RI FRDWLQJV EDWFKHG ZLWK b =U9 ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW $ DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW $ LQFOXGHV b =U&! b =Q2 b 6U2 b DONDOLV 6(0 PLFURJUDSKV RI D FRDWLQJ EDWFKHG ZLWK IULW $ DQG b =U9 DQG ILUHG WR r& 0DJQLILFDWLRQ LV Df ; DQG Ef ; 6KRZQ DUH ODUJH =U9 SDUWLFOHV VXUURXQGHG E\ ILQH ]LUFRQ SUHFLSLWDWHV ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW % DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW % LQFOXGHV b =U&! b =Q2 b 6U2 b DONDOLV 6(0 PLFURJUDSKV RI D FRDWLQJ EDWFKHG ZLWK IULW % DQG b =U9 DQG ILUHG WR r& 0DJQLILFDWLRQ LV Df ; DQG Ef ; 6KRZQ DUH ODUJH =U9 SDUWLFOHV ILQH ]LUFRQ SUHFLSLWDWHV DQG ODUJH ]LUFRQ ILEHUV ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW & DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW & LQFOXGHV b =U b =Q2 b 6U2 b DONDOLV ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW LQFOXGHV b =U b =Q2 b 6U2 b DONDOLV 6(0 PLFURJUDSKV RI FRDWLQJV EDWFKHG ZLWK b =U9 Df IULW & DQG Ef IULW DQG ILUHG WR && PDJQLILFDWLRQ ; f 3DUWLFOHV VKRZQ DUH =U9 SLJPHQW ;UD\ GLIIUDFWLRQ SURILOH IRU ]LUFRQ =U6L [Y

PAGE 16

6(0 PLFURJUDSK RI ]LUFRQYDQDGLXP =U9f SLJPHQW &HUHGHF $ ; PDJQLILFDWLRQ ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW ( DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW ( LQFOXGHV b =U&! b =Q2 b 6U2 DQG b DONDOLV 6(0 PLFURJUDSK RI D FRDWLQJ EDWFKHG ZLWK IULW ( DQG b =U9 DQG ILUHG WR r& 0DJQLILFDWLRQ LV ; 6KRZQ DUH ODUJH =U9 SDUWLFOHV VXUURXQGHG E\ GLVSHUVHG GLRSVLGH ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW ) DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW ) LQFOXGHV b =U&! b =Q2 b 6U2 DQG b DONDOLV 6(0 PLFURJUDSKV RI D FRDWLQJ EDWFKHG ZLWK IULW ) DQG b =U9 DQG ILUHG WR r& 0DJQLILFDWLRQ LV Df ; DQG Ef $OO SDUWLFOHV VKRZQ DUH KDUG\VWRQLWH ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW LQFOXGHV b =U b =Q2 b 6U2 DQG b DONDOLV 6(0 PLFURJUDSK RI D FRDWLQJ EDWFKHG ZLWK IULW DQG b =U9 ILUHG WR r& PDJQLILFDWLRQ ; f 6KRZQ LV FU\VWDOOL]HG 6U&D6J ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW + DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW + LQFOXGHV b =U b =Q2 b 6U2 DQG b DONDOLV 6(0 PLFURJUDSKV RI D FRDWLQJ EDWFKHG ZLWK IULW + DQG b =U9 DQG ILUHG WR r& 0DJQLILFDWLRQ LV Df ; DQG Ef ; 6KRZQ LV FU\VWDOOL]HG 6U&D6L [YL

PAGE 17

5HODWLRQVKLS EHWZHHQ ;5' > @ LQWHJUDWHG LQWHQVLW\ DQG ZHLJKW SHUFHQW ]LUFRQ LQ XQILUHG FRDWLQJV :HLJKW SHUFHQW ]LUFRQ LQ ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 &KDQJHV LQ UHIOHFWDQFH GLVWULEXWLRQV DW QP DQG QP ZDYHOHQJWKV GXH WR LQFUHDVHV LQ SHDN ILULQJ WHPSHUDWXUH IRU FRDWLQJV EDWFKHG ZLWK b b DQG b =U9 ,QIOXHQFH RI IULW GHQVLW\ RQ WKH FRORU VWDELOLW\ RI FRDWLQJV EDWFKHG ZLWK b =U9 SLJPHQW 3LJPHQW DEVRUSWLRQ IDFWRUV .6f IRU FRDWLQJV EDWFKHG ZLWK =U9 SLJPHQW DQG IULWV FRQWDLQLQJ b =U DQG ILUHG WR r& r& RU r& SHDN WHPSHUDWXUH 9LVXDO OLJKWQHVV /rf JUHHQQHVV Drf DQG EOXHQHVV Erf DV D IXQFWLRQ RI ZHLJKW SHUFHQW ]LUFRQ LQ WKH ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 3LJPHQW DEVRUSWLRQ IDFWRU .6f DV D IXQFWLRQ RI =U9 SLJPHQW DQG ]LUFRQ FRQWHQWV LQ FRDWLQJV EDWFKHG ZLWK b =U9 &KDQJHV LQ FRORU VWDELOLW\ GHQRWHG E\ .6 DQG 'HOWD (r DVD IXQFWLRQ RI ZHLJKW SHUFHQW ]LUFRQ LQ FRDWLQJV EDWFKHG ZLWK b =U9 &RORU OLJKWQHVV /rf DQG EOXHQHVV Erf SURJUHVVLRQ ZLWK GLRSVLGH FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ LQ FRDWLQJV EDWFKHG ZLWK IULW ( DQG b =U9 &RORU OLJKWQHVV /rf DQG EOXHQHVV f§Erf SURJUHVVLRQ ZLWK KDUG\VWRQLWH FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ LQ FRDWLQJV EDWFKHG ZLWK IULW ) DQG b =U9 [YL

PAGE 18

&RORU OLJKWQHVV /rf DQG EOXHQHVV Erf SURJUHVVLRQ ZLWK 6U&D6J FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ LQ FRDWLQJV EDWFKHG ZLWK b =U9 ,QWHJUDWHG ORJ YLVFRVLW\ IURP r& WR r& YHUVXV ILUHG FRORU VWUHQJWK .6f DQG VWDELOLW\ 'HOWD (rf LQ FRDWLQJV EDWFKHG ZLWK b =U9 6ORSH LQ WKH ORJ YLVFRVLW\ YHUVXV WHPSHUDWXUH QHDU WKH VRIWHQLQJ SRLQW YHUVXV ILUHG FRORU VWUHQJWK .6f DQG VWDELOLW\ 'HOWD (rf LQ FRDWLQJV EDWFKHG ZLWK b =U9 (O )ULW $ VSHFWUDO UHIOHFWDQFH FXUYHV ( )ULW % VSHFWUDO UHIOHFWDQFH FXUYHV ( )ULW & VSHFWUDO UHIOHFWDQFH FXUYHV ( )ULW VSHFWUDO UHIOHFWDQFH FXUYHV ( )ULW ( VSHFWUDO UHIOHFWDQFH FXUYHV ( )ULW ) VSHFWUDO UHIOHFWDQFH FXUYHV ( )ULW VSHFWUDO UHIOHFWDQFH FXUYHV ( )ULW + VSHFWUDO UHIOHFWDQFH FXUYHV [YLLL

PAGE 19

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f EOXH SLJPHQWV 7KH UHVXOWV ZRXOG DOVR EH DSSOLFDEOH WR RWKHU FHUDPLF SLJPHQWV WKDW XWLOL]H WKH VDPH ]LUFRQ VWUXFWXUH WR LQFRUSRUDWH [L[

PAGE 20

FRORUDQW PHWDO LRQV $ VHFRQGDU\ JRDO ZDV WR UHODWH WKH IULW R[LGH FRPSRVLWLRQ SLJPHQW FRQWHQW ILULQJ WHPSHUDWXUH PHOW YLVFRVLW\ DQG PLFURVWUXFWXUDO GHYHORSPHQW WR WKH ILUHG FRORU $ FHUDPLF WLOH SURFHVV ZDV DSSOLHG WR IDEULFDWH VDPSOH FRDWLQJV IRU WKH VWXG\ $ FRDWLQJnV FRORU ZDV TXDQWLILHG DQG UHODWHG WR KXPDQ SHUFHSWLRQ ZLWK &,( /r Dr DQG Er YDOXHV DQG SLJPHQW DEVRUSWLRQ IDFWRUV .6f FDOFXODWHG EDVHG RQ VSHFWUDO UHIOHFWDQFH GDWD 7KH UHVHDUFK ZDV VXFFHVVIXO LQ TXDQWLI\LQJ WKH LQIOXHQFH RI LQGLYLGXDO JODVV IULW R[LGHV RQ WKH ILUHG FRORU VWUHQJWK DQG KLJKWHPSHUDWXUH VWDELOLW\ RI WKH FRDWLQJV 2SDTXH DQG WUDQVSDUHQW JORVV\ IULW FRPSRVLWLRQV ZKLFK \LHOG H[FHOOHQW FRORU VWUHQJWK DQG VWDELOLW\ ZHUH IRUPXODWHG 0DWKHPDWLFDO PRGHOV IRU SUHGLFWLQJ D FRDWLQJnV FRORU VWUHQJWK DQG VWDELOLW\ JLYHQ WKH IULW R[LGH FRPSRVLWLRQ =U9 SLJPHQW ORDGLQJ DQG SHDN ILULQJ WHPSHUDWXUH ZHUH GHULYHG )ULW R[LGHV RI =U 6U2 =Q2 $ 1D DQG ZHUH IRXQG WR KDYH D VLJQLILFDQW LQIOXHQFH RQ FU\VWDOOL]DWLRQ SLJPHQW GLVVROXWLRQ DQG FRORU GHYHORSPHQW 7KH SURSHUWLHV VL]HV PRUSKRORJLHV DQG TXDQWLWLHV RI FU\VWDOOLQH SKDVHV WKDW SUHFLSLWDWHG LQ WKH FRDWLQJV GXULQJ ILULQJ ZHUH UHODWHG WR WKH FRORU $ WHFKQLTXH IRU SURGXFLQJ XQLTXHO\ OLJKW \HW KLJK [[

PAGE 21

FKURPD FRORUV WKURXJK FRQWURO RI ]LUFRQ SUHFLSLWDWH SDUWLFOH VL]H ZDV GHPRQVWUDWHG ,Q DGGLWLRQ D VWDWLVWLFDO PRGHO ZDV GHYHORSHG IRU FDOFXODWLQJ WKH FRDWLQJ PHOW YLVFRVLW\ DV D IXQFWLRQ RI WKH IULW R[LGH FRPSRVLWLRQ DQG WHPSHUDWXUH 0HOW YLVFRVLW\ YHUVXV WHPSHUDWXUH FXUYHV ZHUH DSSOLHG WR SUHGLFW D IULWnV SRWHQWLDO IRU SURGXFLQJ VWURQJ DQG VWDEOH FRORU [[L

PAGE 22

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

PAGE 23

YDULDWLRQV LQ ILUHG FRDWLQJV GXH WR QRUPDO IOXFWXDWLRQV LQ PDQXIDFWXULQJ SURFHVVLQJ FRQGLWLRQV UHVXOW LQ QRQVWDQGDUG VFUDS SURGXFWV DQG ORZHU SURGXFWLYLW\ )RU H[DPSOH D PDMRU SRUWLRQ RI WKH PLOOLRQ ZRUWK RI SURGXFWV VFUDSSHG LQ E\ 86 FHUDPLF WLOH PDQXIDFWXUHUV UHVXOWHG IURP RXWRI WROHUDQFH FRORUV 6LQFH WKH 86 FHUDPLF WLOH PDUNHW RQO\ FRPSULVHV DERXW b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f FRPSOLFDWH WKH IRUPDWLRQ RI

PAGE 24

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

PAGE 25

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b LQ WKH PLGV WR RYHU b LQ WKH V > (DU@ *OD]H UDZ PDWHULDO FRVWV FRPSULVH URXJKO\ b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

PAGE 26

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

PAGE 27

6ROXWLRQ FRORUDQWV DUH VRPHWLPHV LQWURGXFHG LQWR WKH JOD]H EDWFK DV R[LGHV WKHQ GLVVROYHG DQG SUHFLSLWDWHG DV PHWDO LRQV GXULQJ WKH ILULQJ SURFHVV 2[LGHV VXFK DV &U JUHHQf &R2 EOXHf &X2 JUHHQ WR UHGf )H \HOORZ WR EURZQf DQG 0Q2 SXUSOH WR EURZQf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

PAGE 28

KLJK WHPSHUDWXUH WKDQ RWKHU FRORUDQWV FRORU YDULDWLRQV VWLOO RFFXU 'XULQJ ILULQJ ]LUFRQ PD\ UHPDLQ VWDEOH DQG SURWHFW WKH PHWDO LRQV SDUWLDOO\ GLVVROYH ZLWK WKDW SRUWLRQ UHSUHFLSLWDWLQJ RU UHPDLQLQJ LQ VROXWLRQ RU WRWDOO\ GLVVROYH &RORU FRQWURO LV IXUWKHU FRPSOLFDWHG EHFDXVH PRVW GHFRUDWHG FHUDPLF ZKLWHZDUHV FRQWDLQ PXOWLSOH OD\HUV RI FRDWLQJV )RU H[DPSOH D W\SLFDO VWRQH ORRN IORRU WLOH SURGXFHG E\ )ORULGD 7LOH ,QGXVWULHV UHTXLUHV QLQH GLIIHUHQW FRDWLQJV RYHU WKH SUHVVHG ERG\ 7DEOH f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

PAGE 29

7DEOH ([DPSOH RI &HUDPLF &RDWLQJV $SSOLFDWLRQV RQ D 'HFRUDWHG )ORRU 7LOH *OD]H &RDWLQJ &RORU2SDFLW\ $SSOLFDWLRQ 0HWKRG $SSOLFDWLRQ :HLJKW JFPf (QJREH SULPHU FRDWf :KLWH2SDTXH 5RWDWLQJ GLVN )XPH 7DXSH2SDJXH 6SUD\ JXQ f§ f§ %UXVKLQJ PDFKLQH f§ 0LGGOH FRDW 2IIZKLWH 7UDQVOXFHQW 5RWDWLQJ GLVN )XPH *UH\2SDTXH 6SUD\ JXQ ,QN /LJKW *UH\ 2SDTXH 6FUHHQ SULQWHU ,QN *ROG2SDTXH 6FUHHQ SULQWHU ,QN /LJKW %HLJH 2SDTXH 6FUHHQ SULQWHU ,QN :KLWH2SDTXH 6FUHHQ SULQWHU 7RSFRDW 7UDQVSDUHQW 6SUD\ JXQ

PAGE 30

7DEOH 9DULDEOHV 7KDW ,QIOXHQFH &HUDPLF *OD]H &RORU %DWFK &RPSRVLWLRQ D %DVH LQJUHGLHQWV R[LGH FRPSRVLWLRQ DQG SKDVHV SUHVHQWf E 3LJPHQWV FRPSRVLWLRQ VWUXFWXUH DQG ORDGLQJf F &KHPLFDO DGGLWLYHV LQIOXHQFH DSSOLFDWLRQ GU\LQJ UDWH DQG VPRRWKQHVVf ,, *OD]H 3UHSDUDWLRQ ZLWK %DOO 0LOOLQJ D3DUWLFOH VL]H GLVWULEXWLRQ LQIOXHQFHV PHOWLQJ SRLQWf ,,, $SSOLFDWLRQ D 7KLFNQHVV KLGLQJ SRZHUf E 'U\LQJ UDWH FDQ DIIHFW FRPSRVLWLRQ JUDGLHQW FDXVHG E\ GLIIHUHQFHV LQ SDUWLFOH VHWWOLQJ UDWHVf F 6PRRWKQHVV LQIOXHQFHV JORVV RU GHJUHH RI VSHFXODU UHIOHFWDQFHf ,9 )LULQJ D 7LPH YV WHPSHUDWXUH SKDVH GLVVROXWLRQ DQG SUHFLSLWDWLRQf E .LOQ DWPRVSKHUH R[LGDWLRQUHGXFWLRQ UHDFWLRQV VXOIXU VFXPPLQJ HWFf

PAGE 31

LPSURYHG VORZILUH JOD]H FRPSRVLWLRQV PD\ UHGXFH FRORU FKDQJHV UHVXOWLQJ IURP YDULDWLRQV LQ JOD]H SUHSDUDWLRQ DSSOLFDWLRQ DQG ILULQJ FRQGLWLRQV +RZHYHU WKHUH LV QR SXEOLVKHG FRPSUHKHQVLYH RU TXDQWLWDWLYH UHVHDUFK RQ WKH VXEMHFW 7KHUH LV DOVR D JUHDW ODFN RI UHVHDUFK RQ IDVWILUH JOD]H V\VWHPV )ULWV DUH FHUDPLF FRPSRVLWLRQV WKDW KDYH EHHQ IXVHG TXHQFKHG WR IRUP D JODVV DQG JUDQXODWHG > 'RG@ 7KH\ DUH WKH SULPDU\ LQJUHGLHQWV LQ IDVWILUH FHUDPLF JOD]HV DQG LQ PRVW FDVHV DUH WKH PRVW UHDFWLYH DQG FRUURVLYH SDUW RI WKH IRUPXOD )ULWV IRU ZKLWHZDUHV FRDWLQJV DUH FODVVLILHG DV HLWKHU RSDTXH RSDFLILHG FRQWDLQLQJ =URf RU WUDQVSDUHQW XQRSDFLILHG QR =Uf DQG JORVV\ RU PDWWH %HVLGHV =U IULWV DOVR QRUPDOO\ HPSOR\ 6L DV WKH SULPDU\ JODVV IRUPHU DONDOLV . DQG 1D&f % DQG =Q& RU 6U2 DV WKH PDLQ IOX[HV DQG RWKHU R[LGHV VXFK DV &D2 $ DQG 0J2 7KHVH R[LGHV DUH FRVW HIIHFWLYH HQYLURQPHQWDOO\ VDIH DQG SURYLGH WKH GHVLUHG JOD]H SURSHUWLHV &RPSRVLWLRQV ZLWK %D2 RU 3E2 DUH DYRLGHG EHFDXVH WKHVH HOHPHQWV DUH GHHPHG KD]DUGRXV E\ WKH (3$ 7KHUH LV DOVR DQ LQFUHDVLQJ LQWHUHVW LQ UHSODFLQJ =Q2 ZLWK 6U2 EHFDXVH =Q2 LV FODVVLILHG DV D UHJXODWHG FKHPLFDO E\ (3$ 6HFWLRQV DQG GHWDLO WKH

PAGE 32

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

PAGE 33

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nV FRORU &RUUHODWH WKH HYROXWLRQ RI WKH FRDWLQJnV VWUXFWXUH DQG SURSHUWLHV WR WKH RULJLQDO IULW R[LGH FRPSRVLWLRQ DQG WKH ILUHG FRORU

PAGE 34

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f EOXH SLJPHQW 2[LGH FRPSRVLWLRQV RI WKH IULWV ZHUH GHVLJQHG WR SURYLGH FRVW HIIHFWLYH HQYLURQPHQWDOO\ VDIH IRUPXODV DQG FRPSO\ ZLWK 6HJHUnV UXOHV 6HFWLRQ f IRU HQVXULQJ LQVROXELOLW\ RI WKH IULW DQG ILUHG FRDWLQJ DQG UHDG\ IXVLRQ DW KLJK WHPSHUDWXUH 7KH UDQJH RI R[LGH FRQWHQWV WHVWHG HQFRPSDVVHG DQG H[FHHGHG WKH UDQJH QRUPDOO\ HPSOR\HG IRU JORVV\ FHUDPLF WLOH JOD]HV 6SHFLDO HPSKDVLV ZDV SODFHG RQ FRPSDULQJ IULW FRPSRVLWLRQV ZLWK =U RSDFLILHGf YHUVXV QR =U XQRSDFLILHGf 6U2 YHUVXV =Q2 DV WKH VHFRQGDU\ IOX[ DQG DONDOLVLOLFD UDWLRV 7KH % FRQWHQWV ZHUH NHSW ORZ LQ RUGHU WR DYRLG OLTXLGOLTXLG SKDVH VHSDUDWLRQ 7KH =U9 SLJPHQW WHVWHG ZDV WKH EOXH FRORUDQW PRVW FRPPRQO\ XVHG LQ WKH FHUDPLF WLOH LQGXVWU\ 7KH UHVXOWV ZLOO DOVR EH

PAGE 35

DSSOLFDEOH WR RWKHU FHUDPLF SLJPHQWV ZKLFK XWLOL]H WKH VDPH ]LUFRQ VWUXFWXUH WR LQFRUSRUDWH FRORUDQW PHWDO LRQV 5HVHDUFK JRDOV ZHUH DFKLHYHG E\ SHUIRUPLQJ WKH IROORZLQJ WDVNV (DFK RI WKH HLJKW H[SHULPHQWDO IULWV ZHUH ORDGHG ZLWK IRXU GLIIHUHQW SLJPHQW FRQFHQWUDWLRQV RI b b b DQG b E\ ZHLJKW 7KH\ ZHUH EOHQGHG ZLWK ZDWHU DQG D VXVSHQGLQJ DJHQW WR SURGXFH GLIIHUHQW JOD]H FRDWLQJV &RDWLQJV ZHUH DSSOLHG WR RSDTXH ; ZDOO WLOH ERG\ VXEVWUDWHV XVLQJ D ZHW VSUD\ PHWKRG 6DPSOHV RI HDFK IRUPXOD ZHUH ILUHG WR r& r& DQG r& SHDN WHPSHUDWXUHV XVLQJ D VWDQGDUG IDVWILUH FHUDPLF WLOH LQGXVWULDO KHDWLQJ SURILOH $ WRWDO RI GLIIHUHQW ILUHG FRDWLQJV ZHUH SURGXFHG 7KH VSHFWUDO UHIOHFWDQFH YHUVXV ZDYHOHQJWK DQG WKH JORVV RI HDFK ILUHG FRDWLQJ ZHUH PHDVXUHG 7KH &,( /r Dr Er FRORU YDOXHV SLJPHQW DEVRUSWLRQ IDFWRUV .6f DQG FRORU GLIIHUHQFHV EHWZHHQ WLOHV ILUHG WR r& DQG r& $(rf ZHUH FDOFXODWHG 5HODWLRQVKLSV EHWZHHQ OLJKW DEVRUSWLRQ E\ WKH SLJPHQW DQG FRORU YDOXHV EDVHG RQ KXPDQ SHUFHSWLRQ ZHUH TXDQWLILHG

PAGE 36

'LODWRPHWU\ DQG KHDWLQJ PLFURVFRS\ PHWKRGV ZHUH HPSOR\HG WR PHDVXUH WKH 7J 7V DQG PHOW YLVFRVLW\ YHUVXV WHPSHUDWXUH RI FRDWLQJV EDWFKHG ZLWK b =U9 ,Q FRDWLQJV EDWFKHG ZLWK b =U9 SKDVH FKDQJHV DQG UHVXOWLQJ PLFURVWUXFWXUHV WKDW IRUPHG GXULQJ ILULQJ ZHUH LGHQWLILHG XVLQJ [UD\ GLIIUDFWLRQ VFDQQLQJ HOHFWURQ PLFURVFRS\ DQG HQHUJ\ GLVSHUVLYH [UD\ VSHFWURVFRS\ WHFKQLTXHV )LUHG FRDWLQJV ZHUH TXDQWLWDWLYHO\ DQDO\]HG IRU FRQWHQWV RI =U9 SLJPHQW DQG ]LUFRQ ZKLFK SUHFLSLWDWHG IURP =U DQG 6L LQ WKH IULW 6WDWLVWLFDO PRGHOV ZHUH GHULYHG WR SUHGLFW .6 DQG $(r JLYHQ WKH RULJLQDO IULW R[LGH FRPSRVLWLRQ SLJPHQW ORDGLQJ DQG SHDN ILULQJ WHPSHUDWXUH $Q HTXDWLRQ ZDV DOVR GHYHORSHG IRU FDOFXODWLQJ ORJ U? RI WKH IULW ZLWK b =U9 JLYHQ WKH IULW R[LGH FRPSRVLWLRQ DQG WHPSHUDWXUH &RORU VWUHQJWK DQG VWDELOLW\ ZHUH FRUUHODWHG WR PHOW YLVFRVLW\ FU\VWDOOL]DWLRQ DQG =U9 SLJPHQW GLVVROXWLRQ 7KH PDWKHPDWLFDO PRGHOV DQG H[SHULPHQWDO REVHUYDWLRQV ZHUH UHODWHG WR FXUUHQW VFLHQWLILF OLWHUDWXUH LQ RUGHU WR FROODWH K\SRWKHVHV ZKLFK H[SODLQ WKH UHVXOWV

PAGE 37

7KH IRXQGDWLRQ RI PDWHULDOV VFLHQFH DQG HQJLQHHULQJ UHVHDUFK LV WR JDLQ D EHWWHU XQGHUVWDQGLQJ RI UHODWLRQVKLSV EHWZHHQ SURFHVVLQJ VWUXFWXUH DQG SURSHUWLHV RI PDWHULDOV )LJXUH VXPPDUL]HV FULWLFDO VWHSV WDNHQ GXULQJ WKLV UHVHDUFK WR GHILQH WKH SURFHVVLQJVWUXFWXUHSURSHUWLHV UHODWLRQVKLSV RI LQWHUHVW 7KLV LQYHVWLJDWLRQ IRFXVHG PDLQO\ RQ YDULDEOHV WKDW LQIOXHQFH D FHUDPLF FRDWLQJnV FRORU VWUHQJWK DQG VWDELOLW\ *XLGH IRU 8VLQJ 7KLV 'LVVHUWDWLRQ 7KH SUHVHQW GRFXPHQW LV JUHDWHU LQ OHQJWK WKDQ PRVW GLVVHUWDWLRQV 7KH SULPDU\ REMHFWLYH ZDV QRW RQO\ WR XQYHLO YDOXDEOH LQIRUPDWLRQ IRU EDVLF VFLHQFH LQWHUHVWV EXW DOVR WR FRPSLOH D WH[W WKDW FRXOG EH XVHG DV D UHIHUHQFH E\ HQJLQHHUV ZRUNLQJ LQ LQGXVWU\ 7KXV VRPH VHFWLRQV PD\ EH E\SDVVHG LI RQO\ YHU\ VSHFLILF LQIRUPDWLRQ LV GHVLUHG ,Q WKH %DFNJURXQG FKDSWHU 6HFWLRQ RYHUYLHZV FXUUHQW VFLHQWLILF SULQFLSOHV EHKLQG Df PDWHULDOV LQWHUDFWLRQV ZLWK OLJKW ZLWK D IRFXV RQ FHUDPLFV Ef FRORU SHUFHSWLRQ E\ WKH KXPDQ H\H DQG Ff WKH PRVW FRPPRQ LQGXVWULDO PHWKRG IRU TXDQWLI\LQJ FRORU DQG FRUUHODWLQJ LW WR KXPDQ YLVLRQ 6HFWLRQ VXPPDUL]HV FXUUHQW NQRZOHGJH RI WKH PDWHULDOV SURFHVVLQJ VWUXFWXUHV DQG SURSHUWLHV RI FHUDPLF ZKLWHZDUHV

PAGE 38

)LJXUH )ORZ FKDUW VXPPDU\ RI PDLQ UHVHDUFK YDULDEOHV

PAGE 39

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r Dr Er .6 DQG $(r LRRQRRrFr 6HFWLRQ UHYHDOV WKH

PAGE 40

YLVFRVLW\ YHUVXV WHPSHUDWXUH SURILOHV KHDWLQJ PLFURVFRSH LPDJHV DQG GLODWRPHWULF GDWD IRU FRDWLQJV ORDGHG ZLWK b =U9 6WDWLVWLFDO PRGHOV IRU SUHGLFWLQJ FRORU VWUHQJWK .n6f FRORU VWDELOLW\ $( BQ&_2&f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

PAGE 41

,Q &KDSWHU WKH 'LVFXVVLRQ UHODWLRQVKLSV DUH HVWDEOLVKHG EHWZHHQ FRORU VWUHQJWK DQG VWDELOLW\ VSHFLILF R[LGHV LQ WKH JODVV IULW =U9 ORDGLQJ FU\VWDOOL]DWLRQ =U9 GLVVROXWLRQ DQG PHOW YLVFRVLW\ 6FLHQWLILF H[SODQDWLRQV IRU REVHUYHG SKHQRPHQD DUH SURSRVHG 7KH 'LVFXVVLRQ IRFXVHV RQ EDVLF VFLHQFH LQWHUHVWV DOWKRXJK 6HFWLRQ LV DOVR QRWHZRUWK\ IRU HQJLQHHUV LQ LQGXVWU\ ,W FRUUHODWHV LPSRUWDQW FKDUDFWHULVWLFV RI D FRDWLQJnV ORJ U? YHUVXV WHPSHUDWXUH SORW WR WKH FRORU VWUHQJWK DQG VWDELOLW\

PAGE 42

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f 7KXV LQ RUGHU WR XQGHUVWDQG FRORU ZH PXVW FRQVLGHU WKH WKUHH HOHPHQWV LQYROYHG ZLWK WKH SURGXFWLRQ RI FRORU WKH OLJKW VRXUFHWKH SK\VLFDO PRGLILFDWLRQV RI OLJKW E\ PDWWHU DQG WKH KXPDQ H\H DV D FRORU VHQVRU 9DULRXV XQLWV W\SLFDOO\ DSSOLHG WR TXDQWLI\ OLJKW DQG FRORU DUH OLVWHG LQ $SSHQGL[ $ DORQJ ZLWK HVVHQWLDO FRQYHUVLRQ IDFWRUV

PAGE 43

/LJKW DQG WKH 9LVLEOH 6SHFWUXP /LJKW LV WKH YLVLEOH UDGLDQW HQHUJ\ ZKLFK LQWHUDFWV ZLWK PDWWHU WR SURGXFH ZKDW RXU H\HV GHWHFW DV FRORU ,W LV DQ HOHFWURPDJQHWLF ZDYH WKDW SURSDJDWHV DV HOHFWULF DQG PDJQHWLF ILHOGV 0D[ZHOOnV WKHRU\ > +DO@ SUHGLFWHG WKH H[LVWHQFH RI D VSHFWUXP RI HOHFWURPDJQHWLF ZDYHV GLIIHULQJ RQO\ LQ ZDYHOHQJWK DQG WUDYHOLQJ WKURXJK VSDFH LQ D YDFXXP ZLWK D VSHHG RI F [ PV (OHFWURPDJQHWLF ZDYHV RWKHU WKDQ OLJKW LQFOXGH UDGLR ZDYHV PLFURZDYHV DQG [UD\V 7KH HQHUJ\ RI D OLJKW ZDYH LV TXDQWL]HG LQWR VPDOO EXQGOHV FDOOHG SKRWRQV $FFRUGLQJ WR (LQVWHLQ WKH HQHUJ\ (f RI D SKRWRQ LV > 7LS@ ( KI K f§ $ f ZKHUH K [ f -VHF LV 3ODQFNnV FRQVWDQW I LV LWV IUHTXHQF\ DQG $ WKH ZDYHOHQJWK 2WKHU SKRWRQLF UHODWLRQVKLSV IUHTXHQWO\ DSSOLHG DUH > +XP@ 3 ( PH f§ f P S PH f $S K f ZKHUH P LV WKH PDVV RI D SDUWLFOH DQG S LV LWV PRPHQWXP 7KHVH HTXDWLRQV DOORZ XV WR FRQWHPSODWH D SKRWRQ RU OLJKW DV

PAGE 44

HLWKHU D SDUWLFOH RI HQHUJ\ ( RU D ZDYH ZLWK D FKDUDFWHULVWLF IUHTXHQF\ DQG ZDYHOHQJWK 7KH ZDYHSDUWLFOH GXDOLW\ RI OLJKW FDQ EH GHVFULEHG PDWKHPDWLFDOO\ E\ FRQVLGHULQJ WZR KDUPRQLF ZDYHV ZLWK VOLJKWO\ GLIIHUHQW IUHTXHQFLHV ZKLFK FRQWDLQ WLPH DQG VSDFH GHSHQGHQW FRPSRQHQWV > +XP@ n)W VLQ > N[FRW@ f DQG n) VLQ > N$Nf[ FR$FRfW@ f Q ZKHUH N WKH ZDYH QXPEHU f ; Ne DQJXODU IUHTXHQF\ f 6XSHUSRVLWLRQ RI 9 DQG ?_ DQG FRQVLGHULQJ VLQ D VLQ FRV DIf rVLQD3f \LHOG D QHZ ZDYH > +XP@ 9L A $FR $N $N $FR FRV W W f VLQ N [ &2 W ,Q f LI $FR DQG $N D PRQRFKURPDWLF ZDYH UHVXOWV RI WKH IRUP ??L VLQ N[FRWf f (TXDWLRQ f LOOXVWUDWHV WKH ZDYH FKDUDFWHULVWLFV RI SKRWRQV

PAGE 45

,I $JR DQG$N DUH YHU\ ODUJH WKH FRVLQH SDUW RI f PRGXODWHV WKH DPSOLWXGH RI WKH ZDYH UHVXOWLQJ LQ D VWULQJ RI ZDYH SDFNHWV ,I PDQ\ ZDYHV DUH FRQVLGHUHG ZLWK IUHTXHQFLHV UDQJLQJ EHWZHHQ JR DQG $JR RQH ZDYH SDFNHW UHVXOWV DQG WKH SKRWRQ FDQ EH GHSLFWHG DV D SDUWLFOH $ EHWWHU LQWXLWLYH XQGHUVWDQGLQJ RI PDWHULDOV LQWHUDFWLRQV ZLWK OLJKW FDQ EH DFKLHYHG E\ QRWLQJ OLJKWnV ZDYH DQG SDUWLFOH FKDUDFWHULVWLFV 7KH FRORU RI OLJKW LV UHODWHG WR LWV HQHUJ\ DQG WKXV LWV IUHTXHQF\ RU ZDYHOHQJWK $V VKRZQ RQ WKH HOHFWURPDJQHWLF VSHFWUXP LQ )LJXUH > +XQ@ YLVLEOH OLJKW IDOOV LQ D UDQJH RI QP WR QP LQ ZDYHOHQJWK :KLWH OLJKW LV FRPSULVHG RI DOO WKH YLVLEOH ZDYHOHQJWKV /LJKW FDQ EH SURGXFHG E\ KHDWLQJ REMHFWV WR LQFDQGHVFHQFH RU E\ H[FLWLQJ DWRPV RU PROHFXOHV ZLWK RWKHU IRUPV RI HQHUJ\ ,QFDQGHVFHQW VRXUFHV DUH DSSOLHG WR SURGXFH OLJKW ZLWK D ZDYHOHQJWK HQHUJ\ GLVWULEXWLRQ VLPLODU WR GD\OLJKW )LJXUH f > %LO@ :KHQ D PDWHULDO LV KHDWHG WR LQFDQGHVFHQFH WKH LQFUHDVHG YLEUDWLRQ RI LWV DWRPV UHVXOWV LQ NLQHWLF HQHUJ\ WKDW LV VXIILFLHQW WR H[FLWH HOHFWURQV WR KLJKHU HQHUJ\ OHYHOV 3KRWRQV DUH UHOHDVHG ZKHQ WKH HOHFWURQV GURS EDFN WR WKHLU QRUPDO HQHUJ\ OHYHOV $V DWRP YLEUDWLRQV EHFRPH PRUH HQHUJHWLF WKH IUHTXHQF\ DQG HQHUJ\ RI HPLWWHG

PAGE 46

)5(48(1&< ,1 &<&/(6 3(5 6(&21' f§U f§ *$00$ 5$<6 ; 5$<6 +$5' 62)7 9,2/(7 -B 9$&880 8 9 8/75$9,2/(7 ,1)5$5(' +(57=,$1 :$9(6 1 1($5 )$5 , /BB /,*+7 a BB %/8( *5((1 <(//2: 5(' :$9(/(1*7+ ,1 1$120(7(56 2n / r -r V" L ',5(&7,21$/ 5$',2 5$'$5, )0 7(/(9,6,21 6+257 :$9( %52$'&$67 &" r n :$9(/(1*7+ ,1 0(7(56 f n 32:(5 75$160,66,21 f f )LJXUH (OHFWURPDJQHWLF VSHFWUXP > $GDSWHG IURP +XQ@

PAGE 47

)LJXUH :DYHOHQJWK YV HQHUJ\ GLVWULEXWLRQ RI GD\OLJKW >$GDSWHG IURP %@ 1RWH 5HODWLYH HQHUJ\ GLVWULEXWLRQ SORWV VHW WKH HQHUJ\ DW QP HTXDO WR DQG WKH UHVW RI WKH FXUYH LV UHODWLYH WR WKH GLVWULEXWLRQ UDGLDWHG IURP WKH VRXUFHf

PAGE 48

OLJKW LQFUHDVHV 7KH FRORU RI OLJKW SURGXFHG FKDQJHV IURP UHG DW ORZ WHPSHUDWXUHV WR QHDUO\ ZKLWH DW KLJKHU WHPSHUDWXUHV 7KH FRUUHODWHG FRORU WHPSHUDWXUH RI DQ LQFDQGHVFHQW VRXUFH LV GHILQHG E\ WKH WHPSHUDWXUH DW ZKLFK D EODFN ERG\ ZRXOG RSHUDWH WR SURGXFH D YLVXDO FRORU PDWFK ZLWK WKH LQFDQGHVFHQW VRXUFH > +XQ@ 7KH FRORU RI D UHDO EODFN ERG\ GHSHQGV RQO\ RQ LWV WHPSHUDWXUH QRW LWV FRPSRVLWLRQ $FFRUGLQJ WR :LHQnV GLVSODFHPHQW ODZ > :HD@ WKH SURGXFW RI WKH DEVROXWH WHPSHUDWXUH RI D UDGLDWLQJ EODFN ERG\ DQG WKH ZDYHOHQJWK FRUUHVSRQGLQJ WR WKH PD[LPXP HQHUJ\ LV D FRQVWDQW 7 : f ZKHUH : :LHQn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

PAGE 49

PDWHULDOV > $VN@ ,Q OXPLQHVFHQFH NLQHWLF KHDW HQHUJ\ LV QRW HVVHQWLDO IRU WKH PHFKDQLVP RI H[FLWDWLRQ /XPLQHVFHQFH RFFXUV ZKHQ OLJKW KDV VXIILFLHQW HQHUJ\ WR H[FLWH YDOHQFH EDQG HOHFWURQV WKURXJK WKH HQHUJ\ JDS DQG LQWR WKH FRQGXFWLRQ EDQG :KHQ WKH HOHFWURQV HYHQWXDOO\ IDOO EDFN WR WKH YDOHQFH EDQG SKRWRQV DUH HPLWWHG ,I WKH SKRWRQ HQHUJ\ FRUUHVSRQGV WR ZDYHOHQJWKV EHWZHHQ QP QP YLVLEOH OLJKW LV SURGXFHG ,Q IOXRUHVFHQW PDWHULDOV SKRWRQ ZDYHOHQJWKV FDQ EH FDOFXODWHG ZLWK KH ; f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

PAGE 50

R LQ FRPELQDWLRQ DV LOOXPLQDWHV LQ FRORU PHDVXUHPHQW GHYLFHV VXFK DV VSHFWURSKRWRPHWHUV > %@ 0DWHULDOV ,QWHUDFWLRQV ZLWK /LJKW 3URSHUWLHV RI D PDWHULDO WKDW LQIOXHQFH LWV DSSHDUDQFH LQFOXGH LQGH[ RI UHIUDFWLRQ JORVV WUDQVOXFHQF\ UHIOHFWLYLW\ WUDQVSDUHQF\ DEVRUSWLRQ DQG FRORU 7KHVH SURSHUWLHV UHVXOW SULPDULO\ IURP WKH LQWHUDFWLRQ EHWZHHQ OLJKW DQG D PDWHULDOnV HOHFWURQLF VWUXFWXUH DQG PLFURVWUXFWXUH :KHQ SKRWRQV LQWHUDFW ZLWK D PDWHULDO WKH\ DUH HLWKHU DWWHQXDWHG UHIOHFWHG RU DEVRUEHGf RU WUDQVPLWWHG $V SKRWRQV HQWHU D PDWHULDO WKHLU VSHHG ZLOO DOVR FKDQJH UHVXOWLQJ LQ UHIUDFWLRQ /LJKW PD\ EH SDUWLDOO\ UHIOHFWHG DEVRUEHG RU WUDQVPLWWHG DV UHODWHG WR WKH LQFRPLQJ EHDP LQWHQVLW\ > ,4@ E\ > $VN@ ;U D f ZKHUH ,U ,S DQG ,W DUH SRUWLRQV RI WKH LQFLGHQW EHDP WKDW LV UHIOHFWHG DEVRUEHG DQG WUDQVPLWWHG UHVSHFWLYHO\ 7KH FRPELQDWLRQ RI ZDYHOHQJWKV RI OLJKW UHIOHFWHG IURP DQ RSDTXH PDWHULDO RU UHIOHFWHG DQGRU WUDQVPLWWHG IURP D QRQRSDTXH PDWHULDO SURGXFH LWV FRORU

PAGE 51

D 5HIUDFWLRQ 7KH UHIUDFWLRQ RI OLJKW E\ D PDWHULDO LV HVWLPDWHG E\ 6QHOOnV ODZ I &KH@ VLQ Q f§ Q VLQ W QL ZKHUH  DQJOH RI LQFLGHQFH IURP WKH OLQH QRUPDO WR WKH LUUDGLDWHG VXUIDFH A DQJOH RI WUDQVPLVVLRQ IURP WKH VDPH SODQH Q LQGH[ RI UHIUDFWLRQ RI PDWHULDO Q LQGH[ RI UHIUDFWLRQ RI PHGLXP Q UHODWLYH LQGH[ RI UHIUDFWLRQ RI PDWHULDO WR PHGLXP 1RUPDOO\ Q LV XVHG DV WKH QRWDWLRQ IRU LQGH[ RI UHIUDFWLRQ LI WKH PHGLXP LV DLU RU D YDFXXP 6QHOOnV ODZ FDQ EH GHULYHG IURP )HUPDWnV SULQFLSOH > 7LS@ 7KH SDWK WDNHQ EY OLJKW LQ WUDYHOLQJ IURP RQH SRLQW WR DQRWKHU LV VXFK WKDW WKH WLPH RI WUDYHO LV D PLQLPXP FRPSDUHG ZLWK QHDUE\ SDWKV S f :KHQ OLJKW SDVVHV IURP DLU LQWR D GHQVHU PDWHULDO LWV YHORFLW\ GHFUHDVHV )ROORZLQJ )HUPDWnV SULQFLSOH OLJKW ZLOO PLQLPL]H LWV WUDYHO WLPH E\ LQFUHDVLQJ LWV RSWLFDO SDWK OHQJWK LQ DLU UHODWLYH WR WKH SDWK OHQJWK LQ WKH GHQVHU PDWHULDO 7KLV UHVXOWV LQ D FKDQJH LQ DQG Q 7KXV Q

PAGE 52

LV UHODWHG WR WKH YHORFLW\ DQG ZDYHOHQJWK RI OLJKW ZKLFK LV JLYHQ E\ APDWHULDO ZKHUH W! DQG APDWHULDO DUH WKH YHORFLW\ DQG ZDYHOHQJWK RI OLJKW SDVVLQJ WKURXJK D PDWHULDO DQG ;DLU LV WKH ZDYHOHQJWK RI LQFLGHQW OLJKW LQ DLU $W D FRQVWDQW ZDYHOHQJWK RI LQFRPLQJ OLJKW Q WHQGV WR LQFUHDVH IRU GHQVHU PDWHULDOV :KHQ FRPSDULQJ YDOXHV IRU FHUDPLF YV SRO\PHULF WHIORQ DQG SRO\VW\UHQHf PDWHULDOV LQ 7DEOH > :HD @ LW EHFRPHV HYLGHQW WKDW WKH VWUXFWXUH RI D PDWHULDO DOVR LQIOXHQFHV Q ,I D PDWHULDO LV HDVLO\ SRODUL]HG WKHUH DUH LQFUHDVHG LQWHUDFWLRQV RI LWV HOHFWURQLF VWUXFWXUH ZLWK LQFLGHQW SKRWRQV ,Q GLHOHFWULFV VXFK DV FHUDPLFV WKH LQGH[ RI UHIUDFWLRQ LV UHODWHG WR WKH UHODWLYH GLHOHFWULF FRQVWDQW .n f E\ > .LQ@ Q A.n f ZKHUH LV WKH LQGH[ RI DEVRUSWLRQ 3RODUL]DWLRQ 3f RI WKH HOHFWURQ FORXG DURXQG DQ DWRPLF QXFOHXV LV SURSRUWLRQDO WR WKH HOHFWULF ILHOG VWUHQJWK (f RI WKH LQFRPLQJ OLJKW 3 1 D ( f

PAGE 53

7DEOH ,QGH[ RI 5HIUDFWLRQ RI 6HOHFWHG 0DWHULDOV DW QP :DYHOHQJWK LQ $LU > :HD @ 0DWHULDO 'HQVLW\ JFPf 0HDQ 5HIUDFWLYH ,QGH[ Qf :DWHU 3RO\VW\UHQH 6L JODVVf 7HIORQ 6LOLFRQ 6L TXDUW]f &D&& 'LDPRQG D $ )H KHPDWLWHf

PAGE 54

ZKHUH D LV WKH DYHUDJH GLSROH PRPHQW SHU XQLW ILHOG VWUHQJWK RU WKH SRODUL]DELOLW\ DQG 1 LV WKH QXPEHU RI PDWHULDO SDUWLFOHV SHU XQLW YROXPH %\ WKH /RUHQW]/RUHQW] HTXDWLRQ HOHFWURQLF SRODUL]DWLRQ LV OLQNHG WR WKH UHIUDFWLYH LQGH[ RI D PRQDWRPLF JDV ZKHUH e LV WKH GLHOHFWULF FRQVWDQW LQ D YDFXXP 1R LV DYDJDGURnV QXPEHU DQG Qf LV WKH PRODU UHIUDFWLYLW\ 7KH PRODU UHIUDFWLYLW\ Qf LV GHWHUPLQHG E\ PHDVXULQJ Q DW YDULRXV ZDYHOHQJWKV RI OLJKW DV VKRZQ LQ )LJXUH DQG WKHQ H[WUDSRODWLQJ WR LQILQLWH ZDYHOHQJWK 6LQFH WKH HOHFWURQ GHQVLW\ LV XQLIRUP ZLWKLQ DQ DWRPLF UDGLXV U4f SRODUL]DELOLW\ LV DOVR UHODWHG WR WKH DWRPLF YROXPH RI D PDWHULDO D f (TXDWLRQ f VKRZV WKDW ODUJHU DWRPV ZKLFK FRQWDLQ PRUH HOHFWURQV H[KLELW JUHDWHU SRODUL]DELOLW\ DQG WKXV WHQG WR KDYH D KLJKHU UHIUDFWLYH LQGH[ ,RQLF FKDUJH DOVR SOD\V DQ LPSRUWDQW UROH LQ LQIOXHQFLQJ WKH LQGH[ RI UHIUDFWLRQ $V WKH LRQLF FKDUJH EHFRPHV LQFUHDVLQJO\ QHJDWLYH RXWHU HOHFWURQV DUH PRUH ORRVHO\ ERXQG DQG FDQ LQFUHDVH SRODUL]DELOLW\ ,Q DGGLFLRQ WKH UHIUDFWLYH LQGH[ LV GHSHQGHQW XSRQ FU\VWDO VWUXFWXUH V\PPHWU\ H[FHSW LQ

PAGE 55

:DYHOHQJWK PLFURQVf )LJXUH 5HIUDFWLYH LQGH[ YV ZDYHOHQJWK RI LQFLGHQW OLJKW IRU WKUHH JODVVHV >$GDSWHG IURP .LQ @

PAGE 56

JODVVHV DQG FXELF FU\VWDOV ZKLFK DUH LVRWURSLF &U\VWDOV KDYH D KLJKHU LQGH[ RI UHIUDFWLRQ LQ GHQVHU FORVHSDFNHG GLUHFWLRQV )RU DQLVRWURSLF VXEVWDQFHV WKH PHDQ LQGH[ RI UHIUDFWLRQ LV HVWLPDWHG DV > 3DU@ D \ f Q ZKHUH D DQG \ DUH UHIUDFWLYH LQGLFHV RI WKH SULQFLSOH FU\VWDOORJUDSKLF GLUHFWLRQV 0XOWLSKDVH FU\VWDOOLQH DQG JODVV\ VXEVWDQFHV KDYH VSHFLILF UHIUDFWLYH HQHUJLHV WKDW DUH WKH VXP RI WKH UHIUDFWLYH HQHUJLHV RI WKHLU FRPSRQHQWV > 3DU@ DQG ZKHUH U Q S A Q N 3] f VSHFLILF UHIUDFWLYH HQHUJ\ RI D VXEVWDQFH PHDQ LQGH[ RI UHIUDFWLRQ RI D VXEVWDQFH GHQVLW\ RI D VXEVWDQFH VSHFLILF UHIUDFWLYH HQHUJLHV RI WKH FRPSRQHQWV 3L 3" f f ZHLJKW SHUFHQWDJHV RI WKH FRPSRQHQWV

PAGE 57

5HIHUHQFH YDOXHV IRU VSHFLILF UHIUDFWLYH HQHUJLHV RI R[LGHV FDQ EH XVHG WR HVWLPDWH WKH UHIUDFWLYH LQGLFHV RI JODVVHV JOD]HV DQG FU\VWDOOLQH VSHFLHV E 6XUIDFH 5HIOHFWLRQ DQG *ORVV :KHQ D EHDP RI OLJKW VWULNHV D PDWHULDO D SRUWLRQ RI WKH SKRWRQV PD\ EH UHIOHFWHG 7KH OLJKW UHIOHFWHG DW DQ DQJOH HTXDO WR WKH LQFLGHQW DQJOH LV UHIHUUHG WR DV VSHFXODU UHIOHFWLRQ )LJXUH f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f IURP D SHUIHFWO\ VPRRWK VXUIDFH LQ DQ DLU PHGLXP LV WKH IUDFWLRQ RI

PAGE 58

)LJXUH 5HIOHFWLRQ DQG WUDQVPLVVLRQ RI OLJKW E\ D JODVV\ PDWHULDO FRQWDLQLQJ VXVSHQGHG SDUWLFOHV >$GDSWHG IURP .LQ @

PAGE 59

LQFLGHQW OLJKW UHIOHFWHG DW DQ DQJOH HTXDO WR WKH LQFLGHQW DQJOH 7KLV IUDFWLRQ RI VSHFXODU UHIOHFWDQFH IURP QRUPDO LQFLGHQFH L 2f LV FDOFXODWHG IRU WKH RSWLFDO UHJLRQ RI WKH VSHFWUXP ZLWK )UHVQHOnV IRUPXOD > .LQ@ 5 Q Q c (TXDWLRQ f VKRZV WKDW PDWHULDOV ZLWK D KLJK LQGH[ RI UHIUDFWLRQ KDYH D KLJK UHIOHFWLYLW\ 6LQFH WKH LQGH[ RI UHIUDFWLRQ YDULHV ZLWK WKH ZDYHOHQJWK RI OLJKW VR GRHV 5 )URP )UHVQHOnV ODZ HTXDWLRQV FDQ EH ZULWWHQ WR FRPSDUH WKH UHIOHFWLYLW\ RI SODQH SRODUL]HG OLJKW VR RULHQWHG WR D SODQH PLUURU VXUIDFH WKDW LWV UHIOHFWLRQ UDWLR RI UHIOHFWHG WR LQFLGHQW IOX[f LV PRVW IDFLOLWDWHG 5f RU PRVW KLQGHUHG 5MBf LQ DQ DLU PHGLXP > -XG@ 5 &26 Qa VLQ FRV r-Qa VLQ DQG 5L H R ? Qf FRV AQ VLQ c Qf FRV M r Q VLQ f ZKHUH c LV WKH DQJOH RI LQFLGHQFH )RU XQSRODUL]HG LQFLGHQW OLJKW WKH WRWDO UHIOHFWLYLW\ 5AG LV WKH DYHUDJH RI 5 DQG 5s

PAGE 60

UW 5 5sf f 7KXV WKH UHIOHFWLYLW\ 57 IRU XQSRODUL]HG OLJKW LV WKH DYHUDJH RI UHIOHFWLYLWLHV IRU SODQH SRODUL]HG OLJKW LQ 5 f DQG SHUSHQGLFXODU WR 5BBf WKH SODQH RI WKH LQFLGHQW EHDP ,I c r WKH LQFLGHQW EHDP LV SHUSHQGLFXODU WR WKH VXUIDFH DQG SHUIHFW PLUURU VSHFXODU UHIOHFWLRQ UHVXOWV ,Q WKLV FDVH HTXDWLRQV f DQG f UHGXFH WR HTXDWLRQ f 57 5f ,I [ r RQO\ JUD]LQJ LQFLGHQFH RFFXUV DQG 5 5s 5M %HWZHHQ SHUSHQGLFXODU DQG JUD]LQJ LQFLGHQFH 5 5s DQG 5W YDU\ )RU D SHUIHFW PLUURU VXUIDFH 5 5s 57 UHJDUGOHVV RI A )RU UHDO FRORUDQW OD\HUV ZLWK D JORVV\ PHGLXP PLUURU UHIOHFWDQFH DSSURDFKHV XQLW\ ZLWK JUD]LQJ LQFLGHQFH DV L f§} r ,Q WKLV FDVH WKH LQFLGHQW EHDP FRQWDFWV RQO\ WKDW SDUW RI WKH FRORUDQW OD\HU ZKLFK LV MXVW EHORZ WKH VXUIDFH 7KLV LV WKH KLJKJORVV WUDQVSDUHQW PHGLXP FDUU\LQJ VXVSHQGHG SLJPHQW SDUWLFOHV )LJXUH Df VKRZV WKH UHODWLRQVKLS EHWZHHQ 5 5 DQG 5W IRU A IURP r WR r IRU DQ DLUJODVV ERXQGDU\ ZKHUH WKH UHIUDFWLYH LQGH[ RI WKH JODVV LV )LJXUH Ef SORWV WRWDO UHIOHFWLYLW\ DV D IXQFWLRQ RI DQG LQGH[ RI UHIUDFWLRQ )LJXUH FDQ EH XVHG WR HVWLPDWH WKH JORVV RU

PAGE 61

Ef )LJXUH )UHVQHO UHIOHFWLRQ IRU Df DQ DLUJODVV ERXQGDU\ DQG Ef WRWDO UHIOHFWLYLW\ IRU GLIIHUHQW LQGH[ RI UHIUDFWLRQ JODVVHV

PAGE 62

GHJUHH RI DSSURDFK WR D PLUURU VXUIDFH IRU PHGLXPV VXFK DV SODVWLFV JODVV WH[WLOH ILEHUV DQG SDLQW YHKLFOHV > -XG@ 1RWH LQ )LJXUH Df WKDW IRU D VPRRWK JODVV ZLWK Q 5MB f§! DW r 7KLV IROORZV %UHZVWHUnV ODZ ZKLFK VWDWHV WKDW PLUURU UHIOHFWDQFH LV PRVW IDFLOLWDWHG 5BMB f ZKHQ > -XG@ WDQ Q f 7KXV IRU D JODVV RI Q DW LQFLGHQW DQG YLHZLQJ DQJOHV RI DSSUR[LPDWHO\ r WKHUH LV D ODFN RI SRODUL]HG GLIIXVHO\ UHIOHFWHG OLJKW 3DUDOOHO SRODUL]HG OLJKW LQ WKH UHIOHFWHG EHDP LV WKH RQO\ FRPSRQHQW SUHVHQW IURP WKH XQSRODUL]HG LQFLGHQW OLJKW 7KLV SULQFLSOH KDV EHHQ DSSOLHG WR PHDVXUH JORVV WKURXJK WKH XVH RI D SRODUL]LQJ HOHPHQW ZKLFK VXEWUDFWV WKH VSHFXODU FRPSRQHQW DW %UHZVWHUnV DQJOH ZKHUH 5s %UHZVWHUnV ODZ LV DOVR HPSOR\HG WR PHDVXUH WKH LQGH[ RI UHIUDFWLRQ RI VPRRWK JODVV > )OH@ $ OLJKW SRODUL]HU DQG GHWHFWRU DUH XVHG WR PHDVXUH WKH SRODUL]DWLRQ HIIHFWLYHQHVV RI OLJKW UHIOHFWLQJ IURP WKH JODVV VXUIDFH 7KH DQJOH RI LQFLGHQFH cf ZKHUH UHIOHFWHG SRODUL]DWLRQ LV PRVW HIILFLHQW LV DSSOLHG WR FDOFXODWH Q ZLWK HTXDWLRQ f $V WKH VXUIDFH URXJKQHVV RI DQ RSDTXH FRDWLQJ RU D JODVV PDWUL[ ZLWK VXVSHQGHG FRORUDQW SDUWLFOHV LQFUHDVHV D JUHDWHU SRUWLRQ RI WKH UHIOHFWLRQ EHFRPHV GLIIXVH 7KLV EURDGHQLQJ

PAGE 63

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f WKH SRUWLRQ RI OLJKW GLIIXVHO\ UHIOHFWHG RSDFLW\f DQG WKH

PAGE 64

SRUWLRQV RI OLJKW GLUHFWO\ DQG GLIIXVHO\ WUDQVPLWWHG DUH LPSRUWDQW )LJXUH f 7KHVH DWWULEXWHV DUH LQIOXHQFHG E\ WKH OLJKWVFDWWHULQJ SURSHUWLHV RI SKDVHV RU VPDOO RSDFLI\LQJ SDUWLFOHV VXVSHQGHG LQ WKH JODVV\ PDWUL[ *RRG RSDFLILFDWLRQ LV REWDLQHG ZLWK WKH KLJK OLJKW VFDWWHULQJ DQG UHIOHFWLYLW\ WKLV PHFKDQLVP SURYLGHV &RPPRQ RSDFLILHUV XVHG LQ FHUDPLFV FRDWLQJV LQFOXGH ]LUFRQ =Ur6Lf ]LUFRQLD =Uf 6Q 7L DQG $O 7KH\ DUH VHOHFWHG EDVHG RQ WKHLU RSWLFDO SURSHUWLHV WKH GHVLUHG FRDWLQJ FRPSRVLWLRQ SURFHVVLQJ OLPLWDWLRQV DQG SURSHUWLHV WKDW DIIHFW FRPSDWLELOLW\ ZLWK WKH PDWUL[ SKDVH VXFK DV WKHUPDO H[SDQVLRQ VROXELOLW\ KDUGQHVV DQG PHOWLQJ WHPSHUDWXUH 2SDFLILHUV FDQ EH LQHUW ZLWK UHVSHFW WR WKH KRVW PDWUL[ SKDVH IRUPHG GXULQJ PHOWLQJ RU FU\VWDOOL]HG IURP WKH JODVV PHOW 9HU\ ILQH SDUWLFOHV FDQ EH REWDLQHG IURP PDWHULDOV VXFK DV 7L RU DW KLJKHU ILULQJ WHPSHUDWXUHV =U2 ZKLFK FDQ EH PHOWHG QXFOHDWHG DQG UHFU\VWDOOL]HG GXULQJ WKH KHDWLQJ F\FOH 7KH GHJUHH RI RSDFLW\ RI FHUDPLFV ZKLFK FRQWDLQ SDUWLFOHV VXVSHQGHG LQ D JODVV\ PDWUL[ GHSHQGV RQ ILYH PDLQ IDFWRUV > 3DU@

PAGE 65

7KH GLIIHUHQFH LQ UHIUDFWLYH LQGH[ EHWZHHQ WKH PDWUL[ DQG GLVSHUVHG SDUWLFOHV 7KLV HIIHFW LV GHVFULEHG E\ .HUVWDQ , Q a QLfa Q QA f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f > 5HH@ 7KH VL]H RI GLVSHUVHG SDUWLFOHV ZKLFK VFDWWHU OLJKW 7KH OLJKWVFDWWHULQJ DELOLW\ RI DQ RSWLFDOO\ KHWHURJHQHRXV PDWHULDO FDQ EH HVWLPDWHG ZLWK WKH VFDWWHULQJ RU WXUELGLW\ FRHIILFLHQW 6f 7KLV FRHIILFLHQW LV D

PAGE 66

7DEOH 3URSHUWLHV RI 0DWHULDOV 8VHG IRU 2SDFLI\LQJ &HUDPLF *OD]HV > 3DU@ 2SDFLILHU 0DWHULDO 0HDQ ,QGH[ RI 5HIUDFWLRQ Qf 'LIIHUHQFH LQ Q 2SDFLILHU*OD]H 0HOWLQJ 7HPS &f &RHIILFLHQW RI ([SDQVLRQ ;r& HJ 2 fU? (K =U =U6L 6Q $LU f§ f§ *OD]H f§ f§

PAGE 67

PHDVXUHPHQW RI WKH DWWHQXDWLRQ GXH WR VFDWWHULQJ RI OLJKW DV LW WUDYHUVHV D PHGLXP FRQWDLQLQJ VFDWWHULQJ SDUWLFOHV > .LQ @ L 6 6I9 f f DQG ZKHUH H[S 6I9S H[S 6[f f 6I VFDWWHULQJ IDFWRU WKDW YDULHV EHWZHHQ DQG 9S YROXPH IUDFWLRQ RI VFDWWHULQJ SDUWLFOHV U UDGLXV RI WKH VFDWWHULQJ SDUWLFOH ,4, UDWLR RI OLJKW LQWHQVLW\ VFDWWHUHGLQLWLDO LQWHQVLW\ [ RSULFDO SDWK OHQJWK (TXDWLRQ f VKRZV WKDW VFDWWHULQJ DQG WKXV RSDFLW\ WHQGV WR LQFUHDVH XS WR D SRLQW ZLWK GHFUHDVLQJ SDUWLFOH VL]H 7KH VFDWWHULQJ FRQVWDQW 6If LQFUHDVHV ZLWK SDUWLFOH VL]H Uf DQG LV LQYHUVHO\ SURSRUWLRQDO WR WKH IRXUWK SRZHU RI ZDYHOHQJWK IRU SDUWLFOH VL]HV PXFK VPDOOHU WKDQ WKH ZDYHOHQJWK RI LQFLGHQW OLJKW :KHQ WKH SDUWLFOH VL]H LV DSSUR[LPDWHO\ HTXDO WR WKH ZDYHOHQJWK RI OLJKW 6I UHDFKHV LWV PD[LPXP YDOXH WKHQ GHFUHDVHV ZLWK LQFUHDVLQJ SDUWLFOH VL]H 7KXV PD[LPXP VFDWWHULQJ RFFXUV ZKHQ WKH RSDFLILHU KDV D

PAGE 68

SDUWLFOH VL]H VLPLODU WR WKH ZDYHOHQJWK RI OLJKW XVHG ZKLFK LV LQ WKH UDQJH RI WR PLFURQV LQ GLDPHWHU IRU YLVLEOH OLJKW 7KH QXPEHU RI UHIOHFWLQJ SDUWLFOHV SHU XQLW YROXPH (TXDWLRQV f DQG f LQGLFDWH WKDW RSDFLW\ LQFUHDVHV ZLWK WKH QXPEHU RI SDUWLFOHV $V WKH FRQFHQWUDWLRQ RI RSDFLILHU LQFUHDVHV WKH UDWH RI LQFUHDVH RI RSDFLW\ GHFUHDVHV *OD]HV DQG HQDPHOV W\SLFDOO\ XWLOL]H D PD[LPXP RI b ]LUFRQ E\ nZHLJKW IRU RSDFLILFDWLRQ $SSUR[LPDWHO\ b WR b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

PAGE 69

JODVV GXULQJ WKH KHDWLQJ RU FRROLQJ F\FOH WHQG WR KDYH VKDUS LQWHUIDFH ERXQGDULHV 7KH WKLFNQHVV RI D FRDWLQJ DSSOLHG WR D VXEVWUDWH &RDWLQJ WKLFNQHVV HIIHFWV RQ FRYHULQJ SRZHU FDQ EH UHODWHG ZLWK WKH .XEHOND DQG 0XQN HTXDWLRQV > .LQ@ 5H L 9 5f a 5 5n 5 RR H[S 6[ 5 O5 -5n 5af  A 5n H[S 6[ 5 5f 5RF DQG 5 r g 6 f f ZKHUH 5U FRDWLQJ UHIOHFWDQFH 5n VXEVWUDWH UHIOHFWDQFH 5f WRWDO UHIOHFWLYLW\ RI D FRORUDQW OD\HU VR WKLFN WKDW IXUWKHU LQFUHDVHV LQ OD\HU WKLFNQHVV GR QRW FKDQJH WKH UHIOHFWLYLW\ 6 VFDWWHULQJ FRHIILFLHQW DEVRUSWLRQ FRHIILFLHQW [ FRDWLQJ WKLFNQHVV

PAGE 70

(TXDWLRQ f VKRZV WKDW FRDWLQJ UHIOHFWDQFH LQFUHDVHV XS WR D SRLQW ZLWK WKLFNHU DSSOLFDWLRQV DQG KLJKHU UHIOHFWLRQ VXEVWUDWHV )LJXUH GHPRQVWUDWHV WKH LQFUHDVH LQ UHIOHFWDQFH ZLWK DSSOLFDWLRQ WKLFNQHVV RI D W\SLFDO 7L&S RSDFLILHG ZKLWH JOD]H ILUHG WR r& 7KH PDLQ GLVDGYDQWDJHV RI D WKLFNHU FRDWLQJ LV WKH LQFUHDVHG PDWHULDO UHTXLUHPHQWV DQG FRVW (TXDWLRQV f DQG f LQGLFDWH WKDW RSDFLILHUV ZLWK D KLJK VFDWWHULQJ FRHIILFLHQW 6f DQG ORZ DEVRUSWLRQ FRHIILFLHQW (f DUH PRVW SRZHUIXO 7KH TXDQWLW\ 6[ LV RIWHQ DSSOLHG WR HVWLPDWH WKH VFDWWHULQJ SRZHU RI D FRDWLQJ $QRWKHU PHWKRG JDXJHV RSDFLW\ ZLWK WKH UDWLR RI UHIOHFWDQFH REWDLQHG IURP D FRDWLQJ RYHU D EODFN 5n f YHUVXV D ZKLWH 5n &f EDFNLQJ 7KLV LV WKH 7DSSL 2SDFLW\ 0HWKRG IRU GHWHUPLQLQJ WKH FRQWUDVW UDWLR & 5n5nU J /LTXLGOLTXLG SKDVH VHSDUDWLRQ GXULQJ JODVV IRUPDWLRQ LV DQRWKHU PHWKRG E\ ZKLFK JOD]H RSDFLILFDWLRQ FDQ RFFXU )RU H[DPSOH WKH /L 6L 7L V\VWHP VHSDUDWHV DW KLJK WHPSHUDWXUH LQWR D ORZ UHIUDFWLYH LQGH[ VLOLFDWH JODVV DQG D KLJK LQGH[ WLWDQLD ULFK JODVV > .LP@ 2QH GLVDGYDQWDJH RI WKLV PHFKDQLVP LV WKDW LW LV YHU\ VHQVLWLYH WR SURFHVVLQJ WHPSHUDWXUH 7KHUHIRUH LQ FXUUHQW PDQXIDFWXULQJ VHWWLQJV ZKHUH KHDWLQJ YDULDWLRQV GXULQJ ILULQJ DUH FRPPRQ WKH

PAGE 71

:DYOHQJWK PLOOLPLFURQr )LJXUH 5HIOHFWDQFH YV ZDYHOHQJWK RI OLJKW IRU D 7L2 RSDFLILHG ZKLWH JOD]H ILUHG WR r& DW YDULRXV JOD]H DSSOLFDWLRQ ZHLJKWV LQ JIW >$GDSWHG IURP 3DU@

PAGE 72

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f RU KLJKHU SRUH VL]H DW D JLYHQ FRQFHQWUDWLRQ LQFUHDVH WUDQVOXFHQF\ G $EVRUSWLRQ 7UDQVPLVVLRQ DQG &RORU /LJKW WKDW LV QRW UHIOHFWHG RU WUDQVPLWWHG E\ D PDWHULDO LV DEVRUEHG 7KH OLQHDU DEVRUSWLRQ FRHIILFLHQW .f LQGLFDWHV WKH SRUWLRQ RI QRUPDOO\ LQFLGHQW UDGLDQW HQHUJ\ DEVRUEHG WKURXJK D XQLW GLVWDQFH [f LQ D VLQJOH SKDVH PDWHULDO E\ >:HD@ 7 f§ H[S.[f f ZKHUH 7 IUDFWLRQ RI OLJKW WUDQVPLWWHG DV LW SDVVHV WKURXJK D PDWHULDO

PAGE 73

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

PAGE 74

H[DPSOH DW 6L KDV D JDS HQHUJ\ RI H9 ZKLOH GLDPRQG KDV D JDS HQHUJ\ RI H9 >+XP@ 7KHUHIRUH VLOLFRQ UHTXLUHV OHVV HQHUJ\ IRU HOHFWURQ WUDQVLWLRQV DQG DSSHDUV RSDTXH LQ GD\OLJKW ZKLOH SXUH GLDPRQG LV WUDQVSDUHQW ,Q WKH YLVLEOH VSHFWUXP (J H9 PDWHULDOV GR QRW DEVRUE DQ\ SKRWRQV ZKHUH (J H9 PDWHULDOV DEVRUE DOO YLVLEOH OLJKW >$VN@ )RU LQWHUPHGLDWH HQHUJ\ JDSV D IUDFWLRQ RI WKH LQFLGHQW YLVLEOH OLJKW LV DEVRUEHG ,Q LRQLF FHUDPLFV ILOOHG VKHOOV RI WLJKWO\ ERXQG HOHFWURQV GR QRW KDYH HQHUJ\ OHYHOV DYDLODEOH IRU HOHFWURQ PRYHPHQW >5LF@ DQG PRVW VLQJOH FU\VWDOV DUH WUDQVSDUHQW &RYDOHQW FHUDPLFV KRZHYHU YDU\ LQ WKH OHYHO RI DEVRUSWLRQ )RU H[DPSOH GLDPRQG DQG JUDSKLWH ERWK KDYH FRYDOHQWO\ ERQGHG FDUERQ DWRPV EXW WKHLU RSWLFDO SURSHUWLHV DUH VLJQLILFDQWO\ GLIIHUHQW 'LDPRQG LV WUDQVSDUHQW ZKLOH JUDSKLWH DSSHDUV EODFN $OWKRXJK WKHUH LV D VWURQJ FRYDOHQW ERQGLQJ ZLWKLQ WKH JUDSKLWH KH[DJRQDO QHWZRUN ZHDN 9DQGHU :DDOnV ERQGLQJ EHWZHHQ WKH OD\HUV DOORZV IRU HOHFWURQ PRYHPHQW 7KLV UHVXOWV LQ HOHFWURQ WUDQVLWLRQV DQG DEVRUSWLRQ RI YLVLEOH OLJKW *RRG LQVXODWRUV ZLWK D ODUJH (J VXFK DV GLDPRQG WHQG WR WUDQVPLW OLJKW $EVRUSWLRQ GXH WR HOHFWURQ WUDQVLWLRQV DQG UHVRQDQFH LV LQWULQVLF ZKLOH H[WULQVLF HIIHFWV LQ FHUDPLFV FDQ DOVR FDXVH

PAGE 75

DEVRUSWLRQ DQG FRORU ([WULQVLF HIIHFWV LQFOXGH JUDLQ ERXQGDULHV SRUHV LQFOXVLRQV DQLVRWURS\ DQG DWRP YDFDQFLHV ,Q FHUDPLFV WKH DEVRUSWLRQ FRHIILFLHQW .f LV UHODWHG WR WKH LQGH[ RI DEVRUSWLRQ NLf DOVR UHIHUUHG WR DV WKH DWWHQXDWLRQ LQGH[ RU H[WLQFWLRQ FRHIILFLHQWf E\ >.LQ@ WN ; f )URP HTXDWLRQV f f DQG f LQFUHDVHV ZLWK NL DQG WKH LQGH[ RI UHIUDFWLRQ RI D FHUDPLF PDWHULDO DQG GHFUHDVHV ZLWK KLJKHU ZDYHOHQJWKV RI LQFLGHQW OLJKW 7KH RYHUDOO IUDFWLRQ RI OLJKW WUDQVPLWWHG 7f DIWHU ERWK UHIOHFWDQFH DQG DEVRUSWLRQ ORVVHV LV 7 7n f§ f§5f f H[S .[ff s & f ZKHUH 5 LV WKH UHIOHFWLYLW\ DQG ,W, LV WKH UDWLR RI WUDQVPLWWHG WR LQFLGHQW OLJKW LQWHQVLWLHV %\ HTXDWLRQV f f f DQG f DLO RI WKH LQFRPLQJ OLJKW FDQ EH DFFRXQWHG IRU E\ UHIOHFWLRQ DEVRUSWLRQ DQG WUDQVPLVVLRQ 7RWDO OLJKW LQWHUDFWLRQ ZLWK D PDWHULDO FDQ WKXV EH ZULWWHQ >$VN@ ,UI ORr f OD ;,UIf >ORL5f H[S .[` @ f ,UE ,5 f§ 5f H[S .[f f

PAGE 76

, W f§ 5f H[S.[f f 2 f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

PAGE 77

7UDQVLWLRQV f f DQG f XVXDOO\ DUH FDXVHG E\ LPSXULWLHV RU GHIHFWV LQ D PDWHULDO ZKLOH f LV D EXON SURSHUW\ 2IWHQ SDUWLFOHV DUH VXVSHQGHG LQ D PDWUL[ VXFK DV D JODVV WR FUHDWH HOHFWURQLF WUDQVLWLRQV DQG FRORU 7KH DEVRUSWLRQ FRHIILFLHQW LV SURSRUWLRQDO WR WKH FRQFHQWUDWLRQ RI DEVRUELQJ LRQ Ff DFFRUGLQJ WR %HHUnV ODZ >5HH@ 7 H[S HF[f f DQG HF f ZKHUH e LV WKH H[WLQFWLRQ FRHIILFLHQW REVHUYHG SHU XQLW FRQFHQWUDWLRQ 7KLV LV WKH IXQGDPHQWDO ODZ RI VLPSOH VXEWUDFWLYH FRORUDQW PL[LQJ 7KH PRVW FRPPRQO\ XVHG FRORUDQW LRQV DUH IURP WUDQVLWLRQ PHWDO FRPSRXQGV RU LPSXULWLHV VXFK DV 9 &U 0Q )H &R 1L DQG &X VKRZQ LQ 7DEOH > 5HH@ 7KH\ SURYLGH FRORU LQ PDQ\ FHUDPLF ERGLHV JOD]HV JODVVHV PLQHUDOV JHPV SLJPHQWV DQG SDLQWV &U\VWDO RU OLJDQG ILHOG WKHRULHV GHVFULEH KRZ WKHVH HOHPHQWV SURGXFH FRORU > 3HW@ 7UDQVLWLRQ PHWDOV KDYH XQILOOHG LQQHU RUELWDOV DYDLODEOH IRU WKH FUHDWLRQ RI VSOLW HQHUJ\ OHYHOV IRU HOHFWURQLF WUDQVLWLRQV ,Q IUHH LRQV RUELWDOV KDYH HTXDO HQHUJLHV EXW GLIIHUHQW VSDWLDO RULHQWDWLRQV DV VKRZQ LQ )LJXUH IRU WKH ILYH G RUELWDOV %XW WKH FRRUGLQDWLRQ RI QHJDWLYHO\ FKDUJHG DQLRQV DERXW WKH

PAGE 78

)LJXUH 7KH G RUELWDOV >$GDSWHG IURP .LQ@

PAGE 79

7DEOH 7UDQVLWLRQ (OHPHQWV DQG 7KHLU 3URSHUWLHV > 3HW@ 9 &U 0Q )H &R 1L &X $WRPLF QXPEHU $WRPLF 5DGLXV $QJVWURPV (OHFWURQLF FRQILJXUDWLRQ ,RQL]DWLRQ HQHUJLHV GV GVGV G V G V GV G V )LUVW 6HFRQG 7KLUG 2[LGDWLRQ SRWHQWLDO 2[LGDWLRQ 6WDWHV 0HOWLQJ SRLQW r& 'HQVLW\ JFF +DUGQHVV f§ f§ f§ (OHFWULFDO FRQGXFWLYLW\ f§ (DFK DWRP KDV DQ DUJRQ LQQHU FRUH FRQILJXUDWLRQ 9DOXHV DUH LQ NFDOPROH )RU WKH R[LGDWLRQ SURFHVV 0Vf 0DJf Ha &RPPRQ R[LGDWLRQ VWDWHV WKH PRVW VWDEOH RQH LV LWDOLF +DUGQHVV YDOXHV DUH RQ WKH 0RKV VFDOH &RPSDUHG WR DQ DUELWUDULO\ DVVLJQHG YDOXH RI IRU VLOYHU

PAGE 80

PHWDO FDWLRQ SURGXFHV DQ HOHFWURVWDWLF ILHOG WKDW UDLVHV LQQHU RUELWDO HQHUJLHV QRQXQLIRUPO\ 7KLV HOHFWURVWDWLF LQWHUDFWLRQ EHWZHHQ DQLRQV DQG PHWDO LRQnV HOHFWURQ FORXGV VSOLWV LQQHU G RU I RUELWDOV LQWR GLIIHUHQW HQHUJ\ OHYHOV 7KH HQHUJ\ DQG FRUUHVSRQGLQJ ZDYHOHQJWK RI OLJKW DEVRUEHG E\ WKH PHWDO ZKLFK SURGXFHV FRORU LV HJXDO WR WKH GLIIHUHQFH LQ WKH VSOLW HQHUJ\ OHYHOV )RU H[DPSOH LQ D WHWUDKHGUDO VWUXFWXUH VXUURXQGLQJ D PHWDO LRQ ZLWK XQILOOHG G RUELWDOV WKH G[\ G[] DQG GB RUELWDOV KDYH PRUH HQHUJ\ WKDQ WKH G DQG Gf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f VR GRHV WKH VWUHQJWK RI WKH OLJDQG ILHOG 7KLV VHFWLRQ UHYLHZHG WKH PRVW FRPPRQ FDXVHV RI FRORU LQ PDWHULDOV DQG WKH PHWKRG E\ ZKLFK FRORU ZLOO EH GHULYHG GXULQJ WKH VXEVHTXHQW UHVHDUFK LQYROYHG ZLWK WKLV

PAGE 81

GLVVHUWDWLRQ $ FRPSUHKHQVLYH OLVW RI DOO RI WKH SRVVLEOH FDXVHV RI FRORU DUH OLVWHG LQ $SSHQGL[ % &RORU 3HUFHSWLRQ EY WKH +XPDQ (YH 7KH H\H LV WKH KXPDQ RSWLFDO V\VWHP )LJXUH f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

PAGE 82

$QWHULRU FKDPEHU &HQWUDO UHWLQDO DUWHU\ DQG YHLQ )LJXUH +XPDQ RSWLFDO V\VWHP >$GDSWHG IURP 7LS@

PAGE 83

2Q WKH RWKHU KDQG FRQHV VHQVH FKURPDWLF DV ZHOO DV QHXWUDO FRORUV &RQHV ZKLFK DUH UHVSRQVLEOH IRU GD\ YLVLRQ FDQ GHWHFW D PXFK KLJKHU GHQVLW\ RI UDGLDQW IOX[ WKDQ URGV EXW DUH OHVV VHQVLWLYH DW YHU\ ORZ OHYHOV RI OLJKW 5RGV UHVSRQG WR PLQXWH TXDQWLWLHV RI OLJKW DV ORZ DV FDQGHODV SHU VTXDUH PHUHU FGnPf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f DQG QLJKWWLPH URG RU

PAGE 84

5(/$7,9( 5(63216( :$9(/(1*7+ QP )LJXUH /XPLQRVLW\ IXQFWLRQV RI WKH URGV QLJKWWLPH VFRWRSLF YLVLRQf DQG FRQHV GD\WLPH SKRWRSLF YLVLRQf RI WKH KXPDQ H\H >$GDSWHG IURP +XQ@

PAGE 85

VFRWRSLF FXUYHf IRU WKH VDPH DPRXQW RI HQHUJ\ DW GLIIHUHQW ZDYHOHQJWKV RI OLJKW LQ WKH YLVLEOH VSHFWUXP 7KH FXUYHV ZHUH GHYHORSHG IURP H[SHULPHQWV ZKHUH KXPDQ REVHUYHUV DGMXVWHG WKH LQWHQVLW\ RI OLJKW DW GLIIHUHQW ZDYHOHQJWKV XQWLO WKH\ DSSHDUHG HTXDOO\ OXPLQRXV RU EULJKW > +XQ@ 7KH SURSHUW\ RI OLJKW E\ ZKLFK ZH GHILQH KRZ HDVLO\ ZH FDQ VHH LW LV UHIHUUHG WR DV OXPLQRVLW\ %RWK WKH
PAGE 86

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

PAGE 87

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f ZKLOH UHIOHFWLQJ EOXH f§ QPf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

PAGE 88

:$9(/(1*7+ 1$120(7(56 )LJXUH 5HIOHFWDQFH YHUVXV ZDYHOHQJWK IRU RSDTXH FRDWLQJV FRORUHG ZLWK SLJPHQWV WKDW DEVRUE D SRUWLRQ RI LQFLGHQW OLJKW >$GDSWHG IURP +XQ@

PAGE 89

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

PAGE 90

)LJXUH %DVLF FRPSRQHQWV RI VSHFWURSKRWRPHWHUV > $GDSWHG IURP +XQ@

PAGE 91

3+2720(7(5 $ $QWLKXQW )LJXUH 6FKHPDWLF RI WKH +DUG\ VSHFWURSKRWRPHWHU > $GDSWHG IURP %@

PAGE 92

PDWKHPDWLFDOO\ FRQYHUWHG WR UHSUHVHQW WKHRUHWLFDO VRXUFHV RU LOOXPLQDQWV EHIRUH FRORU YDOXHV DUH GHULYHG 2XWSXWV IURP VSHFWURSKRWRPHWULF PHDVXUHPHQWV LQFOXGH FRORU YDOXHV GHULYHG IURP UHODWLYH HQHUJ\ GLVWULEXWLRQV RI VWDQGDUG OLJKW VRXUFHV VXFK DV DYHUDJH QRRQ GD\OLJKW IURP WKH WRWDO VN\f LOOXPLQDQW $ LQFDQGHVFHQW ODPSf LOOXPLQDQW % QHDU VXQOLJKWf DQG LOOXPLQDQW & DYHUDJH GD\OLJKW IURP WKH WRWDO VN\f 7KH VWDQGDUG IXOO YLVXDO ILHOG RI YLHZ XWLOL]HG WR GHWHFW OLJKW UHIOHFWLQJ IURP DQ REMHFW LV r DQJXODU VXEWHQVH 2FFDVLRQDOO\ D r REVHUYHU LV XVHG WR SURYLGH D ODUJHU ILHOG RI YLHZ /LJKW IOX[ UHIOHFWHG IURP D VDPSOH LV FROOHFWHG IRU PHDVXUHPHQW E\ D ZKLWHOLQHG LQWHJUDWLQJ VSKHUH (OLPLQDWLRQ RI VXUIDFH JORVV IURP WKH FRORU PHDVXUHPHQW SURYLGHV UHVXOWV ZKLFK EHWWHU FRUUHODWH WR YLVXDO LQVSHFWLRQ 7KLV FDQ EH DFFRPSOLVKHG E\ UHSODFLQJ WKH ZKLWH SOXJ RQ WKH VSKHUHnV VSHFXODU FXS ZLWK D EODFN SOXJ 6LQFH WKH VSHFXODU FXS LV RIIVHW WR EH LOOXPLQDWHG DW DQ DQJOH RI UHIOHFWLRQ HTXDO WR WKH DQJOH RI WKH LQFLGHQW EHDP WKH EODFN SOXJ DEVRUEV SULPDULO\ VSHFXODUO\ UHIOHFWHG OLJKW 'LIIXVHO\ UHIOHFWHG IOX[ LV GLYHUWHG XS WKH VSKHUH WR WKH SKRWRGHWHFWRU 7KH FROOHFWHG SKRWRQ HQHUJ\ GLVWULEXWLRQ LV FRQYHUWHG LQWR DQ

PAGE 93

HOHFWULFDO VLJQDO DQG VHQW WR D FRPSXWHU 7KH FRPSXWHU SURJUDP FRQYHUWV PHDVXUHG VSHFWUDO GDWD LQWR QXPEHUV WKDW FDQ EH PRUH HDVLO\ LQWHUSUHWHG DQG FRUUHODWHG WR WKH UHVSRQVH RI WKH KXPDQ H\H 0HDVXUHPHQWV RI JORVV FDQ EH SHUIRUPHG VHSDUDWHO\ ZLWK JRQLRSKRWRPHWHUV RU JORVV PHWHUV ZKLFK PHDVXUH WKH VSHFWUDO UHIOHFWDQFH RU TXDQWLW\ RI OLJKW HPLWWHG LQ GLUHFWLRQV UHODWHG WR WKH VXUIDFH FKDUDFWHULVWLFV RI WKH REMHFW 7KH JORVV RI FHUDPLF FRDWLQJV LV QRUPDOO\ PHDVXUHG DW D r DQJOH RI LQFLGHQFH ZKHUH PLUURU UHIOHFWDQFH LV PRVW IDFLOLWDWHG DFFRUGLQJ WR %UHZVWHUnV ODZ HTXDWLRQ f E %DVLV IRU &RORU 4XDQWLILFDWLRQ 7KH DYHUDJH VHQVLWLYLW\ RI WKH KXPDQ H\H WR HDFK ZDYHOHQJWK RI OLJKW KDV EHHQ GHWHUPLQHG WKURXJK H[WHQVLYH H[SHULPHQWDWLRQ > % +XQ -XG 0DF$ 0DF$@ +XPDQ REVHUYHUV ZHUH DVNHG WR YLVXDOO\ PDWFK WKH FRORUV RI OLJKW IURP LQGLYLGXDO ZDYHOHQJWKV E\ PL[LQJ WRJHWKHU OLJKWV IURP WKUHH FRORUHG SULPDULHV 7KUHH SULPDULHV ZHUH DSSOLHG EHFDXVH WKH H\H FRQWDLQV WKUHH VSHFWUDOO\ XQLTXH UHFHSWRUV IRU GHWHFWLQJ FRORUV 7KH DPRXQWV RI HQHUJ\ RI HDFK RI WKH WKUHH OLJKWV UHTXLUHG WR PDWFK VLQJOH ZDYHOHQJWK FRORUV ZHUH XVHG WR GHYHORS VWDQGDUG REVHUYHU IXQFWLRQV IRU WKH EDVLV RI

PAGE 94

DOO FRORU PHDVXUHPHQW 7KHVH ZHLJKWLQJ IXQFWLRQV DUH DSSOLHG WR WUDQVIRUP VSHFWURSKRWRPHWULF GDWD LQWR QXPEHUV WKDW EHWWHU FRUUHODWH WR WKH ZD\ WKH KXPDQ H\H SHUFHLYHV FRORUV 7KH ZHLJKWLQJ IXQFWLRQV [ \ DQG ] f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nV UHVSRQVH WR OXPLQRVLW\ RU FRORU EULJKWQHVV 7KH DUHDV XQGHUQHDWK WKH WKUHH FXUYHV ZHUH PDGH HTXDO IRU OLJKW RI HTXDO HQHUJ\ DW DOO ZDYHOHQJWKV 2QH IXQFWLRQ ZDV VHOHFWHG WR EH DV QHDU WR ]HUR DV SRVVLEOH IRU DV PXFK RI WKH VSHFWUXP DV SRVVLEOH ,Q WKLV IRUP [ \ DQG ] GR QRW UHSUHVHQW DQ\ UHDO FRORUHG SULPDULHV EXW FDQ EH FRQYHUWHG WR YDOXHV WKDW DUH

PAGE 95

5(/$7,9( $028176 :$9(/(1*7+ 10 )LJXUH :HLJKWLQJ IXQFWLRQV XVHG IRU WKH VWDQGDUG REVHUYHU DW D r ILHOG RI YLHZ > $GDSWHG IURP %LO@

PAGE 96

HDVLHU WR DSSO\ 7KH ZHLJKWLQJ IXQFWLRQV DUH XVHG WR WUDQVIRUP VSHFWUDO UHIOHFWDQFH FXUYHV LQWR WKUHH QXPEHUV UHIHUUHG WR DV WULVWLPXOXV YDOXHV ; < DQG = 7KHVH YDOXHV VSHFLI\ FRORU LQ WHUPV RI WKH PL[WXUH RI UHG ;f JUHHQ +XQ@ ; [ A 6[ [ [ 6M/[A [A f r QU RU & ; 6[[G; f ZKHUH [[ ZHLJKWLQJ IXQFWLRQ [f YDOXH DW ; ZDYHOHQJWK VSHFWUDO HQHUJ\ DW ; ZDYHOHQJWK

PAGE 97

< \ f DQG = ] f FDQ EH FDOFXODWHG LQ WKH VDPH PDQQHU DV ; [f ZLWK f DQG f 7KH VSHFWUDO HQHUJ\ LQ WXUQ LV D IXQFWLRQ RI WKH SURSHUWLHV RI WKH OLJKW VRXUFH IRU DQ LOOXPLQDWH RU DSHUWXUH FRORU 6AVRXUFHf RU WKH UHIOHFWDQFH IURP D UHIOHFWLQJ REMHFW 6:HULDOf DQG 6 ( A;VRXUFH A; f ZKHUH fAPDWHULDO A; A; f (AB HQHUJ\ RI WKH OLJKW VRXUFH DW ; ZDYHOHQJWK SHUFHQW UHIOHFWDQFH RI OLJKW RI ; ZDYHOHQJWK IURP WKH PDWHULDO 6LQFH REMHFWV DUH YLHZHG LQ UHODWLRQ WR WKHLU VXUURXQGLQJV ; < DQG = DUH QRUPDOO\ H[SUHVVHG UHODWLYH WR WKH OXPLQRVLW\ RI D SHUIHFW ZKLWH RSDTXH PDWHULDO ZKHUH 5 DV ; <  ([5[[[G; ([\[G; ([\[G; f f = ([\G"L f

PAGE 98

7KXV IRU D SHUIHFWO\ ZKLWH PDWHULDO < LV DQG ; DQG = YDU\ GHSHQGLQJ XSRQ WKH OLJKW VRXUFH 7ULFKURPDWLF FRHIILFLHQWV [ \ ]f DUH RIWHQ FDOFXODWHG IURP WULVWLPXOXV YDOXHV [ ; f ; < = \ < f ; < = = f = f§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

PAGE 99

XQLIRUP FRORU VSDFH ZKHUH HTXDO GLVWDQFHV ZLWKLQ WKH VSDFH UHSUHVHQW QHDUO\ HTXDO YLVXDO FRORU GLIIHUHQFHV 7KH FXUUHQW PRVW FRPPRQO\ XVHG FRORU VSDFH LQ WKH ZRUOG LV WKH &,( /rDrEr VFDOH ZKLFK ZDV GHYHORSHG LQ 0DF@ F &,( /r Dr Er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f GLDJUDP LQ )LJXUH f 7KH JUHDWHU WKH GLVWDQFH EHWZHHQ WZR ZDYHOHQJWKV RQ WKH SHULPHWHU RI WKH SORW WKH JUHDWHU WKH UDQJH RI FRORUV WKDW FDQ EH SHUFHLYHG LQ WKH LQWHUYDO 7KH WKLUG GLPHQVLRQ LQGLFDWHG LQ WKH JUDSK LV WKH
PAGE 100

)LJXUH /XPLQRVLW\ RU OLJKWQHVV $GDSWHG IURP %LO@

PAGE 101

UDQJH RI FRORUV WKDW FDQ EH SHUFHLYHG GHFUHDVHV 7KLV KRUVHVKRHVKDSHG SORW H[KLELWV QRQXQLIRUP FRORU VSDFH 7KH &,( /rDrEr VFDOH SURYLGHV IRU PRUH HTXLYDOHQW YLVXDO GLVWDQFHV DQG HDVLHU LQWHUSUHWDWLRQ WKDQ WKH < [ \f VFDOH > -XG +XQ@ ,Q WKLV XQLIRUP V\VWHP FDUWHVLDQ FRRUGLQDWHV DUH HPSOR\HG WR UHSUHVHQW WKUHHGLPHQVLRQDO FRORU VSDFH DV WKH H\H VHHV FRORU ZLWK WKUHH VSHFWUDOO\ XQLTXH UHFHSWRUV 7KH VSDFH LV GHILQHG E\ D VHW RI HTXDWLRQV ZKLFK LQFRUSRUDWH WULVWLPXOXV YDOXHV IURP HTXDWLRQV DQG > +XQ@ /r < ;;Qf <<
PAGE 102

7DEOH )DFWRUV IRU 8QLIRUP &RORU 6FDOHV IRU 1RUPDOL]LQJ WR D 6WDQGDUG 5HIHUHQFH :KLWH r 2EVHUYHU f§ R 2EVHUYHU ,OOXPLQDQWV ;Q =Q ;Q &RRO ZKLWH IOXRUHVFHQW EXOE :HVWLQJKRXVH 8OWUDOXPH LQ DOO FDVHV IRU VWDQGDUG ZKLWH

PAGE 103

LQ HTXDWLRQV f WKURXJK f ZHUH IRXQG WR EHVW HTXDWH WKH ORJDULWKPLF UHODWLRQVKLS EHWZHHQ DFWXDO REMHFW OLJKWQHVV 6DQ@ DV GLVFXVVHG LQ 6HFWLRQ )LJXUH GHPRQVWUDWHV /rDrEr FRORU FRRUGLQDWHV 9DOXHV IRU Dr DQG Er DSSURDFK ]HUR IRU QHXWUDO FRORUV VXFK DV ZKLWHV JUD\V DQG EODFNV 7KH KLJKHU WKH Dr DQG Er YDOXHV WKH PRUH VDWXUDWHG WKH FRORU )RU H[DPSOH LQ FHUDPLF JOD]HV /rDrEr YDOXHV FDQ EH DSSOLHG WR VSHFLI\ FRORUV VXFK DV *OD]H &RORU /r Dr Er /LJKW *UD\ 'DUN *UD\ &U6Q 5HG &R&U *UHHQ =U3U LJ /r I DnDbf Er L Er ?OO= ff ZKHUH $(r FRORU GLIIHUHQFH DV GHILQHG E\ WKH PDJQLWXGH RI WKH SRVLWLRQ YHFWRU EHWZHHQ WZR SRLQWV LQ FRORU VSDFH H[SUHVVHG LQ -XGGV XQLWV

PAGE 104

/ EODFN r LL\UfOL £r ?^;,;DfP^<,<4fnQ@ Er >9$GDSWHG IURP 0RU@

PAGE 105

/rL/r OLJKWQHVV XQLWV RI REMHFWV DQG Dr LDr UHGJUHHQ XQLWV RI REMHFWV DQG ErLEr \HOORZEOXH XQLWV RI REMHFWV DQG ([SHULPHQWV E\ 0DF$GDP > 0DF$@ ZLWK KXPDQ REVHUYHUV DVFHUWDLQHG WKDW D $(r RI -XGGV UHSUHVHQWV D QRWLFHDEOH FRORU GLIIHUHQFH UHJDUGOHVV RI WKH FRORU LQ TXHVWLRQ 6LQFH $(r UHSUHVHQWV WKH PDJQLWXGH EXW QRW WKH GLUHFWLRQ RI FRORU GLIIHUHQFH LQGXVWU\ RIWHQ FRPSDUHV LQGLYLGXDO YDOXHV RI $/r $Dr DQG $Er EHWZHHQ WZR FRORUV 7KLV KHOSV WR XQFRYHU WKH EHVW DSSURDFK IRU DGMXVWLQJ PDWHULDO IRUPXODWLRQV WR PRGLI\ FRORUV 1RUPDOO\ $(r FDOFXODWLRQV DUH FRQILUPHG ZLWK D YLVXDO DVVHVVPHQW ,W LV FXVWRPDU\ WR VSHFLI\ WKH FRORU RI ZKLWHZDUH JOD]HV XVLQJ &,( /rDrEr YDOXHV > (SS@ &HUDPLF WLOH PDQXIDFWXUHUV IROORZ $PHULFDQ 1DWLRQDO 6WDQGDUG ,QVWLWXWH $16,f $ VSHFLILFDWLRQV ZKLFK VWLSXODWH $PHULFDQ 6WDQGDUG 7HVW 0HWKRG $670f & IRU PHDVXULQJ VPDOO FRORU GLIIHUHQFHV EHWZHHQ FHUDPLF WLOHV > 0XU $]D@ $16, VSHFLILHV D WROHUDEOH FRORU GLIIHUHQFH RI -XGGV IRU FRPSDWLEOH JOD]H DSSHDUDQFHV HYHQ WKRXJK WKH GLIIHUHQFH LV QRWLFHDEOH E\ KXPDQ REVHUYHUV *RRG UHVXOWV RI FRORU VSHFLILFDWLRQ DQG FRPSDULVRQ DUH DFKLHYHG IURP WKH /rDrEr V\VWHP LQ SDUW EHFDXVH WKH VWDQGDUG

PAGE 106

REVHUYHU ZHLJKWLQJ IXQFWLRQV DQG UHVSRQVH FXUYHV DUH DSSUR[LPDWHO\ OLQHDU RYHU D VKRUW VHJPHQW 2Q WKH RWKHU KDQG WKH /rDrEr V\VWHP FDQ SURYLGH DQRPDORXV UHVXOWV ZKHQ UHODWLQJ JUHHQ DQG UHG FRORUV WR SLJPHQW FRQFHQWUDWLRQ > (SS@ 7KH Dr LV D IXQFWLRQ RI ERWK ; [f DQG <\f WULVWLPXOXV YDOXHV DQG FRUUHVSRQGLQJ ZHLJKWLQJ IXQFWLRQV HTXDWLRQ f 7KHVH FXUYHV RYHUODS VLJQLILFDQWO\ )LJXUH f HVSHFLDOO\ LQ UHG QPf DQG JUHHQ QPf ZDYHOHQJWK UHJLRQV 7KHUH LV DOVR QRQOLQHDU FRUUHODWLRQ EHWZHHQ ; [f DQG <\f DQG Dr LV KHDYLO\ ZHLJKHG E\ ERWK IXQFWLRQV HVSHFLDOO\ LQ WKH UHG ZDYHOHQJWK UHJLRQ 7KXV UHG DQG JUHHQ FRORU LV D FRPSOH[ IXQFWLRQ RI ERWK /r DQG Dr &DVHV UHVXOW ZKHUH DV UHG RU JUHHQ SLJPHQW FRQFHQWUDWLRQ LV LQFUHDVHG WKH /rYDOXH GURSV EXW WKH DrYDOXH GRHV QRW FKDQJH 7KLV FDQ EH PLVOHDGLQJ ZKHQ HYDOXDWLQJ FRORU VWUHQJWK GXH WR WKH SLJPHQW 7KH .XEHOND0XQN DEVRUSWLRQ IXQFWLRQ GHVFULEHG LQ 6HFWLRQ E LV RIWHQ XVHG WR SURYLGH D EHWWHU TXDQWLWDWLYH FRUUHVSRQGHQFH IRU HYDOXDWLQJ UHG RU JUHHQ FRORUDQWV 7KH DEVRUSWLRQ IDFWRU .6f QRUPDOO\ KDV D OLQHDU GHSHQGHQF\ RQ SLJPHQW FRQFHQWUDWLRQ &RQYHUVHO\ WKH Er YDOXH LV HIIHFWLYH IRU HYDOXDWLQJ EOXH RU \HOORZ FRORU LQWHQVLW\ DV D IXQFWLRQ RI SLJPHQW

PAGE 107

FRQFHQWUDWLRQ 7KHUH LV YHU\ OLWWOH RYHUODSSLQJ RI \ DQG ] ZHLJKWLQJ IXQFWLRQV )LJXUH f LQ WKH EOXH QPf DQG \HOORZ QPf OLJKW UHJLRQV ,Q WKH VPDOO RYHUODSSHG DUHD WKHUH LV D VWURQJ LQYHUVH QHJDWLYH FRUUHODWLRQ DQG Er LV ZHLJKHG KHDYLO\ E\ HLWKHU ] RU \ EXW QRW ERWK 7KXV WKH UHODWLYH LQIOXHQFH RI EOXH DQG \HOORZ SLJPHQWV RQ FRORU LQWHQVLW\ LQ WHUPV RI Er DQG OLJKWQHVV /r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

PAGE 108

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

PAGE 109

QXFOHDWLRQ VLWHV DQG WHPSHUDWXUH 7KHVH YDULDEOHV ZLOO EH IXUWKHU GLVFXVVHG LQ 6HFWLRQ 6LOLFD LV WKH SUHGRPLQDQW JODVV IRUPHU XVHG LQ LQGXVWU\ .QRZOHGJH RI WKH SURSHUWLHV RI VLOLFDWH JODVVHV LV HVVHQWLDO IRU IRUPXODWLQJ JOD]HV DQG HQDPHOV 7KH VLOLFRQR[\JHQ WHWUDKHGURQ LV WKH EDVLF VWUXFWXUDO XQLW LQ DOO VLOLFDWH JODVVHV 7KH WHWUDKHGUD DUH OLQNHG WR HDFK RWKHU DW WKH FRUQHUV WR IRUP D WKUHHGLPHQVLRQDO UDQGRP QHWZRUN 7KLV VWUXFWXUH KDV QR ORQJUDQJH V\PPHWU\ RU SHULRGLFLW\ EH\RQG D IHZ WHWUDKHGUD > %DQ@ 'LUHFWLRQDO FRYDOHQW DQG LRQLF ERQGLQJ LQ DFFRUGDQFH ZLWK 3DXOLQJnV UXOHV SURPRWHV WKH IRUPDWLRQ RI 6Lfa VWUXFWXUDO XQLWV > &KL@ 6KRUW UDQJH RUGHU LV LGHQWLFDO LQ FU\VWDOOLQH DQG JODVV\ 6L ZKHUH IROG FRRUGLQDWLRQ DQG D 26L2 ERQG DQJOH RI r ZLWK HDFK WHWUDKHGUDO XQLW LV SUHVHQW %XW FRQQHFWLYLW\ LV PDLQWDLQHG ZLWKRXW FU\VWDOOLQH ORQJ WHUP RUGHU 'LIIHUHQFHV LQ WKH PHGLXP UDQJH RUGHU RI FU\VWDOOLQH DQG DPRUSKRXV 6L LQFOXGH 6L26L ERQG DQJOHV DQG URWDWLRQDO DQJOHV EHWZHHQ WHWUDKHGUD DQG WKH QXPEHU RI 6L ERQGV WKDW FRPSOHWH HDFK RI WKH ULQJV 7KHVH SDUDPHWHUV YDU\ WKURXJKRXW

PAGE 110

WKH JODVV EXW DUH IL[HG ZLWK FU\VWDOOLQH 6L )LJXUH FRPSDUHV FU\VWDOOLQH DQG DPRUSKRXV SKDVHV 6LOLFDWH JODVVHV IROORZ =DFKDULDVHQn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

PAGE 111

&5<67$/ 6758&785( 9,75(286 6758&785( 63(&7580 2) $ &5<67$//,=(' %2'< 63(&7580 2) $1 $0253+286 %2'< )LJXUH &RPSDULVRQ RI VWUXFWXUHV DQG ;5' SDWWHUQV RI FU\VWDOOLQH DQG YLWUHRXV VLOLFD > 6DF@

PAGE 112

7DEOH OLVWV JODVV IRUPHUV LQWHUPHGLDWHV DQG PRGLILHUV 0RVW JODVV IRUPHUV WHQG WR KDYH D KLJK LRQLF SRWHQWLDO !f ZKLFK LV LQGLFDWLYH RI D VWURQJ ERQG ZLWK R[\JHQ ZKLOH PRGLILHUV KDYH D ORZ LRQLF SRWHQWLDO ,Q PXOWLR[LGH JODVVHV VXEVWLWXWLRQV RI FDWLRQV ZLWK KLJKHU LRQLF SRWHQWLDOV HJ DONDOLQH HDUWKV IRU DONDOLVf WHQGV WR LQFUHDVH VWUHQJWK GXUDELOLW\ PHOWLQJ WHPSHUDWXUH DQG YLVFRVLW\ ,RQV VXFK DV $O =Q =U DQG 3E DUH OLVWHG LQ PRUH WKDQ RQH FDWHJRU\ LQ 7DEOH 'LH]HO > 6DF@ K\SRWKHVL]HG WKDW D FDWLRQnV FDSDFLW\ IRU IRUPLQJ PRGLI\LQJ RU DFWLQJ DV DQ LQWHUPHGLDWH GHSHQGV RQ LWV VLQJOHERQG VWUHQJWK ZLWK R[\JHQ UHODWLYH WR WKH RWKHU FDWLRQV LQ WKH VWUXFWXUH 7KXV DV WKH FRRUGLQDWLYH IRUFH WRZDUGV ERQGLQJ ZLWK R[\JHQ FKDQJHV WKH UROH RI WKH FDWLRQ WUDQVIRUPV *ODVV IRUPHUV KDYH WKH KLJKHVW ERQG VWUHQJWK ZLWK R[\JHQ ZKLOH PRGLILHUV KDYH WKH ORZHVW $QRWKHU FRQYHQLHQW ZD\ WR GHVFULEH WKH QHWZRUN FKDUDFWHU RI VLOLFDWH JODVVHV LV ZLWK 5 > .LQ@ VXP RI R[\JHQ DWRPVf f 5 f§‘ VXP RI QHWZRUN IRUPLQJ LRQVf )RU H[DPSOH JLYHQ D IRUPXOD RI JDWRP b 6L JDWRP b 1D DQG JDWRP b &D2

PAGE 113

7DEOH *ODVV )RUPHUV ,QWHUPHGLDWHV DQG 0RGLILHUV &RPPRQO\ (PSOR\HG LQ :KLWHZDUH *OD]HV > 6KD &KL@ 0 LQ 9DOHQFH ,RQLF &RRUGLQDWLRQ 'LVVRFLDWLRQ 6LQJOH %RQG 0; 3RWHQWLDOr 1XPEHU (QHUJ\ SHU 0; 6WUHQJWK NFDOPROHf NFDOPROHf *ODVV )RUPHUV 6L % % 9 $ =U ,QWHUPHGLDWHV 7L $L =U =Q 3E 0RGLILHUV 0J =Q &D 6U %D /L 1D O& 3E 3E R r ,RQLF 3RWHQWLDO YDOHQFHUDGLXV ƒf 7KH LRQLF SRWHQWLDO LQGLFDWHV WKH LPSRUWDQFH RI FKDUJH DQG VSDFH UHODWLRQVKLSV

PAGE 114

[f 5 *OD]HV DQG HQDPHOV QRUPDOO\ DUH LQ WKH UDQJH RI 5 WR $V 5 LQFUHDVHV D JUHDWHU QXPEHU RI R[\JHQ DUH VLQJOHERQGHG ZKLFK ZHDNHQV WKH JODVV VWUXFWXUH DQG ORZHUV WKH PHOWLQJ WHPSHUDWXUH DQG YLVFRVLW\ )RU JODVVHV ZLWK RQO\ RQH W\SH RI JODVV IRUPLQJ LRQ ; < = f DQG ; < 5 f ZKHUH = LV WKH QXPEHU RI R[\JHQV VXUURXQGLQJ WKH FDWLRQ ; LV WKH QXPEHU RI QRQEULGJLQJ R[\JHQV VLQJO\ ERQGHGf SHU SRO\KHGURQ DQG < LV WKH QXPEHU RI EULGJLQJ R[\JHQV SHU SRO\KHGURQ &RUUHVSRQGLQJ UXOHV WKDW DSSO\ IRU VLOLFDWH JODVVHV DUH 7KH 6L WHWUDKHGUD GLFWDWHV WKDW = 7KH PLQLPXP UHTXLUHPHQW IRU D WKUHHGLPHQVLRQDO QHWZRUN LV < :KHQ DONDOL RU DONDOLQH HDUWKV DUH SUHVHQW ZLWK $OR WKH $O RFFXSLHV WKH FHQWHUV RI $ WHWUDKHGUD ZKLFK FRQYHUWV QRQEULGJLQJ WR EULGJLQJ R[\JHQ +HQFH ZKHQ $O LV D QHWZRUN IRUPLQJ LRQ WKH $ FRQWULEXWLRQ LV 5

PAGE 115

:KHQ FRPSDULQJ WKH PRVW FRPPRQ LQGXVWULDO JODVV IRUPHUV 6L DQG % ERURQ IRUPV HTXLODWHUDO WULDQJOH IROG FRRUGLQDWLRQ ZLWK WKH %-O LRQ SODFHG DW WKH FHQWHU 6LQFH VLOLFRQ DFFRPPRGDWHV IRXU R[\JHQ ERQGV WR ERURQnV ERURQn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n EULGJLQJ R[\JHQ DQG D ORVV RI FRQQHFWLYLW\ 2QH 1D MRLQV DQ XQVDWLVILHG R[\JHQ ERQG RQ RQH VLGH ZKLOH WKH UHPDLQLQJ 1D2a JURXS RFFXSLHV WKH YDFDQF\ DW WKH FRUQHU RI WKH WHWUDKHGUD DQG EUHDNV WKH VWUXFWXUH :KHQ PRQRYDOHQW PRGLILHUV DUH UHSODFHG ZLWK GLYDOHQW LRQV VXFK DV &D WKH EULGJH DFURVV WKH JDS LV FRPSOHWHG 7KLV UHVXOWV LQ

PAGE 116

f 6LOLFRQ LRQ 2 %ULGJLQJ R[\JHQ LRQ 1RQ EULGJLQJ R[\JHQ LRQ Df Ef )LJXUH 7ZRGLPHQVLRQDO UHSUHVHQWDWLRQ RI PRGLILHUV Df 1DL DQG Ef &D LQ WKH VLOLFDWH JODVV VWUXFWXUH >$GDSWHG IURP 7D\ @

PAGE 117

$Q LQFUHDVH LQ GHQVLW\ $Q LQFUHDVH LQ UHIUDFWLYH LQGH[ $ GHFUHDVH LQ FDWLRQ PRELOLW\ DQG HOHFWULFDO FRQGXFWLYLW\ $Q LQFUHDVH LQ 7J DQG PHOWLQJ WHPSHUDWXUH $Q LQFUHDVH LQ YLVFRVLW\ $OXPLQXP LV WKH PRVW FRPPRQ LQWHUPHGLDWH LQFRUSRUDWHG LQ JOD]H IRUPXODWLRQV %\ LWVHOI $O FDQQRW IRUP D JODVV VWUXFWXUH EXW LW FDQ UHSODFH VRPH VLOLFRQ LRQV LQ WKH WHWUDKHGURQV 7KH UHVXOW LV D KLJKHU ODWWLFH ERQGLQJ VWUHQJWK ZKLFK LQFUHDVHV YLVFRVLW\ FKHPLFDO UHVLVWDQFH DQG KDUGQHVV 7KH UROH RI $O GHSHQGV ODUJHO\ RQ WKH FRQFHQWUDWLRQ RI DONDOL LRQV ,I DQ DGMDFHQW DONDOL LRQ LV SUHVHQW DQG DFKLHYHV FKDUJH QHXWUDOLW\ $O FDQ VXEVWLWXWH IRU 6L ,Q 6L$O1D JODVVHV LI WKH UDWLR RI PROHV RI $O1D $O1D FDQ VXEVWLWXWH IRU 6L DQG H[FHVV 1D DFWV LQ LWV QRUPDO PRGLILHU UROH $V $O1D DSSURDFKHV JODVV VWUHQJWK DQG YLVFRVLW\ LQFUHDVH $W $O1D RU WKH HTXLYDOHQF\ SRLQW DOXPLQXP LV WHWUDKHGUDOO\ FRRUGLQDWHG $*f DQG WKH JODVV VWUHQJWK LV PD[LPXP )RU $O1D H[FHVV $ODFWV DV D PRGLILHU 6XEVWLWXWLRQV RI VPDOO

PAGE 118

DPRXQWV RI $O IRU 1D RU &D UDLVHV WKH VWUHQJWK EHFDXVH RI WKH KLJK ERQG VWUHQJWK WULYDOHQF\ SURYLGHV /DUJH VXEVWLWXWLRQV IRU VLOLFRQ ZHDNHQV WKH JODVV VWUXFWXUH GXH WR WKH ORZHU LRQLF SRWHQWLDO RI $OL >&KL@ 7DEOH VXPPDUL]HV SURSHUWLHV DVVRFLDWHG ZLWK WKH SUHVHQFH RI R[LGHV FRPPRQO\ HPSOR\HG LQ FHUDPLF JOD]H DQG JODVV IRUPXODWLRQV &RPSXWHU SURJUDPV KDYH EHHQ GHYHORSHG > 'LQ 'LQD 0DO@ WR XVH WKH DGGLWLYH UXOH IRU HVWLPDWLQJ JODVV OLQHDU SURSHUWLHV DV D IXQFWLRQ RI WKH R[LGH FRPSRVLWLRQ 3 F[ &[ &; FQ[Q f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

PAGE 119

7DEOH 3URSHUWLHV $VVRFLDWHG ZLWK WKH 3UHVHQFH RI 9DULRXV 2[LGHV LQ *ODVV > 0XH /DF 9LF@ 6L +LJK PHOWLQJ SRLQW +LJK YLVFRVLW\ +LJK FKHPLFDO UHVLVWDQFH /RZ H[SDQVLRQ ,QFUHDVHV JORVV %r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

PAGE 120

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r LQFLGHQFHf DQG LQFRUSRUDWH PDWHULDOV WKDW SURYLGH R[LGHV RI 6L SULPDU\ JODVV IRUPHUf $ &D2 0J2 1D % 6U2 =Q2 DQG =U %DWFK LQJUHGLHQWV DUH RIWHQ PXOWLFRPSRQHQW DQG PXOWLSKDVH FRPSRXQGV PDNLQJ LW GLIILFXOW WR SUHGLFW ILUHG SURSHUWLHV IURP WKHVH FRPSRQHQWV *OD]H RU IULW EDWFK FRPSRVLWLRQV DUH QRUPDOO\ FRQYHUWHG LQWR DQ HPSLULFDO IRUPXOD RU PRODU HTXLYDOHQWV XVLQJ 6HJHUnV IRUPXOD LQ RUGHU WR EHWWHU FODVVLI\ WKH V\VWHP > (SSE@ 7R GHULYH DQ HPSLULFDO IRUPXOD PDWHULDOV DUH H[SUHVVHG LQ WHUPV RI PROHV RI LQGLYLGXDO R[LGHV SUHVHQW WKHQ GLYLGHG E\ WKH VXP RI WKH PROHV RI 52 JURXS R[LGHV DV VKRZQ LQ 7DEOH > 6DF@ 7KXV WKH

PAGE 121

7DEOH &RPPRQ &HUDPLF 7LOH *OD]H %DVH 0DWHULDOV > %UD (SSE 6KH@ 0LQHUDOV &RPPRQ 7UDGH 1DPHV 0DMRU )XQFWLRQV &OD\ $O6L+ *OD]H (3. 20 6XVSHQVLRQ ILOP VWUHQJWK UKHRORJ\ FRQWURO )ULW JODVV FRPSRVLWLRQ 3. && 3 ,QWURGXFHV R[LGHV VXFK DV 6U2 DQG % DIIHFWV RYHUDOO JOD]H SURSHUWLHV FRORU GHYHORSPHQW JORVV ILULQJ UDQJH WKHUPDO H[SDQVLRQ Df HWF H[SHQVLYH 6LOLFD 6L 6LO&R6LO 3ULPDU\ JODVV IRUPHU LQFUHDVHV JORVV PS DQG PHOW YLVFRVLW\ ORZHUV D $OXPLQD $ $ & ,QFUHDVHV KDUGQHVV GXUDELOLW\ PS PHOW YLVFRVLW\ DQG ILULQJ VWDELOLW\ ORZHUV JORVV DQG D )HOGVSDU 1D‘$f6L 1& ) ,QH[SHQVLYH VRXUFH RI R[LGHV 1D ORZHUV PS PHOW YLVFRVLW\ ILULQJ VWDELOLW\ DQG GXUDELOLW\ LQFUHDVHV D =LQF 2[LGH =Q2 'HQ]R[ =RFKHP ,QFUHDVHV ILULQJ UDQJH DQG GXUDELOLW\ LPSURYHV JOD]H WH[WXUH DQG HODVWLFLW\ WR DYRLG FUD]LQJ H[SHQVLYH :KLWLQJ &D& ,QFUHDVHV GXUDELOLW\ FDQ PDWWHQ DQG FUHDWH RUDQJH SHHO WH[WXUH :ROODVWRQLWH &D6L 9DQVLO : ,QWHUPHGLDWH SURSHUWLHV EHWZHHQ ZKLWLQJ DQG VLOLFD PDWWHQV ZLWKRXW RUDQJH SHHO WH[WXUH QR YRODWLOHVf 2SDFLILHU =U6L 8OWUR[ 6XSHUSD[ $GGV ZKLWHQHVV LPSURYHV FRORU VWDELOLW\ DQG JOD]H KDUGQHVV H[SHQVLYH &KHPLFDOV 'HIORFFXODQW 6RGLXP +H[DPHWD SKRVSKDWH 4XDGURIRV 9LWURIRV /RZHUV YLVFRVLW\ FDQ DLG LQ VXVSHQVLRQ %LQGHU 6RGLXP FDUER[\PHWK\O FHOOXORVH &0&/ %LQGV JOD]H WR ERG\ LQFUHDVHV ILOP VWUHQJWK VORZV GU\LQJ DLGV LQ VXVSHQVLRQ

PAGE 122

7DEOH ([DPSOHV RI &RPSRVLWLRQV :HLJKW bf RI &RPPHUFLDO *OD]HV > 3DU 7D\ 7H\@ 2[LGH )DVW)LUH 0DWWH :DOO 7LOH )DVW)LUH *ORVV 2SDTXH :DOO 7LOH (OHFWULFDO 3RUFHODLQ 'LQQHUZDUH 6DQLWDU\ :DUH 6L $O %r =U /L 1D" NR 0J2 &D2 =Q2 6U2 %D2 3E2r r 3E2 LV VWLOO XVHG LU GLQQHUZDUH JOD]HV EXW LV UDUHO\ LQFRUSRUDWHG LQ RWKHU ZKLWHZDUHV FRDWLQJV GXH WR LWV SRWHQWLDO KD]DUGV WR WKH HQYLURQPHQW

PAGE 123

7DEOH 6XPPDU\ RI ,PSRUWDQW *OD]H 3URSHUWLHV DQG &KDUDFWHULVWLFV > 5HH@ *OD]H 6WDWH 3URSHUW\ &KDUDFWHULVWLFV 0DLQ ,QIOXHQFHV :HW 5KHRORJ\ $SSOLFDWLRQ VPRRWKQHVV DQG GHIHFWV VXVSHQVLRQ 'HQVLW\ $SSOLFDWLRQ VPRRWKQHVV DQG GHIHFWV 3DUWLFOH 6L]H 5KHRORJ\ DQG PHOWLQJ EHKDYLRU 3DUWLFOH VXUIDFH FKHPLVWU\ 'LVSHUVLRQ VWDELOLW\ DQG UKHRORJ\ *UHHQ RQ 7KLFNQHVV DSSOLHG &RORU JORVV GHIHFWV DQG FRVW VXEVWUDWH $GKHVLRQ WR VXEVWUDWH 'HIHFWV VXFK DV FUDZOV b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f 6XUIDFH VPRRWKQHVV 'HIHFWV &RVW LQIOXHQFHV JORVVf

PAGE 124

7DEOH 6HJHUnV )RUPXOD IRU &ODVVLI\LQJ *OD]HV > 3DU 6DF @ 52 JURXS EDVLFf 5 QHXWUDO RU DPSKRWHULFf 5 DFLGLFf 1D NR /L DLR %2 6 2 U =U 0J2 &D2 %D2 )H &U 6Q 7L 6U2 =Q2 0Q2 %L &2 I 7K 3E2 )H2 %H2 &G2 &X2 1L2 &R2 6DPSOH &DOF XODWLRQV f f f f f &KHPLFDO 2[LGH 6HJHUnV (PSLULFDO $QDO\VLV 0ROHFXODU )RUPXOD RU 0RODU )RUPXOD 2[LGH :W b :HLDKW 0ROHV (DXLYDOHQWV :HLDKWr Off f 6XP 52 0ROHV f [f 6 52 PRODU HTXLYDOHQWV f r2[LGH IRUPXOD ZHLJKWfWRWDO IRUPXOD ZHLJKWef ZW b f

PAGE 125

VXPPDWLRQ RI PRODU HTXLYDOHQWV RI 52 RU DONDOLHV DQG DONDOLQH HDUWK R[LGHV LV DOZD\V EURXJKW WR XQLW\ DQG HTXDOV 7KLV SURYLGHV D VWDQGDUG V\VWHP IRU RUJDQL]LQJ UHSUHVHQWLQJ DQG SUHGLFWLQJ ILUHG SURSHUWLHV DVVRFLDWHG ZLWK WKH SUHVHQFH RI VSHFLILF R[LGHV LQ WKH JODVV VWUXFWXUH 7DEOH f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

PAGE 126

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

PAGE 127

FHUDPLF ZKLWHZDUHV VXFK DV VDQLWDU\ZDUH DQG GLQQHUZDUH DUH DOVR FXUUHQWO\ DWWHPSWLQJ WR FRQYHUW WR WKHVH W\SHV RI SURFHVVHV 7KH\ SURYLGH PXFK JUHDWHU SURGXFWLYLW\ DQG ORZHU HQHUJ\ FRVWV WKDQ WKH VORZILUH WXQQHO NLOQ WHFKQRORJ\ WKDW ZDV SUHGRPLQDQW SULRU WR )DVWILUH WLOHV DUH KHDWHG XWLOL]LQJ WR PLQXWH F\FOHV DW SHDN WHPSHUDWXUHV RI r& WR r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

PAGE 128

)RU WKHVH UHDVRQV PDQXIDFWXUHG JODVV IULWV DUH WKH PDLQ EDWFK LQJUHGLHQWV LQ IDVWILUH JOD]HV > 6DF 7R] (QU@ 0RVW IULW LQ WKH 86 LV FXUUHQWO\ XVHG E\ FHUDPLF WLOH PDQXIDFWXUHUV EXW IULW FRQVXPSWLRQ E\ RWKHU ZKLWHZDUHV LQGXVWULHV ZLOO LQFUHDVH DV WKH\ FRQYHUW WR IDVWILUH UROOHU NLOQ WHFKQRORJ\ )ULWV DUH SUHSDUHG E\ PHOWLQJ D EOHQG RI FU\VWDOOLQH UDZ PDWHULDOV DW DSSUR[LPDWHO\ r& DQG WKHQ IDVW TXHQFKLQJ LQ ZDWHU RU ZLWK FRROHG VWDLQOHVV VWHHO UROOV WR IRUP D JODVV 7KHUH DUH KXQGUHGV RI GLIIHUHQW FRPSRVLWLRQV IRUPXODWHG WR SURYLGH VSHFLILF FRDWLQJV SURSHUWLHV ZKLOH XVLQJ ORZ FRVW DQG HQYLURQPHQWDOO\ VDIH PDWHULDOV > %DU@ )ULW FRPSRVLWLRQV FDQ DOVR LQIOXHQFH ZHW JOD]H UKHRORJ\ WKURXJK VXUIDFH FKDUJH GHYHORSPHQW DQG H[FKDQJH RI VROXEOH LRQV ZLWK K\GUR[\O RU K\GURQLXP LRQV >
PAGE 129

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f DQG % DUH ORZHU ZKLOH R[LGHV VXFK DV &D2

PAGE 130

=Q2 DQG 6U2 DUH KLJKHU >$PR 7HX@ 7KLV DLGV LQ GHOD\LQJ WKH IXVLRQ WHPSHUDWXUH DQG WKXV NHHSV WKH FRDWLQJ SHUPHDEOH IRU D ORQJHU SHULRG RI WLPH DQG IDFLOLWDWHV ERG\ RXWJDVLQJ ZLWKRXW WKH FUHDWLRQ RI JOD]H EXEEOHV 0DMRU FDWHJRULHV RI IDVWILUH IULWV WHQG WR YDU\ PDLQO\ LQ DONDOLVLOLFD UDWLR XVH RI 6U2 YHUVXV =Q2 IRU WKH VHFRQGDU\ IOX[ DQG =U FRQWHQW 7KH =U QRUPDOO\ UDQJHV IURP b WR b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f

PAGE 131

7DEOH OLVWV W\SLFDO HPSLULFDO IRUPXODV IRU IDVWILUH WLOH JORVV DQG PDWWH JOD]HV 7KH ODUJHVW YROXPH RI ZKLWHZDUHV VROG ZRUOGZLGH LV FRDWHG ZLWK RSDTXH JORVV JOD]H 2SDFLW\ DULVHV PDLQO\ IURP ]LUFRQLXP FU\VWDOOLQH SKDVHV ]LUFRQLD DQG ]LUFRQf DGGHG WR WKH EDWFK DV D =U6L PLQHUDO RU GLYLWULILHG IURP WKH IULW 7KH UHVXOWLQJ PLFURKHWHURJHQHLWLHV SPf KDYH D VLJQLILFDQWO\ GLIIHUHQW LQGH[ RI UHIUDFWLRQ f IURP WKH JODVV\ PDWUL[ af§f DQG WKXV UHIOHFW DQG VFDWWHU OLJKW 7UDQVSDUHQW JOD]HV DQG IULWV FRQWDLQ OLWWOH RU QR =UR DQG LQFOXGH PRUH &D2 DQG $ WR PDNH XS IRU WKH ORVW PHFKDQLFDO VWUHQJWK LH 0RKV KDUGQHVV DQG DEUDVLRQ UHVLVWDQFHf 2WKHUZLVH WKHLU FRPSRVLWLRQV DUH VLPLODU +LJK &D2 DQG =Q2 LQ WKHVH JOD]HV FDQ FDXVH LPPLVFLEOH OLTXLGOLTXLG SKDVH VHSDUDWLRQ DQG RSDOHVFHQW RSDFLILFDWLRQ WKDW LV QRW RWKHUZLVH QRWLFHDEOH LQ ]LUFRQRSDFLILHG JOD]HV 3KDVH VHSDUDWLRQ YDULHV ZLWK WKH ILULQJ F\FOH DQG FDQ LPSDUW VKDGH YDULDWLRQV LQ FRORUHG JOD]HV 0DWWH JOD]HV KDYH PDQ\ FU\VWDOOLQH SKDVHV ZKLFK FUHDWH ORFDO GLVFRQWLQXLWLHV LQ UHIUDFWLYH LQGH[ 7KLV \LHOGV KLJK GLIIXVH RSDFLW\f DQG ORZ VSHFXODU JORVVf UHIOHFWDQFHV 7KH

PAGE 132

,OO 7DEOH 7\SLFDO (PSLULFDO )RUPXODV LQ 0RODU (TXLYDOHQWV IRU )DVW)LUH &HUDPLF 7LOH *ORVV DQG 0DWWH *OD]HV >$PR $SD 7R]@ 2SDTXH *ORVV 7UDQVSDUHQW *ORVV &DOFLXP 0DWWH $OXPLQD 0DWWH 6L =Ur DLR 1D &D2 0J2 6U2 =Q2 1RWHV rf )ULWV QRUPDOO\ FRQWDLQ =U LQ WKH UDQJH RI WR $GGLWLRQDO =U LV DGGHG WR WKH JOD]H EDWFK LQ WKH PLQHUDO IRUP =U6L&! f &OD\ LV DGGHG WR JOD]HV DV D VXVSHQGLQJ DJHQW WKHUHIRUH JORVV IULWV QRUPDOO\ LQFRUSRUDWH OHVV $&! WKDQ OLVWHG DERYH

PAGE 133

WKUHH W\SHV RI PDWWH JOD]HV DUH SURGXFHG ZLWK KLJK &D2 =Q2 RU $ FRQFHQWUDWLRQV &DOFLXP PDWWHV GHULYH &D2 IURP WKH IULW ZKLWLQJ &D&2f RU ZROODVWRQLWH &D6Lf 7KH &D LRQ LQ JODVV KDV D KLJK ERQG VWUHQJWK DQG FU\VWDO IRUPDWLRQ LV HQHUJHWLFDOO\ IDYRUDEOH > 7R]@ ,I WKH FRQFHQWUDWLRQ RI LRQV a! b E\ ZHLJKW &D2f DQG FRROLQJ UDWH DUH VXIILFLHQW ZROODVWRQLWH DQG VRPHWLPHV GLRSVLGH &D2f0J2f6Lf FU\VWDOV IRUP > (VF@ =LQF PDWWHV DUH SURGXFHG ZKHQ ODPHOODU KH[DJRQDO ZLOOHPLWH =Q6Lf FU\VWDOV IRUP GXULQJ WKH FRROLQJ F\FOH 7\SLFDOO\ DW OHDVW b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

PAGE 134

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

PAGE 135

0(7$/ 2;,'(6 )H 2L 6SLQHOV &G 3LJP 0HWDOOLFV 0DJQHWLWH 7L2W %D62M &U 5XWLOHV &U %XU @

PAGE 136

DQG WKH ODQWKDQLGH JURXSn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f SLJPHQW V\VWHPV KDYH EHHQ XVHG LQ WKH FHUDPLFV LQGXVWU\ IRU RYHU \HDUV > %ORD@ =LUFRQnV WHWUDJDQRO VWUXFWXUH KDV WKH DELOLW\ WR DFFRPPRGDWH D QXPEHU RI LPSXULW\ LRQV VXEVWLWXWLRQDOO\ DQG LWV KLJK FKHPLFDO DQG WKHUPDO VWDELOLW\ PDNH LW LGHDO IRU XVH LQ FHUDPLFV FRDWLQJV

PAGE 137

E R 4 =U6c3U T (SS@

PAGE 138

7DEOH *OD]H 3LJPHQWV DQG 7KHLU 5HTXLUHPHQWV > 7D\@ &RORU 3LJPHQW 0D[LPXP )LULQJ 7HPSHUDn WXUH r&f *OD]H %DVH &KDUDFWHULVWLFV IRU 2SWLPXP &RORU 6WUHQJWK DQG 6WDELOLW\ %OXH OLJKW =LUFRQ YDQDGLXPr 6XLWDEOH IRU XVH LQ DOO JOD]HV GDUN 2OLYLQH FREDOW VLOLFRQ 6XLWDEOH IRU XVH LQ DOO JOD]HV OLJKW 6SLQHO FREDOW ]LQF DOXPLQLXP 6XLWDEOH IRU XVH LQ DOO JOD]HV *UHHQ OLJKW =LUFRQ YDQDGLXPSUDVHRG\PLXP 6XLWDEOH IRU XVH LQ DOO JOD]HV GDUN 6SLQHO FKURPLXP FREDOW ]LQFf 6XLWDEOH IRU XVH LQ PRVW ]LQFIUHH JOD]HV OLJKW *DUQHW FDOFLXP FKURPLXP VLOLFRQ =LQFIUHH KLJK FDOFLXP FRPSRVLWLRQV 2UDQJH %DGGHOH\LWH LQGLXP \WWULXPf YDQDGLXP 6XLWDEOH IRU XVH LQ DOO JOD]HV HVSHFLDOO\ WKRVH ILULQJ DW KLJK WHPSHUDWXUHV 3\URFKORUH OHDG DQWLPRQ\ +LJK OHDG FRQWHQW JOD]HV 5XWLOH WLWDQLXP FKURPLXP /HDGOHVV JOD]HV
PAGE 139

7DEOH FRQWLQXHG &RORU 3LJPHQW 0D[LPXP )LULQJ 7HPSHUDn WXUH r&f *OD]H %DVH &KDUDFWHULVWLFV IRU 2SWLPXP &RORU 6WUHQJWK DQG 6WDELOLW\ 3LQN =LUFRQ LURQr =LUFRQ FDGPLXP VXOSKRVHOHQLGH 6SKHQH FKURPLXP WLQ VLOLFRQ 6SLQHO FKURPLXP DOXPLQXP ]LQF 6XLWDEOH IRU XVH LQ DOO JOD]HV 6XLWDEOH IRU XVH LQ DOO JOD]HV =LQFIUHH KLJK FDOFLXP FRPSRVLWLRQV +LJK ]LQF DQG DOXPLQLXP JOD]H ZLWK ORZ FDOFLXP FRQWHQW %URZQ GDUN 6SLQHO LURQ FKURPLXP ]LQF LURQ FKURPLXP QLFNHOf 9HU\ VXLWDEOH IRU XVH LQ ]LQFFRQWDLQLQJ JOD]HV 9HU\ VXLWDEOH IRU XVH LQ ]LQFFRQWDLQLQJ JOD]HV OLJKW LURQ DOXPLQLXP FKURPLXP ]LQF 9HU\ VXLWDEOH IRU XVH LQ ]LQFFRQWDLQLQJ JOD]HV %ODFN 6SLQHO FKURPLXP FREDOW LURQ QLFNHO =LQFIUHH FRPSRVLWLRQV *UH\ &DVVLWHULWH DQWLPRQ\ YDQDGLXP 6XLWDEOH IRU XVH LQ DOO JOD]HV r 9DQDGLXP SUDVHRG\PLXP DQG LURQ ]LUFRQV DUH WKH PRVW FRPPRQ WULD[LDO SLJPHQWV RR

PAGE 140

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b WR b E\ ZHLJKW ZLWK PRVW V\VWHPV FRQWDLQLQJ FORVHU WR b WR b =LUFRQEDVHG SLJPHQWV IXUQLVK D ORZHU FRORU VWUHQJWK WKDQ PHWDO R[LGHV EXW SURYLGH PXFK PRUH RSDFLILFDWLRQ DQG KLJK WHPSHUDWXUH VWDELOLW\ 7KH\ DUH LQVHQVLWLYH WR YDULDWLRQV LQ NLOQ DWPRVSKHUH DQG WKHLU VROXELOLW\ YDULHV ZLWK WKH PHOW FRPSRVLWLRQ =LUFRQ LV LVRPRUSKRXV ZLWK D WHWUDJRQDO VWUXFWXUH P P P RQH IRXUIROG DQG WZR WZRIROG D[HV ZLWK SHUSHQGLFXODU V\PPHWU\ SODQHVf DV VKRZQ LQ )LJXUH ,W FRQVLVWV RI D FKDLQ RI DOWHUQDWLQJ HGJHVKDULQJ 6L WHWUDKHGUD DQG =U WULDQJXODU GRGHFDKHGUD H[WHQGLQJ SDUDOOHO WR WKH FD[LV 7KH SULVPDWLF KDELW RI ]LUFRQ FRPSRXQGV UHVXOWV IURP HGJHVKDULQJ

PAGE 141

H )LJXUH 7\SLFDO IRUPV RI ]LUFRQ FU\VWDOV DFf D^ ` P^ ` S^ ` [^ ` DQG ]LUFRQ ODWWLFH VWUXFWXUH G Hf > :\F 0DV@

PAGE 142

GRGHFDKHGUD ZKLFK MRLQ WKH FKDLQV 7KHUH DUH IRXU PROHFXOHV SHU XQLW FHOO > *UH@ 7KH =U SRO\KHGUD DUH WZR LQWHUORFNHG WHWUDJRQDOO\ GLVWRUWHG WHWUDKHGUD 2QH WHWUDKHGURQ LV FRPSUHVVHG WKH RWKHU LV VWUHWFKHG SDUDOOHO WR WKH FD[LV 7KH 6L WHWUDKHGUD DUH HORQJDWHG SDUDOOHO WR WKH FD[LV 7KH WZR DQJOHV ELVHFWHG E\ WKH FD[LV DUH OHVV WKDQ WKH r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b =U b 6L b 9 RU 1+9 DQG b

PAGE 143

1D) E\ ZHLJKW 7KH PDWHULDOV DUH KHDWHG WR r& IRU RQH WR HLJKW KRXUV WR FRPSOHWH WKH VROLG VWDWH UHDFWLRQ ,W LV WKHQ ZDVKHG DQG OHDFKHG WR UHPRYH VROXEOH 9 EHFDXVH 9 SURYLGHV XQZDQWHG JUHHQ FRORU > 3DUM :LWKRXW D PLQHUDOL]HUFDWDO\VW VXFK DV 1D) WKH 9 ZKLFK SURYLGHV WKH EOXH FRORU GRHV QRW IRUP RQ WKH ]LUFRQ ODWWLFH 1D) DLGV LQ WKH GHYHORSPHQW RI WKH VROLG VWDWH VROXWLRQ E\ ORZHULQJ WKH UHTXLUHG VLQWHULQJ WHPSHUDWXUH DQG LQFUHDVLQJ WKH PRELOLW\ RI =U DQG 6L IRU IRUPLQJ ]LUFRQ :LWKRXW D FDWDO\VW PHOWLQJ RFFXUV IRU =U DW r& DQG 6L DW r& ZKLOH =U6L EHJLQV WR GHFRPSRVH DW r& > :HD 6KD@ 6RPH GHWDLOV UHJDUGLQJ WKH RULJLQ RI WKH EOXH FRORU DUH VWLOO XQH[SODLQHG $ YDOHQF\ RI 9 LV SUHVHQW EXW WKH FULWLFDO UROH RI 1D) DQG WKH H[DFW ORFDWLRQ RI 9 LQ WKH ODWWLFH UHPDLQ REVFXUH :LWK DQ LRQLF UDGLXV RI ƒ 9 FRXOG WKHRUHWLFDOO\ RFFXS\ 6L ƒf =U ƒf RU LQWHUVWLFLDO VLWHV > 5LF 6KD@ 'HPLUD\ 1DWK DQG +XPPHO VXUPLVHG WKDW 9 LV VROHO\ UHVSRQVLEOH IRU WKH EOXH FRORU DQG RFFXSLHV WKH GRGHFDKHGUDO =U VLWH =Uf 7KH\ DOVR VWDWHG WKDW 1D) LQFUHDVHG WKH VROXELOLW\ DQG WKHUPDO VWDELOLW\ RI 9 LQ ]LUFRQ EXW GLG QRW DIIHFW FRRUGLQDWLRQ RU YDOHQFH > 3DU@ *UHJRULR *UHHQEODWW 3LIHU DQG 6WXUJH SXEOLVKHG WKDW WKH

PAGE 144

RSWLFDO VSHFWUD VXJJHVWV WKDW 9 HQWHUV WKH WHWUDKHGUDO 6L 6Lf UDWKHU WKDQ GRGHFDKHGUDO VLWH > *UH@ 9DQDGLXP PHWDO KDV DQ RXWHU HOHFWURQLF VWUXFWXUH RI G V 7KH FDWLRQ 9 ZKLFK SURGXFHV D EOXH FRORU KDV D G Vr FRQILJXUDWLRQ :KHQ WKH FRRUGLQDWLRQ RI DQLRQV DERXW D VLQJOH G HOHFWURQ LV IRXUIROG RU HLJKWIROG WKH HOHFWURQnV LQLWLDO ILYHIROG GHJHQHUDF\ VKRZQ LQ )LJXUH f LV GLVUXSWHG 7KLV SKHQRPHQRQ LV GHVFULEHG ZLWK OLJDQG ILHOG WKHRU\ LQ 6HFWLRQ G 7KH DQLRQ FRRUGLQDWLRQ DERXW 9 LQ WKH ]LUFRQ VWUXFWXUH UHVXOWV LQ WKH RULJLQDO G RUELWDO HOHFWURQ HQHUJ\ OHYHO JURXQG VWDWH 'f VSOLWWLQJ LQWR D QHZ GRXEOHW JURXQG VWDWH (f DQG DQ H[FLWHG WULSOHW VWDWH f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

PAGE 145

' G FU 77 ([]\]f %[ \f $W=f [\f )LJXUH 6SOLWWLQJ RI WKH G RUELWDO LQ 9 E\ WHWUDKHGUDO 7Gf DQG WHWUDJRQDO 'Gf FU\VWDO ILHOGV 3RVVLEOH HOHFWURQ WUDQVLWLRQV DUH VKRZQ ZLWK GRWWHG DQG VROLG OLQHVf >$GDSWHG IURP *UH@

PAGE 146

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b =U& b 6L b 1D) DQG b )H E\ ZHLJKW 7KH )H ORVHV DQ HOHFWURQ DQG FRQYHUWV WR )H Gf DQG VXEVWLWXWHV IRU =U LQ ]LUFRQ 7KLV UHVXOWV LQ DSSUR[LPDWHO\ b =U6L ZLWK b )H LQ WKH ODWWLFH 7KH ZDYHOHQJWK RI LQFLGHQW OLJKW DEVRUEHG E\ WKH VWUXFWXUH IDOOV LQ WKH UDQJH RI QP QP > 3DU@ =LUFRQSUDVHRG\PLXP \HOORZ LV PDQXIDFWXUHG E\ UHDFWLQJ b =U b 6L b 1D) b .1D2 DQG b 3U2Q E\ ZHLJKW 7KHUH LV GLVFRUG LQ WKHRULHV UHJDUGLQJ WKH VROLG VWDWH UHDFWLRQV WKDW RFFXU LQ IRUPLQJ WKH =U3U VWUXFWXUH 'HFNHU > 'HF VXJJHVWHG WKDW 3U2Q LV UHGXFHG WR 3U DQG WKHQ 3U FRQYHUWV WR 3U E\ ORVLQJ DQ HOHFWURQ 7KH 3UU VXEVWLWXWHV IRU =U LQ ]LUFRQ DQG VRPH XQUHDFWHG

PAGE 147

3U2O/ LV OHIW GLVSHUVHG LQ WKH PDWHULDO 8QILOOHG I RUELWDOV f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n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

PAGE 148

$OWKRXJK ]LUFRQ GRSHG SLJPHQWV DUH WKH PRVW VWDEOH FHUDPLF FRORUDQWV XS WR DSSUR[LPDWHO\ r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f DQG WKH RWKHU XSZDUG Mf WKURXJK DQ HOHPHQWDU\ FRORUDQW OD\HU RI G[ ZKLFK VWDUWV DW ]HUR IURP WKH XQLOOXPLQDWHG VLGH DV VKRZQ LQ )LJXUH 7KH WKLFNQHVV RI G[ LV VPDOO FRPSDUHG WR WKH WRWDO OD\HU WKLFNQHVV ;f EXW ODUJH FRPSDUHG WR SLJPHQW SDUWLFOH GLDPHWHUV 7KH GRZQZDUG IOX[ RI OLJKW GHFUHDVHV GXH WR DEVRUSWLRQ LQWR SLJPHQWV E\ .LG[ DQG UHYHUVHV GLUHFWLRQ WKURXJK VFDWWHULQJ E\ 6LG[ ZKHUH LV WKH DEVRUSWLRQ

PAGE 149

7RS RI FRORUDQW OD\HU (OHPHQWDU\ OD\HU RI WKLFNQHVV GDQ W 'LVWDQFH [ EHWZHHQ EDFNLQJ &RORUDQW OD\HU DQG DQ HOHPHQWDU\ f§ RI WKLFNQHVV $7 f§ OD\HU G[ I %DFNLQJ RI UHIOHFWDQFH :]]]] )LJXUH 6FKHPDWLF RI EDVLV IRU .XEHOND0XQN DQDO\VLV RI FRORUDQW OD\HUV >-XG@

PAGE 150

FRHIILFLHQW DQG 6 LV WKH VFDWWHULQJ FRHIILFLHQW 8SZDUG UHIOHFWHG OLJKW Mf LV DOVR UHGXFHG E\ DEVRUSWLRQ DQG VFDWWHULQJ 7KH FKDQJHV LQ OLJKW IOX[ FDXVHG E\ DQ HOHPHQWDU\ FRORUDQW OD\HU FDQ EH ZULWWHQ > -XG@ GM 6.fMG[ 6LG[ f GL 6.fLG[ 6MG[ f ZKHUH GM DQG GL DUH WKH WRWDO FKDQJHV LQ XSZDUG DQG GRZQZDUG SURFHHGLQJ IOX[HV UHVSHFWLYHO\ ([SRQHQWLDO VROXWLRQV WR GLIIHUHQWLDO HTXDWLRQV f DQG f LQ WHUPV RI UHIOHFWDQFH ZHUH GHULYHG E\ .XEHOND DQG 0XQN LQ 7KH\ DUH HTXDWLRQV f DQG f LQ 6HFWLRQ F RI WKLV GLVVHUWDWLRQ )URP f WKH UDWLR RI DEVRUSWLRQ WR VFDWWHULQJ FRHIILFLHQWV LV JLYHQ E\ .6 55f f ,PSRUWDQW UHTXLUHPHQWV RI HTXDWLRQV f f DQG f DUH f HDFK FRORUDQW OD\HU PXVW EH HYDOXDWHG RQH ZDYHOHQJWK XQLW DW D WLPH QRUPDOO\ LQ QP LQFUHPHQWVf VLQFH DQG 6 DUH D IXQFWLRQ RI ZDYHOHQJWK DQG f FRPSOHWH KLGLQJ SRZHU RU RSDFLW\ RI WKH VXEVWUDWH EDFNLQJ LV DVVXPHG %HHUnV ODZ HTXDWLRQ f LQ 6HFWLRQ Gf GRHV QRW KROG WUXH IRU FRPSOH[VXEWUDFWLYH FRORUDQW PL[LQJ DQG LQVWHDG WKH .XEHLND0XQN HTXDWLRQV DUH DSSOLHG $EVRUSWLRQ DQG VFDWWHULQJ FRHIILFLHQWV RI PDWHULDOV DUH XVHG WR SUHGLFW

PAGE 151

FRORUDQW FRQFHQWUDWLRQV UHTXLUHG WR SURGXFH D FRORU RU WR HYDOXDWH WKH KLGLQJ SRZHU RI D FRDWLQJ (TXDWLRQ f KDV EHHQ XVHG E\ 6HOOLQJ > 6HO@ IRU IRUPXODWLQJ G\HV DQG E\ 'XQFDQ > 'XQ@ IRU SDLQW IRUPXODWLRQV 0XUGRFN :LVH (SSOHU > 0XU@ DQG %ORQVNL > %OR %ORE@ HYDOXDWHG FHUDPLF SLJPHQWV LQ JOD]H XVLQJ .XEHOND0XQN WKHRU\ $SSOLFDWLRQ RI HTXDWLRQ f WR SLJPHQW FRORUHG JOD]HV SUHVXPHV WKH FRPSRVLWH V\VWHP LV WKH FRQFHQWUDWLRQ RI ZHLJKWHG VXPV RI WKH LQGLYLGXDO FRPSRQHQWV > %ORE@ $Q DEVRUSWLRQ IDFWRU .6f > (SSD@ FDQ EH XVHG WR PHDVXUH WKH FRORU VWUHQJWK RI WKH FRDWLQJ ZKLFK LV D IXQFWLRQ RI WKH SLJPHQW FRQFHQWUDWLRQ &SLJPHQWf .6 5f 5 & 3& ASLJPHQW ASLJPHQW A JOD]H F V V SLJPHQW SLJPHQW JOD]H f 7KH UHTXLUHPHQWV RI f IRU HYDOXDWLQJ SLJPHQWHG FHUDPLF FRDWLQJV DUH 6SHFWURSKRWRPHWULF PHDVXUHPHQWV RI UHIOHFWDQFH 5f DUH VXEVWLWXWHG IRU UHIOHFWLYLW\ 5LQ f LI DQ LQFUHDVH LQ FRDWLQJ WKLFNQHVV GRHV QRW FKDQJH WKH UHIOHFWDQFH YDOXHV 5HIOHFWDQFH LV GLIIXVH WKH VSHFXODU FRPSRQHQW PXVW EH H[FOXGHG

PAGE 152

,Q HTXDWLRQ f 5f FRQYHUWV PHDVXUHG GLIIXVH UHIOHFWDQFH LQWR DEVRUSWLRQ VLQFH OLJKW WKDW LV QRW UHIOHFWHG E\ DQ RSDTXH REMHFW PXVW EH DEVRUEHG 7KH IDFWRU 5 DOORZV WKDW PRUH WKDQ DOO RI WKH LQFLGHQW OLJKW FDQQRW EH DEVRUEHG AJOD]H 6LQFH DV & f QW f§! .6 f§ .6 FDOFXODWHG ZLWK S 6JOD]H WKH PLQLPXP UHIOHFWDQFH 5f f YDOXH PHDVXUHG DFURVV WKH YLVLEOH VSHFWUXP LV ZKHUH DEVRUSWLRQ RI WKH JOD]H PDWUL[ LV PLQLPDO DQG DEVRUSWLRQ GXH WR WKH SLJPHQW LV PD[LPXP 7KLV SURYLGHV WKH EHVW LQGLFDWLRQ RI FRORU VWUHQJWK GXH WR SLJPHQW DEVRUSWLRQ $ KLJKHU .6 DW WKH PLQLPXP UHIOHFWDQFH LQGLFDWHV JUHDWHU SLJPHQW VWUHQJWK &DOFXODWHG DEVRUSWLRQ IDFWRUV .6f RI YDULRXV SLJPHQWV DQG SLJPHQWJOD]H V\VWHPV DUH FRPSDUHG WR ZHLJK FRORU VWUHQJWK 3ORWV RI .6 YHUVXV SLJPHQW FRQFHQWUDWLRQ DUH RIWHQ FROODWHG DQG H[KLELW D OLQHDU UHODWLRQVKLS RYHU PRVW RI WKH UDQJH > (SSEM 7KH YHUWLFDO GLVWDQFHV EHWZHHQ ORJ .6 YHUVXV ZDYHOHQJWK FXUYHV DUH SURSRUWLRQDO WR SLJPHQW FRQFHQWUDWLRQ RU OD\HU WKLFNQHVV LI WKH FRDWLQJ GRHV QRW FRPSOHWHO\ KLGH WKH VXEVWUDWH

PAGE 153

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f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

PAGE 154

FORVHU WRJHWKHU ERQGV DUH VWUHQJWKHQHG DQG FU\VWDO IRUPDWLRQ EHFRPHV PRUH UHDGLO\ DWWDLQDEOH &U\VWDOOL]DWLRQ DW D VXIILFLHQWO\ ORZ WHPSHUDWXUH GHSHQGV RQ WKUHH PDLQ IDFWRUV f QXPEHU RI QXFOHL IRUPHG f FU\VWDO JURZWK UDWH f PHOW YLVFRVLW\ > 3DU@ 1XFOHDWLRQ EHJLQV ZKHQ WZR DWRPV ZLWK VXIILFLHQWO\ ORZ YHORFLWLHV FROOLGH DQG FRPH WRJHWKHU DW UHVW ,I WKHUH LV QRW VXIILFLHQW HQHUJ\ WR DJLWDWH WKH UHVXOWLQJ PDVV LW EHFRPHV DQ HDVLHU WDUJHW WKDQ D VLQJOH DWRP 7KH SDLU LV VWUXFN E\ RWKHU DWRPV DQG D QXFOHXV EHJLQV WR IRUP 7KH IUHH HQHUJ\ FKDQJH $*f UHTXLUHG IRU KRPRJHQHRXV QXFOHDWLRQ LV DVVXPLQJ VSKHULFDO JHRPHWU\f > 6DF@ O" ILWn $* QUf\ f§ QUf$*Y n n 7KH ILUVW WHUP RQ WKH ULJKW VLGH RI f LV WKH IUHH HQHUJ\ UHTXLUHG WR IRUP QHZ LQWHUIDFHV DV D IXQFWLRQ RI LQWHUIDFLDO HQHUJ\ \f DQG QXFOHXV UDGLXV Uf 7KLV WHUP DOZD\V KDV D SRVLWLYH YDOXH 7KH VHFRQG WHUP LV WKH IUHH HQHUJ\ IRU WKH EXON WUDQVLWLRQ DV D IXQFWLRQ RI WKH IUHH HQHUJ\ FKDQJH SHU XQLW YROXPH RI WKH QHZ SKDVH IRUPHG $*9f 7KLV WHUP FDQ EH SRVLWLYH RU QHJDWLYH 2YHUDOO $* PXVW EH QHJDWLYH IRU QXFOHDWLRQ WR EHFRPH WKHUPRG\QDPLFDOO\ IDYRUDEOH 7KH FULWLFDO QXFOHL UDGLXV Uf f ZKHUH $*Y KDV D VXIILFLHQWO\

PAGE 155

QHJDWLYH YDOXH WR RYHUFRPH WKH HQHUJ\ UHTXLUHG WR IRUP QHZ LQWHUIDFHV LV GHWHUPLQHG IURP WKH $*fU PD[LPD DQG LV JLYHQ E\ Ur f $W HPEU\R VL]HV JUHDWHU WKDQ Ur WKH FKHPLFDO SRWHQWLDO RI WKH HPEU\R LV OHVV WKDQ WKDW RI WKH EXON SKDVH DQG LW EHFRPHV D JURZLQJ FU\VWDOOLQH QXFOHXV $Q DSSUR[LPDWH HTXDWLRQ IRU $*Y IRU FRQGHQVHG SKDVHV LV > 'RU@ $*9 $+ I 7797Q f ZKHUH $+I LV WKH KHDW RI IXVLRQ 7P LV WKH PHOWLQJ WHPSHUDWXUH DQG 9 LV WKH PRODU YROXPH RI WKH PDWUL[ SKDVH 7KH DFWLYDWLRQ HQHUJ\ UHTXLUHG IRU QXFOHL IRUPDWLRQ $*rf LV Q\ f $*r DJ 7KH UDWH RI QXFOHXV IRUPDWLRQ ,f LV UHODWHG WR $*r E\ $*r f .H[S f QRFP ,Vf N7 ZKHUH N LV %ROW]PDQQnV FRQVWDQW DQG LV D FRHIILFLHQW DVVXPHG LQGHSHQGHQW RI WHPSHUDWXUH %\ FRPELQLQJ f f DQG f JR R ? f§ P\ 9 a7P $+I 77, .H[S N7 f

PAGE 156

(TXDWLRQ f LQGLFDWHV WKDW EHFDXVH RI WKH 77Pf IDFWRU ZLWKRXW D WUDQVSRUW OLPLWDWLRQ WKHUH LV D VKDUS LQFUHDVH LQ WKH QXFOHDWLRQ UDWH DW VRPH FULWLFDO WHPSHUDWXUH EHORZ 7P 7KH WUDQVSRUW IDFWRU RU YLVFRXV IORZ FRQVWUDLQV WKH PD[LPXP QXFOHDWLRQ UDWH SRVVLEOH ZLWK GHFUHDVLQJ WHPSHUDWXUH 7KXV LV VRPHWLPHV HVWLPDWHG DV FPn VHF SRLVH IRU R[LGH JODVV IRUPHUV > &KL@ 7KH QXFOHDWLRQ UDWH FDQ EH IXUWKHU UHODWHG WR WKH JODVV PDWUL[ YLVFRVLW\ 7_f E\ > 'RU@ OHJ ,7f .777Qf f ZKHUH LV D WHPSHUDWXUH LQGHSHQGHQW IDFWRU (TXDWLRQ f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

PAGE 157

,Q HTXDWLRQV f f KHWHURJHQHRXV QXFOHDWLRQ FDQ EH DFFRXQWHG IRU E\ VXEVWLWXWLQJ DQ HIIHFWLYH LQWHUIDFLDO HQHUJ\ \Hf IRU\ > 'RU@ \H = I L< f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

PAGE 158

9LVFRVLW\ 1XFLHDKnRQ UDWH &U\VWDO JURZWK UDWH )LJXUH 5HODWLRQVKLS EHWZHHQ YLVFRVLW\ DQG WHPSHUDWXUH IDYRULQJ QXFOHDWLRQ DQG JURZWK LQ JOD]HV >$GDSWHG IURP 3DU@

PAGE 159

GULYLQJ IRUFH IRU QXFOHDWLRQ LQFUHDVHV EXW WKH GLIIXVLRQ WUDQVSRUW UDWH ORZHUV HQRXJK WR LQKLELW JURZWK %RWK SURFHVVHV DOVR QHHG WLPH WR RFFXU ,I FRROLQJ LV WRR UDSLG QXFOHDWLRQ DQG FU\VWDOOL]DWLRQ PD\ QRW UHVXOW &U\VWDOOL]DWLRQ IURP VROXWLRQ LV D IXQFWLRQ RI LPSXULW\ DQG QXFOHL W\SHV DQG FRQFHQWUDWLRQV WKH QXPEHU RI SKDVHV VHSDUDWLQJ WLPHWHPSHUDWXUH SURFHVVLQJ PDWUL[ SKDVH YLVFRVLW\ DQG GLIIXVLRQ FRHIILFLHQWV 'f 7KH UDWH RI JURZWK RI D VSKHULFDO SDUWLFOH RI UDGLXV U LV > 'RU@ GU f§f§ D f§ GW U f ZKHUH D LV D FRHIILFLHQW GHSHQGHQW RQO\ RQ FRQFHQWUDWLRQV 7KXV XSRQ LQWHJUDWLQJ VHSDUDEOH HTXDWLRQ f LW EHFRPHV HYLGHQW WKDW WKH FU\VWDO UDGLXV JURZV SURSRUWLRQDO WR WKH VTXDUH URRWV RI ERWK WLPH DQG WKH GLIIXVLRQ FRHIILFLHQW 7KH GLIIXVLRQ FRHIILFLHQW LV UHODWHG WR WKH JUDGLHQW RI FRQFHQWUDWLRQ F LQ WKH [ GLUHFWLRQ > %DQ@ GF GnF f W G[ DQG KDV DQ $UUKHQLXV W\SH UHODWLRQVKLS WR WHPSHUDWXUH 'X H[S 457f f

PAGE 160

ZKHUH 'R LV D PDWHULDO IDFWRU ZLWK WKH VDPH XQLWV DV 4 LV WKH DFWLYDWLRQ HQHUJ\ IRU WKH WUDQVSRUW SURFHVV DQG 5 LV WKH JDV FRQVWDQW :KHQ WKH PDWUL[ SKDVH LV QHDU LWV HTXLOLEULXP FRPSRVLWLRQ GXULQJ WKH ODWHU VWDJHV RI JURZWK WKH FU\VWDO UDGLXV LV SURSRUWLRQDO WR WKH FXEH URRW RI WLPH Wf B <&H39PW ff U Ur 57 ZKHUH U4 LV WKH PHDQ FU\VWDO UDGLXV DW W &H LV WKH FU\VWDO FRQFHQWUDWLRQ DQG 9P LV WKH PRODU YROXPH RI SDUWLFOHV 7KH WKHUPRG\QDPLF GULYLQJ IRUFH IRU FU\VWDOOL]DWLRQ LV UHODWHG WR KHDW IORZ DQG UHDUUDQJHPHQW DW WKH PHOWFU\VWDO LQWHUIDFH ,W LV PHDVXUHG DV WKH GLIIHUHQFH LQ *LEEV IUHH HQHUJ\ EHWZHHQ WKH OLTXLG DQG FU\VWDO 7KLV LV DOVR UHIHUUHG WR DV WKH PRODU IUHH HQHUJ\ RI FU\VWDOOL]DWLRQ RI DQ XQGHUFRROHG OLTXLG > &KL@ $*Pf JLYHQ E\ $*P $+I 7 f§7-7P $+I7P$6P f ZKHUH $+I LV WKH KHDW RI IXVLRQ DW WKH PHOWLQJ WHPSHUDWXUH 7f $W WKH PHOWLQJ SRLQW $*P LV ]HUR 7KH KLJK YLVFRVLWLHV RI JODVV PHOWV DQG FRUUHVSRQGLQJ VPDOO JURZWK UDWHV HQDEOH WKH LQWHUIDFH WHPSHUDWXUH 7f WR EH WDNHQ DV WKH JODVV EDWK RU IXUQDFH WHPSHUDWXUH > .LQ M ZKLFK DUH HDVLHU WR PHDVXUH (TXDWLRQ f LV XVHIXO IRU VPDOO XQGHUFRROLQJV EXW IDLOV

PAGE 161

DW YHU\ ODUJH XQGHUFRROLQJV VLQFH $+I DQG $6P DUH WHPSHUDWXUH GHSHQGHQW )RU ODUJH XQGHUFRROLQJV LW KDV EHHQ SUHVXPHG > &KL@ WKDW WKH KHDW FDSDFLW\ GLIIHUHQFH EHWZHHQ WKH OLTXLG DQG FU\VWDO UHPDLQV FRQVWDQW ZLWK XQGHUFRROLQJ DQG $*P FDQ EH EHWWHU HVWLPDWHG ZLWK +RIIPDQnV HTXDWLRQ $+I 777 f $*r f§ 7P ,Q PDQ\ PXOWLR[LGH JODVV V\VWHPV YLVFRVLW\ WHQGV WR LQFUHDVH UDSLGO\ ZLWK GHFUHDVLQJ WHPSHUDWXUH DQG XQGHUFRROLQJV DUH ODUJH )RU WKLV FDVH +RIIPDQnV HTXDWLRQ SURYLGHV WKH EHVW DSSUR[LPDWLRQV $OO RI WKH HTXDWLRQV LQ WKLV VHFWLRQ DUH OHIW LQ WHUPV RI WKH VWDQGDUG UHODWLRQ f EXW HTXDWLRQ f FDQ EH VXEVWLWXWHG ZKHQ DSSURSULDWHf 7KH QDWXUH RI WKH PDWUL[FU\VWDO LQWHUIDFH KDV D VWURQJ LQIOXHQFH RQ WKH NLQHWLFV RI FU\VWDOOL]DWLRQ 0RGHOV XVHG WR SUHGLFW WKH FU\VWDOOL]DWLRQ SURFHVV DUH EDVHG RQ GLIIHUHQW DVVXPSWLRQV UHJDUGLQJ LQWHUIDFH JHRPHWU\ ,Q R[LGH JODVVHV WKH YHORFLW\ RI FU\VWDOOL]DWLRQ X LV RIWHQ HVWLPDWHG ZLWK $*P 8 N fQL f

PAGE 162

ZKHUH ; LV WKH WKLFNQHVV RI WKH WUDQVLWLRQ OD\HU EHWZHHQ OLTXLG DQG FU\VWDO (TXDWLRQ f GHPRQVWUDWHV WKH UHOLDQFH RI FU\VWDOOL]DWLRQ RQ WHPSHUDWXUH YLVFRVLW\ DQG $+I 2IWHQ FU\VWDOOL]DWLRQ YHORFLW\ YHUVXV WHPSHUDWXUH LV GHULYHG XVLQJ f EXW LQDFFXUDFLHV LQ FDOFXODWHG YDOXHV PD\ UHVXOW IURP XQFHUWDLQWLHV LQ JHRPHWULFDO IDFWRUV 2QH IDFWRU WKDW KDV EHHQ XVHG WR HVWLPDWH WKH VXLWDELOLW\ RI DSSO\LQJ f ZDV IRUPXODWHG E\ -DFNVRQ > 'RU@ D $+I57 f W P ZKHUH 5 LV WKH JDV FRQVWDQW ,I WW URXJK LQWHUIDFHV IRUP DQG f LV QRW DFFXUDWH 9DOXHV RI D DOVR UHVXOW LQ GLVFUHSDQFLHV ZKHQ XVLQJ f IRU UHDVRQV QRW FOHDU > 'RU@ 7KH JURZWK UDWH IRU D W\SLFDO JODVV LV ]HUR DW WKH PHOWLQJ SRLQW LQFUHDVHV ZLWK XQGHUFRROLQJ ULVHV WR D PD[LPXP WKHQ GHFUHDVHV DV WKH YLVFRVLW\ LQFUHDVHV 7KLV LV GHPRQVWUDWHG ZLWK $YUDPLnV HTXDWLRQ > 5DK@ ZKLFK DOVR LQGLFDWHV WKDW WKH YROXPH IUDFWLRQ RI FU\VWDOV E ZLWK D JLYHQ WKHUPDO KLVWRU\ LV DQ H[SRQHQWLDO IXQFWLRQ RI WLPH Wf > 5DK@ E "[S f§ WOX W 7WOXnWn f

PAGE 163

&U\VWDOOLQH SKDVH IRUPDWLRQ FDQ EH REVHUYHG DV H[RWKHUPLF WUDQVIRUPDWLRQV RQ '7$ RU '6& FXUYHV ZKLOH IXVLRQ DQG PHOWLQJ DUH HQGRWKHUPLF $FWLYDWLRQ HQHUJ\ IRU FU\VWDO JURZWK 4f KDV EHHQ GHWHUPLQHG XVLQJ '7$ PHDVXUHPHQWV DQG > 3RS@ ,Q G $7f GW .$7 4 FRQVW 57 f ZKHUH &S LV WKH KHDW FDSDFLW\ RI WKH VDPSOH DQG VDPSOH KROGHU LV WKH KHDW WUDQVIHU FRHIILFLHQW DQG $7 LV WKH WHPSHUDWXUH GLIIHUHQFH EHWZHHQ WKH VDPSOH DQG UHIHUHQFH DW WLPH W 7KH KHDW WUDQVIHU FRHIILFLHQW FDQ EH FDOFXODWHG ZLWK $+ .$ f ZKHUH $+ LV WKH WRWDO KHDW HYROYHG GXULQJ WKH UHDFWLRQ DQG $ LV WKH DUHD XQGHU WKH '7$ H[RWKHUPLF SHDN )RU '6& VWXGLHV ZLWK D FRQVWDQW KHDWLQJ UDWH WKH DFWLYDWLRQ HQHUJ\ FDQ EH GHULYHG IURP WKH WHPSHUDWXUH DW WKH SHDN PD[LPXP SRLQW 7 f > 'RU@ ,Q G7 7, ) GW ,Q 45f ,Q Y 457 f ZKHUH G7GW LV WKH KHDWLQJ RU FRROLQJ UDWH Y LV D FRQVWDQW DQG 5 LV WKH JDV FRQVWDQW :KHQ ,Q >WaG7GWf LV SORWWHG YV 7S IRU YDULRXV KHDWLQJ UDWHV WKH VORSH RI WKH FXUYH LV 4

PAGE 164

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

PAGE 165

&U\VWDO JURZWK PPKUf )LJXUH &U\VWDO JURZWK UDWH DV D IXQFWLRQ RI WHPSHUDWXUH LQ 1D&D$O6L&! JODVV >$GDSWHG IURP .LQ@

PAGE 166

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f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

PAGE 167

=LUFRQ ZDV IRXQG WR FU\VWDOOL]H Ff IURP =U DPG 6L R[LGHV LQ D YLWUHRXV Yf PXOWLR[LGH JOD]H IULW GXULQJ KHDWLQJ DFFRUGLQJ WR VHTXHQFH > (VF@ =UYf =UYf a & =U Yf =U2 PFf a & =U2 PFf =U2aWFf ZKHUH PRQRFOLQLF Pf DQG WHWUDJRQDO Wf ]LUFRQLD FRQWLQXH WR SUHFLSLWDWH DQG FDQ EH SUHVHQW XS WR DV KLJK DV r& > $PR@ 5HDFWLRQV f DQG f DUH FRQWUROOHG E\ GLIIXVLRQ RI =U VWUXFWXUDO XQLWV WR WKH FU\VWDOJODVV LQWHUIDFH )LQDOO\ & f =U Ff 6L Yf =U6L2 Ff ZKLFK LV D GLIIXVLRQ FRQWUROOHG ILUVW RUGHU LUUHYHUVLEOH UHDFWLRQ 7KH WHPSHUDWXUH RI PD[LPXP FU\VWDOOL]DWLRQ RI ]LUFRQ VKLIWV XSZDUG ZLWK DQ LQFUHDVH LQ KHDWLQJ UDWH VLQFH WKH SURFHVV LV WLPH GHSHQGHQW 1RUPDOO\ FU\VWDOOL]DWLRQ RI DOO VSHFLHV LQ JOD]HV VXEMHFWHG WR UDSLG KHDWLQJ F\FOHV RFFXUV DW r& > .LQ $PR@ 6RPH IULWV FU\VWDOOL]H ]LUFRQ GXULQJ WKH ILULQJ SURFHVV SURYLGLQJ RSDFLW\ DQG ZKLWHQHVV &U\VWDOOLWHV IRUPHG DUH

PAGE 168

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b E\ ZHLJKW LQ WKH RULJLQDO IULW DOVR KDG QR HIIHFW RQ WKH UHVXOWV GXH WR WKH SUHGRPLQDWLRQ RI VXUIDFH QXFOHDWLRQ > $PR@ &HUDPLF WLOHV SURFHVVHG LQ LQGXVWU\ DUH VXEMHFWHG WR QRQLVRWKHUPDO ILULQJ F\FOHV $W WHPSHUDWXUHV ZKHUH QXFOHDWLRQ DQG JURZWK FDQ RFFXU W\SLFDO IDVWILUH F\FOH KHDWLQJ UDWHV RI r&PLQ SHDN WHPSHUDWXUHV RI

PAGE 169

r& DQG FRROLQJ UDWHV RI r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b > .LQ &RQ 3DU@ EXW LV DOVR GHSHQGHQW RQ FRPSRVLWLRQ DQG SURFHVVLQJ FRQGLWLRQV =LUFRQ E\ LWVHOI GLVVRFLDWHV LQWR =U DQG 6L DW r& s r& DV VKRZQ LQ WKH SKDVH GLDJUDP LQ )LJXUH > /HY@ 0HOWLQJ SRLQWV RI WKH LQGLYLGXDO R[LGHV DUH r& IRU 6L2 DQG r& IRU =U =LUFRQ EHJLQV WR GLVVROYH LQ D JOD]H XVXDOO\ EHORZ r& EXW WKLV YDULHV VLJQLILFDQWO\ ZLWK WKH PHOW FRPSRVLWLRQ 1R DFFXUDWH PRGHOV KDYH EHHQ GHYHORSHG IRU SUHGLFWLQJ ]LUFRQ GLVVROXWLRQ JLYHQ WKH VWDUWLQJ JOD]H RU JODVV IULW R[LGH FRPSRVLWLRQ

PAGE 170

OA ? OBO L /, =U2U ? ; =U 66 /LTXLG /f rr r =U 66 =U6L2$ LB/ =U6L2 X ,s r ? 6L/ ? r ? ? =U6L 6L b 6LR )LJXUH %LQDU\ SKDVH GLDJUDP RI =U DQG 6L V\VWHP WHPSHUDWXUH LQ GHJUHHV FHOFLXVf >$GDSWHG IURP /HY@

PAGE 171

&DVWLORQH HW DO f DQG &RQFHSFLRQ HW DO f > &DV &RQ@ IRXQG WKDW LQ RQH XQRSDFLILHG JOD]H IULW VORZ ILUHG WR r& b WR b ZHLJKW DGGLWLRQV RI ]LUFRQ FRPSOHWHO\ GLVVROYHG DQG FUHDWHG =U DV D QHWZRUN IRUPHU &RQWLQXRXVO\ OHVV ]LUFRQ GLVVROYHG ZLWK FRQFHQWUDWLRQV LQFUHDVLQJ IURP b WR b )URP b WR b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f 7+ f \9V+$6 f ZKHUH 7+f LV WKH PHOWLQJ WHPSHUDWXUH RI D V\VWHP FRQWDLQLQJ VROLG SDUWLFOHV ZLWK D PHDQ FXUYDWXUH + 7+ f LV WKH EXON PHOWLQJ WHPSHUDWXUH IRXQG RQ PRVW SKDVH GLDJUDPVf \ LV WKH VSHFLILF LQWHUIDFUDO IUHH HQHUJ\ HUJVFPf 9V LV WKH VROLG

PAGE 172

SDUWLFOH PRODU YROXPH DQG $6 LV WKH HQWURS\ RI IXVLRQ (TXDWLRQ f DVVXPHV DOO SDUWLFOHV DUH WKH VDPH VL]H DQG VSKHULFDO *HQHUDOO\ DV D SDUWLFOH VL]H JHWV VPDOOHU LWV PHDQ FXUYDWXUH LQFUHDVHV DQG IURP f WKH PHOWLQJ SRLQW LV ORZHUHG 7KH VROXELOLW\ Ff RI D QRQVSKHULFDO VXUIDFH LV > &KL@ &2 L L mf L U A F F' H[S 57 U U ? ZKHUH U DQG U DUH SULQFLSDO UDGLL RI FXUYDWXUH DQG F LV WKH HTXLOLEULXP VROXELOLW\ $JDLQ DV WKH UDGLXV RI FXUYDWXUH GHFUHDVHV DQG WKH LQWHUIDFLDO IUHH HQHUJ\ JHWV ODUJHU WKH VROXELOLW\ RI WKH VROLG LQ WKH OLTXLG SKDVH LQFUHDVHV ,QWHUIDFLDO IUHH HQHUJ\ LV D IXQFWLRQ RI WKH FKHPLFDO SRWHQWLDO RI WKH VROLG DQG OLTXLG SKDVHV WHPSHUDWXUH DQG WKH VSHFLILF LQWHUIDFLDO H[FHVV RI PROHV SHU XQLW DUHD DW WKH LQWHUIDFH %ORQVNL > %ORD@ IRXQG WKDW WKH ZHLJKW b RI ]LUFRQ GLVVROYHG LQ D VORZILUH FHUDPLF WLOH JOD]H LQFUHDVHG H[SRQHQWLDOO\ ZLWK D GHFUHDVH LQ PHDQ SDUWLFOH GLDPHWHU +H GHWHUPLQHG WKDW IRU ]LUFRQ SLJPHQWV ZLWK DQ DVSHFW UDWLR GLVVROXWLRQ ZDV D IXQFWLRQ RI WKH QDUURZHVW FURVVVHFWLRQDO GLPHQVLRQ

PAGE 173

7KH VL]HV RI ]LUFRQ SLJPHQWV XVHG LQ LQGXVWU\ XVXDOO\ OSP LQ PHDQ GLDPHWHU SURYLGH RSWLPXP GLVSHUVDELOLW\ LQ OLTXLG JOD]HV 7KH VL]HV RI ]LUFRQ RSDFLILHUV XWLOL]HG YDULHV ZLGHO\ GHSHQGLQJ RQ RSDFLILFDWLRQ UHTXLUHPHQWV ILULQJ VFKHGXOHV JOD]H FRPSRVLWLRQV DQG FRVW F /LTXLG/LTXLG 3KDVH 6HSDUDWLRQ 6SHFLDO JOD]H RU JODVV FRPSRVLWLRQV DUH IRUPXODWHG VR WKH PHOW VHSDUDWHV LQWR WZR GLIIHUHQW OLTXLG SKDVHV 0RUH WKDQ RQH DPRUSKRXV SKDVH UHPDLQV LQ WKH FRROHG JODVV DQG PD\ FDXVH RSDOHVFHQW RSDFLILFDWLRQ $ WKHUPRG\QDPLF GHVFULSWLRQ RI OLTXLGOLTXLG SKDVH VHSDUDWLRQ LV JLYHQ ZLWK $*PL[ YV FRPSRVLWLRQ SORWV DW GLIIHUHQW WHPSHUDWXUHV ,I WKH SORW LV FRQFDYH XSZDUG D VLQJOH SKDVH KDV D ORZHU IUHH HQHUJ\ WKDQ DQ\ PHFKDQLFDO PL[WXUH RI WZR SKDVHV ZLWK WKH VDPH DYHUDJH FRPSRVLWLRQ DQG LV WKXV PRVW VWDEOH %XW D PLVFLELOLW\ JDS RFFXUV LQ SKDVHV WKDW KDYH D SRVLWLYH H[FHVV *LEEV IUHH HQHUJ\ DW VXIILFLHQWO\ ORZ WHPSHUDWXUHV ,Q WKLV FDVH D VLQJOH SKDVH LV QRW PRVW HQHUJHWLFDOO\ IDYRUDEOH DQG WKH SORW LV FRQFDYH GRZQZDUG 6SRQWDQHRXV XQPL[LQJ RI DQ LQLWLDOO\ XQLIRUP SKDVH NQRZQ DV XSKLOO GLIIXVLRQ UHGXFHV WKH GLIIHUHQFH LQ FKHPLFDO SRWHQWLDO RI WKH FRPSRQHQWV DQG ORZHUV WKH *LEEV IUHH HQHUJ\

PAGE 174

RI WKH V\VWHP 7KLV WKHUPRG\QDPLF SURFHVV UHIHUUHG WR DV VSLQRGDO GHFRPSRVLWLRQ LV UHVSRQVLEOH IRU OLTXLGOLTXLG SKDVH VHSDUDWLRQ > 'H+@ 3KDVH VHSDUDWLRQ LQ VLOLFDWH JODVVHV UHVXOWV IURP D FRPELQDWLRQ RI VLOLFD QHWZRUNIRUPLQJ OLPLWDWLRQV DQG FRPSHWLWLRQ EHWZHHQ FDWLRQV WR VXUURXQG WKHPVHOYHV ZLWK WKH ORZHVW HQHUJ\ R[\JHQ FRQILJXUDWLRQ > .LQ@ 1RQJODVVn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n DVVXPHV %2A WHWUDKHGUDO DQG VRPH % SODQDU FRRUGLQDWLRQ $Q DGGLWLRQ RI DONDOL DQG DONDOLQH HDUWK R[LGH PRGLILHUV UDLVHV WKH DPRXQW RI % IRUPHG E\ LQFUHDVLQJ

PAGE 175

FRQQHFWLYLW\ ZLWK ORFDO FKDQJH FRPSHQVDWLRQ E\ WKH DONDOL $V WKH DONDOL R[LGH FRQFHQWUDWLRQ LV LQFUHDVHG EH\RQG DSSUR[LPDWHO\ PROH b D ORVV LQ VLPSOH WHWUDKHGUDO FRRUGLQDWLRQ RFFXUV /HVV WHWUDKHGUDO VLWHV DUH DYDLODEOH IRU % DQG $O DQG WKH\ FRPSHWH IRU WHWUDKHGUDO JURXSLQJ ZLWK $ IDYRUHG > 3DZ@ 3ODQDU % FRQWHQW LQFUHDVHV ZLWK KLJKHU DONDOL DQG % 7KH DPRUSKRXV SKDVHV DVVRFLDWHG ZLWK %2 DQG % HTXLOLEULXP WHQG WR VHSDUDWH IURP WKH OLTXLG PHOW &RQVLVWHQW RSWLFDO SURSHUWLHV RI FHUDPLF JOD]HV DUH HDVLHU WR DFKLHYH LI SKDVH VHSDUDWLRQ LV QRW D IDFWRU 2SDFLILFDWLRQ DQG FRORU DUH HDVLHU WR FRQWURO ZLWK LQHUW SDUWLFXODWHV VXVSHQGHG LQ WKH JODVV\ PDWUL[ 7KH IROORZLQJ VWHSV DUH WDNHQ LQ RUGHU WR DYRLG SKDVH VHSDUDWLRQ LQ IDVW ILUH JOD]HV > 7R] 3DU 3DZ 7D\ &KL@ % FRQWHQWV DUH KHOG ZW b 6HJHUnV UXOH LV IROORZHG 6HFWLRQ f ZKHUH WKH UDWLR RI DONDOLV WR RWKHU R[LGHV LQ WKH 52 JURXS LV 7KH RYHUDOO DONDOL FRQWHQW LV NHSW DW ZW b &DUH LV WDNHQ WR DYRLG ZDWHU LQFRUSRUDWLRQ LQ IULWV GXULQJ VPHOWLQJ ,QWURGXFWLRQ RI 2+f JURXSV WHQGV WR GLVUXSW WKH JODVV QHWZRUN DQG FRQYHUW % WR % :DWHU FDQ EH PLQLPL]HG E\ XVLQJ XQK\GUDWHG EDWFK UDZ PDWHULDOV

PAGE 176

DQG VPHOWLQJ ZLWK PHWKDQH IXHO ZKLFK WHQGV WR OLEHUDWH OHVV ZDWHU GXULQJ FRPEXVWLRQ WKDQ QDWXUDO JDV $,2 ZW bf DQG =U DGGLWLRQV LQFUHDVH WKH JODVV YLVFRVLW\ DQG KDYH D PDUNHG HIIHFW RQ VORZLQJ GLIIXVLRQ GULYHQ SURFHVVHV VXFK DV LPPLVFLELOLW\ 3KDVH VHSDUDWLRQ LV VXSSUHVVHG ZLWK UDSLG FRROLQJ IURP WKH PHOW G 9LVFRVLW\ 5HODWLRQVKLSV &U\VWDOOL]DWLRQ DQG PHOWLQJ FRQGLWLRQV DUH PDUNHGO\ DIIHFWHG E\ YLVFRVLW\ *ODVV YLVFRVLW\ YDULHV HQRUPRXVO\ ZLWK FRPSRVLWLRQ DQG LV D VWURQJ IXQFWLRQ RI WHPSHUDWXUH $V GHPRQVWUDWHG ZLWK HTXDWLRQV f f f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

PAGE 177

QXFOHDWLRQ DQG JURZWK E\ LQFUHDVLQJ DWRPLF PRELOLW\ DQG FKHPLFDO FKDQJHV 7KH DGGLWLRQ RI R[LGHV WR VLOLFD DOZD\V GURSV LWV YLVFRVLW\ $V VKRZQ LQ )LJXUH WKH ORZHULQJ LV OHDVW ZLWK DOXPLQD DQG DONDOLQH HDUWK R[LGHV DQG JUHDWHVW ZLWK DONDOL R[LGHV =Q2 DQG % ZKLFK WHQG WR EUHDN XS WKH VLOLFDWH ODWWLFH > 'RU@ $W KLJK WHPSHUDWXUH JODVVHV H[KLELW $UUKHQLXV EHKDYLRU > 'RU@ 7_ H[S 457f f ZKHUH 4 LV WKH DFWLYDWLRQ HQHUJ\ IRU YLVFRXV IORZ DQG S LV D WHPSHUDWXUH LQGHSHQGHQW FRHIILFLHQW ,QYHVWLJDWRUV IRXQG WKDW JODVVHV IROORZ $UUKHQLDQ EHKDYLRU DW YHU\ KLJK WHPSHUDWXUHV DQG EHORZ WKH JODVV WUDQVLWLRQ WHPSHUDWXUH 7 f EXW WKH WUXH YLVFRVLW\ LV JUHDWHU WKDQ SUHGLFWHG E\ f DW LQWHUPHGLDWH WHPSHUDWXUHV DERYH > 6LP7KH 9RJHO)XOFKHU7DPPDQ HTXDWLRQ 9)7f KDV EHHQ DSSOLHG WR FDOFXODWH U_ ZKLOH DFFRXQWLQJ IRU WKH QRQ$UUKHQLDQ VWDJH RU U_ H[S $ 77 2 ORJ $ % 77f ZKHUH 7T $ DQG ( DUH FRQVWDQWV f f

PAGE 178

)LJXUH 9LVFRVLWLHV RI VRPH FRPPHUFLDO VLOLFDWH JODVVHV >'RU@

PAGE 179

)LJXUHV REWDLQHG IURP WKH 9)7 HTXDWLRQ DUH PRUH DFFXUDWH WKDQ HTXDWLRQ f IRU D ZLGH WHPSHUDWXUH UDQJH EXW FDOFXODWHG YDOXHV EHJLQ WR GLYHUJH IURP DFWXDO YLVFRVLWLHV QHDU 7 &HUDPLF ZKLWHZDUHV PDQXIDFWXUHUV DQG UHVHDUFKHUV LQ (XURSH DQG WKH 86 QRUPDOO\ FDOFXODWH WKH YLVFRVLWLHV RI JODVV IULWV DQG JOD]HV XVLQJ WKH 9)7 HTXDWLRQ 7KH FRQVWDQWV 74 $ DQG % DUH GHULYHG ZLWK PDWKHPDWLFDO PRGHOV ZKLFK UHTXLUH SDUDPHWHUV WKDW DUH PHDVXUHG ZLWK GLODWRPHWU\ DQG KHDWLQJ PLFURVFRS\ WHFKQLTXHV > %XU &RQ 6FK 3DJ@ 7KH KHDWLQJ PLFURVFRSH PHDVXUHV GLPHQVLRQDO FKDQJHV RI D VDPSOH DV LW LV KHDWHG 7KH WHFKQLTXH LQ SULQFLSOH LV EDVHG RQ WKH UHODWLRQVKLS EHWZHHQ GLPHQVLRQDO FKDQJHV RI D VDPSOH $//f DQG LWV VXUIDFH WHQVLRQ \f DQG YLVFRVLW\ XQGHU JUDYLWDWLRQDO IRUFH Jf GXULQJ KHDWLQJ $W D IL[HG WHPSHUDWXUH WKH VLQWHULQJ UHODWLRQVKLSV DUH $/ \ D /R ULU DQG IRU QRQLVRWKHUPDO FRQGLWLRQV DUH GD \f U_Uf f DW G7f DWf ZKHUH U LV WKH SULQFLSDO UDGLXV RI FXUYDWXUH RI WKH PDWHULDO $V YLVFRVLW\ GURSV ZLWK LQFUHDVLQJ WHPSHUDWXUH WKH VDPSOHnV

PAGE 180

VXUIDFH WHQVLRQ WHQGV WR SXOO LW LQWR D VSKHULFDO VKDSH DQG WKH IRUFH RI JUDYLW\ SXOOV LW GRZQZDUG &RQVLGHULQJ WKDW YLVFRXV IORZ LV QRW LQ HTXLOLEULXP GXULQJ ILULQJ WKH G\QDPLF FRQGLWLRQV DUH GHVFULEHG ZLWK \ \ f f§ SJK f§ AGYG[f E U ZKHUH E LV D FRQVWDQW S LV VDPSOH GHQVLW\ J LV JUDYLWDWLRQDO IRUFH K LV WKH KHLJKW RI WKH VDPSOH DQG 6G[ LV WKH UDWH RI FKDQJH RI NLQHPDWLF YLVFRVLW\ ZLWK WKLFNQHVV 9)7 HTXDWLRQ f FRQVWDQWV 7 $ DQG % DUH FDOFXODWHG IRU JOD]HV DQG IULWV IURP PRGHOV ZKLFK DVVXPH WKH IROORZLQJ FKDUDFWHULVWLF SRLQWV > %XU &RQ 6FK@ 7_ SRLVH DW WKH GLODWRPHWULF 7J 7f SRLVH DW WKH GLODWRPHWULF VRIWHQLQJ SRLQW 7V 7f SRLVH DW WKH KDOIVSKHUH SRLQW GHWHUPLQHG E\ D KHDWLQJ PLFURVFRSH 7KH PRGHOV DUH 7 7Jf§ 7 7V7Jf 77Jf 7V7JL f§ 7f§7Tf 7V7Jf $ 7V 7J 7 7 7 % 7 f§ 7f $f f f f ZKHUH DOO WHPSHUDWXUHV DUH LQ GHJUHHV &HOVLXV

PAGE 181

7KHVH PRGHOV \LHOG XVHIXO YLVFRVLW\ GDWD IRU FU\VWDOOL]DWLRQ DQG GLVVROXWLRQ VWXGLHV LQYROYLQJ IDVWILUH FHUDPLF JOD]HV DQG IULWV EHFDXVH 7KH FKDUDFWHULVWLF YLVFRVLW\ GDWD DW 7 7V DQG 7 DUH ZHOO NQRZQ IRU WKHVH VSHFLILF PDWHULDOV &U\VWDOOL]DWLRQ DQG GLVVROXWLRQ LQ IDVWILUH V\VWHPV RFFXUV DW WHPSHUDWXUHV H[FHHGLQJ r& >$PR 7R] $SD &RQ@ 7KH 7 ZKHUH DFFXUDF\ RI WKH 9)7 HTXDWLRQ LV UHGXFHG LV IDU EHORZ WKLV WHPSHUDWXUH DW 7 r& J *ODVVHV ZLWK KLJK 7V VXFK DV VLOLFDWH IDVWILUH IULWV WHQG WR IROORZ $UUKHQLDQW\SH EHKDYLRU RYHU PRVW RI WKH PHOWIRUPLQJ WHPSHUDWXUH UDQJH > 'RU@ ZKLFK LPSURYHV 9)7 HTXDWLRQ DFFXUDF\ /HKPDQQ (QJHOO DQG +HOOEUXJJH IRXQG D JRRG FRUUHODWLRQ EHWZHHQ WKH GLPHQVLRQDO FKDQJHV LQ D JOD]H GXULQJ ILULQJ DQG WKH PHDVXUHG YLVFRVLW\ RI EXON JOD]H LQ D FUXFLEOH -DPHV DQG 1RUULV LGHQWLILHG WKDW JOD]H H[KLELWHG 1HZWRQLDQ IORZ DV DQ LGHDO IOXLG DW KLJK WHPSHUDWXUH HYHQ ZKHQ ORDGHG ZLWK b LQHUW SLJPHQW SDUWLFOHV *OD]HV EHKDYH LQ DFFRUGDQFH ZLWK WUDGLWLRQDO JODVV YLVFRVLW\ WKHRULHV DQG YLVFRVLW\ WHPSHUDWXUH FXUYHV > 3DU@

PAGE 182

&+$37(5 (;3(5,0(17$/ 352&('85(6 0DWHULDOV DQG 0HWKRGV $ WRWDO RI WKLUW\WZR EDWFKHV ZHUH SUHSDUHG IURP HLJKW JODVV IULWV ORDGHG ZLWK IRXU GLIIHUHQW ]LUFRQYDQDGLXP SLJPHQW FRQFHQWUDWLRQV RI b b b DQG b 7KH EDWFKHV ZHUH EOHQGHG ZLWK ZDWHU DQG D VXVSHQGLQJ DJHQW DQG DSSOLHG WR ZKLWH RSDTXH FHUDPLF WLOH VXEVWUDWHV XVLQJ D ZHW VSUD\ PHWKRG 7KH JUHHQ FRDWLQJV ZHUH ILUHG WR r& r& DQG r& SHDN WHPSHUDWXUHV XWLOL]LQJ D W\SLFDO LQGXVWULDO IDVWILUH FHUDPLF WLOH ILULQJ FXUYH $ WRWDO RI GLIIHUHQW ILUHG VDPSOHV ZHUH SURGXFHG IRU H[DPLQDWLRQ *ODVV )ULWV DQG =U9 3LJPHQW &RPSRVLWLRQV RI WKH HLJKW JODVV IULWV LQYHVWLJDWHG DUH OLVWHG LQ 7DEOH 7KH\ LQFRUSRUDWH WKH PRVW FRVW HIIHFWLYH DQG HQYLURQPHQWDOO\ VDIH R[LGHV IRU SURGXFLQJ IULWV )RUPXODWLRQV GR QRW LQFOXGH %D2 RU 3E2 ZKLFK DUH QRZ GHVLJQDWHG DV KD]DUGRXV PDWHULDOV E\ WKH (3$ 7KH UDQJH RI R[LGH FRQWHQWV IRUPXODWHG H[FHHGV WKH UDQJH QRUPDOO\ XWLOL]HG

PAGE 183

7DEOH )ULWV ,QYHVWLJDWHG :HLJKW 3HUFHQW 2[LGH &RPSRVLWLRQV $ % & ( ) + 6L ER 1D &D2 $ =U 0J2 6U2 =Q2

PAGE 184

7DEOH OFRQWLQXHG 0RODU (TXLYDOHQWV 6HJHUf $ % & ( ) + 6L ER 1D &D2 $ =U 0J2 6U2 =Q2 0ROHFXODU ZHLJKW 6L $ 52 5 $ONDOLV f2WKHU 52 = 5R = UR &DOFXODWHG ([SDQVLRQ &RHII +DOOf r& ; &

PAGE 185

IRU JORVV\ IDVWILUH JOD]HV FRQIRUPV WR 6HJHUnV HPSLULFDO UXOHV RXWOLQHG LQ 6HFWLRQ DQG ZDV GHVLJQHG WR DYRLG SKDVH VHSDUDWLRQ E\ IROORZLQJ WKH VWHSV RXWOLQHG LQ 6HFWLRQ F )ULWV IRU SURGXFLQJ JORVV\ FRDWLQJV ZHUH HYDOXDWHG EHFDXVH WKH YDVW PDMRULW\ RI ZKLWHZDUH JOD]HV DUH JORVV\ 0DMRU YDULDWLRQV LQ FRPSRVLWLRQV EHWZHHQ WKH IULWV LQYROYHG =U FRQWHQW b DQG bf UHSODFHPHQW RI =Q2 ZLWK 6U2 DQG DONDOL6L UDWLR 7KH DONDOLV DQG 1Df 6U2 DQG =Q2 ZHUH RI VSHFLDO LQWHUHVW EHFDXVH RI WKHLU H[WHQVLYH XVH DV IOX[HV DQG WKHLU PDUNHG LQIOXHQFH RQ PRGLI\LQJ WKH 6L2 JODVV QHWZRUN ,Q IULWV FRQWDLQLQJ QR =U $ DQG &D2 ZHUH LQFUHDVHG WR IROORZ LQGXVWULDO SUDFWLFH 7KH IULWV ZHUH SURGXFHG E\ )HUUR &RUSRUDWLRQ LQ &OHYHODQG 2KLR E\ ODERUDWRU\ VPHOWLQJ UDZ PDWHULDOV DW DSSUR[LPDWHO\ r& ZDWHU TXHQFKLQJ WR DYRLG XQZDQWHG FU\VWDOOL]DWLRQ WKHQ EDOO PLOOLQJ WR DFKLHYH WKH GHVLUHG SDUWLFOH VL]H &RPSRVLWLRQV ZHUH FRQILUPHG XVLQJ $$6 DQG ;5) DQDO\VHV RXWOLQHG LQ 6HFWLRQ $ ]LUFRQYDQDGLXP =U9f EOXH SLJPHQW $ GLVSHUVLEOH VWDLQ IURP &HUHGHF &RUSRUDWLRQ ZDV WKH FRORUDQW WHVWHG 7KLV LV WKH PRVW FRPPRQO\ XVHG =U9 SLJPHQW LQ WKH ZKLWHZDUHV LQGXVWU\ ,WV PHFKDQLVPV IRU FRORU SURGXFWLRQ DUH GHWDLOHG LQ 6HFWLRQV G DQG D 2I WKH WKUHH WULD[LDO

PAGE 186

SLJPHQWV EOXH ZDV FKRVHQ EHFDXVH LW KDV LQWHUPHGLDWH KLJK WHPSHUDWXUH VWDELOLW\ ZKHQ FRPSDUHG WR =U3U \HOORZ ORZHVW VWDELOLW\f DQG =U)H FRUDO KLJKHVW VWDELOLW\f > 'HF@ $OVR FRORU FKDQJHV GXH WR SUHFLSLWDWHG FU\VWDO VSHFLHV DUH PRUH GLVWLQJXLVKDEOH ZKHQ XVLQJ EOXH FRORUDQWV EHFDXVH PRVW GHYLWULILHG SKDVHV WHQG WR VFDWWHU OLJKW RI ZDYHOHQJWKV QHDU EOXH SLJPHQWnV SHDN DEVRUSWLRQ LQ WKH UHG WR \HOORZ UHJLRQ &RDWLQJV 3UHSDUDWLRQ DQG $SSOLFDWLRQ &RDWLQJV IRUPXODWLRQV FRQVLVWHG E\ GU\ ZHLJKW RI b b b RU b =U9 b %HQWRQLWH % IURP 0LOZKLWH ,QF DQG IULW 7KH PDWHULDOV ZHUH EOHQGHG ZLWK ZDWHU DQG DSSOLHG RYHU D [ RSDTXH FHUDPLF WLOH ERG\ VXEVWUDWH XVLQJ D ZHW VSUD\ PHWKRG %HQWRQLWH b 6L b $ b )H b 0J2 b &D2 b 1D b /2,f ZDV QHFHVVDU\ IRU VXVSHQVLRQ DQG DYRLGDQFH RI VHWWOLQJ SULRU WR VSUD\LQJ /HYHOV RI =U9 SLJPHQW FRQFHQWUDWLRQV FRPSULVHG WKH UDQJH DSSOLHG LQ LQGXVWULDO JOD]HV ZKHUH b WR b ]LUFRQ SLJPHQW ORDGLQJ LV PRVW FRPPRQ 7KH VXEVWUDWH ZDV FHUDPLF ZDOO WLOH ERG\ ZKLFK FRQWDLQHG D PP WKLFN OD\HU RI )ORULGD 7LOHnV ZKLWH RSDTXH HQJREH SULPHU 7KH VXEVWUDWH ZDV SUHILUHG EHIRUH

PAGE 187

FRDWLQJV ZHUH DSSOLHG LQ RUGHU WR DYRLG ERG\ YRODWLOL]DWLRQ LQIOXHQFH RQ WKH UHVXOWV 7KH WKLFNQHVV RI D FHUDPLF FRDWLQJ LQIOXHQFHV LWV RSDFLW\ HTXDWLRQV DQG LQ 6HFWLRQ Ff $Q RSDTXH VXEVWUDWH LV UHTXLUHG IRU LPSOHPHQWDWLRQ RI .XEHOND 0XQN HTXDWLRQV 6HFWLRQ Ef IRU FDOFXODWLRQ RI FRORU VWUHQJWK GXH WR WKH SLJPHQW )RU WKHVH UHDVRQV FDUHIXO FRQWURO ZDV PDLQWDLQHG WR HQVXUH WKDW D FRQVWDQW YROXPH RI FRDWLQJV VROLGV ZDV DSSOLHG WR HDFK VXEVWUDWH LQ RUGHU WR QHJDWH WKLV DV D FRORU LQIOXHQFLQJ YDULDEOH $ YROXPH RI FP RI JOD]H VROLGV RQ D [ VXEVWUDWH UHSUHVHQWV DYHUDJH LQGXVWULDO VSHFLILFDWLRQV :DWHU ZDV DGGHG WR SURYLGH D VSHFLILF JUDYLW\ IRU HDFK ZHW JOD]H DQG WKHQ WKH DSSOLFDWLRQ ZHLJKW ZKLFK SURYLGHV FP YROXPH RI VROLGV ZDV FDOFXODWHG ZLWK 9V 0 : f§ 3 DQG 006 3 606 ZKHUH : FRDWLQJ DSSOLFDWLRQ ZHLJKW JUDPVf 9V YROXPH RI VROLGV FP f f 3 ZHLJKW b VROLGV

PAGE 188

0 GHQVLW\ RI VROLGV JFPf 6 VSHFLILF JUDYLW\ RI FRDWLQJ 7KH SURFHGXUH XVHG IRU GHWHUPLQLQJ 0 LV RXWOLQHG LQ 6HFWLRQ DQG WKH UHVXOWV DUH GLVFXVVHG LQ 6HFWLRQ )LULQJ &XUYHV 7LOHV ZHUH ILUHG LQ D UHODWLYHO\ VPDOO P OHQJWKf LQGXVWULDO SLORW UROOHU NLOQ DW )ORULGD 7LOH ,QGXVWULHV 7KH KLJK YROXPH RI DLU DQG JDV IORZ LQ WKH NLOQ HQVXUHV XQLIRUP WHPSHUDWXUH GLVWULEXWLRQ DQG FORVH FRUUHVSRQGHQFH EHWZHHQ WKHUPRFRXSOH UHDGLQJV DQG DFWXDO WLOH WHPSHUDWXUHV )LJXUH GLVSOD\V WKH WLPHWHPSHUDWXUH SURILOHV HPSOR\HG FRQVLVWLQJ RI WKUHH GLIIHUHQW FXUYHV ZLWK SHDN WHPSHUDWXUHV RI DQG GHJUHHV &HOVLXV 7KH VKDSHV RI WKH FXUYHV LH KHDWLQJ DQG FRROLQJ UDWHV KROG DW WKH SHDN WHPSHUDWXUH HWFf HTXDWH WR W\SLFDO LQGXVWULDO IDVWILUH WLOH SURFHVVLQJ $ SUHYLRXV VWXG\ RI IDVWILUH WLOH SURFHVVHV FRQFOXGHG WKDW WKH KHDWLQJ UDWH DERYH WKH JOD]H VRIWHQLQJ WHPSHUDWXUH DSSUR[LPDWHO\ r&f DQG WKH KROG WLPH DW WKH SHDN WHPSHUDWXUH LQIOXHQFHG JOD]H GHYHORSPHQW EXW WKH FRROLQJ UDWH KDG QR HIIHFW > $PR@ )LJXUH VKRZV D KHDWLQJ UDWH DERYH r& RI r&PLQ DQG D PLQXWH KROG WLPH DW WKH SHDN WHPSHUDWXUH

PAGE 189

2 R / 2 &/ ( R K 7LPH PLQf )LJXUH 7LPHWHPSHUDWXUH SURILOHV XVHG WR ILUH WKH WLOHV

PAGE 190

)ROORZLQJ WKH LQLWLDO ILULQJ RI VDPSOHV SUHSDUDWLRQ DQG ILULQJ RI WLOHV ZLWK IULWV $ DQG ( ZHUH UHSHDWHG WR YHULI\ WHVW PHWKRG FRQVLVWHQF\ 1R YLVXDO GLIIHUHQFHV ZHUH REVHUYHG EHWZHHQ LQLWLDO DQG UHSHDWHG WHVWV 0DWHULDOV &KDUDFWHUL]DWLRQ DQG $QDO\WLFDO 7HFKQLTXHV $$6 DQG ;5) 6PHOWHG IULW FRPSRVLWLRQV ZHUH FRQILUPHG XVLQJ D 3HUNLQ (OPHU PRGHO DWRPLF DEVRUSWLRQ VSHFWURPHWHU $$6f IRU DOO R[LGHV H[FHSW % =Q2 6U2 DQG =U 7KH DFFXUDF\ RI WKH GHYLFH LV DSSUR[LPDWHO\ b IRU 6L DQG $ b IRU DONDOLQH HDUWK R[LGHV DQG b IRU DONDOLV $ 5LJDNX ZDYHOHQJWK GLVSHUVLYH [UD\ IOXRUHVFHQFH ;5)f VSHFWURPHWHU FRQILUPHG WKH ZHLJKW SHUFHQW RI WKH RWKHU R[LGHV WR ZLWKLQ b DFFXUDF\ E\ LPSOHPHQWLQJ DQ $JHO FRPSXWHU SURJUDP )ULW 'HQVLW\ 'HWHUPLQDWLRQ :HLJKWV RI DQG JUDPV RI GULHG IULW ZHUH DGGHG WR FF RI ZDWHU DQG EOHQGHG WKRURXJKO\ 7KH VSHFLILF JUDYLW\ RI HDFK PL[WXUH ZDV PHDVXUHG DQG LQSXW DORQJ ZLWK WKH ZHLJKW SHUFHQW VROLGV LQWR HTXDWLRQ 7KH

PAGE 191

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

PAGE 192

7KH VSHFWURSKRWRPHWHU ZDV FRQILJXUHG WR H[FOXGH WKH VSHFXODU UHIOHFWLRQ FRPSRQHQW LQ RUGHU WR HOLPLQDWH YDULDWLRQV LQ JORVV IURP GLIIHUHQFHV LQ VXUIDFH URXJKQHVV GXH WR SRRU FRDWLQJV DSSOLFDWLRQVf DV D YDULDEOH LQ FRORU FDOFXODWLRQV 6SHFXODU UHIOHFWDQFH ZDV DOVR H[FOXGHG VR WKDW (PLQ FRXOG EH DSSOLHG LQ HTXDWLRQ WR FDOFXODWH FRORU VWUHQJWK 7KH GLIIUDFWLRQ JUDWLQJ SKRWRGHWHFWRUV DQG PLFURSURFHVVRU ZHUH DUUDQJHG WR RXWSXW UHVXOWV LQ WHUPV RI D QRRQ GD\OLJKW VWDQGDUG VRXUFH DQG D GHJUHH DQJXODU VXEWHQVH 6SHFWUDO UHIOHFWDQFH FXUYHV ZHUH FRQYHUWHG LQWR &,( /r Dr DQG Er YDOXHV VR FRUUHODWLRQV WR KXPDQ YLVLRQ FRXOG EH DFKLHYHG )RU HDFK WLOH VDPSOH PHDVXUHG UHIOHFWDQFH VWDQGDUG REVHUYHU ZHLJKWLQJ IXQFWLRQV DQG OLJKW HQHUJ\ YDOXHV LQ WHUPV RI ZDYHOHQJWK ZHUH LQWHJUDWHG IURP QP WR QP XVLQJ HTXDWLRQV LQ RUGHU WR FDOFXODWH WULVWLPXOXV YDOXHV ; < DQG = 7KHVH YDOXHV DV ZHOO DV ;Q DQG =Q IDFWRUV VKRZQ LQ 7DEOH IRU D GHJUHH REVHUYHU ZHUH XVHG WR FDOFXODWH /r Dr DQG Er IURP HTXDWLRQV (TXDWLRQ ZDV DSSOLHG WR FDOFXODWH $(r IRU TXDQWLI\LQJ FRORU FKDQJHV UHVXOWLQJ IURP ILULQJ WLOHV WR GLIIHUHQW SHDN WHPSHUDWXUHV $ VPDOOHU $(r LQGLFDWHV WKDW D

PAGE 193

FRORU LV OHVV VHQVLWLYH WR YDULDWLRQV LQ ILULQJ WHPSHUDWXUH DQG WKXV LV PRUH VWDEOH $ $(r RI LV MXVW QRWLFHDEOH DQG $16, VSHFLILFDWLRQV IRU FHUDPLF WLOH JOD]HV OLVW D WROHUDEOH UDQJH RI $(r /r Dr DQG Er YDOXHV DW r& DQG r& ZHUH XVHG WR FDOFXODWH $(r EHFDXVH WKLV LV WKH SHDN WHPSHUDWXUH UDQJH W\SLFDOO\ XWLOL]HG E\ FHUDPLF WLOH PDQXIDFWXUHUV $EVRUSWLRQ IDFWRUV .6f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r IURP SHUSHQGLFXODU ZKLFK LV VWDQGDUG IRU JOD]HV DQG FRUUHVSRQGV WR DQ LQGH[ RI UHIUDFWLRQ RI Q DFFRUGLQJ WR %UHZVWHUnV ODZ HTXDWLRQ 6HFWLRQ Ef

PAGE 194

+HDWLQJ 0LFURVFRS\ 0HOW YLVFRVLW\ DQG IORZ RI FRDWLQJV EDWFKHG ZLWK b =U9 SLJPHQW ZHUH VWXGLHG XVLQJ D 0LVXUD +HDWLQJ 0LFURVFRSH +60f 'U\ EDWFKHV ZHUH SUHVVHG LQWR VPDOO VDPSOHV DSSUR[LPDWHO\ [[ PP LQ GLPHQVLRQ 7KH VDPSOHV ZHUH KHDWHG LQ WKH PLFURVFRSHnV IXUQDFH DW D UDWH RI r&PLQ XS WR r& SHDN WHPSHUDWXUH LQ RUGHU WR VLPXODWH WKH H[SHULPHQWDO ILULQJ FXUYHV 6DPSOH LPDJHV ZHUH DFTXLUHG GXULQJ KHDWLQJ ZLWK D &&' 79 FDPHUD DQG OHQV GLJLWDO V\VWHP ZLWK D UHVROXWLRQ RI SP 3KRWRJUDSKV RI WKH VDPSOHV DW YDULRXV WHPSHUDWXUHV ZHUH JHQHUDWHG DQG WKHLU JHRPHWULHV ZHUH FRPSDUHG DQG UHODWHG WR IORZ EHKDYLRU 3DUDPHWHUV VXFK DV VDPSOH KHLJKW ZLGWK FRQWDFW DQJOH DQG WKHLU WHPSHUDWXUH GHULYDWLYHV DUH DXWRPDWLFDOO\ PHDVXUHG E\ WKH +60 VRIWZDUH WR DV ORZ DV b GLPHQVLRQDO FKDQJH +DOIVSKHUH WHPSHUDWXUHV 7Lf HVWDEOLVKHG XVLQJ KHDWLQJ PLFURVFRS\ DQG GLODWRPHWULF 7 DQG 7B GDWD ZHUH HQWHUHG LQWR J HTXDWLRQV DQG WKHQ LQ RUGHU WR JHQHUDWH ORJ Uf YHUVXV WHPSHUDWXUH SORWV EDVHG RQ WKH PRGHOV GHVFULEHG LQ 6HFWLRQ G

PAGE 195

'LODWRPHWUY 'U\ FRDWLQJV EDWFKHV ZHUH SUHVVHG LQWR PP ORQJ URGV DQG KHDWHG ZLWK D 1HW]VFK GLODWRPHWHU DW D UDWH RI r&PLQXWH WR MXVW DERYH WKH JODVV VRIWHQLQJ WHPSHUDWXUH %RWK 7 DQG 7 ZHUH LGHQWLILHG DQG FRHIILFLHQWV RI WKHUPDO J V H[SDQVLRQ ZHUH FDOFXODWHG ;5D\ 'LIIUDFWLRQ ;5'f ;UD\ GLIIUDFWLRQ VFDQV ZHUH SHUIRUPHG RQ DOO XQILUHG DQG ILUHG FRDWLQJV FRQWDLQLQJ b =U9 SLJPHQW 7KLV FRORUDQW ORDGLQJ OHYHO ZDV VHOHFWHG IRU WKH DQDO\VHV EHFDXVH D UDQJH RI b ]LUFRQ SLJPHQW LV PRVW FRPPRQO\ XVHG LQ LQGXVWU\ EXW WKH b OHYHO ZRXOG EHWWHU FRQFHDO WKH VXEVWUDWH DQG WKXV LPSURYH WKH ;5' TXDQWLWDWLYH DQDO\VLV DFFXUDF\ 7KH ;5' VFDQV DQG -&3'6 FDUGV ZHUH XVHG WR TXDOLWDWLYHO\ LGHQWLI\ DLO FU\VWDOOLQH SKDVHV WKDW GHYHORSHG GXULQJ ILULQJ $ TXDQWLWDWLYH DQDO\VLV RI ]LUFRQ FRQWHQW LQ WKH XQILUHG DQG ILUHG FRDWLQJV VHUYHG WR PHDVXUH DQ\ SLJPHQW GLVVROXWLRQ RU ]LUFRQ FU\VWDOOL]DWLRQ LQ IULWV FRQWDLQLQJ =U 7KHVH UHVXOWV ZHUH HTXDWHG WR WKH FRORU VWUHQJWK DQG VWDELOLW\ RI WKH FRDWLQJV

PAGE 196

$ 6LHPHQV 'LIIUDFWRPHWHU VFDQQHG VDPSOHV DW D UDQJH RI r ZLWK D VWHS VL]H RI r DQG D FRXQWLQJ WLPH RI VHFRQGVVWHS XVLQJ &X.D UDGLDWLRQ ,Q RUGHU WR DFKLHYH PD[LPXP GLIIUDFWLRQ LQWHQVLW\ DQG WR HQVXUH ;UD\V GR QRW GHWHFW SKDVHV XQGHU D FRDWLQJ WKH WKLFNQHVV RI D FRDWLQJ Wf VKRXOG EH > $OH@ S VLQ W XSn ZKHUH SS LV WKH PDVV DEVRUSWLRQ FRHIILFLHQW Sn LV WKH FRDWLQJ GHQVLW\ DQG LV WKH ;5' VFDQ DQJOH &DOFXODWLRQV ZLWK f XVLQJ WKH PDVV DEVRUSWLRQ FRHIILFLHQWV &X.Df RI WKH HOHPHQWV LQ WKH FRDWLQJV DW WKH PLQLPXP DSSOLFDWLRQ WKLFNQHVV WHVWHG UHYHDOHG WKDW FRDWLQJ GHQVLWLHV JFUUU DUH VXIILFLHQW IRU DYRLGLQJ WUDQVPLVVLRQ WKURXJK WR WKH HQJREH VXEVWUDWH 6HFWLRQ SURYLGHV WKH UHVXOWV RI WKH IULW GHQVLW\ GHWHUPLQDWLRQV 0HDVXUHPHQW RI ]LUFRQ FRQWHQW ILUVW LQYROYHG VFDQQLQJ XQILUHG FRDWLQJV ZLWK NQRZQ TXDQWLWLHV RI ]LUFRQ IURP r DQG FRPSXWLQJ WKH LQWHJUDWHG DUHDV XQGHU WKH r UHIOHFWLRQ SHDNV $ RI r IRU ]LUFRQ FRUUHVSRQGV WR > @ 0LOOHU LQGH[ UHIOHFWLRQ $ ]LUFRQ ZHLJKW b YHUVXV LQWHJUDWHG LQWHQVLW\ EDVHOLQH FXUYH ZDV HVWDEOLVKHG IURP WKH XQILUHG VWDQGDUGV XVLQJ UHJUHVVLRQ PHWKRGV 7KLV

PAGE 197

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f DQG (QHUJ\ 'LVSHUVLYH ;5DY 6SHFWURVFRS\ ('6f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

PAGE 198

DVVLVWHG LQ GLVWLQJXLVKLQJ WKHLU SKDVH 6DPSOHV HYDOXDWHG ZLWK ('6 ZHUH FRDWHG ZLWK FDUERQ UDWKHU WKDQ JROGSDOODGLXP EHFDXVH FDUERQV SHDNV RYHUODS OHVV ZLWK WKH SHDNV RI RWKHU HOHPHQWV LQ WKH IULWV 6WDWLVWLFDO 0HWKRGV IRU 'HULYLQJ (TXDWLRQV 5LJRURXV VWDWLVWLFDO DQDO\VHV RI WKH UHVXOWV GDWD ZHUH FRQGXFWHG IRU WKH SXUSRVH RI TXDQWLI\LQJ YDULDEOH UHODWLRQVKLSV 6HYHUDO OLQHDU DQG QRQOLQHDU HTXDWLRQV ZHUH GHYHORSHG IURP WKH H[SHULPHQWV /r SHUFHLYHG OLJKWQHVVf YHUVXV Er SHUFHLYHG \HOORZQHVVEOXHQHVVf /r Dr SHUFHLYHG UHGQHVVJUHHQQHVVf DQG Er YHUVXV .6 SLJPHQW DEVRUSWLRQ IDFWRUf HTXDWLRQVf .6 YHUVXV IULW FRPSRVLWLRQ IRU =U9 ORDGLQJV RI b b b DQG b HTXDWLRQVf Q(r FRORU FKDQJH IURP r& WR r& SHDN WHPSHUDWXUHf YHUVXV IULW FRPSRVLWLRQ IRU =U9 ORDGLQJV RI b b b DQG b HTXDWLRQVf :HLJKW SHUFHQW ]LUFRQ YHUVXV ]LUFRQ > @ ;5' LQWHJUDWHG SHDN LQWHQVLW\

PAGE 199

.6 /r Dr DQG Er YHUVXV ZHLJKW SHUFHQW ]LUFRQ SUHFLSLWDWHG IURP IULW ORDGHG ZLWK b =U9 SLJPHQW HTXDWLRQVf /RJ 7_ YHUVXV WHPSHUDWXUH DQG IULW FRPSRVLWLRQ ZLWK b =U9 ORDGLQJ (TXDWLRQV ZHUH FUHDWHG XVLQJ H[SHULPHQWDO GHVLJQ DQG DQDO\VLV FRPSXWHU VRIWZDUH > %RZ@ ,QGHSHQGHQW YDULDEOHV ZLWK FDOFXODWHG DEVROXWH WYDOXHV ZHUH FRQVLGHUHG WR KDYH VLJQLILFDQW HIIHFWV b FRQILGHQFH OHYHOf RQ WKH GHSHQGHQW YDULDEOH DQG ZHUH LQFOXGHG LQ WKH HTXDWLRQ $ W YDOXH LV JLYHQ E\ > %RZ@ ZKHUH EL LV WKH UHJUHVVLRQ FRHIILFLHQW IRU WKH LWK YDULDEOH DQG 6ELf LV WKH VWDQGDUG GHYLDWLRQ RI WKH LWK FRHIILFLHQW (TXDWLRQ FRHIILFLHQWV ZHUH JHQHUDWHG ZLWK D OHDVW VTXDUHV UHJUHVVLRQ PHWKRG > 'RZ@ 7KH SURFHGXUH LQYROYHV PLQLPL]LQJ WKH VXP RI VTXDUHV RI WKH UHVLGXDOV E\ GLIIHUHQWLDWLQJ ZLWK UHVSHFW WR HDFK XQNQRZQ FRHIILFLHQW VHWWLQJ WKH HTXDWLRQV HTXDO WR ]HUR DQG VROYLQJ IRU WKH FRHIILFLHQWV XVLQJ PDWUL[ PHWKRGV /LQHDU QRQOLQHDU DQG LQWHUDFWLYH UHODWLRQVKLSV ZHUH LGHQWLILHG XVLQJ WKLV PHWKRG

PAGE 200

2WKHU VWDWLVWLFV ZHUH DOVR DSSOLHG WR FRQILUP HTXDWLRQ YDOLGLW\ &RHIILFLHQWV RI PXOWLSOH FRUUHODWLRQ 5f ZHUH FRPSXWHG WR LQGLFDWH WKH WRWDO YDULDELOLW\ RI WKH GHSHQGHQW YDULDEOH H[SODLQHG E\ DQ HTXDWLRQ 7KH 5 LV JLYHQ E\ > %RZ@ QSf 6 f f 5 /f§ Q f 6727f ZKHUH Q LV WKH QXPEHU RI H[SHULPHQWV S LV WKH QXPEHU RI WHUPV LQ WKH HTXDWLRQ 677 LV WKH VWDQGDUG GHYLDWLRQ RI GHSHQGHQW YDULDEOH DFWXDO YDOXHV DQG 6 LV WKH VWDQGDUG GHYLDWLRQ RI WKH UHVLGXDOV 7KH H[SHFWHG DFFXUDF\ RI D VWDWLVWLFDOO\ GHULYHG HTXDWLRQ LV 6 ZLWK D b \ [ FRQILGHQFH OHYHOn 6WDWLVWLFV WKDW ZHUH HYDOXDWHG LQ RUGHU WR GHWHFW RXWOLHU H[SHULPHQWV XQLGHQWLILHG WLPH HIIHFWV DQG PRGHO RYHUILWWLQJ LQFOXGHG H[WHUQDOO\ VWXGHQWL]HG (6f DQG VWDQGDUGL]HG 6WGf UHVLGXDOV &RRN YDOXHV DQG 'XUELQ :DWVRQ ':f VWDWLVWLFV $EVROXWH YDOXHV RI (6 RU 6WG UHVLGXDOV RU !; WKH QH[W ODUJHVW UHVLGXDOf DUH FRQVLGHUHG RXWOLHUV )RU ': XQLGHQWLILHG WLPHUHODWHG HIIHFWV H[LVW ZKLOH ': VXJJHVWV DQ RYHUILWWHG HTXDWLRQ &RPSDULVRQV RI &RRN YDOXHV ZHUH PDGH WR GHWHUPLQH WKH LPSDFW RI HDFK LQGLYLGXDO H[SHULPHQW RQ WKH RYHUDOO HTXDWLRQ

PAGE 201

&+$37(5 5(68/76 )ULW DQG 3LJPHQW 3URSHUWLHV $$6 DQG ;5) DQDO\VHV FRQILUPHG WKH FRPSRVLWLRQV OLVWHG LQ 7DEOH 7UDFH TXDQWLWLHV RI )H bf ZHUH DOVR IRXQG LQ HDFK RI WKH IULWV 'HQVLWLHV PHDQ SDUWLFOH GLDPHWHUV DQG FRDWLQJV DSSOLFDWLRQ ZHLJKWV FDOFXODWHG EDVHG RQ WKH IULW GHQVLWLHV DUH OLVWHG LQ $SSHQGL[ & )ULW PHDQ SDUWLFOH GLDPHWHUV UDQJHG IURP WR PLFURQV ZLWK PHDQPHGLDQ UDWLRV RI 7KHVH YDOXHV FRQIRUP WR LQGXVWULDO VWDQGDUGV 7KH SLJPHQW SDUWLFOH VL]H GLVWULEXWLRQ ZDV W\SLFDO ZLWK D PHDQ RI PLFURQV DQG D s VWDQGDUG GHYLDWLRQV UDQJH RI WR SP )ULW GHQVLWLHV UDQJHG IURP WR JFP ZKLFK ZHUH ZHOO DERYH WKH PLQLPXP RI JFP 6HFWLRQ f UHTXLUHG WR DYRLG [UD\ WUDQVPLVVLRQ WKURXJK WR WKH VXEVWUDWH 7KLV DVVXPHV WKH GHQVLWLHV DUH QRW VLJQLILFDQWO\ ORZHUHG GXULQJ IDVWILUH FHUDPLF WLOH SURFHVVLQJ

PAGE 202

&RORU RI )LUHG &RDWLQDV $SSHQGL[ WDEXODWHV WKH IULW FRPSRVLWLRQ =U9 ORDGLQJ DQG SHDN ILULQJ WHPSHUDWXUH YHUVXV /r Dr Er $(r .6 JORVV DQG JOD]H ILW LH FUD]LQJf REVHUYDWLRQV IRU DOO H[SHULPHQWV 1RQH RI WKH RSWLFDO SURSHUWLHV ZHUH IRXQG WR FRUUHODWH WR GLIIHUHQFHV LQ WKH FRDWLQJV DSSOLFDWLRQ ZHLJKWV ZKLFK ZHUH UHTXLUHG WR VXSSO\ D FRQVWDQW YROXPH RI VROLGV RQ WKH VXEVWUDWH 1RUPDOO\ WKH DSSHDUDQFH RI D FRORUHG RSDTXH FHUDPLF WLOH JOD]H LV XQLIRUP WKURXJKRXW WKH HQWLUH VXUIDFH ZKLOH WUDQVOXFHQW FRDWLQJV DUH GDUNHU DW WKH WLOH HGJHV *UHHQZDUH WLOH HGJHV KDYH D KLJKHU SUHVVHG GHQVLW\ WKDQ WKH FHQWHU RI WKH SLHFH 7KH UHVXOWLQJ ORZHU SRUH YROXPH DW WKH HGJHV FDXVHV ZDWHU WR EH DEVRUEHG IURP WKH DSSOLHG ZHW FRDWLQJ DW D VORZHU UDWH 7KH ORQJHU GU\LQJ WLPH DOORZV IRU ILQH SLJPHQW SDUWLFOHV WR VHJUHJDWH WR WKH VXUIDFH DERYH FRDUVH IULW SDUWLFOHV GXH WR WKH SLJPHQWnV VORZHU VHWWOLQJ UDWH 7KLV VHJUHJDWLRQ DW WKH HGJHV UHVXOWV LQ D GHQVHU FRORU OD\HU DQG D GDUNHU ILUHG DSSHDUDQFH RQO\ LI WKHUH LV QRW HQRXJK GLVSHUVHG SLJPHQW DQGRU FU\VWDOOLQH SKDVHV SUHVHQW WR SURGXFH D FRPSOHWHO\ RSDTXH FRDWLQJ &RDWLQJV ZLWK IULWV $+ ILUHG WR r& DQG r& DQG IULWV $ % & DQG ILUHG WR r& DSSHDUHG WR EH RSDTXH

PAGE 203

7KXV WKHLU FRORUV DUH D IXQFWLRQ RI LQFLGHQW OLJKW UHIOHFWHG E\ WKH SLJPHQW FU\VWDO VWUXFWXUH DQG RWKHU VSHFLHV WKDW SUHFLSLWDWHG LQ WKH FRDWLQJV GXULQJ ILULQJ PLQXV OLJKW DEVRUEHG E\ WKH YDQDGLXP PHWDO DWRPV RQ WKH SLJPHQW ODWWLFH (IIHFWV GXH WR WUDQVPLVVLRQ RI OLJKW WKURXJK WR WKH HQJREH VXEVWUDWH ZHUH QRW HYLGHQW &RDWLQJV ZLWK IULWV ( ) DQG ILUHG WR r& DSSHDUHG WR EH VOLJKWO\ WUDQVOXFHQW 7KXV WKHLU FRORUV DUH PDLQO\ D IXQFWLRQ RI UHIOHFWLRQ E\ WKH =U9 SLJPHQW DQG RWKHU FU\VWDOOLQH SKDVHV LQ WKH JODVVHV PLQXV DEVRUSWLRQ E\ WKH YDQDGLXP ,Q DGGLWLRQ D VPDOO DPRXQW RI UHIOHFWLRQ IURP WKH RSDTXH HQJREH VXEVWUDWH EDFNJURXQG PD\ KDYH LQIOXHQFHG WKH DSSHDUDQFH RI WKHVH FRDWLQJV 7KH HQJREH ZDV YHU\ OLJKW DQG VOLJKWO\ \HOORZ )LJXUH f VR LW ZRXOG WHQG WR OHVVHQ WKH EOXH FRORU IURP WKH =U9 7KH FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ ZKLFK FDXVHG GLIIHUHQFHV LQ WKH RSWLFDO SURSHUWLHV RI WKH FRDWLQJV DUH GHWDLOHG LQ 6HFWLRQ DQG &KDSWHU 6SHFWUDO 5HIOHFWDQFH &XUYHV )LJXUH VKRZV WKH VSHFWUDO UHIOHFWDQFH YHUVXV ZDYHOHQJWK SURILOHV IRU WKH XQILUHG EDWFK PDWHULDOV DQG WKH HQJREH VXEVWUDWH 0D[LPXP DEVRUSWLRQ E\ WKH =U9 SLJPHQW

PAGE 204

5HIOHFWDQFH A$YHUDJH RI )ULW 3RZGHUV $+ A$YHUDJH RI 8QILUHG &RDWLQJV $+ f7LOH (QJREH 6XEVWUDWH % %HQWRQLWH 3RZGHU =U9 3LJPHQW 3RZGHU :DYHOHQJWK QPf )LJXUH 6SHFWUDO UHIOHFWDQFH RI XQILUHG UDZ PDWHULDOV DQG WKH HQJREH VXEVWUDWH EDFNLQJ FRDWLQJV FRQWDLQ b =U9 b EHQWRQLWH DQG b IULW E\ ZHLJKWf

PAGE 205

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

PAGE 206

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f DQG UHGXFHG UHIOHFWDQFH YDOXHV DERYH QP &RPSDULVRQV RI )LJXUH DQG GHPRQVWUDWH WKH H[WUHPH YDULDWLRQ LQ RSWLFDO SURSHUWLHV UHVXOWLQJ IURP GLIIHUHQW IULWV :LWK b =U9 FRQWHQW IULW & ILUHG WR r& KDG WKH KLJKHVW DEVRUSWLRQ DW QP 5 f ZKLOH IULW + ILUHG WR r& \LHOGHG WKH ORZHVW 5 f 6SHFWUDO SURSHUWLHV RI WKH XQJOD]HG HQJREH VXEVWUDWH ZHUH IRXQG WR EH VWDEOH RYHU WKH ILULQJ UDQJH LQYHVWLJDWHG DQG WKXV GLG QRW FDXVH FRORU VKLIWV GXH WR FKDQJHV LQ SHDN WHPSHUDWXUH $SSHQGL[ ( UHYHDOV VSHFWUDO UHIOHFWDQFH FXUYHV IRU DOO ILUHG FRDWLQJV

PAGE 207

5HIOHFWDQFH )LJXUH 6SHFWUDO UHIOHFWDQFH RI FRDWLQJV EDWFKHG ZL IULW & ILUHG WR n&

PAGE 208

5HIOHFWDQFH :DYHOHQJWK QPf b =U9 b =U9 b =U9 b =U9 )LJXUH 6SHFWUDO UHIOHFWDQFH RI FRDWLQJV EDWFKHG ZLWK IULW + ILUHG WR n&

PAGE 209

)LJXUHV DQG LOOXVWUDWH SHDN ILULQJ WHPSHUDWXUH HIIHFWV RQ UHIOHFWDQFH IURP FRDWLQJV EDWFKHG ZLWK b =U9 $W r& IULW $ SURGXFHG WKH ORZHVW DEVRUSWLRQ DW QP DQG IULW & WKH KLJKHVW ZKLOH WKH RYHUDOO VKDSHV RI DOO WKH FXUYHV DUH DOLNH )URP r& WR r& VHYHUDO FKDQJHV RFFXUUHG 3HDN UHIOHFWDQFHV IURP ( ) DQG + WKH IULWV FRQWDLQLQJ QR =U VKLIWHG WR KLJKHU ZDYHOHQJWKV )URP r& WR r& PD[LPXP UHIOHFWDQFH UDLVHG IURP QP WR DSSUR[LPDWHO\ QP IRU ( DQG DQG QP IRU ) DQG + 5HIOHFWDQFH DW QP LQFUHDVHG VOLJKWO\ ZLWK IULWV $ % & DQG WR D PXFK JUHDWHU GHJUHH ZLWK IULWV ( ) DQG + 7KH GDWD VKRZ D VWURQJ GHFUHDVH LQ WKH DPRXQW RI \HOORZ OLJKW DEVRUEHG IURP ILUHG FRDWLQJV EDWFKHG ZLWK IULWV FRQWDLQLQJ QR =U DV WKH SHDN ILULQJ WHPSHUDWXUH ZDV UDLVHG IURP r& WR r& 3LJPHQW $EVRUSWLRQ )DFWRUV .6f $EVRUSWLRQ IDFWRUV .6f FDOFXODWHG XVLQJ 5QLQ YDOXHV DUH SORWWHG YHUVXV ZHLJKW SHUFHQW =U9 LQ )LJXUH $OWKRXJK )LJXUH FRQILUPV D SUHYLRXV VWXG\ ZKLFK IRXQG D OLQHDU FRUUHODWLRQ EHWZHHQ .6 DQG SLJPHQW FRQFHQWUDWLRQ > (SSE@ WKH VWURQJ LQIOXHQFH RI PDWUL[ FRQGLWLRQV RQ WKH VORSH RI WKH FXUYH LV HYLGHQW 7KH UHODWLRQVKLS EHWZHHQ .6

PAGE 210

5HIOHFWDQFH f§Af§)ULW $ )ULW % £ )ULW ) r )ULW + +LH )ULW ( )ULW f§Lf§ )ULW )ULW& )LJXUH 6SHFWUDO UHIOHFWDQFH RI FRDWLQJV EDWFKHG ZLWK b =U9 ILUHG WR f&

PAGE 211

5HIOHFWDQFH )ULW $ )ULW % [ )ULW + f§)ULW ) )ULW ( )ULW )ULW& f§Lf§ )ULW )LJXUH 6SHFWUDO UHIOHFWDQFH RI FRDWLQJV EDWFKHG ZLWK b =U9 ILUHG WR n&

PAGE 212

5HIOHFWDQFH A K :DYHOHQJWK QPf f§L [)ULW+ f§$f§)ULW ) D)ULW $ )ULW % f§ff§)ULW f§[f§ )ULW ( )ULW& f§I )ULW )LJXUH 6SHFWUDO UHIOHFWDQFH RI FRDWLQJV EDWFKHG ZLWK b =U9 ILUHG WR &

PAGE 213

)LJXUH 3LJPHQW DEVRUSWLRQ IDFWRUV YHUVXV ZHLJKW SHUFHQW =U9 EDWFKHG LQ FRDWLQJV ILUHG WR f& n& DQG f& SHDN WHPSHUDWXUH /DEHOHG DUH GDWD SRLQWV DQG OLQHDU UHJUHVVLRQ OLQHV WKDW UHSUHVHQW IULWSHDN WHPSHUDWXUH FRPELQDWLRQV ZKLFK \LHOGHG WKH KLJKHVW & r&f DQG ORZHVW + f&f .6 YDOXHV

PAGE 214

DQG =U9 FRQWHQW UHPDLQV QHDUO\ OLQHDU IRU HDFK IULW FRPSRVLWLRQSHDN ILULQJ WHPSHUDWXUH FRPELQDWLRQ EXW WKH VORSHV RI WKH FXUYHV YDU\ JUHDWO\ ZLWK FKDQJHV LQ IULW FRPSRVLWLRQ DQG WHPSHUDWXUH 7KH WHPSHUDWXUH DQG IULW FRPSRVLWLRQ HIIHFWV RQ .6 DUH TXDQWLILHG ZLWK HTXDWLRQV GHULYHG LQ 6HFWLRQ RI WKLV GLVVHUWDWLRQ &RORU LQ /r Dr DQG Er 9DOXHV )LJXUHV DQG FRPSDUH FRORU YDOXHV ZKLFK UHODWH WR KXPDQ YLVLRQ IRU FRDWLQJV ILUHG WR r& DQG r& UHVSHFWLYHO\ *RRG FRUUHODWLRQ EHWZHHQ WUHQGV DW GLIIHUHQW SLJPHQW ORDGLQJV FRQILUPV WHVW PHWKRG UHSHDWDELOLW\ 7KH Dr YDOXH UHGQHVVJUHHQQHVVf SORW DW b =U9 ZDV WKH RQO\ DPELJXRXV GDWD 2WKHUZLVH FRORU SURJUHVVHG GDUNHU ORZHU /rf JUHHQHU ORZHU Drf DQG EOXHU ORZHU Erf ZLWK LQFUHDVLQJ =U9 SLJPHQW FRQFHQWUDWLRQ $W r& IULWV $ % ) DQG + SURGXFHG WKH OLJKWHVW OHDVW EOXH FRORU ZKLOH & DQG DUH WKH GDUNHVW DQG EOXHVW $W r& ) DQG + RYHUWDNH $ DQG % DV WKH OLJKWHVW FRORUV DQG & DQG UHPDLQ WKH GDUNHVW DQG EOXHVW $OVR DW r& IULWV ( ) DQG + RYHUWDNH $ DQG % IRU JHQHUDWLQJ WKH OHDVW EOXH 7KH EOXH FRORU ORVV LQ IULWV ZLWKRXW =U2a (+f ZLWK

PAGE 215

8 r i e  b =U9 b =U9 b =U9 b =U9 b=U b =Q2 b $ONDOL b=U b =Q2 b $ONDOL b =U b 6U2 b $ONDOL b =I b 6U2 b $ONDOL 1R =U b =Q2 b $ONDOL 1R =U b =Q2 b $ONDOL 1=2 b 6U2 b $ONDOL 1R =U b 6U2 b $ONDOL )LJXUH &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG =U9 SLJPHQW DQG ILUHG WR n&

PAGE 216

$ % & D (O (O 4O WLO b =U b =Q2 b $ONDOL b =U b =Q2 b $ONDOL b =U b 6U2 b $ONDOL b =U b 6U2 b $OFDOL 1R =U b =Q2 b $ONDOL 1F =U b =Q2 b $ONDOL 1R =U b 6U2 b $ONDOL 1R =U b 6U2 b $ONDOL b =U9 b =U9 b =U9 N b =U9 )LJXUH &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG =U9 SLJPHQW DQG ILUHG WR n&

PAGE 217

LQFUHDVLQJ SHDN WHPSHUDWXUH FRUUHODWHV WR WKH VKLIWV LQ WKHLU VSHFWUDO FXUYHV QRWHG LQ 6HFWLRQ )LJXUHV )URP b WR b =U9 WKH REVHUYHG UHYHUVH IURP LQFUHDVLQJ JUHHQ DSSHDUDQFH Drf ZLWK SLJPHQW ORDGLQJ WR D OHVVHQLQJ RI JUHHQ VKRZQ LQ )LJXUHV DQG f FDQ EH H[SODLQHG E\ FRPSDULQJ WKH FRDWLQJV VSHFWUDO UHIOHFWLRQ FXUYHV VWDQGDUG REVHUYHU ZHLJKWLQJ IXQFWLRQV LQ )LJXUH DQG HTXDWLRQV DQG 7KH Dr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b WR b VKLIW D JUHDWHU SRUWLRQ RI UHIOHFWDQFH WRZDUGV ZDYHOHQJWKV EHWZHHQ QP DQG QP DW RQH SRLQW WKH LQFUHDVLQJ [ FXUYH EHJLQV WR GUDPDWLFDOO\ RYHUWDNH WKH \ FXUYH UHVXOWLQJ LQ D UHGGHU DSSHDUDQFH 7KLV H[SODLQV WKH DEUXSW UHYHUVH LQ DrYDOXH WUHQG ZLWK LQFUHDVLQJ SLJPHQW OHYHOV &DUH VKRXOG EH WDNHQ WR DYRLG IRUPXODWLRQV ZKLFK

PAGE 218

FUHDWH WKLV SKHQRPHQRQ RWKHUZLVH VOLJKW EDWFKLQJ HUURUV FRXOG UHVXOW LQ ODUJH GHYLDWLRQV LQ WKH FRORU SHUFHLYHG )LJXUHV DQG SUHVHQW FRORU YDOXHV IRU FRDWLQJV ILUHG WR GLIIHUHQW SHDN WHPSHUDWXUHV ZLWK =U9 FRQFHQWUDWLRQV RI b b b DQG b UHVSHFWLYHO\ )LJXUHV XVH WKH VDPH \D[LV FRORU VFDOHV EXW )LJXUH UHTXLUHG D EORZQ XS VFDOH EHFDXVH XQSLJPHQWHG FRDWLQJV UHVXOWHG LQ OHVV FRORU YDULDWLRQV ,Q )LJXUH ZKHUH QR SLJPHQW ZDV SUHVHQW DOO FRDWLQJV SURJUHVVHG JUHHQHU DW KLJKHU WHPSHUDWXUHV DOWKRXJK WKH Dr VFDOH VKRZQ UHSUHVHQWV D YHU\ VPDOO QRWLFHDEOH GLIIHUHQFH $(r f :LWK LQFUHDVLQJ WHPSHUDWXUH QRWLFHDEOH OLJKWHQLQJ RI IULWV ZLWKRXW =U ( ) DQG + ZHOO DV D ORZHULQJ RI \HOORZ Erf LQ DOO IULWV H[FHSW ( ) DQG + ZDV REVHUYHG $W b =U9 )LJXUH f WKH PRVW VLJQLILFDQW YDULDWLRQV LQ FRORU RFFXU LQ WKH ErYDOXH )URP r& WR r& SHDN WHPSHUDWXUH ( ) DQG + IULWV ZLWKRXW =Uf EHFDPH OHVV EOXH KLJKHU Er f ZKLOH WKH EOXH FRORU RI $' IULWV ZLWK =Uf ZDV VWDEOH )ULWV ) DQG + ORVW DOO EOXH FRORU DW r& DQG DSSHDU VOLJKWO\ \HOORZ 7KH \HOORZLQJ VKLIWV WKH FRORU WRZDUGV WKDW RI XQSLJPHQWHG FRDWLQJV DV )LJXUH VKRZV WKH VDPH WUHQGV IRU b =U9 DV

PAGE 219

£O % 4L f§ 'L ( (O 4L WLO b =U b =Q2 b $ONDOL b =U b =Q2 b $ONDOL b =U b 6U2 b $ONDOL b =U b 6U2 b $ONDOL 1R =U b =Q2 b $ONDOL 1R =U b =Q2 b $ONDOL 1R =U b 6U2 b $ONDOL 1R =U b 6U2 b $ONDOL r& P & r& )LJXUH &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG QR =U9 DQG ILUHG WR f& f& RU r&

PAGE 220

2f UH $ % 4L 'L ( ( 4L WLO b =U b =IO2 b $ONDOL b =U b =Q2 b $ONDOL b =U b 6U2 b $ONDOL b =U b 6U2 b $ONDOL 1R =U b =Q2 b $ONDOL 1R =U b =Q2 b $ONDOL 1R =U b 6U2 b $ONDOL 1R =U b 6U2 b $ONDOL & & & )LJXUH &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG b =U9 DQG ILUHG WR n& f& RU f&

PAGE 221

A $ %O 4O 4 (O ( 4O WLO b =U b =Q2 b $ONDOL b =U b =Q2 b $ONDOL b =U b 6U2 b $ONDOL b=U b 6U2 b $ONDOL 1R =U2 b =Q2 b $ONDOL 1R =U b =Q2 b $ONDOL 1R =U b 6U2 b $ONDOL 1R =UW b 6U2 b $ONDOL f& & $& )LJXUH &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG b =U9 DQG ILUHG WR n& n& RU n&

PAGE 222

)LJXUH &RORU YDOXHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG b =U9 DQG ILUHG WR n& r& RU n&

PAGE 223

GHPRQVWUDWHG ZLWK )LJXUH EXW WR D JUHDWHU GHJUHH &RORU LQWHQVLW\ LQ IULWV ZLWKRXW =U (+f ZDV VLJQLILFDQWO\ UHGXFHG ZLWK LQFUHDVLQJ WHPSHUDWXUH 2I WKRVH IULWV FRPSRVLWLRQV ZLWK b DONDOLV ) DQG +f KDG ORZHU FRORU VWDELOLW\ RYHU WKH UDQJH RI ILULQJ WHPSHUDWXUHV WKDQ IULWV ZKLFK FRQWDLQHG b DONDOLV ( DQG *f )LJXUH H[KLELWV FRORU YDOXHV DW HDFK SHDN WHPSHUDWXUH ZLWK FRDWLQJV FRQWDLQLQJ b =U9 7KH /r DQG Er FKDQJHV ZLWK WHPSHUDWXUH IRU IULWV (+ DUH LQ WKH VDPH GLUHFWLRQ EXW ZLWK OHVV PDJQLWXGH WKDQ FRDWLQJV ZLWK b DQG b =U9 7KLV LQGLFDWHV WKDW WKH KLJKHU SLJPHQW FRQFHQWUDWLRQ SURYLGHV OHVV FRORU VHQVLWLYLW\ WR WKH JODVV PDWUL[ RU SLJPHQW GLVVROXWLRQ FKDQJHV GXH WR WHPSHUDWXUH +RZHYHU WKH HIIHFW RI WHPSHUDWXUH RQ SHUFHLYHG JUHHQQHVV LV VKRZQ WR EH JUHDWHU WKDQ ZLWK ORZHU SLJPHQW FRQFHQWUDWLRQV 7KH LQFUHDVHG VHQVLWLYLW\ RI Dr WR WHPSHUDWXUH LV GXH WR WKH VKLIWLQJ RI D JUHDW SRUWLRQ RI WKH VSHFWUDO FXUYH SDVW WKH [ ZHLJKWLQJ IXQFWLRQ PLQLPD QHDU QP )URP r& WR r& SHDN WHPSHUDWXUH FRORU SURJUHVVHG JUHHQHU LQ IULWV ( ) DQG + ZLWK b =U9 )LJXUH UHODWHV WKH UHODWLRQVKLS EHWZHHQ REVHUYHG OLJKWQHVV DQG EOXHQHVV IRU DOO FRDWLQJV WULDOV 7KH =U9 SLJPHQW FUHDWHV FRORU ZKLFK EHFRPHV H[SRQHQWLDOO\ GDUNHU DV

PAGE 224

Er9DOXH )LJXUH 5HODWLRQVKLS EHWZHHQ OLJKWQHVV /rf DQG EOXHQHVV Erf RI FRDWLQJV EDWFKHG ZLWK =U9 SLJPHQW DQG ILUHG WR n& r& DQG f& SHDN WHPSHUDWXUH 1RWH Er! LQ FRDWLQJV ZLWKRXW =U9

PAGE 225

EOXHQHVV LQFUHDVHV $ VHFRQG RUGHU SRO\QRPLDO UHODWLRQVKLS ZHOO FRQIRUPV 5 f WR WKH H[SHULPHQWDO GDWD /RZHU /r DQG Er YDOXHV UHVXOWLQJ IURP KLJKHU =U9 FRQFHQWUDWLRQV \LHOGHG D EHWWHU ILW WR WKH HTXDWLRQ 3RLQWV RQ WKH JUDSK ZKHUH Er GHSLFW SLJPHQW FRQFHQWUDWLRQV RI b RU b 3RVLWLYH Er YDOXHV UHSUHVHQW FRDWLQJV ZKLFK DFFRPPRGDWH QR =U9 7KH ILJXUH LQGLFDWHV WKDW DW D ORZ =U9 FRQWHQWV WKH UHODWLRQVKLS EHWZHHQ /r DQG Er LV OHVV SUHGLFWDEOH 7KH GLVFXVVLRQ LQ 6HFWLRQ UHJDUGLQJ FU\VWDOOL]DWLRQ DQG =U 9 GLVVROXWLRQ ZLOO SURYLGH LQVLJKWV LQWR WKH FDXVHV RI WKLV REVHUYDWLRQ )LJXUH DOVR VXEVWDQWLDWHV WKH XVH RI HLWKHU /r RU Er IRU HYDOXDWLQJ YLVXDO EOXH FRORU VDWXUDWLRQ ZLWK =U9 SLJPHQWV $OO WKUHH FRORU SDUDPHWHUV ZHUH IRXQG WR EH ORJDULWKPLF IXQFWLRQV RI WKH =U9 SLJPHQW DEVRUSWLRQ IDFWRU )LJXUH f 9DOXHV RI 5 IRU WKH UHODWLRQVKLSV DUH KLJKHVW IRU /r DQG Er 5 f 7KH UHGQHVVJUHHQQHVV RI =U9 V\VWHPV LV OHVV SUHGLFWDEOH IURP 5QP 5 f :LWK WKH HTXDWLRQV LQ )LJXUH VWDWLVWLFDO PRGHOV LQ 6HFWLRQ RI WKLV GLVVHUWDWLRQf IRU SUHGLFWLQJ .6 FDQ EH FRQYHUWHG LQWR QXPEHUV WKDW DUH SURSRUWLRQDO WR KXPDQ YLVLRQ 7KH DEVRUSWLRQ IDFWRU ZDV FKRVHQ DV D SULPDU\ YDULDEOH IRU TXDQWLI\LQJ FRORU VWUHQJWK GXH WR LWV GLUHFW FRUUHODWLRQ ZLWK

PAGE 226

)LJXUH =U9 SLJPHQW DEVRUSWLRQ IDFWRU UHODWLRQVKLSV ZLWK /r Dr DQG Er FRORU YDOXHV RI ILUHG FRDWLQJV

PAGE 227

SLJPHQW FRQFHQWUDWLRQ 7KXV =U9 GLVVROXWLRQ DQGRU UHIOHFWLRQ RI OLJKW IURP FU\VWDOOLQH SUHFLSLWDWHV ZKLFK FRXQWHUV WKH FKDUDFWHULVWLF QP DEVRUSWLRQ SHDN DUH HDVLHU WR HYDOXDWH ZLWK .6 YDOXHV +RZHYHU DV )LJXUH LOOXVWUDWHV FKDQJHV LQ .6 DUH PRVW YLVLEOH IRU .6 DQG OHDVW REVHUYDEOH IRU .6 A $Q\ =U9 FRQWHQWV b \LHOGHG .6 IRU DOO IULWV DQG SURFHVVLQJ FRQGLWLRQV 7KXV FKDQJHV LQ .6 DQG FRUUHVSRQGLQJ SLJPHQW FRQFHQWUDWLRQ LQ IULW V\VWHPV ZKLFK LQFRUSRUDWH =U9 SLJPHQWV DUH PRVW QRWLFHDEOH IRU =U9 b &RORU 6WDELOLW\ &RORU YDULDWLRQV UHVXOWLQJ IURP DQ LQFUHDVH LQ SHDN ILULQJ WHPSHUDWXUH IURP r& WR r& DUH UHSUHVHQWHG ZLWK $(r YDOXHV LQ )LJXUH 6LQFH SUHYLRXV GDWD VKRZHG WKDW .6 QHYHU LQFUHDVHG ZLWK WHPSHUDWXUH $(r LQGLFDWHV WKH PDJQLWXGH RI FRORU ORVV ZKLFK FRUUHODWHV WR KXPDQ YLVLRQ 7KH $(r IURP IULWV FRQWDLQLQJ =U $'f DW DOO SLJPHQW FRQFHQWUDWLRQV DQG XQSLJPHQWHG IULWV ZLWK QR =U&F (+ b =U9f IHOO EHORZ WKH $16, UHTXLUHPHQW RI $(r IRU FRPSDWLEOH FHUDPLF WLOH JOD]HV +RZHYHU PDQXIDFWXUHUV SUHIHU WR FRQIRUP WR $(r IRU PDWFKLQJ VXUIDFH DSSHDUDQFHV 2QO\ IULWV $ DQG % ZKLFK XQOLNH DQ\ RI WKH

PAGE 228

&RORU &KDQJH 'HOWD (rf IURP & WR & 3HDN 7HPSHUDWXUH )LJXUH &RORU FKDQJHV $(rf GXH WR D YDULDWLRQ LQ SHDN ILULQJ WHPSHUDWXUH IURP r& WR f& IRU FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG =U9 SLJPHQW

PAGE 229

RWKHU IULWV FRQWDLQHG ERWK =U bf DQG =Q2 bf PHW WKLV VWDQGDUG 3LJPHQWHG IULWV (+ H[KLELWHG FRQVLGHUDEOH FRORU FKDQJHV GXH WR WKH r& LQFUHDVH LQ SHDN ILULQJ WHPSHUDWXUH 7KHVH GDWD FRQFXU ZLWK WKH ILJXUHV LQ 6HFWLRQV DQG LQ LGHQWLI\LQJ WKDW IULWV ZLWK QR =U \LHOG WKH PRVW WHPSHUDWXUH VHQVLWLYH FRORU )ULW FRPSRVLWLRQ DQG WHPSHUDWXUH LQIOXHQFHV RQ $(r DUH TXDQWLILHG ZLWK VWDWLVWLFDO PRGHOV LQ 6HFWLRQ 6SHFXODU *ORVV ([FHSW IRU FRDWLQJV ZLWK IULW % WKHUH ZDV YHU\ OLWWOH LQIOXHQFH RI SLJPHQW FRQFHQWUDWLRQ RQ JORVV DV VKRZQ LQ )LJXUH $OVR H[FOXGLQJ % IULWV FRQWDLQLQJ =U KDG PRUH VWDEOH JORVV RYHU WKH r& WR r& SHDN WHPSHUDWXUH UDQJH ZKLOH IULWV ( ) DQG H[KLELWHG LQFUHDVHG JORVV DW KLJKHU WHPSHUDWXUHV )ULW + \LHOGHG E\ IDU WKH ORZHVW JORVV LQ DOO FDVHV 2YHUDOO H[FHSW IRU % IULWV ZLWK =U2 SURGXFHG D KLJKHU JORVV 7KH JORVV UHVXOWV ZHUH XVHG LQ ODWHU VHFWLRQV RI WKLV GLVVHUWDWLRQ WR FROODERUDWH ZLWK RWKHU HYLGHQFH RI FU\VWDOOL]DWLRQ LQ VRPH FRDWLQJV

PAGE 230

IW 9f R 2 BUR 2 f R 9f 9f 9f R -6 2 2 D )LUHG WR& 2 ‘ b =U9 ‘b =U9 ‘b =U9 f§;f§ b =U9 $ $ % 4L % ( S ( *L WLO b =U b =Q2 b $ONDOL b =U b =Q2 b $ONDOL b =U b 6U2 b $ONDOL b=U b 6U2 b $ONDOL 1R =U b =Q2 b $ONDOL 1R =[ b =Q2 b $ONDOL 1R =U b 6U2 b $ONDOL 1R =U b 6U2 b $ONDOL )LJXUH 6SHFXODU JORVV RI ILUHG FRDWLQJV DW D n DQJOH RI LQFLGHQFH

PAGE 231

9LVFRVLW\ RI &RDWLQDV 3XULQD +HDWLQJ 5HVXOWV LQ YLVFRVLW\ PHDVXUHPHQWV IRU FRDWLQJV EDWFKHG ZLWK b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b E\ ZHLJKW DW WKH H[SHQVH RI VLOLFD IULWV % ) DQG +f FDXVHG WKH JUHDWHVW LQFUHDVH LQ PROWHQ IORZ DW DOO WHPSHUDWXUHV 2YHUDOO IULWV ZLWK =U $'f IORZHG OHVV DW ORZ WHPSHUDWXUH XS WR DSSUR[LPDWHO\ r& DQG UHTXLUHG KLJKHU WHPSHUDWXUHV WR UHDFK WKH VSKHUH SRLQW FRQWDFW

PAGE 232

r& r& $ b =U b =Q2 b $ONDOL %O b =U b =Q2 b $ONDOL 4L b =U b 6U2 b $ONDOL 'L b =U b 6U2 b $ONDOL (O 1R =U b =Q2 b $ONDOL ( 1R =U b =Q2 b $ONDOL 4L 1R =U b 6U2 b $ONDOL + 1R =U b 6U2 b $ONDOL r& )LJXUH +HDWLQJ PLFURVFRSH LPDJHV RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG b =U9 DW f& f& DQG f&

PAGE 233

6WDJH 6WDJH 6SKHUH 6SKHUDO 6WDJHLO $L b =U b =Q2 b $ONDOL %L b =U b =Q2 b $ONDOL &O b =U b 6U2 b $ONDOL &O b =U b 6U2 b $ONDOL (O 1R =U b =Q2 b $ONDOL (O 1R =U b =Q2 b $ONDOL &O 1R =U b 6U2 b $ONDOL +L 1R =U b 6U2 b $ONDOL )LJXUH +HDWLQJ PLFURVFRSH LPDJHV RI FKDUDFWHULVWLF VWDJHV RI IORZ RI FRDWLQJV EDWFKHG ZLWK IULWV $+f DQG b =U9 7KH WHPSHUDWXUH b RI RULJLQDO VDPSOH KHLJKW DQG FRQWDFW DQJOH DUH OLVWHG XQGHU HDFK LPDJHf

PAGE 234

DQJOH r b RULJLQDO KHLJKWFRQWDFW DQJOH f )ULWV FRQWDLQLQJ ERWK =U DQG KLJK DONDOL FRQWHQW H[KLELWHG WKH JUHDWHVW IORZ DW KLJK WHPSHUDWXUHV QHDU r& 7KH WHPSHUDWXUH DW WKH VSKHUH VWDJH 7f ZDV FRPELQHG ZLWK GLODWRPHWULF GDWD WR JHQHUDWH ORJ UL YHUVXV WHPSHUDWXUH FXUYHV IRU HDFK FRDWLQJ DV GHVFULEHG LQ 6HFWLRQ G 'LODWRPHWULF 7 DQG 'LODWRPHWULF FXUYHV LQFOXGLQJ LGHQWLILFDWLRQ RI JODVV WUDQVLWLRQ 7 f DQG VRIWHQLQJ 7Vf WHPSHUDWXUHV DQG FDOFXODWHG FRHIILFLHQWV RI WKHUPDO H[SDQVLRQ OLVWHG DV &R( XQGHU HDFK JUDSKf DUH GLVSOD\HG LQ )LJXUHV DQG IRU DOO b =U9 FRDWLQJV 7KH FUD]LQJ REVHUYHG RQ WLOH VDPSOHV PDGH ZLWK IULWV ) DQG + FRXOG EH DWWULEXWHG WR WKH KLJK WKHUPDO H[SDQVLRQ RI WKH IULWV DQG ; f .L UHVSHFWLYHO\ IURP r&f ZKHQ FRPSDUHG WR WKH ERG\HQJREH VXEVWUDWH ; .f 1RQH RI WKH RWKHU IULWV FUD]HG

PAGE 235

2 HRR 7HPSHUDWXUD S&M 'DWH &RPPHQW +3 IULW rFB PLQf )ORULGD 7LOH /DE )HUUR ,&& 5W' 6DPSOH )$% PP &R( f .O =HUR =(52 f .O 3ODWRQ 4XDUW]f .O 7HPSHUDWXUH I&, 'DWH &RPPHQW +3 IULW f& PLQf )ORULGD 7LOH /DE )HUUR ,&& 5W' 6DPSOH )%% PP &R( f .O =HUR =(52 f .O 3LVWRQ 48$5= f .O 7HPSHUDWXUH SAO 'DWI &RHVVHQW 03 IULW rF PLQf )ORULGD 7LOH /DE )HUUR ,&& 5W' 6DPSOH )&$ PP &R( f .O =HUR =(52 f .O 3LVWRQ 48$5= f .O 7HPSHUDWXUD 34 'DWH &RPPHQW +3 IULW f& PLQf )ORULGD 7H /DE URUUR ,&& 5W' 6DPSOH )'$ PP &R( f .O =HUR ;(52 f .O 3LVWRQ 48$57=f .O )LJXUH 7KHUPDO GLODWRPHWULF DQDO\VHV RI FRDWLQJV $' EDWFKHG ZLWK b =U9 +HDWLQJ UDWH RI r&PLQf

PAGE 236

62 ,62 262 262 'DWH /DE L 6DPSOH =HUR 3ODWRQ 3HUUR ,& )($ =(52 48$5= & 5,' DRD &RPPHQW &R( 03 IULW r& f f f ‘LQ .O .O .O f )ORULGD 7LOH 'DWH /DE 6DPSOH =HUR 3ODWRQ )HUUR ,&& 5L' ))$ DZ =(52 48$5= &RPPHQW &R( 03 IULW r& f f f PLQ .O .O .O )ORULGD 7LOH ff ff r LWRPD UWFDU )UWWWD &+rWH 0UW OH U7ILH PHL 9 7} PH 4 L ?$7r $ aU 6 n ? 7 W Yf§ 7 f D 7 + ULI U UQ Ua3 / 6 ? U L ‘ _ 6 A 7 7$,/ f WRR 9 2 2 2 : 2 & WR 2 2 7HPSHUDWXUH 7&7HPSHUDWXUH 34 'DWH &RPPHQW 03 nOLW r& %LQ )ORULGD 7LOH 'DWH &RPPHQW 03 IULW & QLQ )ORULGD 7LOH /DE )HUUR W&& 5' /DE )HUUR ,&& 5' )*$ PP &F ( f .O 6DPSOH )+$ PP &F ( f .O =(52 f .O =HUR =(52 f .O 3ODWRQ 48$5= f .O 3ODWRQ 48$5= f .O )LJXUH 7KHUPDO GLODWRPHWULF DQDO\VHV RI FRDWLQJV (+ EDWFKHG ZLWK b =U9 +HDWLQJ UDWH RI n&PLQf

PAGE 237

7KH 7J DQG 7V GDWD GLVFORVH WKH IROORZLQJ $ONDOL FRQWHQW LQ WKH IULWV KDG WKH ODUJHVW LQIOXHQFH RQ 7 7KH DYHUDJH 7V ZLWK IULWV FRQWDLQLQJ b DONDOLV ZDV r& :KHQ WKH DONDOLV ZHUH UDLVHG WR b DW WKH H[SHQVH RI 6L 7V GURSSHG WR r& &KDQJHV LQ 6U2 YHUVXV =Q2 FDXVHG WKH JUHDWHVW VKLIW LQ 7J 7KH DYHUDJH 7J LQ IULWV FRQWDLQLQJ 6U2 ZDV r& DQG ORZHUHG WR r& ZKHQ UHSODFLQJ 6U2 ZLWK =Q2 6XEVWLWXWLQJ 6U2 IRU =Q2 LQ WKH JODVV VWUXFWXUH SURGXFHG D VLJQLILFDQW LQFUHDVH LQ WKH UDWLR RI 7J7V /HVV RI D WHPSHUDWXUH ULVH DIWHU 7J ZDV UHTXLUHG WR UHDFK 7V ZLWK 6U2 SUHVHQW 1R RWKHU HIIHFWV RI WKH R[LGHV RQ 7 DQG 7 ZHUH J V GLVWLQJXLVKHG 9LVFRVLW\ YV 7HPSHUDWXUH 'DWD IURP 6HFWLRQV DQG DV ZHOO DV HTXDWLRQV DQG LQ 6HFWLRQ G ZHUH XVHG WR JHQHUDWH WKH ORJ 7f YV WHPSHUDWXUH FXUYHV LQ )LJXUHV DQG 7KH WHPSHUDWXUH UDQJH RI LQWHUHVW EHJLQV QHDU WKH DYHUDJH VRIWHQLQJ SRLQW RI WKH IULWV ar&f DQG HQGV DW WKH KLJKHVW SHDN ILULQJ WHPSHUDWXUH WHVWHG 7KLV LV WKH WHPSHUDWXUH UDQJH ZKHUH FU\VWDOOL]DWLRQ DQG GLVVROXWLRQ FDQ RFFXU

PAGE 238

/RJ 9LVFRVLW\ SRLVHf 7HPSHUDWXUH &f )LJXUH /RJ 7_ YHUVXV WHPSHUDWXUH RI FRDWLQJV EDWFKHG ZLWK b =U9 )ULWV $' LQFRUSRUDWH b =U&!

PAGE 239

/RJ 9LVFRVLW\ SRLVHf 7HPSHUDWXUH &f )LJXUH /RJ 7_ YHUVXV WHPSHUDWXUH RI FRDWLQJV EDWFKHG ZLWK b =U9 )ULWV (+ FRQWDLQ QR =U&!

PAGE 240

,Q ERWK ILJXUHV D GURS LQ YLVFRVLW\ DFURVV WKH ZKROH WHPSHUDWXUH UDQJH LV VKRZQ IRU SDUDOOHO FRPSRVLWLRQV ZKHUH RQO\ WKH DONDOLV6L UDWLR LV LQFUHDVHG LH $f§!% &f§}' (f§}) *f§!+f 7KH LQFUHDVH LQ QRQEULGJLQJ DONDOL LRQV LQ WKH JODVV KDV D PDUNHG HIIHFW RQ ZHDNHQLQJ WKH VWUXFWXUH DQG ORZHULQJ YLVFRVLW\ $Q LQFUHDVH LQ WKH VORSH RI WKH FXUYHV QHDU WKH VRIWHQLQJ SRLQW RI r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nV WUXH HIIHFWV RQ WKH GHSHQGHQW YDULDEOHV 6HYHUDO HTXDWLRQV ZHUH GHULYHG XVLQJ WKH VWDWLVWLFDO WHFKQLTXHV EULHIHG LQ 6HFWLRQ &RORU VWUHQJWK .6f FRORU VWDELOLW\ LQYHUVH RI $(r f DQG PHOW YLVFRVLW\ UHODWLRQVKLSV ZLWK FRDWLQJ FRPSRVLWLRQ DQG

PAGE 241

WHPSHUDWXUH KDYH EHHQ TXDQWLILHG 7KH HTXDWLRQV DUH XVHIXO IRU LGHQWLI\LQT WKH ZHLJKWV RI LQIOXHQFH RI WKH VLJQLILFDQW LQGHSHQGHQW YDULDEOHV RQ FRORU DQG PHOW YLVFRVLW\ DQG IRU LQWHUSRODWLQJ SUHGLFWLRQV ZLWKLQ WKH H[SHULPHQWDO SDUDPHWHUV RI WKLV GLVVHUWDWLRQ +RZHYHU SUHGLFWLRQV EDVHG RQ H[WUDSRODWLRQV EH\RQG WKH UDQJH RI FRQGLWLRQV WHVWHG GXULQJ WKLV UHVHDUFK LH ILULQJ FXUYHV FRDWLQJ WKLFNQHVVHV VXEVWUDWH FRPSRVLWLRQ HWFf ZLOO EH PXFK OHVV DFFXUDWH 6FLHQWLILF LQWHUSUHWDWLRQV RI WKH SK\VLFDO EDVLV IRU WKH UHVXOWV RI WKLV VHFWLRQ DUH JLYHQ LQ &KDSWHU 'LVFXVVLRQ .6 YV &RDWLQJ &RPSRVLWLRQ DQG 7HPSHUDWXUH 7KH IROORZLQJ PDWKHPDWLFDO PRGHOV ZHUH GHYHORSHG XVLQJ H[SHULPHQWDO UHVXOWV IURP HDFK LQGLYLGXDO SLJPHQW ORDGLQJ f DQG IRU FRPSRVLWLRQV ZLWK =U9 DV DQ LQGHSHQGHQW YDULDEOH f DULDEOH &RHIILFLHQW W9DOXH .6Tb ]UY
PAGE 242

9DULDEOH &RHIILFLHQW W9DOXH .6Rb ]UY
PAGE 243

YDULDEOHV ZKRVH DEVROXWH WYDOXHV ZHUH e 7KHQ WKH RWKHU PRGHOV ZHUH GHULYHG ZLWK WKH VDPH YDULDEOHV LI WKHLU DEVROXWH WYDOXHV ZHUH e &DOFXODWHG 5 YDOXHV RI DQG IRU HTXDWLRQV UHVSHFWLYHO\ YHULI\ WKRURXJK H[SODQDWLRQ RI .6 YDULDWLRQ LQ WHUPV RI WKH LQGHSHQGHQW YDULDEOHV LQ WKH PRGHOV (TXDWLRQ ZKLFK LQFOXGHV DOO =U9 OHYHOV ZDV OHVV DFFXUDWH 5 f WKDQ WKH RWKHU PRGHOV (YDOXDWLRQ RI UHVLGXDOV (6 DQG 6WGf DQG 'XUELQ :DWVRQ VWDWLVWLFV DOO f UHYHDOHG QR RXWOLHU H[SHULPHQWV DQG YDOLGDWHG WKH PRGHOV 7KH SUHFLVLRQ RI DQ HTXDWLRQ LV HVWLPDWHG ZLWK WKH VWDQGDUG GHYLDWLRQ RI WKH UHVLGXDOV 6\[f 9DOXHV RI 6\ [ IRU HTXDWLRQV DUH DQG UHVSHFWLYHO\ 7KH PDJQLWXGHV DQG VLJQV RI FRHIILFLHQWV LQ DOO IRXU HTXDWLRQV FRUURERUDWH WKH IROORZLQJ FRQFOXVLRQV +LJKHU 6U2 LQ WKH IULW WHQGV WR UDLVH WKH FRORU VWUHQJWK .6f XQOHVV WKH SLJPHQW FRQFHQWUDWLRQ LV KLJK bf ,Q WKLV FDVH 6U2 ERRVWV .6 RQO\ LQ IULWV FRQWDLQLQJ =U2 ,QFUHDVHG =U ORZHUV .6 XQOHVV 6U2 LV SUHVHQW LQ SODFH RI =Q2f $Q LQWHUDFWLRQ EHWZHHQ 6U2 DQG =U ERRVWV .6 DW DOO SLJPHQW ORDGLQJV 7KH FXUYLOLQHDU ;;f HIIHFW RI WKH $ODONDOLV UDWLR UHYHDOV WKDW Df DW ORZ OHYHOV LQFUHDVLQJ $ODONDOLV

PAGE 244

WHQGV WF VOLJKWO\ GHFUHDVH .6 EXW Ef DW KLJK OHYHOV .6 LV VLJQLILFDQWO\ UDLVHG DV $ODONDOLV LV LQFUHDVHG $W D ORZ SLJPHQW FRQFHQWUDWLRQ bf KLJKHU WHPSHUDWXUHV FDXVH .6 WR GURS LQ DOO RI WKH IULWV $W b DQG b =U9 RQO\ IULWV ZKLFK LQFOXGH 6U2 H[SHULHQFH D GHFUHDVH LQ .6 ZLWK LQFUHDVLQJ WHPSHUDWXUH )LJXUH UHODWHV WKH JRRG ILW EHWZHHQ DFWXDO GDWD DQG HTXDWLRQ SUHGLFWHG UHVXOWV IRU DOO WULDOV ZLWK b =U9 1RWH .6 LV SUHGLFWHG WR UHPDLQ FRQVWDQW RYHU DOO WKUHH WHPSHUDWXUH SRLQWV IRU IULWV ZKLFK LQFOXGH =Q2 $ % ( DQG )f UDWKHU WKDQ 6U2 & DQG +f LQGLFDWLQJ VXSHULRU FRORU VWDELOLW\ ZLWK =Q2 7KH VWDWLVWLFDOO\ DGMXVWHG HIIHFWV RI LQGLYLGXDO YDULDEOHV IURP HTXDWLRQ b =U9f DUH GLVSOD\HG LQ )LJXUH 'DWD IRU WKH SORWV ZHUH JHQHUDWHG E\ LQSXWWLQJ WKH DFWXDO UDQJH RI FRPSRVLWLRQ WHVWHG RI WKH [D[LV YDULDEOH LQWR HTXDWLRQ ZKLOH KROGLQJ YDOXHV RI DOO RWKHU YDULDEOHV LQ WKH HTXDWLRQ DW WKHLU DYHUDJH OHYHO 7KXV )LJXUH GHSLFWV PDWKHPDWLFDO LQWHUSRODWLRQV EDVHG RQ WKH VWDWLVWLFDO PRGHO $Q LQFUHDVH LQ =U IURP WR PRODU HTXLYDOHQWVf LV VKRZQ WF UDLVH .6 LQ WKH SUHVHQFH RI 6U2 DW DYHUDJH RI HTXLYDOHQWVf EXW GHFUHDVH .6 ZKHQ =Q2 DW

PAGE 245

3LJPHQW $EVRUSWLRQ )DFWRU .6f &RDWLQJV $+f )LUHG WR & & DQG & )LJXUH 3LJPHQW DEVRUSWLRQ IDFWRU DFWXDO DQG HTXDWLRQ f SUHGLFWHG UHVXOWV IRU ILUHG FRDWLQJV EDWFKHG ZLWK b =U9

PAGE 246

=U 0RODU (TXLYDOHQWV $, $ONDOLV 0RODU (TXLYDOHQWV )LJXUH ,QGLYLGXDO HIIHFWV RI IULW R[LGHV RQ .6 EDVHG RQ VWDWLVWLFDO PRGHO f IRU b =U9 DQG D SHDN ILULQJ WHPSHUDWXUH RI r&

PAGE 247

DYHUDJH RI HTXLYDOHQWVf UHSODFHV 6U2 7KH PLGGOH JUDSK LOOXVWUDWHV WKDW KLJKHU 6U2 ZLWK =Q2 VWLOO SUHVHQW DW HTXLYDOHQWVf UDLVHV .6 WR D PXFK JUHDWHU GHJUHH ZKHQ =U2 DW DYHUDJH RI HTXLYDOHQWVf LV LQ WKH IULW 7KH ERWWRP SORW SRUWUD\V WKH FXUYLOLQHDU UHODWLRQVKLS EHWZHHQ .6 DQG $ODONDOLV UDWLR )ULWV PDQXIDFWXUHG ZLWKRXW =U DUH QRUPDOO\ GHVLJQHG ZLWK D KLJKHU $ODONDOLV EHFDXVH PRUH $ LV DGGHG WR PDLQWDLQ JODVV VWUHQJWK DQG GXUDELOLW\ :LWK =U LQ WKH JODVV LQFUHDVLQJ $ODONDOLV PRODU HTXLYDOHQWVf FDXVHG .6 WR ORZHU VOLJKWO\ $V $ODONDOLV OHYHOV ZHUH LQFUHDVHG !f ZKHQ QR =U ZDV SUHVHQW D VLJQLILFDQW LQFUHDVH LQ .6 RFFXUUHG 7KH ZHLJKWV RI LQIOXHQFH RI LQGHSHQGHQW YDULDEOHV RQ .6 GHULYHG ZLWK PRGHOV DUH GLVFORVHG LQ )LJXUH $ YDULDEOHnV SHUFHQW WRWDO LQIOXHQFH b,f ZDV FDOFXODWHG E\ b, r Of ZKHUH .6L WKH PD[LPXP UDQJH RI .6 REWDLQHG E\ LQSXWWLQJ DOO YDOXHV WHVWHG RI LQGHSHQGHQW YDULDEOH Lf LQWR WKH VWDWLVWLFDO HTXDWLRQ

PAGE 248

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b =U9 ORDGLQJ $OWKRXJK OHVV GLVWLQFW WKH SORWV VKRZ WUHQGV ZKLFK FRUUHODWH WR WKH VWDWLVWLFDO PRGHOV

PAGE 249

)LJXUH 9DULDEOHV ZHLJKW RI LQIOXHQFH RQ .6 EDVHG RQ VWDWLVWLFDO PRGHOV b =U9f b =U9f DQG =U9f R?R b =U9

PAGE 250

$, $ONDOLV 0RODU (TXLYDOHQWV )LJXUH 3LJPHQW DEVRUSWLRQ IDFWRU WUHQGV ZLWK IULW $DONDOLV UDWLR RI FRDWLQJV EDWFKHG ZLWK b =U9 DQG ILUHG WR r& r& RU r& )RU HDFK JURXS RI WKUHH GDWD SRLQWV DW D VSHFLILF $O&!DONDOL WKH KLJKHVW SRLQW UHSUHVHQWV D FRDWLQJ ILUHG WR r& SHDN WHPSHUDWXUH WKH ORZHVW SRLQW ZDV ILUHG WR r& /LQHV VKRZQ LQWHUVHFW DYHUDJH .6 YDOXHV IRU HDFK JURXS RI WKUHH GDWD SRLQWV

PAGE 251

)LJXUH 3LJPHQW DEVRUSWLRQ IDFWRU YHUVXV IULW =Q2 DQG 6U2 PRODU HTXLYDOHQWV RI ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 DQG ILUHG WR r& r& RU r& SHDN WHPSHUDWXUH 1f R

PAGE 252

)LJXUH ,QIOXHQFH RI IULW =U&! LQ WKH SUHVHQFH RI 6U2 RU =Q2 RQ .6 RI ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 DQG ILUHG WR r& r& RU r& $OO GDWD SRLQWV IRU r& r& DQG r& SHDN ILULQJ WHPSHUDWXUHV DUH SORWWHG /LQHV VKRZQ LQWHUVHFW DYHUDJH .6 YDOXHV IRU HDFK =U&! DQG 6U2 RU =Q2 FRPELQDWLRQ

PAGE 253

2 R R 7 R R R 7f ( R 2 ff 4f 2f & UR F ‘ b =U ‘ b =U )LJXUH 9DULDWLRQV LQ SLJPHQW DEVRUSWLRQ IDFWRU .6f GXH WR FKDQJHV LQ SHDN ILULQJ WHPSHUDWXUH IRU FRDWLQJV EDWFKHG ZLWK b =U9 DQG IULWV FRQWDLQLQJ b =U&! YHUVXV b =U&! )ULWV DOVR LQFOXGH WKH IROORZLQJ ZHLJKW SHUFHQWDJHV $ DQG ( b DONDOLV b 62 b =Q2 DQG ) b DONDOLV b 62 b =Q2 & DQG b DONDOLV b 62 b 6U2 DQG + b DONDOLV b 62 b 6U2

PAGE 254

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f UHSUHVHQWV D FRDWLQJ ILUHG WR r& ZKLOH WKH ORZHVW .6 UHSUHVHQWV D SHDN WHPSHUDWXUH RI r& 7KH YHUWLFDO VSUHDG LQ HDFK JURXS RI GDWD SRLQWV ZKLFK UHSUHVHQWV FRORU YDULDWLRQ ZLWK WHPSHUDWXUH LV JUHDWHVW IRU WKH FRDWLQJV ZLWKRXW =U DQG OHDVW ZKHQ ERWK =U DQG =Q2 DUH SUHVHQW $ FRPSDULVRQ RI .6 YDOXHV UHVXOWLQJ IURP IULWV ZLWK 6U2 DQG =Q2 DW b =U9 LV PDGH LQ )LJXUH $Q RYHUDOO SDWWHUQ RI LQFUHDVLQJ SLJPHQW DEVRUSWLRQ DQG FRORU VWUHQJWK ZLWK D UHGXFWLRQ LQ =Q2 RU KLJKHU 6U2 LV HYLGHQW &RORU GHJUDGDWLRQ IURP r& WR r& ZDV GUDPDWLFDOO\ UHGXFHG ZLWK IULWV FRQWDLQLQJ KLJK =Q2 DQG QR 6U2 )ULWV ) DQG + JHQHUDWHG GDWD WKDW VOLJKWO\ GHYLDWHG IURP WKH RYHUDOO

PAGE 255

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f ,Q HDFK FDVH .6 FKDQJHV OHVV ZLWK SHDN WHPSHUDWXUH LQ IULWV ZLWK =U 7KH ODUJHVW GURSV LQ .6 VWDELOLW\ ZKHQ =U ZDV UHPRYHG LQFUHDVHG YHUWLFDO GLVWDQFH EHWZHHQ WZR EDUVf RFFXUUHG ZKHQ VKLIWLQJ WR KLJKHU DONDOLV $( WR %) DQG &* WR '+f DQG ZKHQ 6U2 UHSODFHG =Q2 $( WR &* DQG %) WR '+f

PAGE 256

$(r YV &RDWLQD &RPSRVLWLRQ 9DOXHV RI $(r ZHUH FDOFXODWHG DV WKH FRORU GLIIHUHQFH EHWZHHQ FRDWLQJV ILUHG WR r&r YHUVXV r& SHDN WHPSHUDWXUH 6LQFH .6 QHYHU LQFUHDVHG ZLWK WHPSHUDWXUH $(r GHQRWHV WKH PDJQLWXGH RI FRORU ORVV ZKLFK FRUUHODWHV WR KXPDQ YLVLRQ DQG WKH GHJUHH RI LQVWDELOLW\ RI WKH V\VWHP 7KH IROORZLQJ VWDWLVWLFDO PRGHOV ZHUH GHYHORSHG XVLQJ H[SHULPHQWDO UHVXOWV IURP HDFK LQGLYLGXDO SLJPHQW ORDGLQJ f DQG IRU DOO FRPSRVLWLRQV ZLWK =U9 DV DQ LQGHSHQGHQW YDULDEOH f 9DULDEOH &RHIILFLHQW W9DOXH
PAGE 257

9DULDEOH &RHIILFLHQW W9DOXH $(r b =U9 a
PAGE 258

ZLWK WKH LQGHSHQGHQW YDULDEOHV 7KH ORJDULWKPLF GHSHQGHQF\ IRXQG EHWZHHQ .6 DQG /r Dr DQG Er LQ )LJXUH 6HFWLRQ DQG WKH OLQHDU UHODWLRQVKLS EHWZHHQ .6 DQG SLJPHQW FRQFHQWUDWLRQ UHYLHZHG LQ 6HFWLRQV DQG Ef H[SODLQ WKLV UHODWLRQVKLS 6PDOO FKDQJHV LQ SHUFHLYHG FRORU ZLWK SLJPHQW DEVRUSWLRQ DUH QHDUO\ OLQHDU DURXQG b b DQG b =U9 DUHDV /DUJH VZLQJV LQ /r Dr Er DQG FRUUHVSRQGLQJ $(r UHVXOWLQJ IURP VLJQLILFDQW FKDQJHV LQ SLJPHQW FRQWHQW RU JODVV PDWUL[ FRPSRVLWLRQDO FKDQJHV WKDW LQIOXHQFH .6 UHTXLUH D ORJDULWKPLF IXQFWLRQ WR GHVFULEH WKH ZKROH UDQJH )RUPDWLRQ RI HTXDWLRQ DOVR UHTXLUHG WKH RPLVVLRQ RI WZR RXWOLHU GDWD IULW + ILUHG WR r& DW b DQG b =U 9 &DOFXODWHG 'XUELQ :DWVRQ DQG UHVLGXDO VWDWLVWLFV IHOO ZLWKLQ WKH DFFHSWDEOH UDQJH 0RGHO DWWHVWV WKDW SLJPHQW ORDGLQJ KDV DQ LQIOXHQFH RQ FRORU VWDELOLW\ 7KH FXUYLOLQHDU UHODWLRQVKLS EHWZHHQ ,Q$(r DQG =U9 FRQWHQW LQGLFDWHV WKDW DQ LQWHUPHGLDWH SLJPHQW OHYHO LV OHDVW YLVXDOO\ VWDEOH ZLWK WHPSHUDWXUH 7KLV FRQIRUPV ZLWK )LJXUH ZKHUH $(r UHVXOWV ZHUH RYHUDOO JUHDWHU IRU b =U9 WKDQ b RU b $OWKRXJK $(r BrF DQG .6 UHVXOWV GLG QRW FRUUHODWH ZHOO WR HDFK RWKHU OLQHDU FRUUHODWLRQ FRHIILFLHQW U f WKH VDPH LQGHSHQGHQW YDULDEOHV ZHUH IRXQG WR

PAGE 259

VLJQLILFDQWO\ DIIHFW ERWK DOEHLW ZLWK GLIIHUHQW PDJQLWXGH DQG GLUHFWLRQ LQ VRPH FDVHV (YDOXDWLRQ RI WKH FRHIILFLHQWV LQ UHYHDOV +LJKHU =U LQ WKH IULW LPSURYHV FRORU VWDELOLW\ ORZHU $(r f ,QFUHDVHG 6U2 ORZHUV VWDELOLW\ UDLVHV $(rf 7KH ;;f HIIHFW RI $ODONDOLV GLFWDWHV WKDW DW ORZ OHYHOV LQFUHDVLQJ $ODONDOLV WHQGV WR VOLJKWO\ UDLVH $(r EXW DW KLJK OHYHOV VWDELOLW\ LV LPSURYHG )LJXUH FRPSDUHV PHDVXUHG DQG HTXDWLRQ SUHGLFWHG $(r IRU DOO HLJKW ILUHG FRDWLQJV EDWFKHG ZLWKLQ b =U9 'LVSOD\HG LV WKH JRRG ILW RI WKH PRGHO DQG WKH WHQGHQF\ IRU LQFUHDVHG FRORU VHQVLWLYLW\ WR WHPSHUDWXUH ZKHQ =U LV UHPRYHG ( ) +f &RORU YDULDWLRQ LV IXUWKHU LQFUHDVHG LQ (+ ZKHQ WKH DONDODL OHYHO LV UDLVHG ) DQG +f DQG ZKHQ 6U2 UHSODFHV =Q2 DQG +f ,QGLYLGXDO HIIHFWV RI R[LGH FRPSRVLWLRQV RQ $(r DW b =U9 FDOFXODWHG E\ LQSXWWLQJ WKH UDQJH RI FRPSRVLWLRQ WHVWHG LQWR DQG KROGLQJ WKH RWKHU R[LGHV DW WKHLU DYHUDJH OHYHO DUH GLVSOD\HG LQ )LJXUH 1RWH WKH \D[LV VFDOH RI WR UHSUHVHQWV D GUDPDWLFDOO\ QRWLFHDEOH FRORU FKDQJH $ $(r A LV QRWLFHDEOH ZLWK WKH KXPDQ H\H /LQHDU WUHQGV RI LPSURYHG FRORU VWDELOLW\ ZLWK KLJKHU =U WR PRODU

PAGE 260

' D ( f -r UH D! D 2 2 k X! ( R m ; UH R D & UH e 2 N R R R ‘ 0HDVXUHG 9DOXHV ’ (TXDWLRQ 3UHGLFWLRQV + %6 % + )ULW )LJXUH 'HOWD (r DFWXDO DQG HTXDWLRQ f SUHGLFWHG UHVXOWV IRU ILUHG FRDWLQJV EDWFKHG ZLWK b =U9

PAGE 261

( R fr ; 8e R f R D" X 7 2f 4 UH 2 ‘& 2 2 &2 R 6U2 0RODU (TXLYDOHQWV p $, $ONDOLV 0RODU (TXLYDOHQWV )LJXUH ,QGLYLGXDO HIIHFWV RI IULW R[LGHV RQ 'HOWD (r EDVHG RQ VWDWLVWLFDO PRGHO IRU b =U9

PAGE 262

HTXLYDOHQWVf DQG ORZHU 6U2 WR PRODU HTXLYDOHQWVf DUH VKRZQ 7KH PLGGOH JUDSK LOOXVWUDWHV WKDW 6U2 DGGLWLRQV LQFUHDVH $(r EXW WKH RYHUDOO FRORU VWDELOLW\ LV EHWWHU LI =U LV DOVR SUHVHQW 7KH ERWWRP SORW VKRZV LQFUHDVHV LQ $ODONDOLV DW KLJK OHYHOV PRODU HTXLYDOHQWVf SURYLGHG JUHDW LPSURYHPHQWV LQ FRORU VWDELOLW\ 7KH ZHLJKWV RI LQIOXHQFH RI WKH R[LGHV RQ $(r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

PAGE 263

&' WU &' L f D 8f &' & D K f f R FQ &O RR f UW 8f R &' f R9! UW + ,6 &O UW D! +r K 2 + &' f Kr WU L ‘ % R D &' e 0 &' FO + -r WU f U &3L 2 f§ +L R + FQ R?R +L f§WVL & WL f 3 R f§ % 2 2 % f§ 1! UW f &8 2 R?R LQ r 6, n 0 3HUFHQW RI 7RWDO ,QIOXHQFH RQ &RORU &KDQJH 'HOWD (ff 3HUFHQW RI 7RWDO ,QIOXHQFH RQ &RORU &KDQJH 'HOWD (rf 3HUFHQW RI 7RWDO ,QIOXHQFH RQ &RORU &KDQJH 'HOWD (rf

PAGE 264

$, $ONDOLV 0RODU (TXLYDOHQWV )LJXUH &RORU FKDQJHV GXH WR YDULDWLRQV LQ SHDN ILULQJ WHPSHUDWXUH DQG IULW $,2DONDOL UDWLR RI FRDWLQJV EDWFKHG ZLWK b =U9

PAGE 265

EXW KDV DQ LQVLJQLILFDQW LQIOXHQFH DW ORZ OHYHOV ZKHQ =U LV SUHVHQW /RJ 9LVFRVLW\ YV &RDWLQJ &RPSRVLWLRQ DQG 7HPSHUDWXUH )RU FRDWLQJV EDWFKHG ZLWK b =U9 WKHLU YLVFRVLW\ UMf GXULQJ KHDWLQJ IURP r& WR r& ZDV WKRURXJKO\ H[SODLQHG E\ WKH PRGHO 9DULDEOH &RHIILFLHQW W9DOXH
PAGE 266

FRPSRVLWLRQ HIIHFW RI WKH ILUVW IRXU IDFWRUV LQ 7KH % FRQVWDQW LQ LV WKH FRHIILFLHQW IRU 7 LQ 7KH VDPH IULW R[LGHV IRXQG WR VLJQLILFDQWO\ PRGLI\ $(r DQG .6 DOVR FRQWUROOHG YLVFRVLW\ EXW ZLWK GLIIHUHQW PDJQLWXGHV RI LQIOXHQFH 7HPSHUDWXUH RI FRXUVH KDG WKH ODUJHVW HIIHFW ZLWK LWV LQYHUVH UHODWLRQVKLS WR ORJ U? 7KH VHFRQG JUHDWHVW RYHUDOO VKLIW LQ YLVFRVLW\ UHVXOWHG IURP WKH UDQJH LQ IULW $ODONDOL HTXLYDOHQWV WHVWHG 8QOLNH UHODWLRQVKLSV ZLWK .6 DQG $(r WKH $ODONDOLV PRODU HTXLYDOHQWV LQ WKH IULW GLG QRW SURGXFH D FXUYLOLQHDU HIIHFW RQ YLVFRVLW\ /RJ U? OLQHDUO\ LQFUHDVHG ZLWK $ IURP WR PRODU HTXLYDOHQWVf DQG GHFUHDVHG ZLWK DONDOL FRQWHQW IURP WR HTXLYDOHQWVf $W DQ DYHUDJH WHPSHUDWXUH RI r& DQG ZLWK RWKHU R[LGHV LQ KHOG DW WKHLU DYHUDJH PRODU HTXLYDOHQW OHYHO =U =Q2 6U2 f WKH UDQJH RI $ODONDOLV WHVWHG SURGXFHG D FKDQJH LQ ORJ 7f 8VLQJ WKH VDPH SURFHGXUH =U KDG WKH VHFRQG ODUJHVW R[LGH HIIHFW ZKHUH LQFUHDVLQJ =U WKURXJK WKH UDQJH H[SHULPHQWHG WR HTXLYDOHQWVf FDXVHG ORJ U? WR UDLVH E\ 6U2 FRQWHQW KDG WKH OHDVW LQIOXHQFH RI WKH VLJQLILFDQW LQGHSHQGHQW YDULDEOHV ZKHUH LQFUHDVLQJ 6U2 IURP WR HTXLYDOHQWV UHVXOWHG LQ DQ HVWLPDWHG ORZHULQJ RI ORJ 7f E\

PAGE 267

,Q WKH 'LVFXVVLRQ 6HFWLRQ RI WKLV GLVVHUWDWLRQ WKH SUHFHGLQJ YLVFRVLW\ GDWD DUH UHODWHG WR FRORU VWUHQJWK DQG VWDELOLW\ UHVXOWV 7KH SRWHQWLDO IRU DSSO\LQJ YLVFRVLW\ GDWD WR HVWLPDWH ]LUFRQ SLJPHQW FRORU VWDELOLW\ LQ QHZ IULW FRPSRVLWLRQV LV FRQVLGHUHG AB (YROYHG &U\VWDOOLQH 6SHFLHV ;UD\ VFDQV RI WKH XQILUHG IULWV ZLWKRXW SLJPHQW RU EHQWRQLWH H[KLELWHG QR FU\VWDOOLQH SKDVH SHDNV ;5' 6(0 DQG ('6 (YDOXDWLRQV $QDO\VHV RI FU\VWDOOLQH SKDVHV SUHVHQW LQ ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 DUH GHWDLOHG LQ WKLV VHFWLRQ 7KHVH GDWD DUH IXUWKHU OLQNHG WR FRORU SURSHUWLHV LQ WKH 'LVFXVVLRQ 6HFWLRQ D )ULWV ZLWK =U2" ;5' SDWWHUQV IURP ILUHG FRDWLQJV EDWFKHG ZLWK IULWV $ % & DQG DOO RI ZKLFK FRQWDLQHG =U DUH GLVSOD\HG LQ )LJXUHV DQG UHVSHFWLYHO\ &RUUHVSRQGLQJ 6(0 PLFURJUDSKV DUH SLFWXUHG LQ )LJXUHV DQG IRU & DQG 'f

PAGE 268

&SV 7KHWH 6FDOH )LJXUH ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW $ DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW $ LQFOXGHV b =U b =Q2 b 6U2 b DONDOLV

PAGE 269

)LJXUH 6(0 PLFURJUDSKV RI D FRDWLQJ EDWFKHG ZLWK IULW $ DQG b =U9 DQG ILUHG WR r& 0DJQLILFDWLRQ LV Df ; DQG Ef ; 6KRZQ DUH ODUJH =U9 SDUWLFOHV VXUURXQGHG E\ ILQH ]LUFRQ SUHFLSLWDWHV

PAGE 270

FSV 7KDWD 6FDOH )LJXUH ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW % DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW % LQFOXGHV b =U b =Q2 b 6U2 b DONDOLV

PAGE 271

)LJXUH 6(0 PLFURJUDSKV RI D FRDWLQJ EDWFKHG ZLWK IULW % DQG b =U9 DQG ILUHG WR r& 0DJQLILFDWLRQ LV Df ; DQG Ef ; 6KRZQ DUH ODUJH =U9 SDUWLFOHV ILQH ]LUFRQ SUHFLSLWDWHV DQG ODUJH ]LUFRQ ILEHUV

PAGE 272

&SV (7KHWD 6FDOH ,r& = ] = =LUFRQ =U6Lf ] I r& L Af§ IO LRQQrF f§‘f§‘f§Lf§Lf§Lf§} nL nf§Lf§!f§‘f§ff§f§Lf§‘f§ )LJXUH ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW & DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW & LQFOXGHV b =U b =Q2 b 6U2 b DONDOLV

PAGE 273

7KDWD 6FDOH )LJXUH ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW LQFOXGHV b =U b =Q2 b 6U2 b DONDOLV

PAGE 274

Ef : f f A F f m f r A f 9 r W .8 8 U ,,, ;, PP )LJXUH 6(0 PLFURJUDSKV RI FRDWLQJV EDWFKHG ZLWK =U9 Df IULW & DQG Ef IULW DQG ILUHG WR r& PDJQLILFDWLRQ ; f 3DUWLFOHV VKRZQ DUH =U9 SLJPHQW

PAGE 275

3DWWHUQV IURP DOO IRXU FRDWLQJV DOLJQ ZHOO ZLWK WKH ]LUFRQ SURILOH VKRZQ LQ )LJXUH ZKHUH VWURQJ SHDNV H[LVW DW UHIOHFWLRQ SODQHV DQG RI > @ r > @ r > @ r DQG > @ r =LUFRQ ZDV WKH RQO\ SKDVH GHWHFWHG LQ $& DQG ZKLOH DGGLWLRQDO SKDVHV ZHUH REVHUYHG LQ % $ SUHYLRXV LQYHVWLJDWLRQ > %OR@ QRWHG WKDW ]LUFRQ WHQGV WR VWURQJO\ RULHQW DORQJ WKH FD[LV LQ WKH SODQH RI D FRDWLQJnV VXUIDFH :LWK ]LUFRQnV WHWUDJRQDO VWUXFWXUH D $ F ƒf RULHQWDWLRQ DORQJ WKH FD[LV UHVXOWV LQ HQKDQFHG > K22@ OLQHV =LUFRQ LQ % & DQG FDXVHV H[DJJHUDWHG > @ OLQHV ZKLFK WHQG WR ULVH ZLWK SHDN WHPSHUDWXUH $Q LQFUHDVH LQ WHPSHUDWXUH ORZHUV WKH FRDWLQJ YLVFRVLW\ DQG DOORZV IRU IDVWHU WUDQVSRUW RI ]LUFRQ IURP WKH EXON PHOW WR WKH VXUIDFH 7KH ]LUFRQ SURILOH IURP IULW $ )LJXUH f KDV D VWURQJHU > @ OLQH WKDQ % & RU 6(0 PLFURJUDSKV IRU $ LQ )LJXUH UHYHDO QXPHURXV VPDOO SPf ]LUFRQ SUHFLSLWDWHV LQ WKH JODVV PDWUL[ VXUURXQGLQJ D ODUJH SDUWLFOH ZKRVH PRUSKRORJ\ PDWFKHV WKDW RI WKH =U9 SLJPHQW SLFWXUHG LQ )LJXUH 7KH VPDOO URXQG SUHFLSLWDWH FOXPSV DUH UHVSRQVLEOH IRU WKH > @ LQWHQVLW\ UHODWLYH WR WKH H[DJJHUDWHG > @ IURP WKH SLJPHQW 7KXV LQ IULW $ ERWK

PAGE 276

)LJXUH ;UD\ GLIIUDFWLRQ SURILOH IRU ]LUFRQ =U6L -&3'6 FDUG ILOH GDWD IURP )HUUR &RUSRUDWLRQ &OHYHODQG 2+f

PAGE 277

)LJXUH 6(0 PLFURJUDSK RI ]LUFRQYDQDGLXP =U9f SLJPHQW &HUHGHF $ ; PDJQLILFDWLRQ

PAGE 278

]LUFRQ SUHFLSLWDWLQJ IURP =U LQ WKH IULW DQG ]LUFRQ IURP WKH =U9 SLJPHQW VWUXFWXUH ZHUH LGHQWLILHG (QODUJHG ;5' SDWWHUQV IURP IULW % LQ )LJXUH VKRZ WKH SUHVHQFH RI QRW RQO\ ]LUFRQ EXW DOVR =UWHWUDJRQDO D ƒ F $f SHDNV DW > @ r > @ r DQG > @ r DJDLQ ZLWK H[DJJHUDWHG > K22@ ,QWHQVLWLHV RI WKH =UW OLQHV GURS VLJQLILFDQWO\ ZLWK DQ LQFUHDVH LQ ILULQJ WHPSHUDWXUH IURP r& WR r& 9HU\ VPDOO SHDNV RI =UPRQRFOLQLF EDGGHOH\LWHf DOVR DSSHDU DW r& DW H[DJJHUDWHG > @ r DQG > @ r 7KH =UPRQRFOLQLF FRPSOHWHO\ GLVDSSHDUV DW r& 7KXV ERWK ]LUFRQLD SKDVHV FU\VWDOOL]HG EHORZ r& DQG GLVVROYHG IURP r& WR r& 6(0 IRFXV RQ WKH IULW % FRDWLQJ ILUHG WR r& )LJXUH f UHYHDOHG =U9 SLJPHQW VPDOOHU SDUWLFOHV ZKLFK UHVHPEOH WKH ]LUFRQ SUHFLSLWDWHV LQ IULW $ DV ZHOO DV ODUJH ILEHUV RI ]LUFRQ 2YHUDOO 6(0 VKRZHG WKHUH ZDV PXFK OHVV ILQH ]LUFRQ SUHFLSLWDWHG IURP % ZKHQ FRPSDUHG WR $ ,Q DGGLWLRQ WR VWURQJ ;5' SHDNV DW > @ SLJPHQWf DQG >@ VSKHULFDO ]LUFRQ SUHFLSLWDWHVf WKH ILEURXV ]LUFRQ UHVXOWHG LQ VWURQJ >@ r f LQWHQVLWLHV 6PDOO FRPSRVLWLRQDO GLIIHUHQFHV EHWZHHQ IULWV $ DQG % SURGXFHG YHU\ GLIIHUHQW FU\VWDOOL]DWLRQ WHQGHQFLHV 0DLQO\ WKH 6LDONDLLV UDWLR YDULHG ZKHUH $ FRQVLVWHG RI b 6L2

PAGE 279

DQG b .1D DQG % FRQWDLQHG b 6L DQG b .1D 7KH b LQFUHDVH LQ DONDOLV DQG b UHGXFWLRQ RI VLOLFD PD\ GHWHU ]LUFRQ SUHFLSLWDWLRQ DQG LQGXFH IRUPDWLRQ RI ORZHU WHPSHUDWXUH =U SKDVHV REVHUYHG LQ % IRU WKH IROORZLQJ UHDVRQV /HVV 6L LV DYDLODEOH IRU IRUPDWLRQ RI =Ur6L $Q LQFUHDVH LQ QRQEULGJLQJ DONDOL LRQV LQ WKH JODVV UHVXOWV LQ ORZHU HQHUJ\ UHTXLUHPHQWV IRU EUHDNLQJ XS WKH QHWZRUN )LJXUHV DQG GLVFORVH WKDW UDLVLQJ WKH DONDOLV IURP b WR b GURSSHG WKH VRIWHQLQJ WHPSHUDWXUH IURP r& WR r& /RJ U? ZDV DOVR UHGXFHG E\ D IDFWRU RI WR DFURVV WKH WHPSHUDWXUH UDQJH 0RUH UDSLG GLIIXVLRQ WUDQVSRUW GXH WR D GURS LQ ORJ 7_ WR DW WKH WHPSHUDWXUHV ZKHUH =U SKDVHV KDYH EHHQ IRXQG WR EHJLQ SUHFLSLWDWLQJ r&f UHVXOWV LQ VLJQLILFDQW =U FU\VWDOOL]DWLRQ (QRXJK =U LV SUHFLSLWDWHG DW ORZ WHPSHUDWXUH WR ZKHUH WKHUH LV VWLOO D VPDOO DPRXQW RI XQGLVVROYHG FU\VWDOV UHPDLQLQJ DW r& 0RUH HQHUJ\ LV UHTXLUHG WR IRUP ]LUFRQ IURP FU\VWDOOLQH =U DQG YLWUHRXV VLOLFD WKDQ ZKHQ ERWK R[LGHV DUH LQ WKH YLWUHRXV VWDWH &U\VWDOOL]DWLRQ GDWD IRU % DQG )LJXUH LQGLFDWH WKDW YDULDWLRQV LQ JORVV LQ WLOHV

PAGE 280

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f /RZHU YLVFRVLWLHV ZRXOG WHQG WR IDYRU UDWKHU WKDQ LQKLELW FU\VWDOOL]DWLRQ VR WKLV GRHV QRW SURYLGH DQ H[SODQDWLRQ IRU WKH ODFN RI FU\VWDOOL]DWLRQ ZKLFK UHVXOWHG LQ LQFUHDVHG FRORU VWUHQJWK ZLWK 6U2 7KH UROHV RI 6U2 YHUVXV =Q2 LQ ]LUFRQ SUHFLSLWDWLRQ GLVVROXWLRQ DQG FRORU GHYHORSPHQW ZLOO EH IXUWKHU GLVFXVVHG LQ 6HFWLRQ F

PAGE 281

E )ULWV ZLWKRXW =U2" ;5' SDWWHUQV IURP ILUHG FRDWLQJV EDWFKHG ZLWK IULWV ( ) DQG + DUH GLVSOD\HG LQ )LJXUHV DQG UHVSHFWLYHO\ &RUUHVSRQGLQJ 6(0 PLFURJUDSKV DUH SLFWXUHG LQ )LJXUHV DQG (DFK RI WKH FRDWLQJV (+ DW DOO WKUHH SHDN ILULQJ WHPSHUDWXUHV VWLOO FRQWDLQHG VRPH ]LUFRQ SLJPHQW ZKLFK LV HYLGHQW E\ > @ OLQHV ZKLFK DUH HQKDQFHG GXH WR VXUIDFH RULHQWDWLRQ $OO RI WKH ]LUFRQ OLQHV GURSSHG ZLWK LQFUHDVHG SHDN WHPSHUDWXUH LQGLFDWLQJ SLJPHQW GLVVROXWLRQ LQ WKH IULWV ZLWKRXW =U =LUFRQ SUHFLSLWDFLQ DQG GLVVROXWLRQ DUH GHWDLOHG ZLWK D TXDQWLWDWLYH DQDO\VLV LQ 6HFWLRQ 'LVVLPLODU WR FRDWLQJV $' SUHFLSLWDWLRQ RI QRQn ]LUFRQLXP VSHFLHV ZDV SURQRXQFHG LQ (+ DQG WKH DPRXQW RI FU\VWDO SKDVH LQFUHDVHG ZLWK KLJKHU IULW DONDOL OHYHOV 6(0 DQG ;5' DOVR LQGLFDWHG WKDW XQOLNH $' ZKHUH FU\VWDOOL]DWLRQ ZDV QHDUO\ FRPSOHWH E\ r& WKH WRWDO DPRXQW RI SUHFLSLWDWH LQFUHDVHG LQ (+ IURP r&r WR r& (QKDQFHG FU\VWDOOL]DWLRQ LQ WKH IULWV ZLWKRXW =U FRXOG QRW EH H[SODLQHG ZLWK FKDQJHV LQ YLVFRVLW\ RU LRQLF SRWHQWLDOV RI WKH FDWLRQV 2YHUDOO QR VLJQLILFDQW GLIIHUHQFHV EHWZHHQ YLVFRVLW\ FXUYHV IRU $' ILJXUH f YHUVXV (+ )LJXUH

PAGE 282

&SD 7KHWD 6FDOH )LJXUH ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW ( DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW ( LQFOXGHV b =U b =Q2 b 6U2 DQG b DONDOLV

PAGE 283

( f } r m n W-P .8 ;,f PP )LJXUH 6(0 PLFURJUDSK RI D FRDWLQJ EDWFKHG ZLWK IULW ( DQG b =U9 DQG ILUHG WR r& 0DJQLILFDWLRQ LV ; 6KRZQ DUH ODUJH =U9 SDUWLFOHV VXUURXQGHG E\ GLVSHUVHG GLRSVLGH

PAGE 284

2SD % 7KHWD 6FDOH r& f§ f§ + S O } +' ',, + = =LUFRQ =U6Lf 'LRSVLGH &D0J6Lf ,, ,ODUG\VWRQLWH &D=Q6Lf + r& L / r& f§/ER U f§cDU f + *2 )LJXUH ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW ) DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW ) LQFOXGHV b =U b =Q2 b 6U2 DQG b DONDOLV

PAGE 285

)LJXUH 6(0 PLFURJUDSKV RI D FRDWLQJ EDWFKHG ZLWK IULW ) DQG b =U9 DQG ILUHG WR r& 0DJQLILFDWLRQ LV Df ; DQG Ef $OO SDUWLFOHV VKRZQ DUH KDUG\VWRQLWH

PAGE 286

7KDWD 6FDOH )LJXUH ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW LQFOXGHV b =U2" b =Q2 b 6U2 DQG b DONDOLV

PAGE 287

* .8 .} XQ f ;, PP )LJXUH 6(0 PLFURJUDSK RI D FRDWLQJ EDWFKHG ZLWK IULW DQG b =U9 ILUHG WR r& PDJQLILFDWLRQ ; f 6KRZQ LV FU\VWDOOL]HG 6U&D6J

PAGE 288

&SV 7KHWD 6FDOH r& M f§ V V 6=6 L ] LN r L = =LUFRQ =U6L2Af i 6 6U&D6LM 6 $ r& X / f r& ‘ f§f§ r f§ $MO EP\ n9" Wf§Lf§? f§L Uf§UALf§W   f§L }L f} Uf§f§I fPAf§? rrr"rf! UUn L )LJXUH ;5' SDWWHUQV IURP FRDWLQJV EDWFKHG ZLWK IULW + DQG b =U9 DQG ILUHG WR r& r& DQG r& )ULW + LQFOXGHV b =U b =Q2 b 6U2 DQG b DONDOLV

PAGE 289

)LJXUH 6(0 PLFURJUDSKV RI D FRDWLQJ EDWFKHG ZLWK IULW + DQG b =U9 DQG ILUHG WR r& 0DJQLILFDWLRQ LV Df ; DQG Ef ; 6KRZQ LV FU\VWDOOL]HG 6U&D6J

PAGE 290

f DUH GLVFHUQDEOH $OWKRXJK =U KDV D UHODWLYHO\ KLJK LRQLF SRWHQWLDO RI VXPPDWLRQV RI WKH PROHV RI HDFK FDWLRQ PXOWLSOLHG E\ LWV LRQLF SRWHQWLDO IURP 7DEOH f UHYHDO WKDW WKH RYHUDOO ERQG VWUHQJWKV RI (+ UDQJH IURP IRU + WR IRU (f DUH QRW VLJQLILFDQWO\ GLIIHUHQW IURP $' UDQJH IURP IRU WR IRU $f 7KH LQFUHDVH LQ $ LRQLF SRWHQWLDO f LQ (+ PDNHV XS IRU WKH ORVV RI =U 7KH VWURQJ SURJUHVVLRQ RI FU\VWDOOL]DWLRQ LQ (+ FDXVHG VLJQLILFDQW YDULDWLRQV LQ WKH RSWLFDO SURSHUWLHV RI WKH FRDWLQJV 7KLV LV HYLGHQW E\ WKH ODUJH VKLIWV LQ VSHFWUDO FXUYHV )LJXUHV f /r Dr DQG Er YDOXHV )LJXUHV f DQG $(r )LJXUH f RYHU WKH UDQJH RI SHDN WHPSHUDWXUHV WHVWHG 2YHUDOO ORZHU JORVV ZDV DOVR H[KLELWHG ZKHQ QR =U ZDV SUHVHQW LQ WKH IULW )LJXUH f 'LRSVLGH &D0J 6Lf ZLWK D PRQRFOLQLF VWUXFWXUH ZDV LGHQWLILHG LQ ( DW r&r DQG r& ZKHUH VHYHUDO ;5' OLQHV LQ )LJXUH PDWFK WKH VWDQGDUG SDWWHUQ 'LRSVLGH SHDNV ORZHUHG IURP r& WR r& WKHQ GLVDSSHDUHG DW r& $ 6(0 PLFURJUDSK RI WKH VXUIDFH RI ( ILUHG WR r& )LJXUH f VKRZV ODUJH SLJPHQW SDUWLFOHV VXUURXQGHG E\ D YHU\ IDLQW SDWWHUQ RI FU\VWDOV 7KH IDLQW FU\VWDOV DSSHDU JUDQXODU LQ PRUSKRORJ\ ZLWK SDUWLFOH VL]HV QHDU WR SP

PAGE 291

$Q ('6 RI WKH IDLQW FU\VWDOV FRQILUPHG WKH SUHVHQFH RI GLRSVLGH E\ GHWHFWLQJ PDLQO\ WKH HOHPHQWV &D 0J 6L DQG 1R IDLQW FU\VWDOV ZHUH YLHZHG XQGHU 6(0 IRU ( ILUHG WR r& RQO\ WKH SLJPHQW ZDV SUHVHQW +DUG\VWRQLWH &D=Q6L ZLWK D WHWUDJRQDO VWUXFWXUH ZDV WKH SUHGRPLQDQW FU\VWDOOLQH SKDVH LQ ) )LJXUH VKRZV WKDW KDUG\VWRQLWH SHDN LQWHQVLWLHV DW >@ r& DQG > @ r ULVH GUDPDWLFDOO\ IURP r&r WR r& SHDN WHPSHUDWXUH $Q LQFUHDVH LQ WKH QXPEHU RI KDUG\VWRQLWH FU\VWDOV ZLWK EDVDO > N@ SODQHV VWURQJO\ DOLJQHG DW WKH VXUIDFH ZRXOG DFFRXQW IRU WKH H[DJJHUDWHG SHDNV 6PDOO SHDNV IRU GLRSVLGH DUH DOVR SUHVHQW DW r& EXW QRW DW r& RU r& $SSDUHQWO\ GLRSVLGH GLVVROYHV ZLWK WHPSHUDWXUHV H[FHHGLQJ r& EXW KDUG\VWRQLWH FRQWLQXHV WR IRUP &U\VWDOV SPf RI KDUG\VWRQLWH DUH GLVSOD\HG LQ WKH 6(0 PLFURJUDSKV LQ )LJXUH $Q LQFUHDVH LQ DONDOLV IURP b LQ ( WR b LQ ) DW WKH H[SHQVH RI VLOLFD UHVXOWHG LQ Df PRUH SUHFLSLWDWH LQ WKH JODVV DQG Ef =Q UHSODFLQJ 0J DV WKH FDWLRQ ERQGLQJ ZLWK &D 6L DQG 2 LQ WKH FU\VWDO VWUXFWXUH 3RVVLEOH H[SODQDWLRQV IRU WKHVH REVHUYDWLRQV DUH )URP ( WR ) WKH JODVV VRIWHQLQJ WHPSHUDWXUH GURSSHG VLJQLILFDQWO\ IURP r& WR r& )LJXUH f 7KH

PAGE 292

LQFUHDVH LQ DONDOLV DOVR UHGXFHG WKH PHOW YLVFRVLW\ E\ D ORJ 7f RI DW r& DQG DW r& 7KLV LQFUHDVHG WKH VORSH RI WKH ORJ 7_ YHUVXV WHPSHUDWXUH FXUYH (QKDQFHG GLIIXVLRQ WUDQVSRUW GXH WR WKH ODUJH YLVFRVLW\ GHFUHDVH DW ORZHU WHPSHUDWXUHV ZKHUH QXFOHDWLRQ DQG JURZWK FDQ RFFXU ZRXOG WHQG WR UDLVH WKH QXPEHU DQG VL]H RI FU\VWDOV IRUPHG 6PDOO GURSV LQ YLVFRVLW\ DW KLJK WHPSHUDWXUHV PD\ DLG LQ LQKLELWLQJ GLVVROXWLRQ $GGLWLRQV RI KLJKHU UDGLXV LRQV 1D ƒ ƒf RSHQ XS WKH VWUXFWXUH DQG DOORZ IRU HDVLHU PRYHPHQW RI =Q UDGLXV ƒf WR UHDUUDQJH DQG UHSODFH VPDOOHU 0J UDGLXV ƒf GXULQJ FU\VWDO IRUPDWLRQ %HFDXVH OHVV 6L ZDV DYDLODEOH DV D JODVV IRUPHU DQG WKH DGGLWLRQ RI QRQEULGJLQJ LRQV ZHDNHQV WKH VWUXFWXUH 0J LV UHWDLQHG LQ WKH JODVV RYHU =Q EHFDXVH LWV LRQLF SRWHQWLDO f LV KLJKHU WKDQ ]LQFnV f 6LQFH ERWK IULWV FRQWDLQHG RQO\ b 0J2 YHUVXV b =Q2 WKH UHGXFHG YLVFRVLW\ UHVXOWLQJ IURP KLJKHU DONDOL FRQWHQW ZRXOG HQDEOH IDVWHU FRPSOHWLRQ RI IRUPDWLRQ RI GLRSVLGH ZLWK DLO RI WKH 0J2 SUHVHQW /RZHU WHPSHUDWXUH GLVVROXWLRQ RI GLRSVLGH FRXOG RFFXU ZKLOH KDUG\VWRQLWH

PAGE 293

FRXOG IRUP WKURXJKRXW WKH F\FOH GXH WR WKH ODUJH DPRXQW RI =Q2 SUHVHQW *UHDWHU IOXFWXDWLRQV LQ WKH RSWLFDO SURSHUWLHV RI ) ZKHQ FRPSDUHG WR ( )LJXUHV f FDQ EH DWWULEXWHG WR WKH FU\VWDOOL]DWLRQ GLIIHUHQFHV )LJXUH DOVR VKRZV WKH ODUJH YDULDWLRQ LQ JORVV H[KLELWHG ZLWK ) EXW QRW ( )RU IULW ILUHG WR r& DQG r& YHU\ VPDOO SHDNV IRU ZDLODVWRQLWH &D6L2 ZLWK D WULFOLQLF VWUXFWXUH DUH REVHUYHG DW YDOXHV RI r r DQG r RYHUODSV ZLWK ]LUFRQf LQ )LJXUH 7KH SHDNV GLVDSSHDU DW r& %\ IDU WKH SUHGRPLQDQW FU\VWDOOLQH VSHFLHV LQ DQG WKH RQO\ SUHFLSLWDWH GHWHFWHG LQ + LV UHSUHVHQWHG E\ VWURQJ UHIOHFWLRQV DW YDOXHV RI r DQG r LQ )LJXUHV DQG $OWKRXJK D -&3'6 FDUG VHDUFK LQLWLDOO\ UHYHDOHG QR PDWFKHV WKH IROORZLQJ VWHSV ZHUH WDNHQ WR LGHQWLI\ WKH SKDVH SUHVHQW > % %ORQVNL SHUVRQDO FRPPXQLFDWLRQ HQJLQHHU DW )HUUR &RUSRUDWLRQ &OHYHODQG 2KLR 0D\ @ ;5' VFDQV LQGLFDWHG D OLQH SKDVH UDWKHU WKDQ D VROLG VROXWLRQ EHFDXVH WKH SDWWHUQV GLG QRW VKLIW IURP GLIIHUHQW ILULQJ WHPSHUDWXUHV RU DFURVV WKH VXUIDFH 7KH GDWD ZHUH LQSXW LQWR -DGH 7H[W )LOH %URZVHU 0DWHULDOV 'DWD ,QF FRPSXWHU VRIWZDUH ZKLFK FDOFXODWHV

PAGE 294

GVSDFLQJV IRU GLIIHUHQW VWUXFWXUHV 7KH GVSDFLQJV REVHUYHG IRU DQG + PDWFKHG ZHOO ZLWK D KH[DJRQDO V\VWHP D E r F D r \ rf ZLWK D ƒ F $ DQG D FHOO YROXPH RI $ 6Q\GHU DQG 'H:ROI ILJXUHV RI PHULW EDVHG RQ D VWDWLVWLFDO FRPSDULVRQ RI REVHUYHG YHUVXV FDOFXODWHG GVSDFLQJV ZHUH 0f DQG )f UHVSHFWLYHO\ )LJXUHV RI PHULW A DUH FRQVLGHUHG D JRRG PDWFK WKXV WKHUH ZDV VLJQLILFDQW LQGH[LQJ EHWZHHQ DFWXDO DQG FDOFXODWHG SDWWHUQV $ -&3'6 FDUG VHDUFK RI KH[DJRQDO VLOLFDWHV VKRZHG WKDW WKH FODWWLFH FRQVWDQW IRXQG ZDV ZLWKLQ QRUPDO UDQJH EXW WKH DFRQVWDQW JLYHV D VPDOOHU FHOO YROXPH WKDQ H[SHFWHG $ VWDQGDUG SURFHGXUH RI URWDWLQJ WKH FRRUGLQDWH V\VWHP E\ r ZDV SHUIRUPHG 7KH K2Of FHOO GLUHFWLRQ ZDV PRYHG WR KNOf 7KLV \LHOGHG ODWWLFH FRQVWDQWV RI D $ DQG F $ ZKLFK ZHUH LQ D UHDVRQDEOH UDQJH DQG DOVR SURYLGHG GVSDFLQJV WKDW FORVHO\ PDWFKHG WKH ;5' VFDQV $Q ('6 XQGHU 6(0 GHWHFWHG SULPDULO\ 6U &D DQG 6L LQ DSSUR[LPDWH DWRP UDWLRV RI UHVSHFWLYHO\ 7KH SURSRUWLRQ RI 6L ZDV SUREDEO\ RYHUHVWLPDWHG GXH WR LWV

PAGE 295

KLJK FRQFHQWUDWLRQ LQ WKH EDFNJURXQG PDWUL[ 6LPLODU KH[DJRQDO VWUXFWXUHV IRXQG LQ WKH -&3'6 FDUG ILOH ZHUH %D=U6L D ƒ F ƒf DQG =U.6L D ƒ F ƒf 7KH SUHVHQW VWUXFWXUH ZDV FRPSDUHG WR WKHVH SKDVHV LQ RUGHU WR VXUPLVH WKH FRUUHFW VWRFKLRPHWU\ RI &D 7KH HVWLPDWHG FHOO YROXPH RI WKH SUHVHQW SKDVH LV EHWZHHQ WKH WZR FRPSDUDEOH VWUXFWXUHV $WRPLF UDGLL RI &D ƒf DQG 6U ƒf DOVR IDOO EHWZHHQ =U $f DQG %D ƒf RU ƒf 7KH 6U DQG &D UDGLL DUH UHODWLYHO\ FORVH DQG LQWHUFKDQJHDEOH LQ WKH VLOLFDWH VWUXFWXUH 8VLQJ 3DXOLQJnV UXOHV > .LQ @ FRRUGLQDWLRQV RI DQLRQV DERXW WKH FDWLRQV DUH %D DQG 6U &D DQG =U 7KXV EDVHG RQ FRPSDULVRQV RI FHOO YROXPHV DQG DWRPLF UDGLL VXEVWLWXWLRQV RI 6U DQG &D IRU RU %D DQG =U DUH DOO SRVVLEOH $ VWRFKLRPHWU\ ZLWK WZR PROHV RI FDOFLXP UHVXOWV LQ 6U&D6L 7KH F\FORVLOLFDWH 6Lf VXEFODVV RI VLOLFDWH PLQHUDOV FRQVLVWV RI WKUHH FORVHG ULQJV RI WHWUDKHGUD HDFK VKDULQJ WZR R[\JHQV ,W SURYLGHV IROG WHWUDKHGUDO FRRUGLQDWLRQ EHWZHHQ 6L DQG 2 DQG ƒ ZLGH LQWHUVWLFHV EHWZHHQ HDFK PHPEHU ULQJ 7KH 6L FRQILJXUDWLRQ ZKLFK UHVXOWV LQ 6U&D6L2J LV QRW

PAGE 296

HQHUJHWLFDOO\ DV IDYRUDEOH DQG LV QRW D VWUXFWXUDO FODVVLILFDWLRQ IRU VLOLFDWH OLQNDJHV > 0DV@ *LYHQ WKHVH EDFNJURXQG GDWD WKH SKDVH LQ DQG + ZDV LGHQWLILHG DV 6U&D6L2T 7KLV KDV D VLPLODU VWRFKLRPHWU\ WR KDUG\VWRQLWH =Q&D6Lf IRXQG LQ ) ZKLFK FRQWDLQHG =Q2 UDWKHU WKDQ 6U2 7KH 6U&D6L LV D QHZ OLQH SKDVH WKDW LV QRW FXUUHQWO\ LQ WKH [UD\ -&3'6 GDWD ILOH 3HDN LQWHQVLWLHV IRU 6U&D6L LQ ERWK DQG + LQFUHDVH VXEVWDQWLDOO\ ZLWK SHDN WHPSHUDWXUHV (VSHFLDOO\ QRWLFHDEOH LQ )LJXUHV DQG DUH WKH > @ r DQG > @ UHIOHFWLRQV ZKLFK DUH H[DJJHUDWHG GXH WR VXUIDFH RULHQWDFLQ 6(0 PLFURJUDSKV LQ )LJXUHV DQG GLVSOD\ YHU\ ODUJH SPf FU\VWDOV RI 6U&D6L LQ FRDWLQJV ILUHG WR r& SHDN WHPSHUDWXUH &RDWLQJ + LV VKRZQ WR KDYH PRUH SUHFLSLWDWHG PDWHULDO DW WKH VXUIDFH ,Q DFFRUGDQFH ZLWK SULRU HYDOXDWLRQV KLJKHU DONDOLV LQ + b .&.1Df UHVXOWHG LQ PRUH FU\VWDO IRUPDWLRQ DQG JUHDWHU IOXFWXDWLRQV LQ FRORU ZLWK WHPSHUDWXUH WKDQ REVHUYHG ZLWK b .1Df 7KH IROORZLQJ SRLQWV VXPPDUL]H WKH UHVXOWV RI 6HFWLRQV D DQG E

PAGE 297

)ULWV ZLWK =U f =LUFRQ DQG D VPDOO DPRXQW RI =U ZHUH WKH RQO\ SKDVHV WKDW SUHFLSLWDWHG GXULQJ ILULQJ &U\VWDOOL]DWLRQ ZDV FRPSOHWH E\ r& f 5HSODFHPHQW RI =Q2 ZLWK 6U2 LQ WKH IULW SUHYHQWHG ]LUFRQ FU\VWDOOL]DWLRQ UHJDUGOHVV RI WKH DONDOL OHYHO WHVWHG f ,QFUHDVLQJ WKH DONDOLV .1Df LQ IULWV ZLWK =Q& ZLWKRXW 6U2f UHGXFHG WKH TXDQWLW\ RI ]LUFRQ SUHFLSLWDWH DQG FDXVHG LWV PRUSKRORJ\ WR EH ODUJH DQG ILEURXV UDWKHU WKDQ VPDOO DQG VSKHULFDO )ULWV ZLWKRXW =U f &U\VWDOOL]DWLRQ DQG ]LUFRQ SLJPHQW GLVVROXWLRQ ZHUH SURQRXQFHG DQG FDXVHG ODUJH VKLIWV LQ FRORU IURP r& WR r& SHDN WHPSHUDWXUH f 7KH DPRXQW RI FU\VWDOOL]DWLRQ DQG GLVVROXWLRQ LQFUHDVHG IURP r& WR r& f ,Q IULWV ZLWK =Q2 &D"=Q6L WHQGHG WR SUHFLSLWDWH ZKLOH 6U&D6L FU\VWDOV IRUPHG LQ IULWV ZLWK 6U2

PAGE 298

f +LJKHU DONDOLV .1Df DQGRU UHSODFHPHQW RI =Q2 ZLWK 6U2 LQ WKH IULWV LQFUHDVHG GRWK FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ 3URSRVHG PHFKDQLVPV DQG WKHLU HIIHFWV RQ RSWLFDO SURSHUWLHV ZLOO EH GHWDLOHG LQ &KDSWHU =LUFRQ 4XDQWLWDWLYH $QDO\VLV (DFK ILUHG FRDWLQJ EDWFKHG ZLWK b =U9 ZDV UHVFDQQHG ZLWK ;5' DW YDOXHV QHDU r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

PAGE 299

(VWLPDWHG :HLJKW 3HUFHQW =LUFRQ ,QWHJUDWHG ;5D\ 3HDN ,QWHQVLW\ RI =LUFRQ >@ 5HIOHFWLRQ )LJXUH 5HODWLRQVKLS EHWZHHQ ;5' > @ LQWHJUDWHG LQWHQVLW\ DQG ZHLJKW SHUFHQW ]LUFRQ LQ XQILUHG FRDWLQJV

PAGE 300

L R 8 “ fr! F R D R 4 e 2f nD! )LJXUH :HLJKW SHUFHQW ]LUFRQ LQ ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 t 4L 4 (O (O 6L + b =U b =U b =U b =U 1R =U 1R =U 1R =U 1R =U b =Q2 b =Q2 b 6U2 b 6U2 b =Q2 b =Q2 b 6U2 b 6U2 b $ONDOL b $ONDOL b $ONDOL b $ONDOL b $ONDOL b $ONDOL b $ONDOL b $ONDOL

PAGE 301

D )ULWV ZLWK =U2" DQG =Q2 $PRXQWV RI ]LUFRQ PXFK JUHDWHU WKDQ WKH b SLJPHQW DGGLWLRQ ZHUH IRXQG LQ $ DQG % =LUFRQ LQ $ ZDV IRXQG WR EH EHWZHHQ b DQG b $OO RI WKH =U LQ WKH IULW ZRXOG EH UHTXLUHG WR SUHFLSLWDWH b ]LUFRQ ,W LV XQOLNHO\ WKDW SLJPHQW ORVV RFFXUUHG LQ $ EHFDXVH ]LUFRQ FU\VWDOOL]DWLRQ ZDV HQHUJHWLFDOO\ IDYRUDEOH DQG WKH $(r IURP r&r& )LJXUH f ZDV H[WUHPHO\ ORZ L f =LUFRQ FRQWHQW LQ IULW % LQFUHDVHG IURP b DW r& WR b DW r& DQG r& 7KLV LV WKH WHPSHUDWXUH UDQJH ZKHUH WKH [UD\ VFDQV VKRZHG WKDW WKH VPDOO DPRXQW RI ]LUFRQLD FU\VWDOV GLVDSSHDUHG DQG ZHUH SUREDEO\ FRQYHUWHG WR ]LUFRQ 7KLV PD\ DFFRXQW IRU WKH GURS LQ JORVV GLVFORVHG LQ )LJXUH 6LQFH ]LUFRQ SUHFLSLWDWLRQ ZDV HQHUJHWLFDOO\ IDYRUDEOH DQG $(r ZDV QR SLJPHQW ORVV ZDV HYLGHQW LQ % E )ULWV ZLWK =U&K DQG 6U2 =LUFRQ SLJPHQW ZDV WKH RQO\ FU\VWDOOLQH SKDVH IRXQG LQ & DQG ZLWK ;5' DQG 6(0 )ULW & \LHOGHG =U9 QHDU b DW DOO ILULQJV ZLWK QR WUHQG GRZQZDUG ZLWK LQFUHDVLQJ WHPSHUDWXUH $ VOLJKW GHFUHDVH LQ ]LUFRQ FRQWHQW IURP r& bf WR r& bf ZDV PHDVXUHG ZLWK :KHQ SUHYLRXV GDWD DUH

PAGE 302

FRPSDUHG & DQG ZLWK 6U2f H[SHULHQFHG ODUJHU $(r YDOXHV )LJXUH DQG PRGHO HTXDWLRQ f VKLIWV LQ .6 ZLWK WHPSHUDWXUH )LJXUH f DQG HQHUJHWLFDOO\ XQIDYRUDEOH ]LUFRQ SUHFLSLWDWLRQ ZKHQ FRPSDUHG WR $ DQG % ZLWK =Q2f 7KLV SURYLGHV HYLGHQFH RI VOLJKW SLJPHQW GLVVROXWLRQ LQ & DQG WR ZLWKLQ WKH DSSUR[LPDWH ]LUFRQ WHVW PHDVXUHPHQW DFFXUDF\ RI sb F )ULWV ZLWKRXW =U2" 6LJQLILFDQW GLVVROXWLRQ RI WKH =U9 SLJPHQW RFFXUUHG ZLWK LQFUHDVLQJ WHPSHUDWXUH LQ DOO RI WKH IULWV ZLWKRXW =U2 1HDUO\ DOO RI WKH SLJPHQW LV SUHVHQW DW r& EXW DW r& =U9 OHYHOV IRU ( ) DQG + GURS WR DQG b UHVSHFWLYHO\ *UHDWHU GLVVROXWLRQ RFFXUUHG ZKHQ DONDOLV LQFUHDVHG IURP b ( DQG *f WR b ) DQG +f 1R VLJQLILFDQW GLIIHUHQFH LQ SLJPHQW ORVV LV HYLGHQW EHWZHHQ IULWV ZLWK =Q2 ( DQG )f RU 6U2 DQG +f 7KXV WKH HQHUJHWLFV RI WKH IULWV ZLWKRXW =U ZHUH IDYRUDEOH IRU Df GLVVROYLQJ =U9 SLJPHQW WR DGG =U WR WKH JODVV VWUXFWXUH DQG Ef UHPRYLQJ &D =Q 6U DQG 6L WR IRUP FU\VWDOOLQH VSHFLHV

PAGE 303

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f ZKHUH FU\VWDOOL]DWLRQ LQFUHDVHG ZLWK ILULQJ WHPSHUDWXUH FRORUV QRW RQO\ SURJUHVVHG OLJKWHU DQG OHVV VDWXUDWHG ORZHU FKURPDf DW KLJKHU

PAGE 304

WHPSHUDWXUHV EXW DOVR VKLIWHG LQ KXH 7KLV ZDV HYLGHQW E\ FKDQJHV LQ WKH /rDrEr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

PAGE 305

&RDWLQJ &RPSRVLWLRQ D =U9 /RDGLQJ &RORU VWUHQJWK GXH WR SLJPHQW DEVRUSWLRQ LQ WHUPV RI .6 ZDV IRXQG WR LQFUHDVH OLQHDUO\ ZLWK SLJPHQW FRQWHQW )LJXUH f ZKLOH YLVXDO FRORU LQWHQVLW\ LQ WHUPV RI /r Dr DQG Er YDOXHV YDULHG ORJDULWKPLFDOO\ )LJXUH f +LJKHU =U9 FRQFHQWUDWLRQV H[KLELWHG OHVV SHUFHLYDEOH FRORU FKDQJHV SHU ZHLJKW SHUFHQW SLJPHQW DGGLWLRQ 7KXV LQGXVWULDO EDWFKLQJ HUURUV DUH OHVV YLVLEOH LQ ILUHG FRDWLQJV ZLWK VDWXUDWHG FRORUV ZKHUH =U9 b ,Q WKHVH FRDWLQJV PRUH WKDQ HQRXJK YDQDGLXP DWRPV DUH SUHVHQW IRU FRQWDFWLQJ QHDUO\ DOO RI WKH LQFLGHQW OLJKW DQG DFKLHYLQJ PD[LPXP DEVRUSWLRQ RI QP UDGLDWLRQ 3UHFLSLWDWLRQ RI W\SLFDO PLQHUDO VWUXFWXUHV LQ WKH JODVV PDWUL[ VXUURXQGLQJ WKH SLJPHQW WHQGV WR LQFUHDVH UHIOHFWLRQ DW QP DQG ORZHU UHIOHFWLRQ DW QP DV VKRZQ IRU EHQWRQLWH LQ )LJXUH 7KH UHVXOWLQJ VSHFWUDO FXUYH UHGLVWULEXWLRQ LV UHODWHG LQ )LJXUH /DUJHU VKLIWV LQ WKH VKDSHV RI WKH FXUYHV RFFXU IRU b DQG b =U9 ZKHQ FRPSDUHG WR b 7KH b VDWXUDWHG FRORU WHQGV WR EH OHDVW VHQVLWLYH WR FKDQJHV LQ JODVV PDWUL[ FRQGLWLRQV GXH WR

PAGE 306

b =U9 f r f f b =U9 b =U9 )LJXUH &KDQJHV LQ UHIOHFWDQFH GLVWULEXWLRQV DW QP DQG QP ZDYHOHQJWKV GXH WR LQFUHDVHV LQ SHDN ILULQJ WHPSHUDWXUH IRU FRDWLQJV EDWFKHG ZLWK b b DQG b =U9

PAGE 307

FU\VWDOOL]DWLRQ LQ (+ 7KXV KLJKHU SLJPHQW FRQFHQWUDWLRQV DUH UHFRPPHQGHG IRU IULWV ZKLFK GR QRW LQFRUSRUDWH =U 9LVLEOH FRORU FKDQJHV GXH WR YDULDWLRQV LQ ILULQJ WHPSHUDWXUH ZHUH JUHDWHVW IRU DQ LQWHUPHGLDWH SLJPHQW ORDGLQJ RI b )LJXUH HTXDWLRQ f LQ IULWV ZKHUH SURQRXQFHG FU\VWDOOL]DWLRQ RFFXUUHG (+f ,Q IULWV ZLWK =U ZKLFK H[KLELWHG OLWWOH RU QR FU\VWDOOL]DWLRQ RU SLJPHQW ORVV IURP r& WR r& SLJPHQW ORDGLQJ ZDV QRW D VLJQLILFDQW YDULDEOH IRU FRORU VWDELOLW\ $Q LQWHUPHGLDWH =U9 FRQWHQW H[KLELWHG WKH JUHDWHVW YLVLEOH FRORU FKDQJH GXH WR D VKLIW LQ SHDN ILULQJ WHPSHUDWXUH SRVVLEO\ EHFDXVH 9HU\ OLJKW FRORUV KDYH Df OHVV SLJPHQW DQG FRORU WR ORVH GXULQJ SURFHVVLQJ DQG Ef WKHUH LV D ORJDULWKPLF UHODWLRQVKLS EHWZHHQ WKH DPRXQW RI OLJKW UHIOHFWHG DQG RXU DELOLW\ WR VHH LW )LJXUH f WKXV WKH KXPDQ H\H LV OHVV VHQVLWLYH WR FKDQJHV LQ OLJKW FRORUV 'DUN VDWXUDWHG FRORUV FRQWDLQ PRUH SLJPHQW SDUWLFOHV WKDQ UHTXLUHG IRU PD[LPXP DEVRUSWLRQ RI LQFLGHQW OLJKW WKHUHIRUH SLJPHQW GLVVROXWLRQ LV OHVV QRWLFHDEOH E =U&Q 2I DOO WKH IULW R[LGHV HYDOXDWHG =U KDG WKH ODUJHVW LQIOXHQFH RQ LPSURYLQJ WKH FRORU VWDELOLW\ RI WKH FRDWLQJV DW

PAGE 308

DQ\ =U9 FRQWHQW )LJXUHV DQG HTXDWLRQV f 7KH VWDWLVWLFDO DQDO\VLV IRXQG DQ LQYHUVH OLQHDU UHODWLRQVKLS EHWZHHQ $(r B DQG =U PRODU HTXLYDOHQWV ,Q IULWV ZLWKRXW =U ODUJH SUHFLSLWDWHV IRUPHG GXULQJ ILULQJ DQG FU\VWDOOL]DWLRQ LQFUHDVHG DW KLJKHU WHPSHUDWXUHV FDXVLQJ D UHGXFWLRQ LQ FRORU VWUHQJWK DQG VWDELOLW\ %HVLGHV 6L DQG % ZKLFK DUH VWULFWO\ JODVV IRUPHUV ]LUFRQLXP KDV D JUHDWHU LRQLF SRWHQWLDO DQG VLQJOH ERQG VWUHQJWK ZLWK R[\JHQ WKDQ DQ\ RWKHU FDWLRQ LQ WKH IULWV H[FHSW $O *ODVVHV ZLWK KLJK =U FRQWHQWV WHQG WR KDYH D VWURQJ KLJK PHOWLQJ YLVFRXV DQG GHQVH VWUXFWXUH +HDWLQJ PLFURVFRSH LPDJHV )LJXUH f VRIWHQLQJ WHPSHUDWXUH 7Vf GDWD )LJXUHV DQG f DQG YLVFRVLW\ SURILOHV )LJXUHV DQG f GLVFORVH WKDW LQFUHDVHV LQ PROWHQ IORZ DQG GHFUHDVHV LQ 7V DQG YLVFRVLW\ GXH WR LQFUHDVHV LQ DONDOL FRQWHQW ZHUH RI PXFK OHVV PDJQLWXGH ZKHQ =U ZDV SUHVHQW LQ WKH IULW $ KLJKHU 6L UDWLR GXH WR D UHGXFWLRQ LQ JODVV IRUPLQJ 6L DQG PRUH QRQEULGJLQJ DONDOL LRQV ZDV SDUWLDOO\ FRPSHQVDWHG IRU E\ =U EULGJLQJ WKH H[FHVV R[\JHQ LQ WKH VWUXFWXUH )ULW % b DONDOLVf SUHFLSLWDWHG OHVV ]LUFRQ WKDQ IULW $ b DONDOLVf EHFDXVH PRUH =U ZDV UHWDLQHG LQ WKH JODVV VWUXFWXUH WR ERQG ZLWK WKH H[WUD 2n IURP WKH KLJKHU DONDOL FRQWHQW ,Q IULWV & DQG ZKHUH 6U2 UHSODFHG =Q2

PAGE 309

DOO RI WKH =U UHPDLQHG LQ WKH JODVV VWUXFWXUH DW ERWK DONDOL OHYHOV WHVWHG IRU UHDVRQV WKDW ZLOO EH GLVFXVVHG LQ WKH QH[W VHFWLRQ 7KH HIIHFW RI =U RQ $(r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f ZKHQ =Q2 ZDV SUHVHQW )LJXUH HTXDWLRQV f DQG UDLVHG .6 ZKHQ 6U2 UHSODFHG =Q2 =LUFRQ SUHFLSLWDWLRQ ZKLFK RFFXUUHG LQ IULWV ZLWK =Q2 DQG =U LQFUHDVHG WKH RYHUDOO UHIOHFWDQFH IURP WKH FRDWLQJ DQG OLJKWHQHG LWV FRORU =LUFRQ IRUPDWLRQ ZDV FRPSOHWH E\ r& DQG UHVXOWHG LQ JRRG FRORU VWDELOLW\ IURP r& WR r& :KHQ 6U2 UHSODFHG =Q2 LQ IULWV ZLWK =U WKH VROXELOLW\ RI =U LQFUHDVHG WR WKH SRLQW ZKHUH QR ]LUFRQ FU\VWDOOL]HG IURP WKH IULW 6LQFH QR ]LUFRQ

PAGE 310

)ULW 'HQVLW\ JFPf R b $ONDOLV A b $ONDOLV )LJXUH ,QIOXHQFH RI IULW GHQVLW\ RQ WKH FRORU VWDELOLW\ RI FRDWLQJV EDWFKHG ZLWK b =U9 SLJPHQW

PAGE 311

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a DQG ]LUFRQ VROXELOLW\ DQG $* RI IRUPDWLRQ RI ]LUFRQf LV JLYHQ LQ )LJXUH $ JUDSKLFDO PHWKRG ZDV HPSOR\HG WR HVWLPDWH KRZ PXFK RI WKH ]LUFRQ SUHVHQW LQ ILUHG FRDWLQJV $' ZDV SLJPHQW UDWKHU WKDQ ]LUFRQ SUHFLSLWDWHG IURP =U LQ WKH IULW 7KH TXDQWLWDWLYH [UD\ DQDO\VLV RI ]LUFRQ FRQWHQW LQ 6HFWLRQ F LGHQWLILHG KRZ PXFK SLJPHQW ZDV LQ (+ ZKHUH WKH IULWV FRQWDLQHG QR =U DQG =U9 ZDV WKH RQO\ IRUP RI ]LUFRQ LQ WKH FRDWLQJVf 6LQFH .6} DV =U9! SRVLWLYH [ D[LV LQWHUFHSWV HVWLPDWH TXDQWLWLHV RI GLVVROYHG SLJPHQW

PAGE 312

)ULW $ &2 + :HLJKW b =U9 3LJPHQW )ULW& A & f /LQHDU &f f§ & /LQHDU &f & ‘/LQHDU &f )LJXUH 3LJPHQW DEVRUSWLRQ IDFWRUV .6f IRU FRDWLQJV EDWFKHG ZLWK =U9 SLJPHQW DQG IULWV FRQWDLQLQJ b =U DQG ILUHG WR r& r& RU r& SHDN WHPSHUDWXUH

PAGE 313

,QWHUFHSWV IRU IULWV $ DQG % DUH DW ]HUR ZKLOH IULW & YDULHV IURP WR DQG LV WR )LJXUH LQGLFDWHV WKDW D VPDOO DPRXQW RI SLJPHQW GLVVROXWLRQ RFFXUV LQ IULWV ZLWK =U ZKHQ 6U2 LV SUHVHQW EXW QRW ZKHQ =Q2 UHSODFHV 6U2 7KH GLVVROXWLRQ KRZHYHU LV VWLOO PXFK OHVV WKDQ IRXQG LQ IULWV ZLWKRXW =U (+f 7KH LQWHUDFWLRQ YDULDEOH 6U2 ; 7f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f DQG R[LGH GHQVLWLHV =Q2 JFP =U JFPf RI ]LQF DQG ]LUFRQLXP DUH FRPSDUDEOH 7KH UDGLXV RI 6U= LV KLJKHU ƒf DQG WKH GHQVLW\ RI 6U2 LV ORZHU JFPf 7KHUHIRUH IDYRUDEOH VXEVWLWXWLRQ RI =Q DV DQ LQWHUPHGLDUH LQ WKH JODVV VWUXFWXUH IRU VRPH =U LV OLNHO\ WR ORZHU WKH DFWLYDWLRQ HQHUJ\ DQG $* RI IRUPDWLRQ RI

PAGE 314

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rf ZKHQ 6U2 UHSODFHV =Q2 EXW QR HIIHFW RQ .6 ZDV IRXQG 7KH RYHUDOO FRORU VWUHQJWK FDQQRW EH UDLVHG ZLWK 6U2 WKURXJK D UHGXFWLRQ LQ ]LUFRQ SUHFLSLWDWLRQ EHFDXVH QR =U LV SUHVHQW LQ WKH IULW WR IRUP ]LUFRQ 3LJPHQW GLVVROXWLRQ KRZHYHU DQG WKH FRUUHVSRQGLQJ KLJK FRORU YDULDWLRQ ZKLFK WHPSHUDWXUH VWLOO LQFUHDVH ZKHQ 6U2 UHSODFHV =Q2 LQ IULWV ZLWKRXW =UR ,G $OR2A$ONDOLV 7KH GDWD DQG VWDWLVWLFDO PRGHOV UHYHDOHG WKDW WKH $ DONDOLV UDWLR KDG D ODUJHU LQIOXHQFH RQ $(r DQG .6 WKDQ RWKHU R[LGHV LQ IULWV ZLWKRXW =U :KHQ QR =U ZDV SUHVHQW KLJKHU UDWLRV UHVXOWHG LQ LPSURYHG FRORU VWUHQJWK DQG VWDELOLW\ DW KLJK WHPSHUDWXUH &RQYHUVHO\ LQFUHDVHV LQ

PAGE 315

$ODONDOLV VOLJKWO\ ORZHUHG FRORU VWUHQJWK DQG VWDELOLW\ LQ IULWV ZKLFK LQFRUSRUDWHG =U $OVR WKH JUHDWHVW LQFUHDVHV LQ PHOW YLVFRVLW\ RFFXUUHG LQ IULWV ZLWK RU ZLWKRXW =U ZKHQ WKH $ODONDOLV ZDV UDLVHG $OWKRXJK $O KDV D ODUJH LRQLF SRWHQWLDO f DQG LQFUHDVHV WKH VRIWHQLQJ SRLQW YLVFRVLW\ DQG FRUUHVSRQGLQJO\ ORZHUV PDVV WUDQVSRUW DQG GLIIXVLRQ UDWHVf DQG GHQVLW\ RI D JODVV LQ WKH SUHVHQW VWXG\ LW ZDV QRW DV HIIHFWLYH DV =U LQ LQKLELWLQJ SLJPHQW GLVVROXWLRQ DQG FU\VWDOOL]DWLRQ DERYH r& )ULWV ZLWK =U \LHOGHG PXFK PRUH VWDEOH RSWLFDO SURSHUWLHV WKDQ IULWV ZKHUH =U ZDV UHSODFHG ZLWK $ DQG &D2 :KHQ =U0 LV PLVVLQJ IULWV (+f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f FRPELQHG ZLWK ODUJH UDGLXV 1D ƒf RU ƒf RFFXSDWLRQ RI QHDUE\ LQWHUVWLFLDO VLWHV FUHDWHV D ODUJHU

PAGE 316

VWUDLQ RQ WKH VWUXFWXUH WKDQ ZKHQ =U ƒf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

PAGE 317

ERWK .6 DQG $(r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f DQG VLJQLILFDQWO\ DOWHUHG WKH VKDSHV RI WKH VSHFWUDO FXUYHV D =LUFRQ 7KH VPDOO UDGLXV RI ]LUFRQLXP SURYLGHV D KLJK GLIIXVLRQ UDWH IRU IRUPLQJ FU\VWDOV )LJXUHV DQG UHYHDOHG QXPHURXV SP SDUWLFOHV DQG SP ILEHUV RI ]LUFRQ

PAGE 318

SUHFLSLWDWHG IURP =U DQG 6L LQ WKH IULW VXUURXQGLQJ SLJPHQW SDUWLFOHV SP LQ VL]H 7KH SDUWLFOH VL]H DQG LQGH[ RI UHIUDFWLRQ RI D PDWHULDO GHWHUPLQH LWV DELOLW\ WR VFDWWHU OLJKW =LUFRQnV GHQVH WHWUDJRQDO VWUXFWXUH \LHOGV D KLJK LQGH[ RI UHIUDFWLRQ Q f ZKHQ FRPSDUHG WR WKH FRDWLQJ PDWUL[ Q f 0D[LPXP VFDWWHULQJ IURP ]LUFRQ RFFXUV ZLWK SDUWLFOH VL]HV RI WR SP > %ORE@ 7KH SP ]LUFRQ SUHFLSLWDWH SDUWLFOHV LQ IULWV $ DQG % DUH VOLJKWO\ VPDOOHU WKDQ WKH RSWLPXP VL]H IRU VFDWWHULQJ WKH PRVW YLVLEOH OLJKW +RZHYHU VL]HV FORVHU WR SP QP QPf LQFUHDVH WKH DPRXQW RI EOXH OLJKW VFDWWHUHG DQG OHVVHQ WKH TXDQWLW\ RI \HOORZ OLJKW UHIOHFWHG EHFDXVH QHDU QP FRUUHVSRQGV WR WKH ZDYHOHQJWK RI EOXH OLJKW 7KLV HQKDQFHV WKH EOXH FRORU SURGXFHG E\ WKH =U9 SLJPHQW ZKLFK DEVRUEV \HOORZ OLJKW ZLWKRXW JUHDWO\ VFDWWHULQJ RWKHU YLVLEOH ZDYHOHQJWKV GXH WR LWV ODUJH VL]H 7KH VPDOO SP ]LUFRQ SUHFLSLWDWHV DOVR SURGXFHG KLJK JORVV ZKHUH FU\VWDOOL]DWLRQ RI ODUJHU SDUWLFOHV LQ IULWV ) DQG + SPf FDXVHG D VLJQLILFDQW LQFUHDVH LQ GLIIXVH UHIOHFWDQFH DQG ORZHULQJ RI VSHFXODU UHIOHFWDQFH )LJXUH f 7KH JORVV RI IULW % ZDV ORZHU WKDQ $ EHFDXVH VRPH ODUJH ILEHUV RI ]LUFRQ SPf KDG SUHFLSLWDWHG DORQJ

PAGE 319

ZLWK WKH VPDOOHU SDUWLFOHV 7KH LQFUHDVHG DONDOL FRQWHQW LQ IULW % UHVXOWHG LQ D ORZHU YLVFRVLW\ PHGLXP ZKLFK HQKDQFHG FU\VWDO JURZWK )LJXUH LOOXVWUDWHV KRZ ]LUFRQ SUHFLSLWDWHV DQG EOXH SLJPHQW PRGLILHG D FRDWLQJnV FRORU /LQHDU UHODWLRQVKLSV UHSUHVHQW ILUHG FRDWLQJV ZKHUH WKH RQO\ ]LUFRQ SUHVHQW LV SLJPHQW Abf DQG QRQOLQHDU SORWV DUH IRU FRDWLQJV FRQWDLQLQJ b =U9 SOXV FU\VWDOOL]HG ]LUFRQ 7KH WZR FXUYHV FRQYHUJH DW b ]LUFRQ ZKHUH RQO\ SLJPHQW LV SUHVHQW 'DWD IURP DOO WKUHH SHDN ILULQJ WHPSHUDWXUHV DUH JUDSKHG ,QFUHDVHV LQ WKH TXDQWLW\ RI ]LUFRQ SUHFLSLWDWH LQ WKH FRDWLQJV IURP b WR b UHVXOWHG LQ JUHDWHU ORVVHV LQ OLJKWQHVV $/r f WKDQ JUHHQQHVV $Dr f RU EOXHQHVV $Er f 7KH SUHFLSLWDWHV FDXVHG PRUH VFDWWHULQJ RI DOO YLVLEOH ZDYHOHQJWKV DQG OHVV DEVRUSWLRQ RI QP OLJKW ZKLFK UHGXFHG WKH RYHUDOO FRORU VWUHQJWK +RZHYHU WKH EOXH FRORU ZDV SDUWLDOO\ SUHVHUYHG GXH WR WKH HQKDQFHG VFDWWHULQJ RI QP OLJKW IURP WKH SP SDUWLFOHV 7KH UHVXOW LV D YHU\ OLJKW EXW GLVWLQFW KLJKFKURPD EOXH )ULWV $ DQG % SURGXFHG WKH KLJKHVW EOXH Erf YDOXHV RI OLJKW FRORUV ZLWK /r YDOXHV GXH WR WKLV SKHQRPHQRQ $Q DQDO\VLV RI WKH VSHFWUDO UHIOHFWDQFH GDWD DOVR VKRZV WKDW E\ FRPSDULVRQ IULW $ \LHOGHG WKH KLJKHVW UHIOHFWDQFH f DW

PAGE 320

/r9DOXH Dr9DOXH Er9DOXH A 3UHFLSLWDWHG =LUFRQ b =U9 3LJPHQW p 2QO\ =U9 3LJPHQW 3UHVHQW )LJXUH 9LVXDO OLJKWQHVV /rf JUHHQQHVV Drf DQG EOXHQHVV Erf DV D IXQFWLRQ RI ZHLJKW SHUFHQW ]LUFRQ LQ WKH ILUHG FRDWLQJV EDWFKHG ZLWK b =U9

PAGE 321

QP EOXH ZDYHOHQJWKV ZKHQ QP UHIOHFWDQFH YDOXHV ZHUH QHDU 6XSSRUW RI WKLV K\SRWKHVLV LV JLYHQ LQ )LJXUH RI SUHYLRXV 6HFWLRQ 'DWD SRLQWV IRU OLJKW FRORUV ZLWK KLJKHU WKDQ SUHGLFWHG /rYDOXHV DERYH WKH /r YHUVXV Er FXUYH UHSUHVHQW IULWV $ DQG % ZKHUH ]LUFRQ SUHFLSLWDWHG $OO RI WKH RWKHU FRDWLQJV DSSHDUHG WR EH GDUNHU \HOORZHU DQG JUHHQHU WKDQ $ DQG % 7KH UHODWLRQVKLS EHWZHHQ FRORU DQG ZHLJKW SHUFHQW ]LUFRQ SUHFLSLWDWH ZDV IRXQG WR EH ORJDULWKPLF /r ,Q =f f Dr ,Q =f f Er ,Q =f f ZKHUH = LV WKH ZHLJKW SHUFHQW RI ]LUFRQ SUHFLSLWDWH 7KHVH UHODWLRQVKLSV FRUUHVSRQG WR WKH SKHQRPHQRQ QRWHG LQ %DFNJURXQG 6HFWLRQ ZKHUH SHUFHLYHG OLJKWQHVV KDV EHHQ IRXQG WR EH D ORJDULWKPLF IXQFWLRQ RI WKH DFWXDO OLJKW OHYHO UHIOHFWHG IURP DQ REMHFW > +XQ@ $W KLJKHU ]LUFRQ FRQWHQWV DQG UHIOHFWDQFHV D OHVV QRWLFHDEOH FRORU GLIIHUHQFH LV SHUFHLYDEOH SHU XQLW FKDQJH LQ UHIOHFWDQFH (TXDWLRQV DOVR FOHDUO\ UHYHDO WKDW Dr DQG Er FKDQJH OHVV WKDQ /r ZLWK LQFUHDVLQJ ]LUFRQ SUHFLSLWDWH FRQWHQW 7KXV D PHWKRG KDV EHHQ TXDQWLILHG IRU SURGXFLQJ YHU\ RSDTXH JORVV\ OLJKW DQG FOHDQ EOXH FRORU E\ PDLQWDLQLQJ D ORZ Er DV /r

PAGE 322

JRHV XS 6LPLODUO\ FRORU IURP RWKHU SLJPHQWV FDQ EH HQKDQFHG WKURXJK FRQWUROOHG FU\VWDOOL]DWLRQ RI ]LUFRQ WR D SDUWLFOH VL]H WKDW WHQGV WR VFDWWHU PRUH OLJKW RI ZDYHOHQJWKV FRUUHVSRQGLQJ WR WKH GHVLUHG FRORU DQG OHVV OLJKW ZKHUH DEVRUSWLRQ E\ WKH SLJPHQW RFFXUV 6HFWLRQV DQG UHYHDOHG WKDW ]LUFRQ FU\VWDOOL]DWLRQ FRXOG EH FRQWUROOHG E\ YDU\LQJ WKH 6U2=Q2 DQG $DONDOL UDWLRV 3LJPHQW GLVVROXWLRQ ZKHQ QR ]LUFRQ SUHFLSLWDWH LV SUHVHQW DV GHQRWHG ZLWK WKH OLQHDU SORWV LQ \LHOGV QHDUO\ WKH VDPH HIIHFW RQ /r DQG Er $GGLQJ RU VXEWUDFWLQJ =U9 DW TXDQWLWLHV Ab GRHV QRW SURGXFH XQLTXH FRORU $V QRWHG LQ 6HFWLRQ PLQLPXP ]LUFRQ SLJPHQW ORVV DQG FRUUHVSRQGLQJO\ JUHDWHU FRORU VWDELOLW\ LV DFKLHYHG ZLWK IULWV IRUPXODWHG WR SUHFLSLWDWH ]LUFRQ 7KH UHODWLRQVKLS EHWZHHQ OLJKW DEVRUSWLRQ E\ WKH SLJPHQW LQ WHUPV RI .6 YDOXHV DQG ZHLJKW SHUFHQW ]LUFRQ LV GLVSOD\HG LQ )LJXUH 7KH FXUYHV DJDLQ FRQYHUJH DW b ]LUFRQ ZKHUH RQO\ SLJPHQW LV SUHVHQW =LUFRQ DIIHFWHG .6 E\ .6 =fr f &RPSDULVRQV RI HTXDWLRQV DQG )LJXUH ZLWK HTXDWLRQ DQG )LJXUH FRQILUP WKDW WKH SP ]LUFRQ SUHFLSLWDWH FU\VWDOV LQIOXHQFHG OLJKWQHVV /rf WR D JUHDWHU GHJUHH WKDQ WKH DEVRUSWLRQ RI OLJKW QHDU QP ZKLFK

PAGE 323

)LJXUH 3LJPHQW DEVRUSWLRQ IDFWRU .6f DV D IXQFWLRQ RI =U9 SLJPHQW DQG ]LUFRQ FRQWHQWV LQ FRDWLQJV EDWFKHG ZLWK b =U9

PAGE 324

FDXVHV EOXHQHVV Erf 7KH DV\PSWRWLF FXUYH LQ )LJXUH UHYHDOV WKDW UDLVLQJ WKH DPRXQW RI ]LUFRQ IURP b WR b GURSSHG .6 E\ RQO\ &RUUHVSRQGLQJO\ EOXH FRORU ORZHUHG E\ RQO\ $Er ZKLOH OLJKWQHVV FKDQJHV ZHUH PXFK PRUH YLVLEOH ZKHUH $/r 7KXV SUHFLSLWDWLRQ RI SP ]LUFRQ IURP b WR b E\ ZHLJKW GRHV QRW VLJQLILFDQWO\ DOWHU WKH EOXH FRORU VWUHQJWK LI WKH =U9 SLJPHQW LV PDLQWDLQHG 7KH ZLGHU VSUHDG RI GDWD SRLQWV RQ FRORU DQG DEVRUSWLRQ SORWV )LJXUHV DQG f ZKHUH RQO\ SLJPHQW LV SUHVHQW LV LQGLFDWLYH RI WKH VLPXOWDQHRXV =U9 GLVVROXWLRQ DQG SURQRXQFHG FU\VWDOOL]DWLRQ RI ODUJH QRQ]LUFRQ VSHFLHV LQ IULWV (+ ,Q IULWV ZKHUH QR ]LUFRQ SUHFLSLWDWHG ERWK SLJPHQW GLVVROXWLRQ DQG H[WHQVLYH FU\VWDOOL]DWLRQ ZHUH PRUH SUHYDOHQW DQG FDXVHG VLJQLILFDQW FRORU YDULDWLRQV )LJXUH f 3UHFLSLWDWLRQ RI ]LUFRQ ZDV FRPSOHWH EHIRUH r& DQG WKH DPRXQW RI ]LUFRQ LQ V\VWHPV ZKHUH LWV FU\VWDOOL]DWLRQ ZDV HQHUJHWLFDOO\ IDYRUDEOH VWD\V QHDUO\ FRQVWDQW IURP r& WR r& 7KH HQHUJHWLF SUHIHUHQFH IRU FU\VWDOOL]LQJ ]LUFRQ LQ WKHVH JODVVHV LQKHUHQWO\ SUHVHUYHV ]LUFRQ SLJPHQWV DQG FRQVHTXHQWO\ \LHOGV JRRG FRORU VWDELOLW\

PAGE 325

:HLJKW 3HUFHQW =LUFRQ DW & )LJXUH &KDQJHV LQ FRORU VWDELOLW\ GHQRWHG E\ .6 DQG 'HOWD (r DV D IXQFWLRQ RI ZHLJKW SHUFHQW ]LUFRQ LQ FRDWLQJV EDWFKHG ZLWK b =U9

PAGE 326

E 'LRSVLGH $ VPDOO DPRXQW RI GLRSVLGH &D0J6Lf ZLWK D PRQRFOLQLF VWUXFWXUH ZDV LGHQWLILHG LQ FRDWLQJV EDWFKHG ZLWK IULW ( DQG ILUHG WR r& DQG r& )LJXUH f EXW WKH GLRSVLGH ZDV FRPSOHWHO\ GLVVROYHG DW r& 7KLV SURJUHVVLRQ FRPELQHG ZLWK =U9 SLJPHQW GLVVROXWLRQ DOWHUHG WKH FRORUV RI WKHVH FRDWLQJV 'LRSVLGHnV GHQVLW\ S JFPf LQGH[ RI UHIUDFWLRQ Q GHSHQGLQJ RQ FU\VWDO RULHQWDWLRQf DQG FRUUHVSRQGLQJ DELOLW\ WR VFDWWHU OLJKW DUH OHVV WKDQ ]LUFRQnV ,WV SDUWLFOH VL]HV SPf LQ WKH ILUHG FRDWLQJ ZHUH WRR ODUJH WR FDXVH SUHIHUHQWLDO VFDWWHULQJ RI YLVLEOH OLJKW ZDYHOHQJWKV 6WURQJ [UD\ SHDNV DW D RI r IURP >@ UHIOHFWLRQ UHVXOW IURP WKH GLRSVLGH VWUXFWXUH )LJXUH ZHLJKV WKH LPSDFW RI SLJPHQW GLVVROXWLRQ DQG GLRSVLGH FU\VWDOOL]DWLRQ VLJQLILHG E\ > @ UHIOHFWLRQ RQ FRORU FKDQJHV 6LPXOWDQHRXV GLVVROXWLRQ RI GLRSVLGH DQG SLJPHQW RFFXUUHG ZLWK LQFUHDVLQJ WHPSHUDWXUH 7KH FRORU FKDQJHV DUH PDLQO\ GXH WR SLJPHQW GLVVROXWLRQ UDWKHU WKDQ WKH VPDOO DPRXQW RI GLRSVLGH IRUPHG DV GHQRWHG E\ WKH ORZ [UD\ SHDN LQWHQVLWLHV %HFDXVH IULW ( FRQWDLQHG =Q2 UDWKHU WKDQ 6U2 DQG D KLJK $ODONDOL UDWLR f LW KDG WKH PRVW VWDEOH JODVV VWUXFWXUH RI WKH IULWV ZLWKRXW =U 7KHVH FRQGLWLRQV

PAGE 327

)LJXUH &RORU OLJKWQHVV /rf DQG EOXHQHVV Erf SURJUHVVLRQ ZLWK GLRSVLGH FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ LQ FRDWLQJV EDWFKHG ZLWK IULW ( DQG b =U9

PAGE 328

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r IRU ( $(r IRU ) DW b =U9f ZHUH FRQVHTXHQFHV &U\VWDOOL]DWLRQ RI KDUG\VWRQLWH &D=Q6L ZLWK D WHWUDJRQDO VWUXFWXUH LQFUHDVHG ZLWK SHDN WHPSHUDWXUH LQ FRDWLQJV EDWFKHG ZLWK IULW ) )LJXUH f ,WV GHQVLW\ S JFPf UHODWLYHO\ ORZ LQGH[ RI UHIUDFWLRQ QR YDOXHV ZHUH IRXQG LQ WKH OLWHUDWXUH EXW Q LV SURSRUWLRQDO WR Sf DQG ODUJH SUHFLSLWDWH VL]H SPf SURGXFHG OHVV VFDWWHULQJ RI YLVLEOH OLJKW WKDQ ]LUFRQ 3URJUHVVLRQ RI KDUG\VWRQLWH FU\VWDOOL]DWLRQ ZLWK SHDN WHPSHUDWXUH LV GHQRWHG E\ WKH VWURQJ [UD\ > @ UHIOHFWLRQ SHDN LQWHQVLWLHV VKRZQ LQ

PAGE 329

)LJXUH 2YHU WKH UDQJH RI KDUG\VWRQLWH SUHFLSLWDWHG =U 9 SLJPHQW ZHLJKW GURSSHG IURP b WR QHDU b 8QOLNH WKH FDVH ZKHUH ILQH FU\VWDOV RI ]LUFRQ SUHVHUYHG VRPH RI WKH EOXH FRORU LQ IULWV $ DQG % VLPXOWDQHRXV SUHFLSLWDWLRQ RI KDUG\VWRQLWH DQG ORVV RI =U9 LQ ) UHVXOW LQ D ODUJHU UHGXFWLRQ LQ EOXH $Er f WKDQ LQFUHDVH LQ OLJKWQHVV $/r f 7KXV D PXFK JUHDWHU ORZHULQJ RI FKURPD ZLWK LQFUHDVLQJ OLJKWQHVV RFFXUUHG G 6WURQWLXP &DOFLXP 6LOLFDWH 7KH PDJQLWXGH RI FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ LQ IULHV ZLWKRXW =U2 LV IXUWKHU LQFUHDVHG ZKHQ 6U2 UHSODFHV =Q2 DV ZDV WKH FDVH LQ IULWV DQG + $ UHGXFWLRQ LQ PHOW YLVFRVLW\ DQG JUHDWHU ]LUFRQ SLJPHQW VROXELOLW\ IRU UHDVRQV GHWDLOHG LQ 6HFWLRQ Ff DFFRPSDQ\LQJ WKH VXEVWLWXWLRQ UHVXOWHG LQ ORZHU FRORU VWDELOLW\ $(r IRU DQG IRU + DW b =U9f DQG JORVV $ SUHYLRXVO\ XQLGHQWLILHG FU\VWDOOLQH SKDVH 6U&D6L FU\VWDOOL]HG LQ FRDWLQJV EDWFKHG ZLWK IULWV DQG + )LJXUH DQG f 7KH ODUJH SP SUHFLSLWDWHV ZLWK D KH[DJRQDO VWUXFWXUH UHODWLYHO\ ORZ GHQVLW\ DSSUR[LPDWHO\ JFP HVWLPDWHG EDVHG RQ FRPSDULVRQV ZLWK VLPLODU VWUXFWXUHVf DQG VPDOO LQGH[ RI UHIUDFWLRQ SURGXFH OHVV

PAGE 330

)LJXUH &RORU OLJKWQHVV /r f DQG EOXHQHVV f§Erf SURJUHVVLRQ ZLWK KDUG\VWRQLWH FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ LQ FRDWLQJV EDWFKHG ZLWK IULW ) DQG b =U9

PAGE 331

YLVLEOH OLJKW VFDWWHULQJ WKDQ ]LUFRQ 7KH DPRXQW RI FU\VWDOOLQH SKDVH IRUPHG LQ WKH JODVV LQFUHDVHG ZLWK SHDN WHPSHUDWXUH ZKLOH WKH TXDQWLW\ RI ]LUFRQ SLJPHQW GHFUHDVHG )LJXUH LOOXVWUDWHV WKH FRORU FKDQJHV GXH WR 6U&D6L FU\VWDOOL]DWLRQ GHQRWHG E\ [UD\ > @ UHIOHFWLRQ SHDN LQWHQVLW\ YHUVXV SLJPHQW ORVV 7KH SORWV VKRZ JUHDWHU =U9 GLVVROXWLRQ LQ IULW + ZKHUH WKH $ODONDOL UDWLR f DQG WKXV WKH PHOW YLVFRVLW\ DUH ORZHU +LJKHU ;5' SHDN LQWHQVLWLHV UHVXOW IURP IULW $ODONDOLV f SHUKDSV GXH WR SUHIHUUHG RULHQWDWLRQ DW WKH VXUIDFH RI WKH FRDWLQJ 7KH JORVV UHDGLQJV )LJXUH f DQG 6(0 PLFURJUDSKV )LJXUHV DQG f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

PAGE 332

)LJXUH &RORU OLJKWQHVV /rf DQG EOXHQHVV Erf SURJUHVVLRQ ZLWK 6U&D6L FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ LQ FRDWLQJV EDWFKHG ZLWK b =U9

PAGE 333

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f 0RGLI\LQJ 1D DQG DONDOL FDWLRQV HQKDQFH IORZ E\ GLVUXSWLQJ ERQG OLQNDJHV DQG EUHDNLQJ XS WKH JODVV QHWZRUN ZKLOH KLJK LRQLF SRWHQWLDO $O FRXQWHUDFWV WKLV HIIHFW 7R D OHVVHU GHJUHH FRDWLQJ YLVFRVLW\ DOVR FRUUHODWHG GLUHFWO\ WR =U FRQWHQW DQG LQYHUVHO\ ZLWK 6U2 1XFOHDWLRQ SURFHHGV DW D ORZHU WHPSHUDWXUH DQG KLJKHU YLVFRVLW\ WKDQ FU\VWDOOL]DWLRQ ,Q IULWV ZLWK =U DQG =Q2 ZKHUH ]LUFRQ FU\VWDOOL]DWLRQ ZDV PRVW IDYRUDEOH IULWV $ DQG

PAGE 334

%f D JUHDWHU QXPEHU RI ILQHO\ GLVSHUVHG VSKHULFDO SUHFLSLWDWHV UHVXOWHG IURP LQFUHDVHG YLVFRVLW\ FRQGLWLRQV IULW $ )LJXUH f ZKHUH WKH $ODONDOLV ZDV KLJKHU ,Q KLJK YLVFRVLW\ PHGLXPV ZLWK ORZ GLIIXVLRQ UDWHV LQWHUQDO HQHUJ\ GXH WR LQWHUIDFHV LV UHGXFHG ZLWK VSKHULFDO SDUWLFOHV /DUJHU SUHFLSLWDWHV ZLWK D ILEURXV PRUSKRORJ\ DQG D VPDOOHU QXPEHU RI SDUWLFOHV FU\VWDOOL]HG LQ WKH ORZHU YLVFRVLW\ PHGLXP IULW % )LJXUH f ZKHUH WKH JURZWK UDWH ZDV KLJKHU 7KH GHPRQVWUDWHG VL]H DQG PRUSKRORJ\ FRQWURO RI ]LUFRQ SUHFLSLWDWLRQ WKURXJK YLVFRVLW\ UHJXODWLRQ FDQ EH XVHG WR DOWHU FRORU JORVV DQG RSDFLW\ LQ JODVV FRDWLQJV ,Q IULWV ZLWK =U& ZKHUH ]LUFRQ GLG QRW FU\VWDOOL]H & DQG 'f KLJKHU VORSHV LQ WKH ORJ U_ YHUVXV WHPSHUDWXUH FXUYHV DUH QRWHG QHDU WKH VRIWHQLQJ SRLQWV RI DSSUR[LPDWHO\ r& /HVV FU\VWDOOL]DWLRQ WHQGV WR UHVXOW ZKHQ WKHUH LV D KLJK YLVFRVLW\ DW WKH VRIWHQLQJ SRLQW 7Vf DQG D YLVFRVLW\ ZKLFK LQFUHDVHV UDSLGO\ EHORZ 7V > .LQ@ +HQFH WKHUH LV OHVV RSSRUWXQLW\ IRU QXFOHL IRUPDWLRQ DQG WKH VXEVHTXHQW YROXPH RI FU\VWDOV SUHFLSLWDWHG DW KLJK WHPSHUDWXUH LV ORZHU 5HSODFHPHQW RI =Q2 ZLWK 6U2 LQ WKHVH IULWV SURPSWHG WKLV RXWFRPH ,Q IULWV ZLWKRXW =U (+f WKH VL]H DQG YROXPH RI =Q2 EDVHG )LJXUH f RU 6U2EDVHG )LJXUHV DQG f

PAGE 335

SUHFLSLWDWHV LQFUHDVHG ZLWK ORZHULQJ YLVFRVLW\ 7KLV FDXVHG VHYHUH ORVV RI FRORU DQG JORVV LQ WKH FRDWLQJV $OWKRXJK UHSODFHPHQW RI =Q2 ZLWK 6U2 LQ WKHVH V\VWHPV DOVR UDLVHG WKH VORSH RQ WKH ORJ U_ FXUYHV QHDU 7 QR FRUUHVSRQGLQJ UHGXFWLRQ LQ FU\VWDOOL]DWLRQ ZDV REVHUYHG 3UHFLSLWDWLRQ RI ]LUFRQ ZKHUH PRVW RI WKH QXFOHDWLRQ DQG JURZWK RFFXU EHORZ r& LV PRUH LQIOXHQFHG E\ WKH YDOXHV DQG UDWHV RI FKDQJHV RI YLVFRVLW\ QHDU 7 &RPSDULVRQV RI IULWV ZLWK DQG ZLWKRXW =U VXJJHVW WKDW =U9 SLJPHQW GLVVROXWLRQ LQ ERWK V\VWHPV FDQQRW EH GLUHFWO\ OLQNHG WR WKH VDPH YLVFRVLW\ FKDUDFWHULVWLFV )RU H[DPSOH IULWV % b =Uf DQG ) b =Uf KDG VLPLODU YLVFRVLW\ FXUYHV EXW DOO RI WKH SLJPHQW ZDV VWLOO SUHVHQW LQ % DIWHU ILULQJ WR r& ZKHUH OHVV WKDQ b ZDV PHDVXUHG LQ ) +RZHYHU LI WKH IULWV ZLWKRXW =U DUH HYDOXDWHG VHSDUDWHO\ DQ LQYHUVH FRUUHODWLRQ EHWZHHQ YLVFRVLW\ DQG ERWK =U9 GLVVROXWLRQ DQG FU\VWDOOL]DWLRQ EHFRPHV HYLGHQW 7KH PHOW YLVFRVLW\ RI IULW V\VWHPV KDV OHVV RI DQ LQIOXHQFH RQ GLVVROYLQJ ]LUFRQ SLJPHQWV ZKHQ HQRXJK =U LV SUHVHQW LQ WKH IULW WR FDXVH ]LUFRQ WR EH WKH RQO\ HQHUJHWLFDOO\ IDYRUDEOH FU\VWDOOLQH SKDVH :KHQ VXIILFLHQW =U LV SUHVHQW LQ WKH JODVV IRU DFKLHYLQJ QHWZRUN FRQQHFWLYLW\ DQG VWDELOLW\ WKHUH

PAGE 336

LV QR GULYLQJ IRUFH IRU GLVVROYLQJ WKH ]LUFRQ SLJPHQW LQ RUGHU WR IUHH XS =U IRU WKH VWUXFWXUH 9DOXH DV D 3UHGLFWRU RI .6 DQG $(r 3ORWV RI PHOW YLVFRVLWLHV YHUVXV .6 DQG $(r RI ILUHG FRDWLQJV EDWFKHG ZLWK b =U9 DUH JLYHQ LQ )LJXUH ;D[LV YDOXHV ZHUH GHULYHG IRU HDFK FRDWLQJ E\ LQSXWWLQJ WKHLU IULW R[LGH PRODU HTXLYDOHQW YDOXHV LQWR HTXDWLRQ DQG WKHQ LQWHJUDWLQJ WKH UHPDLQLQJ IXQFWLRQ RI WHPSHUDWXUH IURP r& WR r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

PAGE 337

)LJXUH ,QWHJUDWHG ORJ YLVFRVLW\ IURP r& WR r& YHUVXV ILUHG FRORU VWUHQJWK .6f DQG VWDELOLW\ 'HOWD (r f LQ FRDWLQJV EDWFKHG ZLWK b =U9

PAGE 338

UDSLGO\ WR HTXLOLEULXP VRRQ DERYH 7J 7KXV YLVFRVLW\ RYHU WKH ZKROH WHPSHUDWXUH UDQJH LV QRW D VLJQLILFDQW IDFWRU LQ FU\VWDOOL]DWLRQ DQG FRUUHVSRQGLQJ FRORU DOWHUDWLRQV )RU IULWV ZLWK =U WKH VORSH RI WKH ORJ YLVFRVLW\ FXUYH QHDU WKH VRIWHQLQJ SRLQW 7 f ZDV IRXQG WR FRUUHODWH SRVLWLYHO\ ZLWK RYHUDOO FRORU VWUHQJWK .6f EXW LQYHUVHO\ ZLWK VWDELOLW\ )LJXUH UHYHDOV WKDW WKLV UHODWLRQVKLS FDQ EH DSSOLHG WR SUHGLFW .6 5 f DQG $(r 5 f /RZ 5 YDOXHV IRU IULWV ZLWKRXW =U YHULI\ QR VLJQLILFDQW FRUUHODWLRQV )RU IULWV ZLWK =U FRORU VWUHQJWK LV JUHDWO\ LQFUHDVHG ZLWK D KLJKHU VORSH GXH WR WKH FRUUHVSRQGLQJ ORZHU FU\VWDOOL]DWLRQ RI ]LUFRQ WKDW WHQGV WR OLJKWHQ FRORU $ YHU\ VOLJKW KDUGO\ QRWLFHDEOH GHFUHDVH LQ VWDELOLW\ $(r IURP WR f UHVXOWLQJ IURP D KLJKHU VORSH LV GXH WR WKH LQFUHDVH LQ =U9 VROXELOLW\ DV D FRQVHTXHQFH RI KLJKHU 6U2 OHYHOV ,I KLJK FRORU VWUHQJWK LV WKH PDLQ REMHFWLYH WLJKWHU FRQWURO RI ILULQJ WHPSHUDWXUHV FDQ EH XVHG WR FRPSHQVDWH IRU WKH VOLJKW ORVV RI VWDELOLW\ ,Q VXPPDU\ )LJXUHV DQG GHPRQVWUDWH WKDW FRORU VWUHQJWK DQG VWDELOLW\ LQ JODVV IULW]LUFRQ SLJPHQW V\VWHPV FDQ EH FRQWUROOHG E\ UHJXODWLQJ WKH PHOW YLVFRVLW\

PAGE 339

)LJXUH 6ORSH LQ WKH ORJ YLVFRVLW\ YHUVXV WHPSHUDWXUH QHDU WKH VRIWHQLQJ SRLQW YHUVXV ILUHG FRORU VWUHQJWK .6f DQG VWDELOLW\ 'HOWD (rf LQ FRDWLQJV EDWFKHG ZLWK b =U9

PAGE 340

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

PAGE 341

&+$37(5 6800$5< $1' &21&/86,216 7KH SUHVHQW UHVHDUFK LQYHVWLJDWHG WKH GHYHORSPHQW RI EOXH FRORU LQ JODVV IULW]LUFRQYDQDGLXP SLJPHQW V\VWHPV IRU FHUDPLF ZKLWHZDUHV FRDWLQJV 7KH SULPDU\ REMHFWLYHV ZHUH WR Df GHWHUPLQH LI FRVW HIIHFWLYH HQYLURQPHQWDOO\ VDIH JODVV IULWV FRXOG EH IRUPXODWHG WR LPSURYH WKH ILUHG FRORU VWUHQJWK DQG KLJKWHPSHUDWXUH VWDELOLW\ RI FHUDPLF FRDWLQJV DQG Ef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

PAGE 342

PRVW FRUURVLYH WR FHUDPLF FRORUDQWV ([SHULPHQWDO IULW FRPSRVLWLRQV ZHUH GHVLJQHG WR LQFOXGH ORZ FRVW QRQKD]DUGRXV R[LGHV DQG FRQIRUP ZLWK 6HJHUnV UXOHV 7KH UDQJH RI R[LGH FRQWHQWV WHVWHG HQFRPSDVVHG DQG H[FHHGHG WKH UDQJH QRUPDOO\ HPSOR\HG IRU JORVV\ FHUDPLF WLOH JOD]HV 6SHFLDO HPSKDVLV ZDV SODFHG RQ FRPSDULQJ FRPSRVLWLRQV ZLWK =U RSDFLILHGf YHUVXV QR =U XQRSDFLILHGf 6U2 YHUVXV =Q2 DQG ORZ YHUVXV KLJK DONDOLV 1Df 7KH 6U2 DQG =Q2 ZHUH RI VSHFLDO LQWHUHVW EHFDXVH RI WKHLU H[WHQVLYH XVH DV VHFRQGDU\ IOX[HV DQG WKHLU PDUNHG LQIOXHQFH RQ PRGLI\LQJ WKH 6L JODVV QHWZRUN 2WKHU R[LGHV LQ WKH IULW IRUPXODV ZHUH $ &D2 0J2 DQG % 7KH % FRQWHQWV ZHUH NHSW ORZ LQ RUGHU WR DYRLG OLTXLGOLTXLG SKDVH VHSDUDWLRQ $ ]LUFRQYDQDGLXP =U9f EOXH SLJPHQW ZDV WHVWHG EHFDXVH LW EHORQJV WR WKH PRVW FRPPRQ DQG VWDEOH JURXS RI FRORUDQWV XVHG LQ FHUDPLF JOD]HV WKH ]LUFRQ WULD[LDO SLJPHQWV $ FRDWLQJnV ILUHG FRORU VWUHQJWK ZDV TXDQWLILHG E\ LWV SLJPHQW DEVRUSWLRQ IDFWRU .6f FDOFXODWHG ZLWK WKH UHIOHFWDQFH YDOXH DW WKH =U9 SLJPHQW SHDN DEVRUSWLRQ ZDYHOHQJWK RI QP &RORU VWDELOLW\ ZDV H[SUHVVHG DV WKH LQYHUVH RI $(r 7KH $(r ZDV FDOFXODWHG DV WKH PDJQLWXGH RI WKH SRVLWLRQ YHFWRU LQ /r Dr Er FRORU VSDFH EHWZHHQ WKH FRORUV RI WZR FRDWLQJV ILUHG WR r& DQG r& SHDN

PAGE 343

WHPSHUDWXUH UHVSHFWLYHO\ 7KH /r Dr DQG Er YDOXHV FDOFXODWHG EDVHG RQ VSHFWUDO UHIOHFWDQFH FXUYHV DQG VWDQGDUG REVHUYHU WULVWLPXOXV ZHLJKWLQJ IXQFWLRQV FRUUHODWH WR OLJKWQHVV /rf UHGQHVVJUHHQQHVV Drf DQG \HOORZQHVV EOXHQHVV Er f FRORU SHUFHSWLRQ E\ WKH KXPDQ H\H (TXDWLRQV ZHUH GHYHORSHG IURP WKH UHVXOWV WKDW UHODWH SLJPHQW DEVRUSWLRQ RI QP OLJKW WR WKH FRORU YLVXDOL]HG 6LJQLILFDQW LPSURYHPHQWV LQ FRDWLQJ FRORU VWUHQJWK DQG VWDELOLW\ WKURXJK DGMXVWPHQWV LQ IULW R[LGH FRPSRVLWLRQV ZHUH GHPRQVWUDWHG 2YHU WKH UDQJH RI FRPSRVLWLRQV WHVWHG =U 6U2 =Q& $ DQG DONDOL FRQWHQWV KDG YDU\LQJ GHJUHHV RI LQIOXHQFH RYHU D FRDWLQJn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

PAGE 344

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f SUHFLSLWDWHG LQ WZR IULW FRPSRVLWLRQV ZKLFK FRQWDLQHG QR =U RU =Q2 7KLV UHVHDUFK DOVR HVWDEOLVKHG WKDW PHOW YLVFRVLW\ GDWD HVWLPDWHG ZLWK PRGHOV WKDW XWLOL]H KHDWLQJ PLFURVFRS\ DQG GLODWRPHWU\ PHDVXUHPHQWV FDQ EH DSSOLHG DV DQ LQGXVWULDO TXDOLW\ FRQWURO WRRO IRU SUHGLFWLQJ D IULWnV SRWHQWLDO IRU SURGXFLQJ VWURQJ DQG VWDEOH FRORU 7KH DUHD XQGHU D ORJ YLVFRVLW\ YHUVXV WHPSHUDWXUH FXUYH IURP 7V WR WKH SHDN

PAGE 345

WHPSHUDWXUH ZDV SURSRUWLRQDO WR FRORU VWUHQJWK DQG VWDELOLW\ LQ IULWV ZLWKRXW =U ,Q WKHVH IULWV FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ RFFXUUHG WKURXJKRXW WKH ILULQJ F\FOH DERYH 7 7KH VORSH RI WKH FXUYH QHDU 7V ZDV SURSRUWLRQDO WR FRORU VWUHQJWK DQG LQYHUVHO\ SURSRUWLRQDO ZLWK D VPDOO HIIHFWf WR FRORU VWDELOLW\ LQ IULWV ZLWK =U ZKHUH FU\VWDOOL]DWLRQ ZDV FRPSOHWH EHORZ r& 6LJQLILFDQW UHVXOWV RI WKLV UHVHDUFK DUH VXPPDUL]HG LQ WKH IROORZLQJ VHFWLRQV =U9 3LJPHQW DQG &RORU 9DOXHV f 0D[LPXP DEVRUSWLRQ RI YLVLEOH OLJKW E\ WKH ]LUFRQ YDQDGLXP SLJPHQW =U9f RFFXUV DW QP ZDYHOHQJWK H9f LQ WKH UHG\HOORZ UHJLRQ RI WKH VSHFWUXP 7KLV LV WKH GLIIHUHQFH LQ HQHUJ\ EHWZHHQ VSOLW GRUELWDOV LQ 9 RQ WKH ]LUFRQ ODWWLFH f 0D[LPXP UHIOHFWDQFH IURP =U9 RFFXUV ZLWK QP H9f EOXHJUHHQ OLJKW f 7KHUH LV D OLQHDU UHODWLRQVKLS EHWZHHQ WKH SLJPHQW DEVRUSWLRQ IDFWRU .6f FDOFXODWHG DW QP DQG =U9 FRQWHQW +RZHYHU WKH VORSH RI WKH OLQH YDULHV VLJQLILFDQWO\ ZLWK IULW R[LGH FRPSRVLWLRQ DQG ILULQJ WHPSHUDWXUH

PAGE 346

f :LWK LQFUHDVLQJ =U9 FRQWHQW WKH SHUFHLYHG FRORU RI D FHUDPLF FRDWLQJ SURJUHVVHV GDUNHU ORZHU /r f DQG EOXHU ORZHU Erf +RZHYHU WKH WUHQG RI LQFUHDVLQJ JUHHQ ORZHU Drf ZLWK SLJPHQW FRQWHQW IURP b WR b =U9 UHYHUVHV IURP b WR b ZKHUH WKH JUHHQLVK KXH DSSHDUV WR VKLIW GLUHFWLRQ DQG SURJUHVV WRZDUGV UHG KLJKHU Drf $ VKLIW LQ PD[LPXP UHIOHFWDQFH WR ZDYHOHQJWKV EHORZ WKH [ VWDQGDUG ZHLJKWLQJ IXQFWLRQ PLQLPD FRUUHODWHV WR WKH DEUXSW QRQOLQHDU VKLIW LQ KXPDQ SHUFHSWLRQ f 9DULDWLRQV LQ FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ KDYH D OHVVHU HIIHFW RQ WKH SHUFHLYHG FRORU ZKHQ KLJK SLJPHQW ORDGLQJV Ab DUH XWLOL]HG f 3LJPHQW FRQWHQW LQFUHDVHV IURP WR b KDYH QR VLJQLILFDQW LQIOXHQFH RQ ILUHG JORVV f $W D FRQVWDQW SLJPHQW ORDGLQJ PRVW RI WKH FRORU YDULDWLRQ GXH WR FKDQJHV LQ IULW FRPSRVLWLRQ RU ILULQJ WHPSHUDWXUH RFFXUV LQ /r DQG Er f )RU =U9 SLJPHQWV D VHFRQG RUGHU SRO\QRPLDO IXQFWLRQ ZHOO GHVFULEHV 5 f WKH SRVLWLYH FRUUHODWLRQ EHWZHHQ /r DQG Er f 9DOXHV IRU SHUFHLYHG FRORU /r Dr DQG Er DUH DOO ORJDULWKPLF IXQFWLRQV RI .6 DW QP ZKHUH D KLJKHU .6 UHVXOWV LQ ORZHU FRORU YDOXHV KLJKHU FKURPDf

PAGE 347

)ULW 2[LGH &RPSRVLWLRQ f 7KH VWURQJHVW EOXH FRORU RU KLJKHVW .6 LV SURGXFHG ZLWK IULWV FRQWDLQLQJ KLJK =U DQG 6U2 DQG QR =Q2 7KH ZHDNHVW FRORU UHVXOWV LQ IULWV ZLWK QR =U DQG ORZ $ DONDOLV f ,Q IULWV ZLWK =U FRORU VWUHQJWK .6f LV LQFUHDVHG WR WKH JUHDWHVW GHJUHH ZKHQ 6U2 UHSODFHV =Q2 7KLV LQKLELWV SUHFLSLWDWLRQ RI ]LUFRQ IURP WKH IULW ZKLFK XOWLPDWHO\ OLJKWHQV FRORU $ UHODWLYHO\ VWDEOH WUDQVSDUHQW IULW UHVXOWV IURP JODVV FRPSRVLWLRQV ZLWK =U DQG 6U2 EXW QR =Q2 7KH VWDWLVWLFDO PRGHOV SUHGLFW D OLQHDU UHODWLRQVKLS EHWZHHQ 6U2 DQG .6 ZKHQ =Q2 LV DOVR SUHVHQW f ,Q IULWV ZLWKRXW =U 6U2 YHUVXV =Q2 FDXVHV QR VLJQLILFDQW GLIIHUHQFH LQ .6 f ,I 6U2 LV SUHVHQW LQ WKH IULW ZLWKRXW =Q2 DQ LQFUHDVH LQ =U PRODU HTXLYDOHQWV LQ WKH IULW OLQHDUO\ UDLVHV .6 E\ UHGXFLQJ WKH VROXELOLW\ RI WKH ]LUFRQ SLJPHQW ,Q IULWV ZLWK =Q2 ZKHUH ]LUFRQ SUHFLSLWDWLRQ LV IDYRUDEOH KLJKHU =U FRQWHQW OLQHDUO\ ORZHUV .6 f 5DLVLQJ WKH $ODONDOLV PRODU HTXLYDOHQWV VOLJKWO\ ORZHUV FRORU VWUHQJWK LQ IULWV ZLWK =U ZKLFK WHQG WR SUHFLSLWDWH ]LUFRQ 7KLV RFFXUV EHFDXVH $O UHSODFHV VRPH =U LQ WKH JODVV QHWZRUN ZKLFK UHOHDVHV PRUH =U WR IRUP

PAGE 348

]LUFRQ DQG OLJKWHQ WKH FRORU :KHQ QR =U LV SUHVHQW KLJKHU $ODONDOLV UDWLRV VLJQLILFDQWO\ LQFUHDVH .6 DQG FRORU VWDELOLW\ DW KLJK WHPSHUDWXUH E\ LQKLELWLQJ FU\VWDOOL]DWLRQ DQG =U9 GLVVROXWLRQ GXH WR DQ LQFUHDVHG PHOW YLVFRVLW\ f 7KH HIIHFWV RI =U DQG WHPSHUDWXUH RQ FRORU VWUHQJWK DUH SURJUHVVLYHO\ UHGXFHG ZLWK LQFUHDVLQJ =U9 FRQWHQW 7KH VDPH TXDQWLW\ RI SLJPHQW GLVVROXWLRQ LV OHVV QRWLFHDEOH LQ VDWXUDWHG FRORUV f 7KH OHDVW QRWLFHDEOH FRORU FKDQJHV LQ ILUHG FRDWLQJV GXH WR YDULDWLRQV LQ SHDN ILULQJ WHPSHUDWXUH DUH DFKLHYHG ZLWK KLJK =U DQG =Q2 DQG QR 6U2 $(r ZDV A f 7KH ORZHVW FRORU VWDELOLW\ UHVXOWV IURP IULWV ZLWKRXW =U DQG ORZ LQ $ODONDOLV $(r UDQJHG IURP f f &RORU VWDELOLW\ LV LQFUHDVHG WR WKH JUHDWHVW GHJUHH E\ UDLVLQJ WKH =U FRQWHQW ZKLFK LQKLELWV ERWK FU\VWDOOL]DWLRQ QHDU WKH SHDN ILULQJ WHPSHUDWXUH DQG ]LUFRQ SLJPHQW GLVVROXWLRQ 7KHUH LV DQ LQYHUVH OLQHDU UHODWLRQVKLS EHWZHHQ =U PRODU HTXLYDOHQWV DQG $(r f 5DLVLQJ WKH 6U2 PRODU HTXLYDOHQWV LQ IULWV ZLWK RU ZLWKRXW =U OLQHDUO\ UHGXFHV KLJK WHPSHUDWXUH FRORU VWDELOLW\ GXH WR DQ LQFUHDVH LQ ]LUFRQ SLJPHQW VROXELOLW\ f ,Q IULWV ZLWK =U FRORU VWDELOLW\ ORZHUV YHU\ VOLJKWO\ ZLWK KLJKHU $ODONDOLV :KHQ QR =U LV SUHVHQW

PAGE 349

VLJQLILFDQW LPSURYHPHQWV LQ FRORU VWDELOLW\ DUH DFFRPSOLVKHG E\ UDLVLQJ WKH $ODONDOLV ZKLFK LQFUHDVHV YLVFRVLW\ DQG WKXV LQKLELWV FU\VWDOOL]DWLRQ DQG =U9 GLVVROXWLRQ 9LVFRVLW\ &U\VWDOOL]DWLRQ DQG =U9 'LVVROXWLRQ f 'DWD REWDLQHG ZLWK GLODWRPHWU\ DQG KHDWLQJ PLFURVFRS\ DQDO\VHV RI IULWV FDQ EH FRUUHODWHG WR PHOW YLVFRVLW\ FU\VWDOOL]DWLRQ DQG FRORU GHYHORSPHQW ZLWK ]LUFRQ SLJPHQWV f 7KH PHOW YLVFRVLW\ RI IULWV FDQ EH GHVFULEHG ZLWK D PRGHO RI WKH IRUP RI WKH 9)7 HTXDWLRQ ZKHUH WKH $ FRQVWDQW FDQ EH WDNHQ DV WKH WHPSHUDWXUH LQGHSHQGHQW R[LGH FRPSRVLWLRQ HIIHFW f $Q LQFUHDVH LQ WKH PRODU HTXLYDOHQWV UDWLR RI $OR DONDOLV KDV WKH ODUJHVW LQIOXHQFH RI WKH IULW R[LGHV RQ UDLVLQJ WKH JODVV VRIWHQLQJ SRLQW DQG PHOW YLVFRVLW\ f 7R D OHVVHU GHJUHH KLJKHU =U DQG ORZHU 6U2 UHVXOWV LQ LQFUHDVHG YLVFRVLW\ f 5HSODFLQJ =Q2 ZLWK 6U2 KDV WKH ODUJHVW HIIHFW RQ LQFUHDVLQJ 7J DQG WKH UDWLR 7J7V f ,Q IULWV ZLWK =U DQG =Q2 D JUHDWHU QXPEHU RI ILQH SPf VSKHULFDO SDUWLFOHV RI ]LUFRQ SUHFLSLWDWH ZKHQ WKH YLVFRVLW\ LV LQFUHDVHG E\ UDLVLQJ WKH $ODONDOLV UDWLR

PAGE 350

$ PRUH HORQJDWHG ILEURXV PRUSKRORJ\ SUHGRPLQDWHV IURP ORZHU YLVFRVLW\ FRQGLWLRQV f 3UHFLSLWDWLRQ RI ]LUFRQ IURP =U LQ WKH IULW LV LQKLELWHG ZKHQ WKH VORSH RI WKH ORJ 7f YHUVXV WHPSHUDWXUH FXUYH QHDU WKH VRIWHQLQJ SRLQW LV LQFUHDVHG E\ UHSODFLQJ =Q2 ZLWK 6U2 f ,Q IULWV ZLWKRXW =U WKH VL]H DQG YROXPH RI SUHFLSLWDWH LQFUHDVHV ZLWK ORZHULQJ YLVFRVLW\ UHVXOWLQJ LQ ZHDNHU FRORU DQG ORZHU JORVV f ,Q IULWV ZLWKRXW =U WKHUH LV QR FRUUHODWLRQ EHWZHHQ FU\VWDOOL]DWLRQ DQG WKH VORSH RI WKH ORJ Uf FXUYH QHDU WKH VRIWHQLQJ SRLQW f ,Q IULWV ZLWKRXW =U D ORZHU YLVFRVLW\ FDXVHV PRUH SLJPHQW WR GLVVROYH UHVXOWLQJ LQ D ZHDNHU FRORU 9LVFRVLW\ LV QRW D GRPLQDQW IDFWRU LQ =U9 GLVVROXWLRQ LQ IULWV FRQWDLQLQJ =U f $ KLJKHU LQWHJUDWHG DUHD XQGHU WKH ORJ U_ YHUVXV WHPSHUDWXUH FXUYH IURP 7V WR WKH SHDN ILULQJ WHPSHUDWXUH FRUUHODWHV ZHOO ZLWK LQFUHDVHG FRORU VWUHQJWK .6f DQG VWDELOLW\ ORZHU $(rf LQ IULWV ZLWKRXW =U $ JUHDWHU RYHUDOO YLVFRVLW\ LQKLELWV FU\VWDOOL]DWLRQ RI QRQ]LUFRQLXP VSHFLHV LQ WKHVH IULWV ,Q IULWV FRQWDLQLQJ =U2 ZKHUH

PAGE 351

]LUFRQ SUHFLSLWDWLRQ LV FRPSOHWH E\ r& QR VXFK FRUUHODWLRQ H[LVWV f $ KLJKHU VORSH LQ WKH ORJ U_ YHUVXV WHPSHUDWXUH FXUYH QHDU WKH VRIWHQLQJ SRLQW FRUUHODWHV ZHOO ZLWK LQFUHDVHG FRORU VWUHQJWK DQG D VOLJKW GHFUHDVH LQ VWDELOLW\ $(r Af LQ IULWV ZLWK =U EHFDXVH ]LUFRQ QXFOHDWLRQ DQG JURZWK LV LQKLELWHG 1R VXFK FRUUHODWLRQ H[LVWV IRU IULWV ZLWKRXW =U ZKHUH FU\VWDOOL]DWLRQ RFFXUV WKURXJK WKH UDQJH RI ILULQJ WHPSHUDWXUHV f ,Q IULWV ZLWK =U ]LUFRQ LV WKH RQO\ HQHUJHWLFDOO\ IDYRUDEOH SUHFLSLWDWH SUHVHQW DW QRUPDO SHDN WHPSHUDWXUHV RI r& DQG r& f =LUFRQ FU\VWDOV QHDU SP LQ VL]H LQ WKH JODVV ZLWK =U9 DOORZ IRU EOXH FRORU ZLWK D YHU\ KLJK OLJKWQHVV /rf GXH WR HQKDQFHG VFDWWHULQJ RI EOXH OLJKW f 9DOXHV IRU WKH FRORU SHUFHLYHG /r Dr DQG Er DUH DOO ORJDULWKPLF IXQFWLRQV RI WKH ZHLJKW SHUFHQW ]LUFRQ SUHFLSLWDWHG LQ WKH JODVV +LJKHU ]LUFRQ FDXVHV PRUH OLJKWHQLQJ KLJKHU /rf WKDQ ORVVHV RI JUHHQ KLJKHU Drf DQG EOXH KLJKHU Erf f 7KH SLJPHQW DEVRUSWLRQ IDFWRU .6f LQFUHDVHV H[SRQHQWLDOO\ ZLWK D GHFUHDVH LQ WKH DPRXQW RI ]LUFRQ SUHFLSLWDWH

PAGE 352

f 3HUFHSWLEOH FRORU DOWHUDWLRQV GXH WR YDULDWLRQV LQ ILULQJ WHPSHUDWXUHV DUH UHGXFHG LQ V\VWHPV WKDW FU\VWDOOL]H ]LUFRQ IURP =U LQ WKH IULW &U\VWDOOL]DWLRQ LV FRPSOHWH EHIRUH r& DQG WKH DPRXQW RI ]LUFRQ OHIW LQ FRDWLQJV ILUHG WR r& r& DQG r& UHPDLQV QHDUO\ WKH VDPH f ,Q IULWV ZKHUH ]LUFRQ GRHV QRW FU\VWDOOL]H =U9 SLJPHQW GLVVROXWLRQ WHQGV WR RFFXU ZLWK WKH JUHDWHVW ORVV UHVXOWLQJ LQ IULWV ZLWKRXW =U DQG D ORZ $ODONDOLV UDWLR ,Q IULWV ZLWK =U ]LUFRQ FU\VWDOOL]DWLRQ GRHV QRW RFFXU ZKHQ 6U2 UHSODFHV =Q2 f /r Dr DQG Er LQFUHDVH DQG .6 GHFUHDVHV OLQHDUO\ ZLWK WKH ZHLJKW SHUFHQWDJH RI =U9 SLJPHQW GLVVROYHG IURP WR bf LQ WKH JODVV f ,Q WKH IULWV WHVWHG ZLWKRXW =U WKH PDLQ SKDVHV WKDW FU\VWDOOL]HG ZHUH GLRSVLGH DQG KDUG\VWRQLWH ZKHQ =Q2 ZDV SUHVHQW DQG 6U&D6L ZKHQ 6U2 UHSODFHG =Q2 7KH 6U&DA6L LV D QHZ OLQH SKDVH XQLGHQWLILHG RQ WKH VWDQGDUG [UD\ -&3'6 FDUG ILOH f 7KH TXDQWLW\ RI KDUG\VWRQLWH RU 6U&D6L2J SUHFLSLWDWH LQFUHDVHV ZLWK WHPSHUDWXUH IURP r& WR r& UHVXOWLQJ LQ SRRU FRORU VWUHQJWK DQG VWDELOLW\ &U\VWDOOL]DWLRQ RI WKH WZR VSHFLHV FDXVHV D VKLIW LQ WKH PD[LPXP UHIOHFWDQFH IURP QP DW r& WR QHDU QP DW r& 7KLV UHVXOWV LQ

PAGE 353

KLJKHU /r DQG Er YDOXHV DQG JUHDWO\ SHUFHSWLEOH FKDQJHV LQ FRORU KLJK $(rf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

PAGE 354

&+$37(5 )8785( :25. 7KH SUHVHQW GLVVHUWDWLRQ XQFRYHUHG VHYHUDO SRWHQWLDO UHVHDUFK WRSLFV ZKLFK DUH RI LQWHUHVW IRU ERWK EDVLF VFLHQFH DQG LQGXVWULDO SXUSRVHV 3RVVLEOH IXWXUH LQYHVWLJDWLRQV LQFOXGH f ([SDQG WKH PDWKHPDWLFDO PRGHOV HVWDEOLVKHG LQ WKLV VWXG\ IRU SUHGLFWLQJ .6 $(r /r Dr Er DQG ORJ U? E\ DGGLQJ DV LQGHSHQGHQW YDULDEOHV Df WKH RWKHU WZR ]LUFRQ WULD[LDO SLJPHQWV ]LUFRQ SUDVHRG\PLXP \HOORZ DQG ]LUFRQLURQ FRUDO Ef FRPSOHWH WLPH YHUVXV WHPSHUDWXUH ILULQJ SURILOHV UDWKHU WKDQ WKH SHDN WHPSHUDWXUH RI D VSHFLILF F\FOH DQG Ff D ZLGHU UDQJH RI JODVV R[LGH FRPSRVLWLRQV f 'HULYH VWDWLVWLFDO PRGHOV WR GHVFULEH WKH LQIOXHQFHV RI IULW R[LGH FRPSRVLWLRQ DQG FU\VWDOOL]DWLRQ RQ JORVV f 'HYHORS PDWWH ZKLWHZDUHV FRDWLQJV JORVV f DW r LQFLGHQFHf ZLWK LPSURYHG FRORU VWUHQJWK DQG VWDELOLW\ ,QYHVWLJDWH ZKHWKHU EDWFKHV FRQWDLQLQJ IULWWHG =U DQG KLJK

PAGE 355

PLQHUDO DGGLWLRQV RI UHODWLYHO\ FRDUVH $ DUH RSWLPXP %DVHG RQ WKH ILQGLQJV RI WKLV UHVHDUFK IULWWHG =U ZRXOG WHQG WR LQKLELW FU\VWDOOL]DWLRQ DQG SLJPHQW GLVVROXWLRQ WKDW FDXVH FRORU YDULDWLRQ ZKLOH $ PLQHUDOV PD\ SURYLGH IRU DQ XQGHUILUHG PDWWH FU\VWDOOLQH WH[WXUH f &RQWURO WKH TXDQWLW\ VL]H DQG PRUSKRORJ\ RI ]LUFRQ SUHFLSLWDWLQJ IURP =U LQ WKH JODVV IULW E\ YDU\LQJ UDWLRV RI 6U2=Q2 DQG $DONDOLV ([SORUH WKH UDQJH RI FRORUV DQG RSDFLILFDWLRQ WKDW FDQ EH DFKLHYHG ZLWK ]LUFRQ SLJPHQWV WKURXJK VHOHFWLYH VFDWWHULQJ RI OLJKW E\ WKH ]LUFRQ SUHFLSLWDWHV f 6WXG\ WKH LQIOXHQFH RI ]LUFRQ SLJPHQW SDUWLFOH VL]H RQ WKH FRORU RI D ILUHG FRDWLQJ ,QYHVWLJDWH LI QP =U9 EOXH QP =U3U \HOORZ DQG QP =U)H FRUDO SLJPHQW VL]HV SURYLGH RSWLPXP FRORU VWUHQJWK DQG RSDFLW\ E\ HQKDQFLQJ WKH UHIOHFWLRQ RI WKH GHVLUHG FRORU RI OLJKW 'HYHORS VXUIDFWDQWV WKDW DOORZ IRU HDV\ GLVSHUVLRQ RI VXFK ILQH SDUWLFOHV LQ DQ LQGXVWULDO FRDWLQJ EDVH 'HYLVH PHWKRGV IRU DYRLGLQJ GLVVROXWLRQ RI WKH VPDOO ]LUFRQ SDUWLFOHV f &RPSDUH WKH HIIHFWV RI EDWFK DGGLWLRQV RI ]LUFRQ PLQHUDOV YHUVXV IULWWHG =U RQ WKH PHOW YLVFRVLW\ FU\VWDOOL]DWLRQ DQG FRORU GHYHORSPHQW RI FHUDPLF FRDWLQJV 'HWHUPLQH ZKDW PD[LPXP ]LUFRQ PLQHUDO SDUWLFOH VL]H SUHVHUYHV

PAGE 356

WKH FRORU E\ GLVVROYLQJ LQWR WKH PHOW EHIRUH WKH ]LUFRQ SLJPHQW LV DWWDFNHG f 'HWHUPLQH ZKHWKHU RWKHU LQWHUPHGLDWH RU JODVV IRUPHU R[LGHV WHQG WR VWLPXODWH RU LQKLELW ]LUFRQ SUHFLSLWDWLRQ IURP =U LQ WKH IULW ,Q WKH SUHVHQW VWXG\ =Q2 HQKDQFHG ]LUFRQ FU\VWDOOL]DWLRQ DQG 6U2 SUHYHQWHG LW )RU H[DPSOH LQYHVWLJDWH %D2 ZKLFK KDV D GHQVLW\ JFPf FDWLRQ UDGLXV ƒf DQG FDWLRQ FKDUJH f WKDW DUH FRPSDUDEOH WR 6U2 JFP ƒ FKDUJHf f 3HUIRUP TXDQWLWDWLYH WKHUPRG\QDPLF DQG NLQHWLF VWXGLHV RI FU\VWDOOL]DWLRQ DQG ]LUFRQ SLJPHQW GLVVROXWLRQ LQ IULWV ZLWKRXW =U 'HWHUPLQH WKH $* RI IRUPDWLRQ RI HDFK FU\VWDOOLQH VSHFLHV f 6\QWKHVL]H WKH QHZ 6U&D6L FU\VWDOOLQH SKDVH DQG WKRURXJKO\ VWXG\ LWV VWUXFWXUH DQG SURSHUWLHV $GG WKH QHZ LQIRUPDWLRQ WR WKH H[LVWLQJ [UD\ -&3'6 FDUG ILOH f &RPSDUH PLFURZDYH YHUVXV FRQYHQWLRQDO KHDWLQJ PHWKRGV IRU ILULQJ FRORUHG FHUDPLF FRDWLQJV (YDOXDWH LI PHOWLQJ EHKDYLRU FU\VWDOOL]DWLRQ ]LUFRQ SLJPHQW GLVVROXWLRQ DQG WKH FRUUHVSRQGLQJ FRORU VWDELOLW\ DQG VWUHQJWK DUH DOWHUHG GXH WR SRVVLEOH FKDQJHV LQ GLIIXVLRQ WUDQVSRUW SURFHVVHV f 6\QWKHVL]H ]LUFRQ SLJPHQWV XVLQJ DOWHUQDWLYH PHWKRGV VXFK DV PLFURZDYH SURFHVVLQJ 'HWHUPLQH LI D JUHDWHU QXPEHU

PAGE 357

RI GRSDQW FRORU LRQV FDQ EH LQFRUSRUDWHG RQ WKH ]LUFRQ ODWWLFH WKDQ FXUUHQWO\ DFKLHYHG ZLWK FRQYHQWLRQDO WHFKQLTXHV

PAGE 358

$33(1',; $ 81,76 )25 '(6&5,%,1* /,*+7 $1' &2/25 > 1DV@ $OO 8QLWV 5HODWHG WR (QHUJ\ &RORU :DYHOHQJWK ;f LQ 1DQRPHWHUV QPf )UHTXHQF\ +]f [O2f :DYH 1XPEHU Y FPf r (QHUJ\ LQ H9 r (QHUJ\ LQ FDO PRO [O2f ,QIUDUHG IDU ,QIUDUHG QHDU 5HG 2UDQJH
PAGE 359

:DYH 1XPEHU FP f X ; )UHTXHQF\ +]f K [ F $Of $f ZKHUH F WKH VSHHG RI OLJKW [ FPVHF r(QHUJ\ H9f ( K [ [ A(QHUJ\ FDO PROf ( 8 [ r :DYHOHQJWK QPf [ (QHUJ\ H9f $f $f $f $O 3KRWRPHWULF 8QLWV &DQGHOD FGf 6, XQLW IRU OXPLQRXV LQWHQVLW\ 2QH VTXDUH PHWHU RI EODFNERG\ DW HPLWV FG 2QH FDQGHOD SURGXFHV RQH OXPHQ RI IOX[ SHU VWHUDGLDQ RI VROLG DQJOH PHDVXUHG IURP WKH VRXUFH /XPHQ OPf XQLW RI OXPLQRXV IOX[ RU OLJKW LQWHQVLW\ HPLWWHG E\ D SRLQW VRXUFH RI FG WKURXJK D XQLW VROLG DQJOH VWHUDGLDQf /XPLQDQFH /f XQLW RI EULJKWQHVV LQ FG Paf 7KH VXQ KDV / [ FG P DQG D IOXRUHVFHQW ODPS DSSUR[LPDWHO\ FG P ,OOXPLQDQFH LQ OX[ O[f 2QH OPP 2QH OX[ IRRW FDQGOHV r 8QLWV DSSO\ WR D VLQJOH HOHFWURQ

PAGE 360

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

PAGE 361

9LEUDWLRQV DQG 6LPSOH ([FLWDWLRQV ,QFDQGHVFHQFH )ODPHV ODPSV FDUERQ DUF OLPHOLJKW *DV ([FLWDWLRQV 9DSRU ODPSV OLJKWQLQJ DXURUDV VRPH ODVHUV 9LEUDWLRQV DQG 5RWDWLRQV :DWHU LFH LRGLQH EOXH JDV IODPH *HRPHWULFDO DQG 3K\VLFDO 2SWLFV 'LVSHUVLYH 5HIUDFWLRQ 3RODUL]DWLRQ HWF 5DLQERZ KDORV VXQ GRJV JUHHQ IODVK RI VXQ ILUH LQ JHPVWRQHV 6FDWWHULQJ %OXH VN\ UHG VXQVHW EOXH PRRQ PRRQVWRQH 5DPDQ VFDWWHULQJ EOXH H\HV DQG VRPH RWKHU ELRORJLFDO FRORUV ,QWHUIHUHQFH 2LO VOLFN RQ ZDWHU VRDS EXEEOHV FRDWLQJ RQ FDPHUD OHQVHV VRPH ELRORJLFDO FRORUV 'LIIUDFWLRQ $XUHROH JORU\ GLIIUDFWLRQ JUDWLQJV RSDO VRPH ELRORJLFDO FRORUV PRVW OLTXLG FU\VWDOV

PAGE 362

$33(1',; & '(16,7< 3$57,&/( 6,=( $1' $33/,&$7,21 :(,*+7 '$7$ &OO 'HQVLW\ DQG 3DUWLFOH 6L]H %DVHG RQ 9ROXPH b )5,7 $%&'()*+ $ =U9 3LJPHQW 'HQVLW\ f§ JFPf 0HDQ 3DUWLFOH 'LDPHWHU SPf &O &RDWLQDV $SSOLFDWLRQ :HLJKWV JUDPVf DW 6S*U )5,7 $ % & ( ) + b =U9 b =U9 b =U9 b =U9

PAGE 363

$33(1',; '$7$ )520 &2$7,1*6 %$7&+(' :,7+ )5,7 b %(1721,7( $1' =U9 3,*0(17 $1' ),5(' 86,1* $ 0,187( &(5$0,& 7,/( &<&/(

PAGE 364

'$7$ )520 &2$7,1*6 %$7&+(' :,7+ )5,7 b %(1721,7( $1' =U9 3,*0(17 $1' ),5(' 86,1* $ 0,187( &(5$0,& 7,/( &<&/( 4 %4 r )ULW &RPSRVLWLRQ :HLJKW bf 1D4 .4 eD4 $,4 =U4 0J4 6U2 =Q2 rr :Wb =U9 3LJPHQW 3HDN )LULQJ 7HPS &O / &,( D E 'HOWD (r & 'HJUHH $QJOH *ORVV .6 DW QP rrr *OD]H (OO f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. D f§ 2. 2. 2. 2. 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. 2. 2. 2. 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f f§ 2. f f§ 2. f§ 2. 2. 2. 2. 2.

PAGE 365

64 % r )ULW &RPSRVLWLRQ :HLJKW bf 1D .4 &D4 $, =I4 0J2 6I2 =Q4 rr :/b =U9 3LJPHQW 3HDN )LULQJ 7HPS &f / &,( D E 'HOWD (r & 'HJUHH $QJOH *ORVD .6DW QP *OD]H (f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG &UD]HG &UD]HG &UD]HG &UD]HG f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. 2. 2. 2. 2. f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG &UD]HG &UD]HG &UD]HG &UD]HG X\ A

PAGE 366

r )ULW &RPSRVLWLRQ :HLJKW bf rr :/b =U9 3HDN )LULQJ &,( 'HOWD (r 'HJUHH .6 DW rrf *OD]H 64 %4 1V4 .4 &D2 $4 =U 0J4 6U2 =Q2 3LJPHQW 7HPS &O / } E T H[ R k k $QJOH *ORVV 4QP )LW f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. f§ 2. 2. 2. 2. 2. f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG f§ &UD]HG &UD]HG &UD]HG &UD]HG &UD]HG r &RDWLQJV ZHUH DSSOLHG DW FP YROXPH SHU ; WLOH DYHUDJH ZHW DSSOLFDWLRQ ZHLJKW RI JUDPV DW D VSHFLILF JUDYLW\f rr 3LJPHQW ZDV &HUHGHF $ WXUTXRLVHEOXH GLVSHUVLEOH VWDLQ rrr 7KHUPDO H[SDQVLRQ RI WKH ZDOO WLOH ERG\HQJREH VXEVWUDWH ZDV ; H&

PAGE 367

$33(1',; ( )5,7 63(&75$/ 5()/(&7$1&( '$7$ $1' &859(6 $7 ($&+ 7(03(5$785( $1' 3,*0(17 /2$',1*

PAGE 368

7DEOH (O (QJREH DQG )ULWV $ DQG % 5HIOHFWDQFH 'DWD SHUFHQWDJHV DUH ZWb =U9 SLJPHQW DGGHG WR IULWf :DYHOHQJWK & 3($. 7(03(5$785( LPQ(QDREH $ $ b $ A b $ b % A b % b % b & 3($. 7(03(5$785( (QDREH $ $ b $ b $ b % % b % b % b  6 & 3($. 7(03(5$785( (QJREH $ $ b $ b $ b % b % b % b _ n

PAGE 369

7DEOH ( )ULWV & DQG 5HIOHFWDQFH 'DWD SHUFHQWDJHV DUH ZWb =U9 SLJPHQW DGGHG WR IULWf :DYHOHQJWK & 3($. 7(03(5$785( DP@ F & A b & b & A b ' b b b & 3($. 7(03(5$785( 4 & b & b & b 4 b b A b L ,& 3($. 7(03(5$785( & & b & b & b ' b b A b , c _

PAGE 370

7DEOH ( )ULWV ( DQG ) 5HIOHFWDQFH 'DWD SHUFHQWDJHV DUH ZWb =U9 SLJPHQW DGGHG WR IULWf :DYHOHQJWK L & 3($. 7(03(5$785( QPO ( ( b ( b ( b ( ) b ) b ) A b r & 3($. 7(03(5$785( ( ( A b ( b ( A b ( ) b ) b ) A b & 3($. 7(03(5$785( ( ( b ( b ( b ( ) A b ) A b ) b

PAGE 371

7DEOH ( )ULWV DQG + 5HIOHFWDQFH 'DWD SHUFHQWDJHV DUH ZWb =U9 SLJPHQW DGGHG WR IULWf :DYHOHQJWK & 3($. 7(03(5$785( ,QPO 6 b A b b + + b + A b + r b & 3($. 7(03(5$785( H b b A b 8 + b + b + A b & 3($. 7(03(5$785( 6 b b b + + b + b + b ,

PAGE 372

)5,7 $ & 3($. 7(03(5$785( :DYHOHQJWK QPf f§rb =U9 f§ff§b =U9 rb =U9 mb =U9M )5,7 $ & 3($. 7(03(5$785( :DYHOHQJWK QPf b =U9 b =U9 b =U9 b =U9, )5,7 $ & 3($. 7(03(5$785( :DYHOHQJWK QPf f§b =U9 rb =U9 f§b =U9 b =U9_ m O )LJXUH (O )ULW $ VSHFWUDO UHIOHFWDQFH FXUYHV

PAGE 373

)5,7 % & 3($. 7(03(5$785( :DYHOHQJWK QPf f§b =U9 rb =U9 b =U9 b =U9 )5,7 % & 3($. 7(03(5$785( :DYHOHQJWK QPf f§b =U9 rb =U9 Hb =U9 rb =U9 )5,7 % & 3($. 7(03(5$785( :DYHOHQJWK QPf rb =U9 P b =U9 Db =U9 rb =U9 )LJXUH ( )ULW % VSHFWUDO UHIOHFWDQFH FXUYHV

PAGE 374

)5,7 & & 3($. 7(03(5$785( :DYHOHQJWK QPf !b =U9 b =U9 b =U9 b =U9 L )5,7 & & 3($. 7(03(5$785( :DYHOHQJWK QPf c b =U9 b =U9 b =I9 !} b =U9 )5,7 & & 3($. 7(03(5$785( :DYHOHQJWK QPf b =U9 b =U9 b =U9 b =U9 )LJXUH ( )ULW & VSHFWUDO UHIOHFWDQFH FXUYHV

PAGE 375

)5-7 & 3($. 7(03(5$785( )5,7 & 3($. 7(03(5$785( )5,7 & 3($. 7(03(5$785( :DYHOHQJWK QPf rb =U9 f§b =U9 f§b =U9 mb =U9, )LJXUH ( )ULW VSHFWUDO UHIOHFWDQFH FXUYHV

PAGE 376

)5,7 ( & 3($. 7(03(5$785( :DYHOHQJWK QPf f§b =U9 f§b =U9 rb =U9 rb =U9 )5,7 ( & 3($. 7(03(5$785( :DYHOHQJWK QPf Hb =U9 mb =U9 rb =U9 rb =U9_ )5,7 ( & 3($. 7(03(5$785( :DYHOHQJWK QPf rb =U9 rb =U9 rb =U9 [b =U9 )LJXUH ( )ULW ( VSHFWUDO UHIOHFWDQFH FXUYHV

PAGE 377

)5,7 ) & 3($. 7(03(5$785( :DYHOHQJWK QPf b =U9 rb =I9 b =U9 [ b =I9 )5,7 ) & 3($. 7(03(5$785( :DYHOHQJWK QPf rb =U9 b =U9 b =U9 N b =U9 )5,7 ) & 3($. 7(03(5$785( )LJXUH ( )ULW ) VSHFWUDO UHIOHFWDQFH FXUYHV

PAGE 378

)5,7 & 3($. 7(03(5$785( :DYHOHQJWK QPf b =U9 b =U9 b =U9 b =U9 c )5,7 & 3($. 7(03(5$785( :DYHOHQJWK QPf M f§b =U9 mb =U9 rf§b =U9 Zb =U9 )5,7 & 3($. 7(03(5$785( :DYHOHQJWK QPf L }f§b =U9 f§ff§b =U9 rb =U9 rmb =U9 )LJXUH ( )ULW VSHFWUDO UHIOHFWDQFH FXUYHV

PAGE 379

)5,7+ & 3($. 7(03(5$785( :DYHOHQJWK QPf M b =U9 r b =U9 H b =U9 [ b =U9 c )5,7+ & 3($. 7(03(5$785( :DYHOHQJWK QPf f§b =U9 rb =U9 rb =U9 b =U9 )5,7+ & 3($. 7(03(5$785( :DYHOHQJWK QPf b =U9 b =U9 b =U9 b =U9 )LJXUH ( )ULW + VSHFWUDO UHIOHFWDQFH FXUYHV

PAGE 380

5()(5(1&(6 $OH / $OH[DQGHU DQG + .OXJ ;UDY 'LIIUDFWLRQ 3URFHGXUHV QG (G -RKQ :LOH\ DQG 6RQV 1HZ
PAGE 381

%ORE 5 %ORQVNL &HUDPLF (QJLQHHULQJ DQG 6FLHQFH 3URFHHGLQJV f f %RD 0 %RDV 0DWKHPDWLFDO 0HWKRGV LQ WKH 3K\VLFDO 6FLHQFHV QG (G -RKQ :LOH\ DQG 6RQV 1HZ
PAGE 382

'HF & 'HFNHU &HUDPLF (QJLQHHULQJ DQG 6FLHQFH 3URFHHGLQJV f f 'H+ 5 'H+RII 7KHUPRG\QDPLFV LQ 0DWHULDOV 6FLHQFH 0F*UDZ+LOO 1HZ
PAGE 383

(SSE 5 (SSOHU DQG (SSOHU &HUDPLF (QJLQHHULQJ DQG 6FLHQFH 3URFHHGLQJV f f (VF $ (VFDUGLQR 2XDOLFHU f )HU 7KH 86 7LOH ,QGXVWU\ 3DPSKOHW GLVWULEXWHG E\ )HUUR &RUSRUDWLRQ &OHYHODQG 2+ f )OH )OHPLQJ ([SHULPHQWDO 7HFKQLTXHV LQ *ODVV 6FLHQFH & 6LPPRQV DQG (O%D\RXPL HGVf $PHULFDQ &HUDPLF 6RFLHW\ :HVWHUYLOOH 2+ f *UH 6 *UHJRULR 0 *UHHQEODWW 3LIHU DQG 0 6WXUJH -RXUQDO RI &KHPLVWU\ DQG 3KYVLFLV f f *UX *UXP*U]KLPDLOR *ODVV &HUDPLF f f +DO +DOOLGD\ DQG 5 5HVQLFN )XQGDPHQWDO RI 3K\VLFV QG (G -RKQ :LOH\ DQG 6RQV 1HZ
PAGE 384

.LQ : .LQJHU\ + %RZHQ DQG 8KOPDQQ ,QWURGXFWLRQ WR &HUDPLFV QG (G -RKQ :LOH\ DQG VRQV 1HZ
PAGE 385

0XU 0XUUD\ $PHULFDQ 1DWLRQDO 6WDQGDUG 6SHFLILFDWLRQV IRU &HUDPLF 7LOH $ 7LOH &RXQFLO RI $PHULFD 3ULQFHWRQ 1f 0XU 6 0XUGRFN 7 :LVH DQG (SSOHU $PHULFDQ &HUDPLF 6RFLHW\ %XOOHWLQ f f 1DV 1DVVDX 7KH 3K\VLFV DQG &KHPLVWU\ RI &RORU -RKQ :LOH\ DQG VRQV 1HZ
PAGE 386

6DF 0 6DFNV (0$ &HUDPLF 3URFHVVLQJ &ODVV 1RWHV 8QLYHUVLW\ RI )ORULGD f 6DQ & 6DQGHUV DQG :\V]HFNL -RXUQDO RI 2SWLFDO 6RFLHW\ RI $PHULFD f 6FK + 6FKRO]H *ODVWHFK %HU f 6HO + 6HOOLQJ -RXUQDO RI WKH 6RFLHW\ RI '\HUV &RORULVWV f 6H] 6H]]L 7LOH ,WDOLD f 6KD 6KDFNHOIRUG DQG : $OH[DQGHU &5& 0DWHULDOV 6FLHQFH DQG (QJLQHHULQJ +DQGERRN &5& 3UHVV %RFD 5DWRQ )/ f 6KH / 6KHSSDUG &HUDPLF ,QGXVWU\ f f 6LP 6LPPRQV ([SHULPHQWDO 7HFKQLTXHV LQ *ODVV 6FLHQFH & 6LPPRQV DQG (O%D\RXPL (GVf $PHULFDQ &HUDPLF 6RFLHW\ :HVWHUYLOOH 2+ f 7D\ 7D\ORU DQG $ %XOO &HUDPLFV *OD]H 7HFKQRORJ\ 3HUJDPRQ 3UHVV (OPVIRUG 1< f 7HX 7HXFKHUW 5 7KRPDV DQG $ 0XOOHU=HOO ,QWHUFHUDP f f 7LS 3 7LSOHU 3K\VLFV :RUWK 3XEOLVKHUV ,QF 1HZ
PAGE 387

:\F 5 :\FRII &U\VWDO 6WUXFWXUHV UG (G -RKQ :LOH\ DQG 6RQV 1HZ
PAGE 388

%,2*5$3+,&$/ 6.(7&+ 'DYLG (DUO ZDV ERUQ LQ %XIIDOR 1HZ
PAGE 389

DGPLWWHG WR GRFWRUDO FDQGLGDF\ DW WKH 8QLYHUVLW\ RI )ORULGD LQ 'HFHPEHU RI +LV SURIHVVLRQDO DIILOLDWLRQV LQFOXGH PHPEHUVKLSV LQ WKH $PHULFDQ &HUDPLF 6RFLHW\ $&HU6f 1DWLRQDO ,QVWLWXWH IRU &HUDPLF (QJLQHHUV 1,&(f 3KL .DSSD 3KL LQWHUQDWLRQDO DFDGHPLF KRQRU VRFLHW\ &HUDPLF 0DQXIDFWXULQJ &RXQFLO &0&f &HUDPLF (GXFDWLRQDO &RXQFLO &(&f DQG .HUDPRV +H LV DOVR D 'LVWLQJXLVKHG 0HQWRU IRU $&HU6

PAGE 390

, FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ & ; 'DYLG ( 6ODUN &KDLUPDQ 3URIHVVRU RI 0DWHULDOV 6FLHQFH DQG (QJLQHHULQJ ( 'RZ :KLWQH\ 3URIHVVRU RI 0DWO DQG (QJLQHHULQJ DV FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI SMLORVRSK\ ,V 6FLHQFH FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ 3URIHVVRU RI 0DWHULDOV 6FLHQFH DQG (QJLQHHULQJ

PAGE 391

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

PAGE 392

/2 LV/ L ; 81,9(56,7< 2) )/25,'$


INFLUENCES OF COMPOSITION, MELT VISCOSITY AND CRYSTALLIZATION
ON THE COLOR STRENGTH AND STABILITY OF MULTI-OXIDE GLASS
FRIT/ZIRCON-VANADIUM PIGMENT SYSTEMS FOR CERAMIC
WHITEWARES COATINGS APPLICATIONS
By
DAVID A. EARL
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1998

ACKNOWLEDGMENTS
I am grateful for the guidance and inspiration provided
by Dr. David E. Clark, chairman of my supervisory committee.
My association with Dr. Clark has greatly enhanced my
academic, professional and personal growth over the past five
years.
I would like to thank Dr. Joseph Simmons, Dr. E. Dow
Whitney, Dr. Jack Mecholsky and Dr. Dinesh Shah for
participating on my supervisory committee. Thanks also go to
Kristie Leiser, Mark Moore, Robert DiFiori, Diane Folz and
Rebecca Schulz of Dr. Clark's research group for their
advice.
In addition I would like to acknowledge the industrial
support of this research. I would like to thank Florida Tile
Industries; Bob Blonski, Klaus Meinssen, Bruno Burzacchini
and Marzia Barrattini of Ferro Corporation; and Dan Swiler,
Hong Chen and Pam Lucas of Ceredec Corporation.
Finally, and most importantly, I am grateful for the
patience and encouragement of my wife, Jacquie. This
n

research effort would not have been possible without her
support.
in

TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ii
LIST OF TABLES viii
LIST OF FIGURES x
ABSTRACT xix
CHAPTERS
1. INTRODUCTION 1
1.1 Color in the Ceramics Industry 1
1.2 Glaze Colorants 5
1.3 Potential Influence of Frit 7
1.4 Overview of Dissertation Goals 12
1.5 Guide for Using This Dissertation 16
2. BACKGROUND 21
2.1Color Theory 21
2.1.1 Light and the Visible Spectrum 22
2.1.2 Materials Interactions with Light 29
2.1.2a Refraction 30
2.1.2b Surface Reflection and Gloss 36
2.1.2c Opacity and Translucency 42
2.1.2d Absorption, Transmission and
Color 51
2.1.3 Color Perception by the Human Eye 60
2.1.4 Color Measurement 65
2.1.4a Spectrophotometry 65
2.1.4b Basis for Color Quantification.... 72
2.1.4c CIE L*a*b* Measurement Scale 78
IV

2.2 Color in Ceramic Glazes 86
2.2.1 Silicate Glass Structures and
Properties 87
2.2.2 Glaze Base Materials and Formulas 99
2.2.3 Fast-Fire Whiteware Glazes and Frits... 105
2.2.4 Ceramic Colorants 113
2.2.4a Zircon Triaxial Pigments 115
2.2.4b Kubelka-Munk Analysis of
Colorant Layers 127
2.2.5 Frit Influence on Color Development.... 132
2.2.5a Crystallization Mechanisms 132
2.2.5b Zircon Crystallization and
Dissolution 145
2.2.5c Liquid-Liquid Phase Separation.... 152
2.2.5d Viscosity Relationships 155
3. EXPERIMENTAL PROCEDURES 161
3.1 Materials and Methods 161
3.1.1 Glass Frits and Zr-V Pigment 161
3.1.2 Coatings Preparation and Application... 165
3.1.3 Firing Curves 167
3.2 Materials Characterization and Analytical
Techniques 169
3.2.1 AAS and XRF 169
3.2.2 Frit Density Determination 169
3.2.3 Laser Diffraction Particle Size
Analysis 170
3.2.4 Spectrophotometry and Color
Calculations 170
3.2.5 Gloss Measurements 172
3.2.6 Heating Microscopy 173
3.2.7 Dilatometry 174
3.2.8 X-Ray Diffraction (XRD) 174
3.2.9 Scanning Electron Microscopy (SEM) and
Energy Dispersive X-Ray Spectroscopy (EDS).. 176
3.3 Statistical Methods for Deriving Equations.... 177
4. RESULTS 180
4.1 Frit and Pigment Properties 180
4.2 Color of Fired Coatings 181
4.2.1 Spectral Reflectance Curves 182
4.2.2 Pigment Absorption Factors (K/S) 188
v

4.2.3 Color in L*, a* and b* Values 193
4.2.4 Color Stability 206
4.2.5 Specular Gloss 208
4.3 Viscosity of Coatings During Heating 210
4.3.1 Heating Microscope Images 210
4.3.2 Dilatometric Tg and T2 213
4.3.3 Viscosity vs. Temperature 216
4.4 Derived Statistical Models 219
4.4.1 K/S vs. Coating Composition and
Temperature 220
4.4.2 AE* vs. Coating Composition 235
4.4.3 Log Viscosity vs. Coating Composition
and Temperature 244
4.5 Evolved Crystalline Species 246
4.5.1 XRD, SEM and EDS Evaluations 24 6
4.5.1a Frits with ZrC>2 246
4.5.1b Frits without ZrC>2 260
4.5.2 Zircon Quantitative Analysis 277
4.5.2a Frits with ZrÜ2 and ZnO 280
4.5.2b Frits with Zr02 and SrO 280
4.5.2c Frits without ZrÜ2 281
5. DISCUSSION 282
5.1 Color Strength and Stability Dependency 282
5.1.1 Coating Composition 284
5.1.1a Zr-V Loading 284
5.1.1b Zr02 286
5.1.1c SrO vs. ZnO 290
5.1. Id Al203/Alkalis 293
5.1.2 Crystalline Species 296
5.1.2a Zircon 296
5.1.2b Diopside 305
5.1.2c Hardystonite 307
5.1.2d Strontium Calcium Silicate 308
5.2 Melt Viscosity 312
5.2.1 Influence on Crystallization and
Zr-V Dissolution 312
5.2.2 Value as a Predictor of K/S and AE*.... 315
vi

6. SUMMARY AND CONCLUSIONS 320
6.1 Zr-V Pigment and Color Values 324
6.2 Frit Oxide Composition 326
6.3 Viscosity, Crystallization and Zr-V
Dissolution 328
7. FUTURE WORK 333
APPENDICES
A UNITS FOR DESCRIBING LIGHT AND COLOR 337
B THE 15 CAUSES OF COLOR 339
C DENSITY, PARTICLE SIZE AND APPLICATION WEIGHT DATA. 341
D DATA FROM COATINGS BATCHED WITH FRIT, 2.5%
BENTONITE AND Zr-V PIGMENT, AND FIRED USING A
45-MINUTE CERAMIC TILE CYCLE 343
E FRIT SPECTRAL REFLECTANCE DATA AND CURVES AT EACH
TEMPERATURE AND PIGMENT LOADING 34 7
REFERENCES 359
BIOGRAPHICAL SKETCH 367
vi 1

LIST OF TABLES
Table Page
1.1 Example of Ceramic Coatings Applications on a
Decorated Floor Tile 8
1.2 Variables That Influence Ceramic Glaze Color 9
2.1 Index of Refraction of Selected Materials at
589 nm Wavelength in Air 32
2.2 Properties of Materials Used for Opacifying
Ceramic Glazes 45
2.3 Transition Elements and Their Properties 58
2.4 Factors for Uniform Color Scales for Normalizing
to a Standard Reference White 81
2.5 Glass Formers, Intermediates and Modifiers
Materials Commonly Employed in Whiteware
Glazes 92
2.6 Properties Associated with the Presence of
Various Oxides in Glass 98
2.7 Common Ceramic Tile Glaze Base Materials 100
2.8 Examples of Compositions (Weight %) of Commercial
Glazes 101
2.9 Summary of Important Glaze Properties and
Characteristics 102
2.10 Seger's Formula for Classifying Glazes 103
2.11 Typical Empirical Formulas, in Molar Equivalents,
for Fast-Fire Gloss and Matte Glazes Ill
viii

2.12 Glaze Pigments and Their Requirements 117
3.1 Frits Investigated 162
E.l Engobe and Frits A and B Reflectance Data 347
E.2 Frits C and D Reflectance Data 348
E.3 Frits E and F Reflectance Data 349
E.4 Frits G and H Reflectance Data 350
IX

LIST OF FIGURES
Figure Page
1.1 Flow chart summary of main research variables 17
2.1 Electromagnetic spectrum 25
2.2 Wavelength vs. energy distribution of daylight.... 26
2.3 Refractive index vs. wavelength of incident
light for three glasses 34
2.4 Reflection and transmission of light by a glassy
material containing suspended particles 37
2.5 Fresnel reflection for a) an air/glass boundary
and b) total reflectivity for different index
of refraction 40
2.6 Reflectance vs. wavelength of light for a TÍO2
opacified white glaze fired to 1000°C, at
various glaze application weights in g/ft2 50
2.7 The 5 d orbitals 57
2.8 Human optical system 61
2.9 Luminosity functions of the rods (nighttime
scotopic vision) and cones (daytime photopic
vision) of the human eye 63
2.10 Reflectance versus wavelength for opaque coatings
colored with pigments that absorb a portion of
incident light 67
2.11 Basic components of spectrophotometers 69
x

2.12 Schematic of the Hardy spectrophotometer
2.13 Weighting functions used for the standard
observer at a 2° field of view 74
2.14 Luminosity or lightness (Y) and chromaticity
(x, y) MacAdam limits for colors viewed in
daylight 7 9
2.15 Schematic of L*a*b* color space 83
2.16 Comparison of structures and XRD patterns of
crystalline and vitreous silica 90
2.17 Two-dimensional representation of modifiers
(a) Na+1 and (b) Ca+2 in the silicate glass
structure 95
2.18 Inorganic pigment family 114
2.19 CIE a* and b* chroma of ceramic pigments 116
2.20 Typical forms of zircon crystals. (a-c): a{ 100} ,
m{ 110} , p{ 101} , x{ 211} ; and zircon lattice
structure (d,e) 120
2.21 Splitting of the d orbital in V+4 by tetrahedral
(Td) and tetragonal (D2ci) crystal fields 124
2.22 Schematic of basis for Kubelka-Munk analysis of
colorant layers 128
2.23 Relationship between viscosity and temperature
favoring nucleation and growth in glazes 137
2.24 Crystal growth rate as a function of temperature
in Na20-Ca0-Al203-SiC>2 glass 144
2.25 Binary phase diagram of Zr02 and Si02 system 14 9
2.26 Viscosities of some commercial silicate glasses... 157
XI

3.1 Time-temperature profiles used to fire the tiles.. 168
4.1 Spectral reflectance of unfired raw materials and
the engobe substrate backing 183
4.2 Spectral reflectance of coatings batched with
frit C, fired to 100CTC 186
4.3 Spectral reflectance of coatings batched with
frit H, fired to 1100°C 187
4.4 Spectral reflectance of coatings batched with
2.0% Zr-V, fired to 1000°C 189
4.5 Spectral reflectance of coatings batched with
2.0% Zr-V, fired to 1050°C 190
4.6 Spectral reflectance of coatings batched with
2.0% Zr-V, fired to 1100oC 191
4.7 Pigment absorption factors versus weight percent
Zr-V batched in coatings fired to 1000°C, 1050"C
and 1100 °C peak temperature 192
4.8 Color values of coatings batched with frits
(A-H) and Zr-V pigment, and fired to 1050°C 194
4.9 Color values of coatings batched with frits
(A-H) and Zr-V pigment, and fired to 1100'C 195
4.10 Color values of coatings batched with frits
(A-H) and no Zr-V, and fired to 1000°C, 1050°C
or 1100*C 198
4.11 Color values of coatings batched with frits
(A-H) and 0.5% Zr-V, and fired to 1000°C, 1050°C
or 1100“C 199
4.12 Color values of coatings batched with frits
(A-H) and 2.0% Zr-V and fired to 1000°C, 1050°C
or U00°C 200
xi i

4.13 Color values of coatings batched with frits
(A-H) and 5.0% Zr-V, and fired to 1000°C, 1050°C
or 1100°C 201
4.14 Relationship between lightness (L*) and blueness
(-b*) of coatings batched with Zr-V pigment and
fired to 1000°C, 1050°C and 1100°C peak
temperature 203
4.15 Zr-V pigment absorption factor relationships
with L*, a* and b* color values of fired
coatings 205
4.16 Color changes (AE*) due to a variation in peak
firing temperature from 1050°C to 1100°C for
coatings batched with frits (A-H) and Zr-V
pigment 207
4.17 Specular gloss of fired coatings at a 60° angle
of incidence 209
4.18 Heating microscope images of coatings batched
with frits (A-H) and 2.0% Zr-V, at 900°C, 1000°C
and 1100°C 211
4.19 Heating microscope images of characteristic
stages of flow of coatings batched with frits
(A-H) and 2.0% Zr-V 212
4.20 Thermal dilatometric analyses of coatings A-D
batched with 2.0% Zr-V 214
4.21 Thermal dilatometric analyses of coatings E-H
batched with 2.0% Zr-V 215
4.22 Log r| versus temperature of coatings batched with
2.0% Zr-V. Frits A-D incorporate 8% Zr02 217
4.23 Log r| versus temperature of coatings batched with
2.0% Zr-V. Frits E-H contain no Zr02 218
xiii

4.24 Pigment absorption factor actual and equation
(4.2) predicted results for fired coatings
batched with 2.0% Zr-V 224
4.25 Individual effects of frit oxides on K/S based on
statistical model (4.2) for 2.0% Zr-V and a
peak firing temperature of 1050°C 225
4.26 Variables weight of influence on K/S, based on
statistical models 4.1 (0.5% Zr-V), 4.2
(2% Zr-V) and 4.3 (5% Zr-V) 228
4.27 Pigment absorption factor trends with frit
AI2O3:alkalis ratio of coatings batched with
2.0% Zr-V and fired to 1000°C, 1050°C or 1100°C. 229
4.28 Pigment absorption factor versus frit ZnO and
SrO molar equivalents of fired coatings batched
with 2.0% Zr-V and fired to 1000°C, 1050°C or
1100°C peak temperature 230
4.29 Influence of frit ZrC>2, in the presence of SrO or
ZnO, on K/S of fired coatings batched with 2.0%
Zr-V and fired to 1000°C, 1050°C or 1100°C 231
4.30 Variations in pigment absorption factor (K/S) due
to changes in peak firing temperature, for
coatings batched with 2.0% Zr-V and frits
containing 8% Zr02 versus 0% Zr02 232
4.31 Delta E* actual and equation (4.7) predicted
results for fired coatings batched with 2.0%
Zr-V 239
4.32 Individual effects of frit oxides on Delta E*,
based on statistical model 4.7 for 2.0% Zr-V.... 240
4.33 Variables weight of influence on Delta E*, based
on statistical models 4.6 (0.5% Zr-V), 4.7
(2.0% Zr-V) and 4.8 (5.0% Zr-V) 242
xiv

4.34 Color changes due to variations in peak firing
temperature and frit AI2O3: alkali ratio of
coatings batched with 2.0% Zr-V 243
4.35 XRD patterns from coatings batched with frit A
and 2.0% Zr-V, and fired to 1000°C, 1050°C and
1100°C. Frit A includes 8% ZrC>2, 12% ZnO,
0% SrO, 5% alkalis 247
4.36 SEM micrographs of a coating batched with frit A
and 2.0% Zr-V, and fired to 1100°C.
Magnification is (a) X 1,000 and (b) X 6,000.
Shown are large Zr-V particles surrounded by-
fine zircon precipitates 248
4.37 XRD patterns from coatings batched with frit B
and 2.0% Zr-V, and fired to 1000°C, 1050°C and
1100°C. Frit B includes 8% ZrC>2, 12% ZnO,
0% SrO, 10% alkalis 249
4.38 SEM micrographs of a coating batched with frit B
and 2.0% Zr-V, and fired to 1100°C.
Magnification is (a) X 1,000 and (b) X 6,000.
Shown are large Zr-V particles, fine zircon
precipitates and large zircon fibers 250
4.39 XRD patterns from coatings batched with frit C
and 2.0% Zr-V, and fired to 1000°C, 1050°C and
1100°C. Frit C includes. 8% ZrÜ2, 0% ZnO,
12% SrO, 5% alkalis 251
4.40 XRD patterns from coatings batched with frit D
and 2.0% Zr-V, and fired to 1000°C, 1050°C and
1100°C. Frit D includes 8% Zr02, 0% ZnO,
12% SrO, 10% alkalis 252
4.41 SEM micrographs of coatings batched with 2.0%
Zr-V, (a) frit C and (b) frit D, and fired to
1100CC. (magnification X 1,000). Particles
shown are Zr-V pigment 253
4.42 X-ray diffraction profile for zircon, ZrSi04 255
xv

4.43 SEM micrograph of zircon-vanadium (Zr-V) pigment,
Ceredec 41715A, X 1,000 magnification 256
4.44 XRD patterns from coatings batched with frit E
and 2.0% Zr-V, and fired to 1000°C, 1050°C and
1100°C. Frit E includes 0% ZrC>2, 12% ZnO,
0% SrO and 5% alkalis 261
4.45 SEM micrograph of a coating batched with frit
E and 2.0% Zr-V, and fired to 1050°C.
Magnification is X 1,000. Shown are large
Zr-V particles surrounded by dispersed diopside. 262
4.46 XRD patterns from coatings batched with frit
F and 2.0% Zr-V, and fired to 1000°C, 1050°C
and 1100°C. Frit F includes 0% ZrC>2, 12% ZnO,
0% SrO and 10% alkalis 263
4.47 SEM micrographs of a coating batched with frit
F and 2.0% Zr-V, and fired to 1100°C.
Magnification is (a) X 1,000 and (b) 6,000.
All particles shown are hardystonite 264
4.48 XRD patterns from coatings batched with frit G
and 2.0% Zr-V, and fired to 1000°C, 1050°C and
1100°C. Frit G includes 0% Zr02, 0% ZnO, 12%
SrO and 5% alkalis 265
4.49 SEM micrograph of a coating batched with frit
G and 2.0% Zr-V, fired to 1100°C (magnification
X 1, 000). Shown is crystallized SrCa2SÍ30g 266
4.50 XRD patterns from coatings batched with frit
H and 2.0% Zr-V, and fired to 1000°C, 1050°C
and 1100°C. Frit H includes 0% ZrÜ2, 0% ZnO,
12% SrO and 10% alkalis 267
4.51 SEM micrographs of a coating batched with frit
H and 2.0% Zr-V, and fired to 1100°C.
Magnification is (a) X 1,000 and (b) X 6,000.
Shown is crystallized SrCa2Si309 268
xvi

4.52 Relationship between XRD [ 312] integrated
intensity and weight percent zircon in unfired
coatings 278
4.53 Weight percent zircon in fired coatings batched
with 2.0% Zr-V 279
5.1 Changes in reflectance distributions at 400 nm
and 640 nm wavelengths due to increases in peak
firing temperature for coatings batched with
0.5%, 2.0% and 5.0% Zr-V 285
5.2 Influence of frit density on the color stability
of coatings batched with 2.0% Zr-V pigment 289
5.3 Pigment absorption factors (K/S) for coatings
batched with Zr-V pigment and frits containing
8% Zr02, and fired to 1000°C, 1050°C or 1100°C
peak temperature 291
5.4 Visual lightness (L* ) , greenness (-a*) and
blueness (-b*) as a function of weight percent
zircon in the fired coatings batched with 2.0%
Zr-V 299
5.5 Pigment absorption factor (K/S) as a function of
Zr-V pigment and zircon contents in coatings
batched with 2.0% Zr-V 302
5.6 Changes in color stability denoted by K/S and
Delta E* , asa function of weight percent
zircon in coatings batched with 2.0% Zr-V 304
5.7 Color lightness (L*) and blueness (-b*)
progression with diopside crystallization
and pigment dissolution in coatings batched
with frit E and 2.0% Zr-V 306
5.8 Color lightness (L*) and blueness (—b*)
progression with hardystonite crystallization
and pigment dissolution in coatings batched
with frit F and 2.0% Zr-V 309
xvi 1

5.9 Color lightness (L*) and blueness (-b*)
progression with SrCa2SÍ30g crystallization
and pigment dissolution in coatings batched
with 2.0% Zr-V 311
5.10 Integrated log viscosity from 700°C to 1100°C
versus fired color strength (K/S) and stability
(Delta E*) in coatings batched with 2.0% Zr-V... 316
5.11 Slope in the log viscosity versus temperature
near the softening point versus fired color
strength (K/S) and stability (Delta E*) in
coatings batched with 2.0% Zr-V 318
E.l Frit A spectral reflectance curves 351
E.2 Frit B spectral reflectance curves 352
E.3 Frit C spectral reflectance curves 353
E.4 Frit D spectral reflectance curves 354
E.5 Frit E spectral reflectance curves 355
E.6 Frit F spectral reflectance curves 356
E.7 Frit G spectral reflectance curves 357
E.8 Frit H spectral reflectance curves 358
xviii

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
INFLUENCES OF COMPOSITION, MELT VISCOSITY AND CRYSTALLIZATION
ON THE COLOR STRENGTH AND STABILITY OF MULTI-OXIDE GLASS
FRIT/ZIRCON-VANADIUM PIGMENT SYSTEMS FOR CERAMIC
WHITEWARES COATINGS APPLICATIONS
By
David A. Earl
December 1998
Chairman: David E. Clark
Major Department: Materials Science and Engineering
Color control is becoming increasingly important in the
industrial processing of ceramics coatings. Multi-oxide
glass frits are the predominant materials in ceramic
whitewares coatings, and zircon doped pigments are the most
commonly used colorants.
The primary objective of this research was to determine
if glass frits could be formulated to improve the fired color
strength and high-temperature stability of ceramic coatings
colored with zircon-vanadium (Zr-V) blue pigments. The
results would also be applicable to other ceramic pigments
that utilize the same zircon structure to incorporate
xix

colorant metal ions. A secondary goal was to relate the frit
oxide composition, pigment content, firing temperature, melt
viscosity and microstructural development to the fired color.
A ceramic tile process was applied to fabricate sample
coatings for the study. A coating's color was quantified and
related to human perception with CIE L*, a* and b* values and
pigment absorption factors (K/S), calculated based on
spectral reflectance data.
The research was successful in quantifying the influence
of individual glass frit oxides on the fired color strength
and high-temperature stability of the coatings. Opaque and
transparent glossy frit compositions which yield excellent
color strength and stability were formulated. Mathematical
models for predicting a coating's color strength and
stability given the frit oxide composition, Zr-V pigment
loading and peak firing temperature were derived. Frit
oxides of Zr02, SrO, ZnO, A1203, Na20 and K20 were found to
have a significant influence on crystallization, pigment
dissolution and color development. The properties, sizes,
morphologies and quantities of crystalline phases that
precipitated in the coatings during firing were related to
the color. A technique for producing uniquely light yet high
xx

chroma colors through control of zircon precipitate particle
size was demonstrated.
In addition, a statistical model was developed for
calculating the coating melt viscosity as a function of the
frit oxide composition and temperature. Melt viscosity
versus temperature curves were applied to predict a frit's
potential for producing strong and stable color.
xxi

CHAPTER 1
INTRODUCTION
1.1 Color in the Ceramics Industry
The color of a product often determines its aesthetic
and monetary value. Selling prices of ceramics such as tile,
brick, artware, dinnerware, cookware, porcelain enamel,
concrete, bathroom fixtures and decorative glass are directly
related to their appearance.
Customers often select ceramic products based on viewing
sample chips or prototypes at stores or retail distributors.
Sometimes buyers wish to color coordinate with the fixtures,
appliances or paint of an existing building. Occasionally,
customers purchase more of the same product in the future,
for example, adding the same color floor tile to an enlarged
room. In all of these cases, it is imperative that
manufacturers strictly conform to tight color tolerances,
year after year, in order for goods to comply visually with
customer expectations.
Inconsistencies in manufactured colors lead to
significant increases in industrial costs. Noticeable color
1

2
variations in fired coatings due to normal fluctuations in
manufacturing processing conditions result in nonstandard,
scrap products and lower productivity. For example, a major
portion of the $120 million worth of products scrapped in
1997 by U.S. ceramic tile manufacturers resulted from out-of-
tolerance colors. Since the U.S. ceramic tile market only
comprises about 0.6% of the world ceramic whitewares sector,
problems with color variations result in significant
industrial waste worldwide [ Sez98] . In addition, product
inventory levels must be increased in order to accommodate
multiple color shades per item.
Color consistency in manufacturing has become
increasingly difficult due to the rapid evolution of
processing technology to improve productivity and
profitability. Often during the transition, accessories for
the same product line must match in color but are made with
two different processes. For example, in the ceramic tile
industry, flat "field" tile is predominately manufactured
with completely automated fast-fire roller kiln technology,
while difficult to handle trim and angle shapes are
constrained to high-labor slow-fire tunnel kiln processes.
Firing differences between the processes (30-minute fast-fire
vs. 14-hcur siow-fire cycles) complicate the formation of

3
matching glaze colors, but customers demand no noticeable
difference between coordinating trim and field glazes. Color
incompatibility between these products is currently the
largest customer complaint in the ceramic tile industry.
Other whitewares companies, including some manufacturers of
sanitaryware and dinnerware, are planning to convert to the
same fast-fire roller kiln technology and will encounter
similar problems with color variations.
Undesirable color differences between coatings batched
with the same formula occur even in facilities that utilize
the same type of firing technology throughout the factory.
Firing variations result from different kiln models, sizes
and shapes, the debugging of new kiln eguipment, kiln fan and
burner wear over time, and changes in ambient conditions
which cause the kiln intake air density, humidity and
temperature to vary. Gaps in the product load entering a
kiln, normally due to breakdowns in machines upstream in the
production line, also cause firing temperature versus time
profiles to fluctuate. Shifts in firing conditions result in
variations in crystallization and pigment dissolution in
ceramic coatings, which alter their fired gloss, opacity and
color. This burdens industry with low yields, customer
complaints and potential loss of business.

4
Another source of color variation can be attributed to
the continued effort to reformulate ceramic coatings in order
to lower raw material costs. Unfortunately, some of the
cheaper systems have also yielded iower color strength and
stability during high-temperature processing. Due to
increasing foreign competition, domestic companies have been
compelled to lower manufacturing costs to enable selling
price reductions and gain a competitive edge. For example,
in the ceramic tile industry, import market share rose from
about 20% in the mid-1970s to over 60% in the 1990s [ Ear94] .
Glaze raw material costs comprise roughly 10% of all tile
manufacturing expenses or approximately $100 million per year
[ Fer96, Sez98]. In order to stay in business, it is crucial
for whitewares companies to optimize color consistency while
minimizing glaze costs.
There is an increasing interest in the ceramics industry
to develop low cost color systems which are reproducible over
a range of processing conditions. High-temperature
interactions between multi-oxide glass frits and pigments in
ceramic glazes greatly affect the fired color. The influence
of frit is most significant since it is a relatively
expensive, carefully manufactured material usually added at
the highest weight percent of ingredients in fast-fire glaze

batches. Compared to other glaze components, frit normally
has the lowest melting temperature and is the most corrosive
5
to ceramic colorants. Most frit in the U.S. is used by
ceramic tile manufacturers, but frit consumption by other
whitewares industries will increase as they convert to fast-
fire roller kiln technology. Zircon doped pigments are the
most commonly used colorants for whitewares coatings because
they have the best high-temperature stability.
The overall objective of this research was to determine
the influence of various oxides in glass frits on color
development with zircon doped pigments during firing. The
results could be used to improve the color strength and
stability of industrial ceramic glaze systems.
This chapter very briefly introduces the reader to color
in ceramic glazes, the potential influence of frit on color
and an overview of the dissertation goals. More detailed
discussions are included in succeeding chapters.
1.2 Glaze Colorants
Ceramic glazes provide an impervious glassy decorative
coating for whitewares. Colors are produced with selective
scattering or absorption of incident light by colloidal-sized
particles suspended in the glassy matrix.

6
Solution colorants are sometimes introduced into the
glaze batch as oxides, then dissolved and precipitated as
metal ions during the firing process. Oxides such as Cr203
(green), CoO (blue), CuO (green to red), Fe203 (yellow to
brown) and MnO, (purple to brown) were common sources of metal
ion solution colorants.
They are now rarely used in high volume whitewares
processes such as ceramic tile manufacturing because their
solubility and reprecipitation, and thus the color produced,
are extremely sensitive to the glaze composition, particle
size distribution, firing time and kiln atmosphere.
More typically, manufactured inert pigment particles are
added to glazes to obtain color. The most widely used for
industrial glazes are zircon crystal lattices doped with
metal ions. They provide a lower chroma than other pigments,
but the zircon structure is less soluble at high temperature.
Besides, most of the demand in the whitewares market is for
relatively weak, light colors.
The most common zircon-based pigments are zircon-
vanadium blue, zircon-iron coral and zircon-praseodymium
yellow. They are referred to as the triaxial pigments and
can be blended to achieve most glaze colors desired by
customers. Even though zircon crystals are more stable at

high temperature than other colorants, color variations still
occur. During firing, zircon may remain stable and protect
the metal ions, partially dissolve with that portion
reprecipitating or remaining in solution, or totally
dissolve.
Color control is further complicated because most
decorated ceramic whitewares contain multiple layers of
coatings. For example, a typical "stone look" floor tile
produced by Florida Tile Industries requires nine different
coatings over the pressed body (Table 1.1). Some
applications overlap, while others are distinctly separate on
rhe surface to create a more natural appearance and depth of
design. Both physical and chemical interactions between
layers influence pigment dissolution and the nucleation and
growth of new phases which affect opacity, gloss and color.
1.3 Potential Influence of Frit
Many variables influence color development in ceramic
glazes, as outlined in Table 1.2. The strength and high
temperature stability of ceramic pigments are highly
dependent upon the base glaze composition. Limited
preliminary studies [ Dec93, Byr94, Blo93] indicate that

Table 1.1. Example of Ceramic Coatings Applications on a
Decorated Floor Tile.
Glaze Coating
Color/Opacity
Application
Method
Application
Weight
(g/cm2)
Engobe
(primer coat)
White/Opaque
Rotating disk
0.05
Fume
Taupe/Opague
Spray gun
0.007
—
—
Brushing machine
—
Middle coat
Off-white/
Translucent
Rotating disk
0.09
Fume
Grey/Opaque
Spray gun
0.003
Ink 1
Light Grey/
Opaque
Screen printer
<0.001
Ink 2
Gold/Opaque
Screen printer
<0.001
Ink 3
Light Beige/
Opaque
Screen printer
<0.001
Ink 4
White/Opaque
Screen printer
0.001
Topcoat
Transparent
Spray gun
0.02

9
Table 1.2. Variables That Influence Ceramic Glaze Color.
I. Batch Composition
a. Base ingredients (oxide composition and phases
present)
b. Pigments (composition, structure and loading)
c. Chemical additives (influence application drying
rate and smoothness)
II. Glaze Preparation with Ball Milling
a.Particle size distribution (influences melting
point)
III. Application
a. Thickness (hiding power)
b. Drying rate (can affect composition gradient caused
by differences in particle settling rates)
c. Smoothness (influences gloss or degree of specular
reflectance)
IV Firing
a. Time vs. temperature (phase dissolution and
precipitation)
b. Kiln atmosphere (oxidation/reduction reactions,
sulfur "scumming," etc.)

10
improved slow-fire glaze compositions may reduce color
changes resulting from variations in glaze preparation,
application and firing conditions. However, there is no
published comprehensive or quantitative research on the
subject. There is also a great lack of research on fast-fire
glaze systems.
Frits are ceramic compositions that have been fused,
quenched to form a glass and granulated [ Dod94] . They are
the primary ingredients in fast-fire ceramic glazes and in
most cases are the most reactive and corrosive part of the
formula.
Frits for whitewares coatings are classified as either
opaque (opacified; containing Zr0o) or transparent
(unopacified; no Zr02) , and glossy or matte. Besides Zr02,
frits also normally employ Si02 as the primary glass former,
alkalis (K¿0 and Na2C) , B203 and ZnC or SrO as the main
fluxes, and other oxides such as CaO, A1203 and MgO. These
oxides are cost effective, environmentally safe and provide
the desired glaze properties. Compositions with BaO or PbO
are avoided because these elements are deemed hazardous by
the EPA. There is also an increasing interest in replacing
ZnO with SrO because ZnO is classified as a regulated
chemical by EPA. Sections 2.2.1-2.2.3 and 2.2.5 detail the

theoretical effects of individual frit oxides on the
properties of glasses and ceramic glazes.
High temperature properties of frits influence crystal
growth and dissolution rates in glazes. It was observed
[ Jam85, Dor94] that dissolution and crystallization
velocities in multi-oxide glasses are diffusion transport
related and inversely proportional to the glass melt
viscosity, although no accurate models have been developed.
Glass viscosity, in turn, varies with composition and overall
has an Arrhenius-type inverse exponential relationship to
temperature. Phase changes in the glass resulting from these
phenomena alter the optical properties and color.
During fast-fire ceramic processing, a glaze is
typically in the molten stage at the peak temperature for
only 3 to 5 minutes. The rapid changes in heating and
cooling rates create a complex thermodynamic system where
phases are often not brought to equilibrium at high
temperature. Frit compositions which reduce the sensitivity
of crystallization and pigment dissolution to processing
variations would be beneficial for color control. The
"ideal" frit would preserve the pigment and precipitate the
same quantity and morphology of desired crystalline phases
over a wide range of firing conditions. The frit should also

12
produce a coating with enough opacity to hide the substrate
without significantly concealing the pigment and achieve the
desired surface gloss without defects. Currently, it is
not known if zircon pigments dissolve during fast-fire
cycles, and there is uncertainty regarding what phases
precipitate.
1.4 Overview of Dissertation Goals
The main goals of this investigation were to
1. Determine if glass frit oxide compositions could be
formulated to improve the fired color strength and high-
temperature stability of industrial whitewares coatings
colored with zircon pigments.
2. Relate the optical properties resulting from zircon-
vanadium pigment in a glass matrix to the color
perceived.
3. Quantify the influence of individual frit oxides,
pigment loading and peak firing temperature on a
coating's color.
4. Correlate the evolution of the coating's structure and
properties to the original frit oxide composition and
the fired color.

13
5. Ascertain whether frit melt viscosity data can be
applied as an industrial quality control tool for
predicting a frits potential for producing strong and/or
stable color with zircon pigments.
"Fast-fire" ceramic tile manufacturing constitutes a
major portion of the whitewares industry and consumes most of
the frit produced in the U.S. This was the chosen processing
method for preparing and firing coating samples. Materials
selected for the study were eight laboratory-smelted frits
and a zircon-vanadium (Zr-V) blue pigment. Oxide
compositions of the frits were designed to provide cost
effective, environmentally safe formulas and comply with
Seger's rules (Section 2.2.2) for ensuring insolubility of
the frit and fired coating, and ready fusion at high
temperature. The range of oxide contents tested encompassed
and exceeded the range normally employed for glossy ceramic
tile glazes. Special emphasis was placed on comparing frit
compositions with Zr02 (opacified) versus no ZrCu
(unopacified), SrO versus ZnO as the secondary flux, and
alkali/silica ratios. The B203 contents were kept low in
order to avoid liquid-liquid phase separation. The Zr-V
pigment tested was the blue colorant most commonly used in
the ceramic tile industry. The results will also be

14
applicable to other ceramic pigments which utilize the same
zircon structure to incorporate colorant metal ions.
Research goals were achieved by performing the following
tasks:
1. Each of the eight experimental frits were loaded with
four different pigment concentrations of 0%, 0.5%, 2.0%
and 5.0% by weight. They were blended with water and a
suspending agent to produce 32 different glaze coatings.
2. Coatings were applied to opaque 2"X6" wall tile body
substrates using a wet spray method. Samples of each
formula were fired to 1000°C, 1050°C and 1100°C peak
temperatures using a standard "fast-fire" ceramic tile
industrial heating profile. A total of 96 different
fired coatings were produced.
3. The spectral reflectance versus wavelength and the gloss
of each fired coating were measured. The CIE L* a* b*
color values, pigment absorption factors (K/S), and
color differences between tiles fired to 1050°C and
1100°C (AE*) were calculated. Relationships between
light absorption by the pigment and color values based
on human perception were quantified.

15
4. Dilatometry and heating microscopy methods were employed
to measure the Tg, Ts and melt viscosity versus
temperature of coatings batched with 2.0% Zr-V.
5. In coatings batched with 2.0% Zr-V, phase changes and
resulting microstructures that formed during firing were
identified using x-ray diffraction, scanning electron
microscopy and energy dispersive x-ray spectroscopy
techniques. Fired coatings were quantitatively analyzed
for contents of Zr-V pigment and zircon which
precipitated from Zr02 and Si02 in the frit.
6. Statistical models were derived to predict K/S and AE*
given the original frit oxide composition, pigment
loading and peak firing temperature. An equation was
also developed for calculating log r\ of the frit with
2.0% Zr-V given the frit oxide composition and
temperature.
7. Color strength and stability were correlated to melt
viscosity, crystallization and Zr-V pigment
dissolution.
8. The mathematical models and experimental observations
were related to current scientific literature in order
to collate hypotheses which explain the results.

16
The foundation of materials science and engineering
research is to gain a better understanding of relationships
between processing, structure and properties of materials.
Figure 1.1 summarizes critical steps taken during this
research to define the processing-structure-properties
relationships of interest. This investigation focused mainly
on variables that influence a ceramic coating's color
strength and stability.
1.5 Guide for Using This Dissertation
The present document is greater in length than most
dissertations. The primary objective was not only to unveil
valuable information for basic science interests but also to
compile a text that could be used as a reference by engineers
working in industry. Thus, some sections may be bypassed if
only very specific information is desired.
In the Background chapter, Section 2.1 overviews current
scientific principles behind (a) materials interactions with
light with a focus on ceramics, (b) color perception by the
human eye and (c) the most common industrial method for
quantifying color and correlating it to human vision.
Section 2.2 summarizes current knowledge of the materials,
processing, structures and properties of ceramic whitewares

17
Figure 1.1.
Flow chart summary of main research variables.

18
coatings and their influences on color. A special emphasis
is placed on ceramic tile glazes. A review of common frit
compositions, ceramic colorants, crystallization, phase
separation and melt viscosity relationships is given.
Throughout the Procedure, Results and Discussion
chapters, references are made to specific principles and
equations outlined in Chapter 2.
Chapter 3 details the experimental procedures applied
for the research. This includes descriptions of typical
industrial "fast-fire" ceramic glaze frit compositions, wet
coating application methods and firing profiles. The two
main categories of frits investigated were with and without
Zr02. Materials characterization and analytical techniques
and procedures typically applied by industry to evaluate
whitewares coatings are also reviewed.
Chapter 4 shows the results of the research performed
for this dissertation. The particle size distributions,
densities and chemical analyses of the starting frits and
Zr-V blue pigment are given in Section 4.1. Section 4.2
details the optical properties of each of the 96 fired
coatings samples, as indicated by spectral reflectance
curves, gloss measurements and calculated color values of L*,
a*, b* , K/S and AE1* io50-noo°c* Section 4.3 reveals the

19
viscosity versus temperature profiles, heating microscope
images and dilatometric data for coatings loaded with 2.0%
Zr-V. Statistical models for predicting color strength
(K/’S) , color stability (AE+ 1050_n0C|OC) anc^ melt viscosity as a
function of the frit oxide composition and firing temperature
are given in Section 4.4. These equations provide a method
for engineers in the whitewares industry to estimate the
potential color strength, color stability and melt viscosity
resulting from various frit compositions when utilizing a
typical "fast-fire" heating profile. If the reader is only
interested in specifying frit compositions to obtain certain
color or viscosity results but is not concerned with the
crystallization or pigment dissolution processes responsible
for the optical properties, then it is not necessary to read
Sections 4.5 and 5.1.
Section 4.5 details the structures, compositions and
morphologies of crystalline species that precipitated in the
coatings during firing. An analysis for zircon present in
the coatings quantifies the amount of zircon precipitation
from fritted Si02 and Zr02, and Zr-V pigment dissolution that
occurred during firing. Results from XRD, SEM and EDS
analyses are shown.

20
In Chapter 5, the Discussion, relationships are
established between color strength and stability, specific
oxides in the glass frit, Zr-V loading, crystallization, Zr-V
dissolution and melt viscosity. Scientific explanations for
observed phenomena are proposed. The Discussion focuses on
basic science interests, although Section 5.2.2 is also
noteworthy for engineers in industry. It correlates
important characteristics of a coating's log r\ versus
temperature plot to the color strength and stability.

CHAPTER 2
BACKGROUND
2.1 Color Theory
The color of an object is its most apparent attribute.
Other properties such as gloss and opacity also contribute to
appearance.
Color is a term that can be used to describe the
reflection or transmission of light in visible wavelengths,
the properties of an object and the perception of the eye.
As one pioneer of color science, Dean B. Judd, stated
[ Nas83] ,
Color is that aspect of the appearance of objects
and lights which depends upon the spectral
composition of the radiant energy reaching the
retina of the eye and upon its temporal and spatial
distribution thereon, (p. 3)
Thus, in order to understand color, we must consider the
three elements involved with the production of color: the
light source,the physical modifications of light by matter
and the human eye as a color sensor. Various units typically
applied to quantify light and color are listed in Appendix A
along with essential conversion factors.
21

22
2.1.1 Light and the Visible Spectrum
Light is the visible radiant energy which interacts with
matter to produce what our eyes detect as color. It is an
electromagnetic wave that propagates as electric and magnetic
fields. Maxwell's theory [ Hal86] predicted the existence of
a spectrum of electromagnetic waves differing only in
wavelength and traveling through space in a vacuum with a
speed of c = 3 x 108 m/s. Electromagnetic waves other than
light include radio waves, microwaves and x-rays.
The energy of a light wave is quantized into small
bundles called photons. According to Einstein, the energy
(E) of a photon is [ Tip80]
E = hf = h —
A
(2.1)
where h = 6.63 x 30“34 J-sec is Planck's constant, f is its
frequency and A, the wavelength.
Other photonic relationships frequently applied are
[ Hum93]
2 P"
E = me" = — (2.2)
2m
p = me
(2.3)
Ap = h
(2.4)
where m is the mass of a particle and p is its momentum.
These equations allow us to contemplate a photon or light as

23
either a particle of energy E or a wave with a characteristic
frequency and wavelength.
The wave-particle duality of light can be described
mathematically by considering two harmonic waves with
slightly different frequencies which contain time and space
dependent components [ Hum93] :
'Ft = sin [ kx-cot] (2.5)
and
'F2 = sin [ (k+Ak)x -(co+Aco)t] (2.6)
, 2 n
where k = = the wave number (2.7)
X
co = 2nf = angular frequency (2.8)
Superposition of i)/1 and \|/2 and considering sin a + sin
(3 = 2cos2 (a-f3) *sin(a+P) yield a new wave [ Hum93] :
= Vi + ^ =
Aco
Ak
Ak
Aco
cos
t -
2
2 t
• sin
k +
2
x -
(0 +
2
t
In (2.9), if Aco = 0 and Ak = 0, a monochromatic wave results
of the form
\\i = 2 sin (kx-cot) (2.10)
Equation (2.10) illustrates the wave characteristics of
photons.

24
If Ago andAk are very large, the cosine part of (2.9)
modulates the amplitude of the wave, resulting in a string of
wave packets. If many waves are considered with frequencies
ranging between go and Ago, one wave packet results and the
photon can be depicted as a particle. A better intuitive
understanding of materials interactions with light can be
achieved by noting light's wave and particle characteristics.
The color of light is related to its energy and thus its
frequency or wavelength. As shown on the electromagnetic
spectrum in Figure 2.1 [ Hun87] , visible light falls in a
range of 380 nm to 760 nm in wavelength. White light is
comprised of all the visible wavelengths.
Light can be produced by heating objects to
incandescence, or by exciting atoms or molecules with other
forms of energy.
Incandescent sources are applied to produce light with a
wavelength energy distribution similar to daylight (Figure
2.2) [ Bil67] . When a material is heated to incandescence,
the increased vibration of its atoms results in kinetic
energy that is sufficient to excite electrons to higher
energy levels. Photons are released when the electrons drop
back to their normal energy levels. As atom vibrations
become more energetic, the frequency and energy of emitted

FREQUENCY IN CYCLES PER SECOND
10
24
10
22
10
20
10
18
10
16
10
14
10
12
10
10
—r
106 104 102 1
—1 1 1 1 1 1 1
GAMMA RAYS
X RAYS
HARD , SOFT
VACUUM U V.
HERTZIAN WAVES
✓N
ULTRAVIOLET
INFRARED
NEAR.
FAR
I I
I
I L__
LIGHT ~ __
VIOLET
BLUE GREEN YELLOW
J -L
RED
300 400
500 600
WAVELENGTH IN NANOMETERS
700
760
DIRECTIONAL
RADIO (RADARI
FM
TELEVISION
SHORT WAVE
BROADCAST
WAVELENGTH IN METERS
POWER
TRANSMISSION
A. I I l_
106 I08
Figure 2.1
Electromagnetic spectrum. [ Adapted from Hun87]

26
Figure 2.2. Wavelength vs. energy distribution of daylight.
[Adapted from BÍ167] (Note: Relative energy distribution
plots set the energy at 555 nm equal to 100, and the rest of
the curve is relative to the distribution radiated from the
source.)

27
light increases. The color of light produced changes from
red at low temperatures to nearly white at higher
temperatures.
The correlated color temperature of an incandescent
source is defined by the temperature at which a black body
would operate to produce a visual color match with the
incandescent source [ Hun87] . The color of a real black body
depends only on its temperature, not its composition.
According to Wien's displacement law [ Wea79] , the product of
the absolute temperature of a radiating black body and the
wavelength corresponding to the maximum energy is a constant:
T = W (2.11)
where W = Wien's displacement constant.
A common incandescent light source which operates at a
high enough temperature to emit a spectrum close to daylight
involves heating a tungsten filament enclosed in an evacuated
fused silica bulo. Tungsten filaments are close
approximations to black bodies. Some tungsten filament lamps
used for light sources in color measurement devices
incorporate glass filters to provide a more accurate match to
the daylight spectral energy distribution.
Light can also be generated from luminescence or outer
electron shell interactions in fluorescent and phosphorescent

28
materials [ Ask94] . In luminescence, kinetic heat energy is
not essential for the mechanism of excitation. Luminescence
occurs when light has sufficient energy to excite valence
band electrons through the energy gap and into the conduction
band. When the electrons eventually fall back to the valence
band, photons are emitted. If the photon energy corresponds
to wavelengths between 380 nm-760 nm, visible light is
produced. In fluorescent materials, photon wavelengths can
be calculated with
he
X = (2.12)
E Q
where Eg is the energy gap. Fluorescent lamps are
electrically excited to produce light. Normally with such
light, the spectral energy distribution is not as continuous
as with incandescent sources. Spectral lines are typically
narrow at specific wavelengths. In order to broaden the
spectral curve, sometimes more than one source is used in
combination. For example, in common household fluorescent
lamps, a spectrum from mercury vapor, electrically stimulated
inside the bulb, interacts with fluorescent powder on the
inside of the glass tube to generate a "cool white" light.
It is common to use from one to several light sources acting

o
z.
in combination as illuminates in color measurement devices
such as spectrophotometers [ BÍ167] .
2.1.2 Materials Interactions with Light
Properties of a material that influence its appearance
include index of refraction, gloss, translucency,
reflectivity, transparency, absorption and color. These
properties result primarily from the interaction between
light and a material's electronic structure and
microstructure. When photons interact with a material, they
are either attenuated (reflected or absorbed) or transmitted.
As photons enter a material, their speed will also change,
resulting in refraction. Light may be partially reflected,
absorbed, or transmitted as related to the incoming beam
intensity [ IQ] by [ Ask94]
I0 = Xr + !a + (2.13)
where Ir, Ia and It are portions of the incident beam that is
reflected, absorbed and transmitted, respectively. The
combination of wavelengths of light reflected from an
opaque material or reflected and/or transmitted from a
nonopaque material produce its color.

30
2.1.2a Refraction
The refraction of light by a material is estimated by
Snell's law f Che83] :
sin 9 n.
= — = n,, (2.14:
sin 9t ni
where
0¿ = angle of incidence from the line normal to the
irradiated surface.
0^ = angle of transmission from the same plane,
n = index of refraction of material,
n, = index of refraction of medium.
n91 - relative index of refraction of material to
medium.
Normally, n is used as the notation for index of refraction
if the medium is air or a vacuum. Snell's law can be derived
from Fermat's principle [ Tip80] :
The path taken bv light in traveling from one point
to another is such that the time of travel is a
minimum compared with nearby paths, (p. 612)
When light passes from air into a denser material, its
velocity decreases. Following Fermat's principle, light will
minimize its travel time by increasing its optical path
length in air relative to the path length in the denser
material. This results in a change in 0., 0, and n. Thus, n

31
is related to the velocity and wavelength of light, which is
given by
^"material
where u and ^material are the velocity and wavelength of light
passing through a material and ^air is the wavelength of
incident light in air.
At a constant wavelength of incoming light, n tends to
increase for denser materials. When comparing values for
ceramic vs. polymeric (teflon and polystyrene) materials in
Table 2.1 [ Wea7 9] , it becomes evident that the structure of a
material also influences n.
If a material is easily polarized, there are increased
interactions of its electronic structure with incident
photons. In dielectrics such as ceramics, the index of
refraction is related to the relative dielectric constant
(K' ) by [ Kin76]
n = ^K' + K: (2.16)
where is the index of absorption.
Polarization (P) of the electron cloud around an atomic
nucleus is proportional to the electric field strength (E) of
the incoming light:
P
N a E
(2.17)

32
Table 2.1. Index of Refraction of Selected Materials
at 589 nm Wavelength in Air [ Wea7 9] .
Material
Density
(g/cm3)
Mean Refractive Index
(n)
Water
1.0
1.33
Polystyrene
1.06
1.60
Si02 (glass)
1.41-1.46
1.46
Teflon
2.17
1.35
Silicon
2.33
3.49
Si02 (quartz)
2.64-2.66
1.55
CaCC3
2.93
1.60
Diamond
3.51
2.41
a A12Q3
3.97
1.77
Fe203 (hematite)
5.24
2.95

33
where a is the average dipole moment per unit field strength
or the polarizability and N is the number of material
particles per unit volume.
By the Lorentz-Lorentz equation, electronic polarization
is linked to the refractive index of a monatomic gas:
where £0 is the dielectric constant in a vacuum, No is
avagadro's number, and n„, is the molar refractivity. The
molar refractivity n„ is determined by measuring n at various
wavelengths of light as shown in Figure 2.3 and then
extrapolating to infinite wavelength.
Since the electron density is uniform within an atomic
radius (rQ) , polarizability is also related to the atomic
volume of a material:
a - (2.19)
Equation (2.19) shows that larger atoms, which contain more
electrons, exhibit greater polarizability and thus tend to
have a higher refractive index.
Ionic charge also plays an important role in influencing
the index of refraction. As the ionic charge becomes
increasingly negative, outer electrons are more loosely bound
and can increase polarizability. In addicion, the refractive
index is dependent upon crystal structure symmetry, except in

34
Wavelength (microns)
Figure 2.3. Refractive index vs. wavelength of incident
light for three glasses. [Adapted from Kin7 6]

35
glasses and cubic crystals which are isotropic. Crystals
have a higher index of refraction in denser, close-packed
directions. For anisotropic substances, the mean index of
refraction is estimated as [ Par73]
a + (3 + y (2.20)
n =
3
where a, (3 and y are refractive indices of the principle
crystallographic directions.
Multiphase crystalline and glassy substances have
specific refractive energies that are the sum of the
refractive energies of their components [ Par73] :
and
where
K
r
n
p
1 ^2
n - 1
K
+ k2
P;
100
+ . . .
(2.22)
= specific refractive energy of a
substance.
= mean index of refraction of a substance.
= density of a substance.
= specific refractive energies of the
components.
Pi» P? • - • - weight percentages of the components.

36
Reference values for specific refractive energies of oxides
can be used to estimate the refractive indices of glasses,
glazes and crystalline species.
2.1.2b Surface Reflection and Gloss
When a beam of light strikes a material, a portion of
the photons may be reflected. The light reflected at an
angle equal to the incident angle is referred to as specular
reflection (Figure 2.4). This "mirror-like" scattering
occurs at one angle from the point of reflection on a smooth,
nonmetallic surface. If the surface has some roughness, some
light may be scattered at all angles from the point of
reflection. This diffuse reflection is common in ceramics
where surfaces are not normally perfectly smooth. The total
reflectivity is the fraction of incident light specularly and
diffusely reflected. The gloss of a surface is related to
the relative amounts of specular and diffuse reflections. It
can be defined as the degree of approach to a mirror surface.
A perfect mirror surface has a maximum gloss and reflects all
visible light. This surface itself is invisible and has no
apparent microstructure.
Within the visible spectrum, the reflectivity (R) from a
perfectly smooth surface in an air medium is the fraction of

37
Figure 2.4. Reflection and transmission of light by a glassy
material containing suspended particles. [Adapted from
Kin7 6]

38
incident light reflected at an angle equal to the incident
angle. This fraction of specular reflectance from normal
incidence (0i=O) is calculated for the optical region of the
spectrum with Fresnel's formula [ Kin76] :
R =
n-1
n+1
¡2.23:
Equation (2.22) shows that materials with a high index of
refraction have a high reflectivity. Since the index of
refraction varies with the wavelength of light, so does R.
From Fresnel's law, equations can be written to compare
the reflectivity of plane polarized light so oriented to a
plane mirror surface that its reflection (ratio of reflected
to incident flux) is most facilitated (R.() or most hindered
(Ri) in an air medium [ Jud65] :
cos
0i - ~
sin" 0,
cos
0,+ *Jn~ - sin2 0,
(2.24:
and
Ri =
e, -
I 2 •> \
n‘ cos
/^n" - sin" 0¡
n“ cos
0j + *,
/n“ - sin" 0,
(2.25)
where 0¡ is the angle of incidence.
For unpolarized incident light, the total reflectivity
(Rqd is the average of R and R±:

39
rt= (R, + R±) / 2 (2.26)
Thus, the reflectivity, RT, for unpolarized light is the
average of reflectivities for plane polarized light in (R )
and perpendicular to (R_,_) the plane of the incident beam. If
0¡ = 0°, the incident beam is perpendicular to the surface and
perfect mirror specular reflection results. In this case,
equations (2.24) and (2.25) reduce to equation (2.23) (RT =
R) . If 0; = 90°, only grazing incidence occurs and R = R± =
Rp = 1. Between perpendicular and grazing incidence, R , R,
and R^ vary.
For a perfect mirror surface R = R± = RT = 1.0
regardless of 0i. For real colorant layers with a glossy
medium, mirror reflectance approaches unity with grazing
incidence as 0i —» 90°. In this case, the incident beam
contacts only that part of the colorant layer which is just
below the surface. This is the high-gloss transparent medium
carrying suspended pigment particles.
Figure 2.5(a) shows the relationship between R , R, and
Rt for 0^ from 5° to 90° for an air/glass boundary where the
refractive index of the glass is 1.5. Figure 2.5(b) plots
total reflectivity as a function of 0, and index of
refraction. Figure 2.5 can be used to estimate the gloss or

40
b)
Figure 2.5. Fresnel reflection for a) an air/glass boundary
and b) total reflectivity for different index of refraction
glasses.

41
degree of approach to a mirror surface for mediums such as
plastics, glass, textile fibers and paint vehicles [ Jud65] .
Note in Figure 2.5(a) that for a smooth glass with n =
1.5, Rj_ —> 0 at 0, = 56°. This follows Brewster's law, which
states that mirror reflectance is most facilitated {R_j_ =0)
when [ Jud65]
tan 0.; = n (2.27)
Thus, for a glass of n = 1.5 at incident and viewing angles
of approximately 56°, there is a lack of polarized diffusely
reflected light. Parallel polarized light in the reflected
beam is the only component present from the unpolarized
incident light. This principle has been applied to measure
gloss through the use of a polarizing element which subtracts
the specular component at Brewster's angle where R± = 0.
Brewster's law is also employed to measure the index of
refraction of smooth glass [ Fle93] . A light polarizer and
detector are used to measure the polarization effectiveness
of light reflecting from the glass surface. The angle of
incidence (0¡) where reflected polarization is most efficient:
is applied to calculate n with equation (2.27).
As the surface roughness of an opaque coating or a glass
matrix with suspended colorant particles increases, a greater
portion of the reflection becomes diffuse. This broadening

42
of the reflection band and lowering of the specular intensity
tends to lower the gloss. Gloss is greater for smoother,
higher index of refraction surfaces. Surface roughness in
ceramic glazes caused by crystal structures, defects,
interfaces, or a uneven application usually lowers the gloss
by increasing the amount of diffuse reflection.
The relative amounts of specular and diffuse reflection,
and the gloss of the material, have an effect on the color
revealed. For example, glossy paint will appear to lose its
color in daylight glare, but "flat" paint will look nearly
the same. Most color measurement devices take reflection and
gloss into account when generating values for color.
2.1.2c Opacity and Translucencv
Subsurface reflections can occur in materials that are
not completely opague. The reflectivity of opaque metals is
typically 0.9 to 0.95, where most clear glasses are closer to
0.05 [ Ask94] .
Optical characteristics of ceramic glazes, glasses and
enamels can be developed by modifying their internal
reflective and light-scattering properties. Characteristics
such as the portion of light specularly reflected (gloss),
the portion of light diffusely reflected (opacity), and the

43
portions of light directly and diffusely transmitted are
important (Figure 2.4). These attributes are influenced by
the light-scattering properties of phases or small opacifying
particles suspended in the glassy matrix. Good opacification
is obtained with the high light scattering and reflectivity
this mechanism provides.
Common opacifiers used in ceramics coatings include
zircon (Zr02*Si02), zirconia (Zr02), Sn02, TiCh and Al203.
They are selected based on their optical properties, the
desired coating composition, processing limitations and
properties that affect compatibility with the matrix phase
such as thermal expansion, solubility, hardness and melting
temperature.
Opacifiers can be inert with respect to the host matrix
phase, formed during melting, or crystallized from the glass
melt. Very fine particles can be obtained from materials
such as Ti02, or at higher firing temperatures ZrO-,, which can
be melted, nucleated and re-crystallized during the heating
cycle.
The degree of opacity of ceramics which contain
particles suspended in a glassy matrix depends on five main
factors [ Par73] :

44
1.
The difference in refractive index between the matrix
and dispersed particles. This effect is described by
Kerstan:
I, = I,
(n: ~ ni)~
(n2 + n^"
(2.28)
where
I = intensity of diffuse reflected light or intensity
of opacification.
I = incident beam intensity.
n2 = index of refraction of the glassy matrix.
n2 = index of refraction of the opacifier.
In ceramic coatings, this mechanism provides the
greatest influence on opacity. As the difference between the
refractive indices of the matrix and particle phases
increases, more light scattering and higher opacification is
achieved. Generally, glazes and enamels have refractive
indices ranging from 1.50 to 1.70 while opacifiers range from
2.0 to 2.8 (Table 2.2) [ Ree83] .
2. The size of dispersed particles which scatter light.
The light-scattering ability of an optically
heterogeneous material can be estimated with the scattering
or turbidity coefficient (S) . This coefficient is a

45
Table 2.2. Properties of Materials Used for Opacifying
Ceramic Glazes [ Par73] .
Opacifier
Material
Mean Index of
Refraction
(n)
Difference in n:
Opacifier-Glaze
Melting
Temp.
(C)
Coefficient
of Expansion
X107/°C
eg
O
•r\
Eh
2.50
0.95
1,830
88
Zr02
2.40
0.85
2,715
70
Zr02-Si02
2.05
0.50
2,430
41
Sn02
2.04
0.49
1,625
41
Air
1.00
0.55
—
—
Glaze
1.50-
1.70
0
—
—

46
measurement of the attenuation due to scattering of light as
it traverses a medium containing scattering particles
[ Kin7 6] :
3 -i
s = J SfVpr (2.29)
and
where
exp
-3SfVp
exp ( - Sx)
(2.30)
Sf = scattering factor that varies between 0 and 4.
V = volume fraction of scattering particles,
r = radius of the scattering particle.
Ia/I = ratio of light intensity scattered/initial
intensity.
x = oprical path length.
Equation (2.29) shows that scattering, and thus opacity,
tends to increase up to a point with decreasing particle
size. The scattering constant (Sf) increases with particle
size (r) and is inversely proportional to the fourth power of
wavelength for particle sizes much smaller than the
wavelength of incident light. When the particle size is
approximately equal to the wavelength of light, Sf reaches its
maximum value, then decreases with increasing particle size.
Thus, maximum scattering occurs when the opacifier has a

47
particle size similar to the wavelength of light used, which
is in the range of 0.38 to 0.76 microns in diameter for
visible light.
3. The number of reflecting particles per unit volume.
Equations (2.29) and (2.30) indicate that opacity
increases with the number of particles. As the concentration
of opacifier increases, the rate of increase of opacity
decreases. Glazes and enamels typically utilize a maximum of
17% zircon by 'weight for opacification. Approximately 3% to
4% becomes an intermediate part of the amorphous glass
structure, while the rest serves as light-scattering
particles.
4. Higher opacification is obtained when there is a
distinct boundary and steep concentration gradient
between the matrix and particle phases.
Diffusion and the consequent reduction in concentration
gradient between the matrix and particles results in lower
opacification. Opacifier/glass systems with relatively low
diffusion coefficients at the required processing
temperatures are most effective. Crystals precipitated in a

48
glass during the heating or cooling cycle tend to have sharp
interface boundaries.
5. The thickness of a coating applied to a substrate.
Coating thickness effects on covering power can be
related with the Kubelka and Munk equations [ Kin76] :
Re =
1
i V
R.) ~ R.
R'
R_
oo
exp
Sx
R.
lR- JJ
(R' -R~)“
( 1
1 'll
R'
exp
Sx
-—R«
. R” .
Roc
00 / J
and
K
R- = 1 + ¥
+
2K
S
/2
(2.31)
(2.32)
where
Rr, = coating reflectance.
R' = substrate reflectance.
R„ = total reflectivity of a colorant layer so thick
that further increases in layer thickness do not
change the reflectivity.
S = scattering coefficient
K = absorption coefficient,
x = coating thickness.

49
Equation (2.31) shows that coating reflectance increases
up to a point with thicker applications and higher reflection
substrates. Figure 2.6 demonstrates the increase in
reflectance with application thickness of a typical TiCp
opacified white glaze fired to 1000°C. The main
disadvantages of a thicker coating is the increased material
requirements and cost.
Equations (2.31) and (2.32) indicate that opacifiers
with a high scattering coefficient (S) and low absorption
coefficient (E) are most powerful. The quantity Sx is often
applied to estimate the scattering power of a coating.
Another method gauges opacity with the ratio of reflectance
obtained from a coating over a black (R' = 0) versus a white
(R' = C.89) backing. This is the Tappi Opacity Method for
determining the contrast ratio C0 89 = R'0/R'r, 8g.
Liquid-liquid phase separation during glass formation is
another method by which glaze opacification can occur. For
example, the Li20 - Si02 - Ti02 system separates at high
temperature into a low refractive index silicate glass and a
high index titania rich glass [ Kim59] . One disadvantage of
this mechanism is that it is very sensitive to processing
temperature. Therefore, in current manufacturing settings
where heating variations during firing are common, the

Wavalength, millimicron*
Figure 2.6. Reflectance vs. wavelength of light for a TiO
opacified white glaze fired to 1000°C, at various glaze
application weights in g/ft2. [Adapted from Par73]

51
relative proportions of glassy phases and opacification would
be inconsistent and cause product appearance variations.
If translucent rather than opacified appearances are
desired, particles in the glassy phase must create diffuse
transmission. Translucency is important for products such as
opal glass, where opaque substrates are not utilized. It is
most common to achieve translucency by dispersing a
particulate phase with a slightly different index of
refraction than the glassy matrix. Translucency is also
often controlled with porosity, where lower pore
concentrations (higher material density) or higher pore size
at a given concentration increase translucency.
2.1.2d Absorption, Transmission and Color
Light that is not reflected or transmitted by a material
is absorbed. The linear absorption coefficient (K) indicates
the portion of normally incident radiant energy absorbed
through a unit distance (x) in a single phase material by
[Wea79]
T = — = exp(-Kx) (2.33)
where
T = fraction of light transmitted as it passes through
a material.

52
I /Ii = transmitted intensity of light/initial intensity
of beam after reflection.
In the Raleigh scattering mechanism of absorption,
photons are deflected from electrons orbiting an atom without
any change in energy. This mechanism is more common for high
atomic number atoms and low photon energies [Ask94], If an
electron is ejected from an atom, consuming some of the
photon energy, this is referred to as Compton scattering.
Resonance occurs when the frequency of material oscillations
is close to the frequency of incident radiation, which
results in the absorption of radiation.
If incident light stimulates electrons to change their
energy level, the photons are absorbed and the material is
opaque to this particular wavelength of light. Because there
is no energy gap in metals, electron movement into higher
energy levels occurs at almost any photon energy. Therefore,
metals have a high absorption coefficient and are opaque to
most wavelengths of electromagnetic radiation.
The energy gap in semiconductors is greater than metals
and smaller than insulators. Semiconductors with small band
gaps can transmit: photons with energies below the energy gap
Eg or become opaque and absorb photons of higher energy. For

53
example, at 300 K, Si has a gap energy of 1.12 eV, while
diamond has a gap energy of 5.47 eV [Hum93]. Therefore,
silicon requires less energy for electron transitions and
appears opaque in daylight, while pure diamond is
transparent. In the visible spectrum, Eg > 3.1 eV materials
do not absorb any photons, where Eg < 1.8 eV materials absorb
all visible light [Ask94]. For intermediate energy gaps, a
fraction of the incident visible light is absorbed.
In ionic ceramics, filled shells of tightly bound
electrons do not have energy levels available for electron
movement [Ric92], and most single crystals are transparent.
Covalent ceramics, however, vary in the level of absorption.
For example, diamond and graphite both have covalently bonded
carbon atoms, but their optical properties are significantly
different. Diamond is transparent while graphite appears
black. Although there is a strong covalent bonding within
the graphite hexagonal network, weak Vander Waal's bonding
between the layers allows for electron movement. This
results in electron transitions and absorption of visible
light. Good insulators with a large Eg such as diamond tend
to transmit light.
Absorption due to electron transitions and resonance is
intrinsic, while extrinsic effects in ceramics can also cause

54
absorption and color. Extrinsic effects include grain
boundaries, pores, inclusions, anisotropy and atom vacancies.
In ceramics, the absorption coefficient (K) is related
to the index of absorption (ki) (also referred to as the
attenuation index or extinction coefficient) by [Kin76]
K = (2.34)
From equations (2.15), (2.16) and (2.34), K increases with ki
and the index of refraction of a ceramic material and
decreases with higher wavelengths of incident light.
The overall fraction of light transmitted (T1) after
both reflectance and absorption losses is
T
T' = — = ( (1 — R) ‘ exp (-Kx))
± C
(2.35)
where R is the reflectivity and It/I0 is the ratio of
transmitted to incident light intensities. By equations
(2.13), (2.23),
(2.28) and (2.35),
ail of the incoming
light
can be accounted
for by reflection,
absorption and
transmission.
Total light
interaction with a
material can thus
be
written [Ask94]
Irf = IcR
(2.36
la = (
X-Irf) - [Io d-R)
exp ( - Kx} ]
(2.37
Irb = I0R (1 — R)
exp (- Kx)
(2.38

55
I
t
10 (1 — R) 2 exp(-Kx)
(2.39)
I
O
(2.40)
where
I f = intensity of light reflected at the incident
surface.
I = intensity of light absorbed by the material.
a
I b = intensity of light reflected at the back face.
Color is produced in many materials through selective
scattering and absorption of incident light. This
selectivity often results from variations in the absorption
coefficient with wavelength. Four electron transitions
concurrent with this type of absorption are common causes of
color [Ric92]:
1. Internal transitions with rare-earth or transition
metals or other ions with incomplete inner electron
shells.
2. Charge transfer, where electrons are transferred
from one ion to another.
3. Electronic transition caused by crystal
imperfections.
4. Bad gap transitions found in many semiconducting
compounds, as discussed earlier in this section.

56
Transitions (1), (2) and (3) usually are caused by impurities
or defects in a material, while (4) is a bulk property.
Often, particles are suspended in a matrix such as a glass to
create electronic transitions and color. The absorption
coefficient is proportional to the concentration of absorbing
ion (c), according to Beer's law [Ree83]:
T = exp (-ecx) (2.41)
and
K = ec (2.42)
where £ is the extinction coefficient observed per unit
concentration. This is the fundamental law of simple
subtractive colorant mixing.
The most commonly used colorant ions are from transition
metal compounds or impurities such as V, Cr, Mn, Fe, Co, Ni
and Cu shown in Table 2.3 [ Ree83] . They provide color in
many ceramic bodies, glazes, glasses, minerals, gems,
pigments and paints.
Crystal or ligand field theories describe how these
elements produce color [ Pet72] . Transition metals have
unfilled inner orbitals available for the creation of split
energy levels for electronic transitions. In "free" ions,
orbitals have equal energies but different spatial
orientations, as shown in Figure 2.7 for the five d orbitals.
But the coordination of negatively charged anions about the

57
Figure 2.7. The 5 d orbitals. [Adapted from Kin76]

Table 2.3. Transition Elements and Their Properties [ Pet72].
V
Cr
Mn
Fe
Co
Ni
Cu
Atomic number
23
24
25
26
27
28
29
Atomic Radius,
Angstroms
1.31
1.25
1.37
1.24
1.25
1.25
1.28
Electronic
configuration1
Ionization
energies2
3d34s2
3d54sJ
3d54s2
3d6 4 s2
3d7 4 s2
3d84s2
3d104 s1
First
155
156
171
182
181
176
Second
338
334
361
373
393
418
Third
676
713
777
706
772
810
Oxidation
potential3
+ 1.2
+ 0.91
+ 1.18
+ 0.44
+ 0.28
+ 0.25
-0.34
Oxidation
States4
2,3,4,5
2,3,6
2,3,4,7
2,3
2,3
2
1,2
Melting point,
°C
1710
1930
1220
1535
1495
1455
1083
Density, g/cc
5.96
7.20
7.20
7.86
8.9
8.90
8.92
Hardness5
—
9.0
5.0
4.5
—
—
2.5-3.0
Electrical
conductivity6
—
62
32
16
17
24
96
1Each atom has an argon inner core configuration.
2Values are in kcal/mole.
3For the oxidation process: M(s) = M2+(ag) + 2e~.
4Common oxidation states; the most stable one is italic.
5Hardness values are on the Mohs scale.
6Compared to an arbitrarily assigned value of 100 for silver.

59
metal cation produces an electrostatic field that raises
inner orbital energies nonuniformly. This electrostatic
interaction between anions and metal ion's electron clouds
splits inner d or f orbitals into different energy levels.
The energy and corresponding wavelength of light absorbed by
the metal which produces color is egual to the difference in
the split energy levels.
For example, in a tetrahedral structure surrounding a
metal ion with unfilled d orbitals, the dxy, dxz and d.,_
orbitals have more energy than the d 2 2 and d„2 orbitals.
x y z
Color-producing transitions are allowed between these two
split groups, and the wavelengths of light absorbed depend
upon the magnitude of the splitting. Thus, only a limited
range of colors can be produced by any given ion.
The oxidation state of the metal also has an influence
on the magnitude of splitting and resulting spectral
properties. For example, Cu+ is colorless in solution while
Cu+2 has a strong blue color [ Pot67] . When the valence of a
given element increases (e.g., smaller d occupancy), so does
the strength of the ligand field.
This section reviewed the most common causes of color in
materials and the method by which color will be derived
during the subsequent research involved with this

60
dissertation. A comprehensive list of all of the possible
causes of color are listed in Appendix B.
2.1.3 Color Perception bv the Human Eve
The eye is the human optical system (Figure 2.8) [ Tip80]
that allows us to perceive the color, gloss, opacity and
dimensions of an object. The eye is sensitive to light
between wavelengths of 400 and 700 nanometers.
Light enters the eye through the pupil and is focused by
the cornea-lens system on the retina. As the distance of an
object from the eye varies, the ciliary muscle changes the
lens shape to improve focus of the image on the retina.
The photosensitive parts of the eye are the rods and
cones of the retina. These tiny structures receive images
and transmit information along the optic nerve to the brain.
The size of an image on the retina increases with the number
of rods and cones activated, which is proportional to the
apparent size of the object being viewed.
Rods respond to very small amounts of radiant energy
and, thus, serve for night vision. Rods do not detect hue or
chromatic colors but only perceive neutral colors such as
white, gray and black.

Anterior
chamber
Central retinal
artery and vein
Figure 2.8. Human optical system. [Adapted from Tip80]

62
On the other hand, cones sense chromatic as well as
neutral colors. Cones, which are responsible for day vision,
can detect a much higher density of radiant flux than rods
but are less sensitive at very low levels of light. Rods
respond to minute quantities of light as low as 10-6 candelas
per square merer (cd/'m2) , while cones require at least 1CT3
cd/m2 [ Jud65] .
At illumination commonly referred to as twilight, both
rods and cones are active. The approximate range of
luminances which correspond to twilight or the mesopic region
is from 1CT3 cd/m2 to 10 cd/m2 [ Jud65] . In this range, color
judgments are extremely unreliable because the relative
degree of rod and cone vision continually changes.
Therefore, color inspections in manufacturing should not be
carried our in this condition, although many factories and
industrial inspection areas are dimly lit. Luminance at
approximately 10e cd/m2 is the maximum level where the human
eye can perceive color with cone vision.
The eye is not equally sensitive to all wavelengths of
light. It has been demonstrated that 555 nm light is viewed
more easily than other wavelengths. Figure 2.9 graphically
shows the sensitivity or relative response of the human eye
at daytime (cone or photopic curve) and nighttime (rod or

RELATIVE RESPONSE
63
400 5 00 6 0 0 700
WAVELENGTH nm
Figure 2.9. Luminosity functions of the rods (nighttime
scotopic vision) and cones (daytime photopic vision) of the
human eye. [Adapted from Hun87]

64
scotopic curve) for the same amount of energy at different
wavelengths of light in the visible spectrum. The curves
were developed from experiments where 52 human observers
adjusted the intensity of light at different wavelengths
until they appeared equally luminous or bright [ Hun87] . The
property of light by which we define how easily we can see it
is referred to as luminosity.
Both the Young-Helmholtz and Hering experiments of the
mid-1800s confirmed that human observers see colors with
three spectrally unique receptors which detect black-white,
red-green and yellow-blue. Subsequent numeric scales
developed for quantifying color contain three values; one for
each opponent-color pair.
Color sensory responses of the eye are not linear with
the amount of stimulus. For example, there is a logarithmic
relationship between the actual light level reflected from an
object and its perceived lightness. Color perception is also
a function of the light source and directions of illumination
and view. For these reasons standardized observation
conditions are required for industrial inspection. Even so,
differences between individuals are great enough to affect
visual color quality control. Most often, human observations
are coupled with color measurements performed by machines

65
such as spectrophotometers in order to make final
determinations.
2.1.4 Color Measurement
Color measurement, like human eye perception, depends on
the light source, the sample being viewed and the observing
apparatus. Properly operated color measurement equipment,
however, can provide more repeatable results than subjective
human perception. Through a series of calculations, measured
colorimetric data can be converted into values that better
relate to human vision.
2.1.4a Spectrophotometry
The spectral characteristics of an object determine its
perceived color. Spectral characteristics are defined by the
reflectance or transmittance of light from a material as a
function of its wavelength. Spectrophotometers are used to
measure reflectance or transmittance from a sample as a
percentage of incident light at each wavelength in the
visible spectrum, normally in 0.5 nm increments [ Mac91a] .
Typically, reflectance is measured for opaque materials and
transmittance for transparent materials where the color of
light after it passes through a material is important.

66
Figure 2.10 reveals reflectance curves for opaque
coatings colored with pigments which absorb a portion of the
incident light [ Hun87] . The white coating reflects a high
portion of the incident light across the whole visible
spectrum while the black coating absorbs most of the light
flux over the wavelength range. Colors of blue, green,
yellow and red are created through selective absorption and
reflection of different light wavelengths by the pigment.
For example, the blue coating is shown to absorb primarily
yellow to red light (550-700 nm) while reflecting blue (450—
550 nm). The color of a ceramic whiteware glaze coating
results from this mechanism. Glazes are applied over an
opaque white substrate, and the reflectance curve produced is
the sum of reflectances from the pigment particles, other
crystalline species present, and the white substrate minus
specific wavelengths of light absorbed by the pigment
colorant. Normally, the background substrate and undoped
pigment crystal strongly reflect all visible wavelengths and
appear white or light yellow without the light absorbing
metal ion incorporated in the pigment. In contrast, the
color of light transmitted through a nonopaque glass results
from the incident beam minus both light absorbed by the
structure and reflected from the irradiated side.

WAVELENGTH. NANOMETERS
Figure 2.10. Reflectance versus wavelength for opaque
coatings colored with pigments that absorb a portion of
incident light. [Adapted from Hun87] .

68
The basic components of spectrophotometers are outlined
in Figure 2.11. Reflectance factors are measured one
wavelength at a time, normally at 0.5 nm increments, by
isolating wavelengths with gratings, prisms or interference
filters and slits. This is the monochromator device in
Figure 2.11 [ Hun87] . Current spectrophotometers are similar
to original mechanism developed by A. C. Hardy in 1928, shown
in Figure 2.12. The position of mirror slit #2 is adjusted
for wavelength isolation.
Typical light sources include a tungsten filament lamp
or a pulsed xenon bulb, which, in conjunction with a prism,
produce white light. Illumination is normally near 10J cd/m2
where only cone vision occurs. For accurate color
comparisons, the relative energy versus wavelength
distribution from the source must exactly match the desired
standard, but the total energy or illuminance from the source
can range from 10 cd/m2 to 106 cd/m2 where the rods are
inactive.
Real light sources are difficult to standardize, and it
is often useful to compare the color of objects viewed under
various wavelength energy distributions. Normally
reflectance values from the real light source are

69
Figure 2.11. Basic components of spectrophotometers.
[ Adapted from Hun87]

70
PHOTOMETER
A
Anti-hunt
Figure 2.12. Schematic of the Hardy spectrophotometer.
[ Adapted from BÍ167]

71
mathematically converted to represent theoretical sources or
illuminants before color values are derived. Outputs from
spectrophotometric measurements include color values derived
from relative energy distributions of standard light sources
such as D65 (average noon daylight from the total sky) ,
illuminant A (incandescent lamp), illuminant B (near
sunlight) and illuminant C (average daylight from the total
sky) .
The standard full visual field of view utilized to
detect light reflecting from an object is 2° angular
subtense. Occasionally, a 10° observer is used to provide a
larger field of view.
Light flux reflected from a sample is collected for
measurement by a white-lined integrating sphere. Elimination
of surface gloss from the color measurement provides results
which better correlate to visual inspection. This can be
accomplished by replacing the white plug on the sphere's
specular cup with a black plug. Since the specular cup is
offset to be illuminated at an angle of reflection equal to
the angle of the incident beam, the black plug absorbs
primarily specularly reflected light. Diffusely reflected
flux is diverted up the sphere to the photodetector. The
collected photon energy distribution is converted into an

72
electrical signal and sent to a computer. The computer
program converts measured spectral data into numbers that can
be more easily interpreted and correlated to the response of
the human eye.
Measurements of gloss can be performed separately with
goniophotometers or gloss meters, which measure the spectral
reflectance or quantity of light emitted in directions
related to the surface characteristics of the object. The
gloss of ceramic coatings is normally measured at a 60° angle
of incidence, where mirror reflectance is most facilitated,
according to Brewster's law (equation 2.27).
2-..1-, 4 b Basis for Color Quantification
The average sensitivity of the human eye to each
wavelength of light has been determined through extensive
experimentation [ BÍ167, Hun87, Jud65, MacA35, MacA42] . Human
observers were asked to visually match the colors of light
from individual wavelengths by mixing together lights from
three colored primaries. Three primaries were applied
because the eye contains three spectrally unique receptors
for detecting colors. The amounts of energy of each of the
three lights required to match single wavelength colors were
used to develop standard observer functions for the basis of

73
all color measurement. These weighting functions are applied
to transform spectrophotometric data into numbers that better
correlate to the way the human eye perceives colors.
The weighting functions ( x , y and z ) are plotted in
Figure 2.13. Mathematical functions for describing colors
obtained by mixing different sets of primary colors have been
shown to always be related by a set of linear transformations
[ Hun87] . Therefore, there was some flexibility for selecting
the three standard primaries which provide x , y and z
the most user-friendly set of weighting functions. The
curves in Figure 2.13 were derived with the following useful
properties:
1. One of the functions, the y curve, was made to equal
the photopic plot shown previously in Figure 2.9, which
indicates the eye's response to luminosity or color
brightness.
2. The areas underneath the three curves were made equal
for light of equal energy at all wavelengths.
3. One function was selected to be as near to zero as
possible for as much of the spectrum as possible.
In this form x , y and z do not represent any real
colored primaries but can be converted to values that are

RELATIVE AMOUNTS
400 500 600 700
WAVELENGTH. NM
Figure 2.13. Weighting functions used for the standard
observer at a 2° field of view. [ Adapted from Bil67]

75
easier to apply. The weighting functions are used to
transform spectral reflectance curves into three numbers
referred to as tristimulus values, X, Y and Z. These values
specify color in terms of the mixture of red (X), green (Y)
and blue (Z) primary light that would produce the same color.
The Y value also still includes brightness detected. If two
materials are found to have the same measured X, Y and Z,
they will appear to have the same color under that specific
viewing condition.
At any one wavelength, X = x , Y = y and Z = z . For
example, at a wavelength of 450 nm in Figure 2.13, the light
detected would consist of proportions 32 : 5 : 175 of red ( x
or X) : green (y or Y) : blue (z or Z) light. For all
wavelengths in the visible spectrum, the contribution of each
tristimulus value can be calculated with [ Hun87]
X = x ^ + Sx x x + Xi + ... + S x ^ (2.43)
1 * 22 33 nr.
or
70C
X = J Sxx.dX (2.44)
400
where
xx = weighting function ( x) value at X wavelength.
3x= spectral energy at X wavelength.

76
Y ( y ) and Z ( z ) can be calculated in the same manner as X
( x) with (2.43) and 2.44). The spectral energy, in turn, is
a function of the properties of the light source for an
illuminate or aperture color (S^source) or the reflectance from
a reflecting object (SWerial) :
and
S = E
^Xsource
(2.45)
where
“^material
(2.46)
E^_ = energy of the light source at A wavelength.
= percent reflectance of light of A wavelength from
the material.
Since objects are viewed in relation to their surroundings,
X, Y and Z are normally expressed relative to the luminosity
of a perfect white opaque material where R = 100, as
X
Y
100
100
100
Í
ExR*x,dA
J
Ex y*dA
J
ExRx Y^dA
J
E„ yxdA
J
E*.R)l 2 x.dA
(2.47)
(2.48)
Z
J E, yxdA
(2.49)

77
Thus, for a perfectly white material, Y is 100 and X and Z
vary depending upon the light source.
Trichromatic coefficients
(x, y, z)
are often calculated
from tristimulus values:
x =
X
(2.50)
X
+
Y + Z
y =
Y
(2.51)
X
+
Y + Z
Z
(2.52)
Z —
X
+
Y + Z
where x+v+z=1.0.
The
X
and y are
coefficients used to
indicate chromaticity or color, while tristimulus Y is
normally kept to represent luminosity. MacAdam in 1935
[ MacA35] proposed the first color measurement space with Y,
x, y cartesian coordinates.
Most current color measurement systems apply tristimulus
values rather than trichromatic coefficients. Tristimulus
values are further converted to allow for easier
interpretation of color in three-dimensional black-white,
red-green and yellow-blue space. Since the tristimulus
system was sanctioned by the International Commission on
Illumination or CIE in 1931, there have been over 30 three-
dimensional color spaces developed through transformations of
X, Y and Z values. The best scales provide an approximately

78
uniform color space where equal distances within the space
represent nearly equal visual color differences. The
current most commonly used color space in the world is the
CIE L*a*b* scale which was developed in 1976 (Mac96] .
2.1.4c CIE L* a* b* Color Measurement Scale
Tristimulus data are converted into scales which, based
on visual discrimination experiments, correlate to perceived
color differences. Approximately uniform scales have been
developed where differences between measured colors
throughout the color space are proportional to visual
distinotion.
Human eye sensitivity for detecting color differences
varies across the visible spectrum of wavelengths observed.
Visual color discrimination is greatest near 485 nm and 590
nm and least around 425 nm and 650 nm. This is represented
in the MacAdam color limits (Y, x, y) diagram in Figure
2.14). The greater the distance between two wavelengths on
the perimeter of the plot, the greater the range of colors
that can be perceived in the interval. The third dimension
indicated in the graph is the Y-value from 0 through 95.
Note as the Y-value or lightness increases, the potential

79
Figure 2.14. Luminosity or lightness (Y) and chromaticity
(x, y) MacAdam limits for colors viewed in daylight.
[ Adapted from Bil67]

80
range of colors that can be perceived decreases. This
horseshoe-shaped plot exhibits nonuniform color space.
The CIE L*a*b* scale provides for more equivalent visual
distances and easier interpretation than the (Y, x, y) scale
[ Jud65, Hun87] . In this uniform system, cartesian
coordinates are employed to represent three-dimensional color
space as the eye sees color with three spectrally unique
receptors. The space is defined by a set of equations which
incorporate tristimulus values from equations 2.47, 2.48 and
2.49 [ Hun87] :
L* = 116 (Y / Yn)1/3
- 16
(2.53)
a* = 500 [(X/Xn)1/3 -
(Y/Y„)I/3]
(2.54)
b* = 200 [(Y / Yn)1/3 -
(z/z„)‘'j
(2.55)
(with constraints of Z/Xn, Y/Yn, Z/Zn > 0.01)
where
L* = Lightness dimension (Orblack to 100:white).
a* = redness (positive a*) or greenness (negative a*),
b* = yellowness (positive b*) or blueness (negative b*).
Xn, Yn and Zn = Tristimulus values of a standard
reference white object.
The Xn and Zn values vary depending on the illuminant and
observer angle, as shown in Table 2.4. Cubed root functions

81
Table 2.4. Factors for Uniform Color Scales for Normalizing
to a Standard Reference White.
2°
Observer
1
H»
o
0
Observer
Illuminants
Xn
Zn
Xn
D65
95.021
108.849
94.825
107.399
Cool white
fluorescent bulb
98.166
68.073
102.158
69.623
Westinghouse
Ultralume
108.354
34.352
111.350
35.352
= 100.00 in all cases for standard white.

82
in equations (2.53) through (2.55) were found to best equate
the logarithmic relationship between actual object lightness
(Y) and human perception [ San57] as discussed in Section
2.1.3.
Figure 2.15 demonstrates L*a*b* color coordinates.
Values for a* and b* approach zero for neutral colors such as
whites, grays and blacks. The higher the a* and b* values,
the more saturated the color. For example in ceramic glazes,
L*a*b* values can be applied to specify colors such as
Glaze Color
L*
a*
b*
Light Gray
83.7
-0.5
0.5
Dark Gray
57.0
0.0
0.5
Cr-Sn Red
29.9
22.9
4.8
Co-Cr Green
44.1
-18.0
-9.8
Zr-Pr Yellow
86.9
-8.5
21.5
Co-Si Blue
52.5
6.2
-28.1
The measurement
of
color difference
between
two objects using
CIE L*a*b* uniform
color space is calculated
from
AE' » [(L,
- L* 2
f + (a',-a%)2 +
(b* i “ b* 2
)J (2.56)
where
AE*
color difference as
defined
by the magnitude
of the position vector between two points in
color space, expressed in Judds units.

83
L=0
black
¿*=ii6-(y/r0)l/3-i6
á*= 5oo-[(X'A'0)'/,-<7/r0),/’]
f?*=200-[(V/Yoy,3-(Z/Zn)m]
Figure 2.15. Schematic of L*a*b* color space. [Adapted from
Mor96]

84
L*i,L*2 = lightness units of objects 1 and 2.
a* i,a* 2 = red-green units of objects 1 and 2.
b*i,b*2 = yellow-blue units of objects 1 and 2.
Experiments by MacAdam [ MacA42] with human observers
ascertained that a AE* of > 1.0 Judds represents a noticeable
color difference, regardless of the color in question. Since
AE* represents the magnitude but not the direction of color
difference, industry often compares individual values of AL* ,
Aa* and Ab* between two colors. This helps to uncover the
best approach for adjusting material formulations to modify
colors. Normally AE* calculations are confirmed with a
visual assessment.
It is customary to specify the color of whiteware glazes
using CIE L*a*b* values [ Epp96] . Ceramic tile manufacturers
follow American National Standard Institute (ANSI) A137.1
specifications, which stipulate American Standard Test Method
(ASTM) C609-90 for measuring small color differences between
ceramic tiles [ Mur89, Aza97] . ANSI specifies a tolerable
color difference of 3.0 Judds for compatible glaze
appearances, even though the difference is noticeable by
human observers.
Good results of color specification and comparison are
achieved from the L*a*b* system in part because the standard

85
observer weighting functions and response curves are
approximately linear over a short segment.
On the other hand, the L*a*b* system can provide
anomalous results when relating green and red colors to
pigment concentration [ Epp96] . The a* is a function of both
X( x) and Y(y) tristimulus values and corresponding
weighting functions (equation 2.54). These curves overlap
significantly (Figure 2.13), especially in red (650-680 nm)
and green (530-550 nm) wavelength regions. There is also
nonlinear correlation between X( x) and Y(y), and a* is
heavily weighed by both functions, especially in the red
wavelength region. Thus, red and green color is a complex
function of both L* and a*. Cases result where as red or
green pigment concentration is increased, the L*-value drops
but the a*-value does not change. This can be misleading
when evaluating color strength due to the pigment. The
Kubelka-Munk absorption function, described in Section
2.2.4b, is often used to provide a better quantitative
correspondence for evaluating red or green colorants. The
absorption factor (K/S) normally has a linear dependency on
pigment concentration.
Conversely, the b* value is effective for evaluating
blue or yellow color intensity as a function of pigment

86
concentration. There is very little overlapping of y and z
weighting functions (Figure 2.13) in the blue (450-500 nm)
and yellow (550-600 nm) light regions. In the small
overlapped area, there is a strong inverse negative
correlation and b* is weighed heavily by either z or y , but
not both. Thus, the relative influence of blue and yellow
pigments on color intensity in terms of b* and lightness L*
can be easily separated and correlated to visual
interpretation.
2,2 Color in Ceramic Glazes
Glazes are thin glassy layers formed on the surface of
ceramic products by firing an applied coating. They are
utilized mainly to improve strength and chemical durability
and provide a readily cleanable decorative surface for many
products including tile, sanitary ware, dinnerware and
electrical porcelain. Glazes are formulated to achieve
desired mechanical and optical properties of the final
product through careful control of processing parameters.
Since ceramic whitewares glazes are normally applied over an
opaque white substrate, the color produced is a function of
the reflectance of the incoming beam from the pigment and
other crystalline species in the glass matrix, plus any

87
reflectance from the substrate, minus absorption from the
metal ion on the doped pigment structure. Ceramic glazes are
usually opaque; therefore, reflection by the substrate is
insignificant.
The glaze composition and behavior during processing
determines the color developed by dissolved metallic atoms or
suspended inert pigments. Glass is the major phase present
in glazes; therefore, glazes are classified as amorphous or
glassy in their physical and chemical nature. Thus, current
glass science theories should apply in defining the
mechanisms involved with their color development.
The research of this dissertation focuses on the pigment
system most widely used in the ceramic tile industry, the
doped zircon structure.
2,2.1 Silicate Glass Structures and Properties
A glass is an amorphous solid, where no long range order
of molecular constituents is observed [ Dor94] . Glass is
usually formed by rapidly cooling a melt and solidifying
materials from the supercooled liquid. The cooling rate must
be high enough to avoid the nucleation and growth of
crystals. Nucleation and crystallization rates are dependent
upon glass composition, viscosity, number of heterogeneous

88
nucleation sites and temperature. These variables will be
further discussed in Section 2.2.5.
Silica is the predominant glass former used in
industry. Knowledge of the properties of silicate glasses
is essential for formulating glazes and enamels. The
silicon-oxygen tetrahedron is the basic structural unit in
all silicate glasses. The tetrahedra are linked to each
other at the corners to form a three-dimensional random
network. This structure has no long-range symmetry or
periodicity beyond a few tetrahedra [ Ban86] . Directional
covalent and ionic bonding in accordance with Pauling's
rules promotes the formation of (Si04)~4 structural units
[ Chi97] .
Short range order is identical in crystalline and glassy
Si02, where 4-fold coordination and a O-Si-O bond angle of
109.5° with each tetrahedral unit is present. But
connectivity is maintained without crystalline long term
order.
Differences in the medium range order of crystalline and
amorphous Si02 include Si-O-Si bond angles and rotational
angles between tetrahedra, and the number of Si-0 bonds that
complete each of the rings. These parameters vary throughout

89
the glass but are fixed with crystalline Si02. Figure 2.16
compares crystalline and amorphous phases.
Silicate glasses follow Zachariasen1s four rules for the
formation of oxide glasses based on a random network
[ Kin7 6] :
1. Each oxygen ion should be linked to two or less cations.
2. The coordination number of oxygen ions about the central
cation is not more than four.
3. Oxygen polyhedra share corners only.
4. For each polyhedron, at least three corners should be
shared.
Cations which tend to form oxygen in polyhedra of
triangles or tetrahedra are glass formers. In multi-oxide
silicate glasses, other constituent ions refill the voids
left by silicon and oxygen. These network modifiers, such as
the alkalis, provide additional oxygen and occupy random
positions to furnish local charge neutrality. Modifiers tend
to break up the network structure and often cause the glass
to crystallize or devitrify. Intermediates are cations of
higher valence and lower coordination number than modifiers.
They do not form a glass by themselves but may be
incorporated in part into the network structure.

90
CRYSTAL STRUCTURE
VITREOUS STRUCTURE
SPECTRUM OF A CRYSTALLIZED BODY
SPECTRUM OF AN AMORPHOUS BODY
Figure 2.16. Comparison of structures and XRD patterns of
crystalline and vitreous silica. [ Sac86]

91
Table 2.5 lists glass formers, intermediates and
modifiers. Most glass formers tend to have a high ionic
potential (>7.0) which is indicative of a strong bond with
oxygen, while modifiers have a low ionic potential. In
multi-oxide glasses, substitutions of cations with higher
ionic potentials (e.g., alkaline earths for alkalis) tends to
increase strength, durability, melting temperature, and
viscosity. Ions such as Al, Zn, Zr and Pb are listed in more
than one category in Table 2.5. Diezel [ Sac86] hypothesized
that a cation's capacity for forming, modifying or acting as
an intermediate depends on its single-bond strength with
oxygen relative to the other cations in the structure. Thus,
as the coordinative force towards bonding with oxygen
changes, the role of the cation transforms. Glass formers
have the highest bond strength with oxygen while modifiers
have the lowest.
Another convenient way to describe the network character
of silicate glasses is with R [ Kin76]
(sum of oxygen atoms) (2.57)
R = —■
(sum of network forming ions)
For example, given a formula of 80 g-atom % Si02, 10 g-atom %
Na20 and 10 g-atom % CaO,

92
Table 2.5. Glass Formers, Intermediates and Modifiers
Commonly Employed in Whiteware Glazes [ Sha92, Chi97] .
M in
Valence
Ionic
Coordination
Dissociation
Single Bond
M0X
Potential*
Number
Energy per M0K
Strength
(kcal/mole)
(kcal/mole)
Glass
Formers
Si
4
10.3
4
424
106
B
3
15.0
3
356
119
B
3
15.0
4
356
89
V
5
12.5
4
449
90-112
A1
3
5.3
4
317-402
108
Zr
4
4.6
6
485
81
Intermediates
Ti
4
6.3
6
435
73
Ai
3
5.3
6
317-402
53-67
Zr
4
4.6
8
485
61
Zn
2
2.4
2
144
72
Pb
2
1.5
2
145
7 3
Modifiers
Mg
2
2.6
6
222
37
Zn
2
2.4
4
144
36
Ca
2
1.9
8
257
32
Sr
9
1.6
8
256
32
Ba
2
1.5
8
260
33
Li
1
1.3
4
144
36
Na
1
l.C
6
120
20
K
1
0.8
9
115
13
Pb
4
4.3
6
232
39
Pb
o
1.5
4
145
36
* Ionic Potential = valence/radius (Á)
The ionic potential indicates the importance of charge and space
relationships.

93
(80x2) +10+10
R = = 2.25
80
Glazes and enamels normally are in the range of R = 2.25
to 2.75. As R increases, a greater number of oxygen are
single-bonded which weakens the glass structure and lowers
the melting temperature and viscosity.
For glasses with only one type of glass forming ion,
X + Y = Z (2.58)
and
X + 0.5 Y = R (2.59)
where Z is the number of oxygens surrounding the cation, X is
the number of nonbridging oxygens (singly bonded) per
polyhedron and Y is the number of bridging oxygens per
polyhedron. Corresponding rules that apply for silicate
glasses are
1. The Si04 tetrahedra dictates that Z = 4.0.
2. The minimum requirement for a three-dimensional network
is Y > 2.0.
3. When alkali or alkaline earths are present with Al,,0o,
the Al+3 occupies the centers of A104 tetrahedra which
converts nonbridging to bridging oxygen. Hence, when
Al+3 is a network forming ion, the A1203 contribution is
R = 1.5.

94
When comparing the most common industrial glass formers,
Si+4 and B+3, boron forms equilateral triangle 3-fold
coordination with the B+J ion placed at the center. Since
silicon accommodates four oxygen bonds to boron's 3, boron's
oxygen dissociation energy and corresponding viscosity and
melting temperature are lower.
The most common modifiers employed in ceramic glazes are
Na+1 and Ca+2. Increasing the number of modifier ions causes
more breaks in the silicon-oxygen tetrahedron and lowers the
glass viscosity. This increases the freedom of the
tetrahedrons to assume a crystalline structure and can
consequently lead to devitrification. Higher valence
modifiers such as Ca+2 are more tightly bound to the lattice
and tend to strengthen the glass structure when replacing
Na+1. Figure 2.17 demonstrates that Na+, like other
monovalent modifiers, causes the formation of two non¬
bridging oxygen and a loss of connectivity. One Na+ joins an
unsatisfied oxygen bond on one side, while the remaining NaO~
group occupies the vacancy at the corner of the tetrahedra
and breaks the structure. When monovalent modifiers are
replaced with divalent ions such as Ca+2, the bridge across
the gap is completed. This results in

95
• Silicon ion
O Bridging oxygen ion
0 Non - bridging oxygen ion
(a)
(b)
Figure 2.17. Two-dimensional representation of modifiers (a)
Na+i and (b) Ca+2 in the silicate glass structure. [Adapted
from Tay8 6]

96
1. An increase in density.
2. An increase in refractive index.
3. A decrease in cation mobility and electrical
conductivity.
4. An increase in Tg and melting temperature.
5. An increase in viscosity.
Aluminum is the most common intermediate incorporated in
glaze formulations. By itself Al+3 cannot form a glass
structure, but it can replace some silicon ions in the
tetrahedrons. The result is a higher lattice bonding
strength which increases viscosity, chemical resistance and
hardness.
The role of Al+3 depends largely on the concentration of
alkali ions. If an adjacent alkali ion is present and
achieves charge neutrality, Al+3 can substitute for Si-14. In
Si02-Al203-Na20 glasses, if the ratio of moles of Al/Na<1.0,
Al+3+Na+1 can substitute for Si+4, and excess Na+ acts in its
normal modifier role. As Al/Na approaches 1.0, glass
strength and viscosity increase. At Al/Na = 1.0, or the
equivalency point, aluminum is tetrahedrally coordinated
(A1G4) and the glass strength is maximum. For Al/Na > 1.0,
excess Al+J acts as a modifier. Substitutions of small

97
amounts of Al+3 for Na+ or Ca+2 raises the strength because of
the high bond strength trivalency provides. Large
substitutions for silicon weakens the glass structure due to
the lower ionic potential of Al+i [Chi97] .
Table 2.6 summarizes properties associated with the
presence of oxides commonly employed in ceramic glaze and
glass formulations. Computer programs have been developed
[ Din96, Din96a, Mal96] to use the additive rule for
estimating glass linear properties as a function of the oxide
composition:
P = c.x, + C,X2 4 C3X3 + ... 4 cnxn (2.60)
where P is the property, c is the weight percentage of an
oxide component and x2, x2, x3 and xn are appropriate factors.
Linear properties such as coefficient of thermal expansion,
surface tension and cost can be assessed, although results
can vary considerably with processing parameters. Results
are even less accurate for multiphase glass-ceramics such as
glazes.
Algorithms for accurately calculating more useful
properties and characteristics such as gloss, color and
viscosity which has an exponential temperature dependence,
have not been developed.

98
Table 2.6. Properties Associated with the Presence of
Various Oxides in Glass [ Mue8 3, Lac97, Vic97] .
Si02
High melting point. High viscosity. High chemical resistance.
Low expansion. Increases gloss.
B2°3
Low melting point. Low viscosity. High setting rate. High
surface hardness. Very low expansion up to a certain limit.
Low surface tension. High chemical resistance.
Na20
Low melting point. Low viscosity. Slow setting. Low chemical
resistance. High expansion. Low mechanical strength. Excess
causes devitrification.
k2o
Similar to Na20 but slightly higher viscosity.
PbO
Low melting point. Slow setting. High density. High
refractive index. Low surface tension. High elasticity.
BaO
Assists melting. Some fluxing action. Can replace lead oxide.
CaO
Fluxing action at high temperatures. Very quick setting.
Excess causes drvitrification. Increases viscosity and
strength when replacing Na20.
Ti02
Fluxing action when replacing silica. Good chemical
resistance. Reduces viscosity. Slow setting.
ZnO
Generally has a fluxing action. High viscosity. Decreases
chemical resistance. Low expansion.
MgO
Refractory. High viscosity. Slew setting. Low expansion.
High mechanical strength. High elasticity. High chemical
resistance. Can matten and opacify in large amounts.
A1203
Refractory. Very high viscosity. Slow setting. Prevents
devitrification. High chemical resistance. Low expansion.
High mechanical strength. Can lower gloss.
SrO
High expansion. High surface tension. Low durability.
Relatively high melting point and low viscosity compared to
other fluxes. Can increase gloss.
Li 2O
Low expansion. High chemical durability. High hardness. Low
viscosity. High surface tension. Can increase gloss.
Zr02
High opacification and whiteness. Increases melting point,
viscosity and surface tension. High mechanical strength. Low
expansion.

99
2.2.2 Glaze Base Materials and Formulas
Materials for glazes include processed and beneficated
minerals, specially formulated glass frits and industrial
grade chemicals [ Tay86, Wor82, Mei97] . They are selected
based on their cost, safety to the environment and ability to
yield desired glaze properties. Table 2.7 lists materials
used for batching ceramic tile glazes, and Table 2.8 shows
oxide compositions of some commercial glazes bases. Table
2.9 outlines important properties and characteristics to
consider when formulating glazes. Most whitewares glazes
appear to be glossy (gloss > 60.0 at 60° incidence) and
incorporate materials that provide oxides of Si02 (primary
glass former) , A1203, CaO, MgO, Na20, K20, B203, SrO, ZnO and
Zr02.
Batch ingredients are often multicomponent and
multiphase compounds, making it difficult to predict fired
properties from these components. Glaze or frit batch
compositions are normally converted into an empirical formula
or molar equivalents using Seger's formula in order to better
classify the system [ Epp97b] . To derive an empirical formula,
materials are expressed in terms of moles of individual
oxides present, then divided by the sum of the moles of RO
group oxides, as shown in Table 2.10 [ Sac86] . Thus, the

100
Table 2.7. Common Ceramic Tile Glaze Base Materials [ Bra86
Epp97b, She97] .
Minerals
Common
Trade Names
Major Functions
Clay;
Al203-2Si02-2H20
Glaze #1,
EPK, OM-4
Suspension, film strength, rheology
control
Frit; glass
composition
PK493, 3336
CC261, P930
Introduces oxides such as SrO and B203;
affects overall glaze properties, color
development, gloss, firing range,
thermal expansion (a), etc; expensive.
Silica; Si02
Sil-Co-Sil
Primary glass former; increases gloss,
m.p. and melt viscosity; lowers a.
Alumina; A1203
A-2, C-31
Increases hardness, durability, m.p.,
melt viscosity and firing stability;
lowers gloss and a.
Feldspar;
Na20■A1203•6Si02
NC-4, F-4
Inexpensive source of oxides; Na20
lowers m.p., melt viscosity, firing
stability and durability; increases a.
Zinc Oxide; ZnO
Denzox,
Zochem
Increases firing range and durability;
improves glaze texture and elasticity
to avoid crazing; expensive
Whiting; CaC03
#10
Increases durability; can matten and
create "orange peel" texture.
Wollastonite;
CaSi03
Vansil W-20
Intermediate properties between whiting
and silica; mattens without "orange
peel" texture (no volatiles).
Opacifier;
ZrSi04
Ultrox,
Superpax
Adds whiteness; improves color
stability and glaze hardness;
expensive.
Chemicals
Deflocculant;
Sodium Hexameta-
phosphate
Quadrofos,
Vitrofos
Lowers viscosity; can aid in
suspension.
Binder; Sodium
carboxymethyl-
cellulose
CMC-7L2
Binds glaze to body; increases film
strength; slows drying; aids in
suspension.

101
Table 2.8. Examples of Compositions (Weight %) of Commercial
Glazes [ Par73, Tay86, Tey95] .
Oxide
Fast-Fire
Matte Wall
Tile
Fast-Fire
Gloss Opaque
Wall Tile
Electrical
Porcelain
Dinnerware
Sanitary
Ware
Si02
44.5
47.3
70.0
55.9
59.7
Al203
16.7
8.5
14.3
9.6
18.6
B2°3
3.0
1.6
6.0
Zr02
13.7
13.2
Li20
Na?0
3.4
2.6
3.1
2.1
k2o
1.8
2.9
5.0
1.7
3.0
MgO
1.6
CaO
11.3
9.5
10.7
7.7
11.2
ZnO
4.0
8.7
5.4
SrO
5.7
BaO
PbO*
16.0
* PbO is still used ir. dinnerware glazes but is rarely incorporated in
other whitewares coatings due to its potential hazards to the
environment.

102
Table 2.9. Summary of Important Glaze Properties and
Characteristics [ Ree95] .
Glaze
State
Property/
Characteristics
Main Influences
Wet
Rheology
Application smoothness and defects.
suspension
Density
Application smoothness and defects.
Particle Size
Rheology and melting behavior.
Particle surface
chemistry
Dispersion stability and rheology.
Green on
Thickness applied
Color, gloss, defects and cost.
substrate
Adhesion to
substrate
Defects such as "crawls."
% Volatiles
Defects such as "pinholes."
During
Thermal expansion
Fired strength, warpage and
firing
crazing defects.
Melting temperature
All fired properties/characteristics.
Firing range
All fired properties/characteristics.
Viscosity
All fired properties/characteristics.
Surface tension
All fired properties/characteristics.
Stability from
reacting with
kiln gases
All fired properties/characteristics.
Fired
Strength
Physical durability
Elasticity
Chemical resistance
Color
Gloss
Opacity
(for avoiding crazing defects)
Surface smoothness
Defects
Cost
(influences gloss)

103
Table 2.10. Seger's Formula for Classifying Glazes [ Par73,
Sac8 6] .
RO group
(basic)
R203 (neutral or
amphoteric)
R02 (acidic)
Na20,
k2o,
Li20,
ai2o3/
B2O3
S ÍO2 r
Zr02
MgO,
CaO,
BaO,
Fe203(
Cr203,
Sn02,
Ti02
SrO,
ZnO,
MnO,
Bi203
C0O2 r
Th02
PbO,
FeO,
BeO,
CdO,
CuO,
NiO,
CoO
Sample Calc
ulations:
(1)
(2)
(3)
(4)
(5)
Chemical
Oxide
Seger's Empirical
Analysis
Molecular
Formula or Molar
Formula
Oxide Wt
% Weiaht
Moles
Eauivalents
Weiaht*
(l)+(2)
(3)-5- Sum RO Moles
(4) x (2)
(I
RO molar equivalents
= 1.0)
*Oxide formula weight(5)+total formula weight(£5) = wt % (1)

104
summation of molar equivalents of RO or alkalies and alkaline
earth oxides is always brought to unity and equals 1.0. This
provides a standard system for organizing, representing and
predicting fired properties associated with the presence of
specific oxides in the glass structure (Table 2.6).
For simplicity, glazes and frits are further classified
by ranges of molar equivalents in the amphoteric and acidic
categories. For example, ceramic tile glazes total
amphoterics are typically in the range of 0.1-0.3 and acidics
of 2.0-4.0, while porcelain glazes contain 0.5-1.2
amphoterics and 6.0-12.0 acidics.
Certain empirical rules have been derived to ensure
correct proportioning of batch ingredients [ Par73] :
1. The R0:R02 ratio should fall between 1:1 and 1:3.
2. The ratio of the alkalies to the other oxides in the RO
group should be < 1:1.
3. Txhe B203:Si02 ratio should be < 1:2.
4. The A1203 content should be < 0.2.
Rule 1 achieves proper fusion of the materials. The RO
group or basic ingredients are the primary fluxes, with an
order of decreasing effectiveness of Li20, Na20, K70, BaO,
CaO, SrO, MgO and ZnO. The R02 of interest for rule 1 is
Si02, the primary glass former.

105
The second and third rules ensure that the fired glaze
or frit is insoluble. High alkali content results in a
solubility approaching that of sodium silicate, while high
boric oxide can render a stability similar to boric acid.
The fourth rule is also intended to ensure ready fusion.
High A1203 content tends to increase the glass melting
temperature and viscosity and thus inhibits fusion.
Most commonly, the glaze melting temperature is lowered
by increasing alkali or B203 content and reducing Si02 and
A1203 while staying within the constraints of the four
empirical rules. Other methods employed to lower the fusion
point include
1. Milling the materials to a finer particle size to
increase local curvature and capillarity effects [ DeH93,
Rah 9 5] .
2. Introducing oxides as frits rather than the raw state,
because a glass lattice structure has a lower binding
energy to overcome by heat than its corresponding
crystalline phase.
2-2,3 Fast-Fire Whiteware Glazes and Frits
Fast-fire roller kiln processes are the most common
employed to produce ceramic tile. Manufacturers of other

106
ceramic whitewares such as sanitaryware and dinnerware are
also currently attempting to convert to these types of
processes. They provide much greater productivity and lower
energy costs than the slow-fire tunnel kiln technology that
was predominant prior to 1990. Fast-fire tiles are heated
utilizing 30 to 90 minute cycles at peak temperatures of
1050°C to 1200°C. These rapid schedules require precision in
formulating glazes to achieve the desired optical and
mechanical properties.
Shorter firing cycles have reduced the time available
for
1. Evolution of gases from the body before the glaze fuses
so that glaze bubbles are avoided.
2. Dissolution of glaze ingredients and reaching molten
equilibrium at the peak temperature.
3. Crystallization of phases.
Thus, it is necessary to formulate glazes which
1. Afford the highest fusion temperature possible while
rapidly maturing in a short period of time.
2. Incorporate as much of the final desired phases and
properties in the starting materials as possible.
3. Allow for precise control of the formation of new phases
during firing.

107
For these reasons, manufactured glass frits are the main
batch ingredients in fast-fire glazes [ Sac86, Toz86, Enr96] .
Most frit in the U.S. is currently used by ceramic tile
manufacturers, but frit consumption by other whitewares
industries will increase as they convert to fast-fire roller
kiln technology. Frits are prepared by melting a blend of
crystalline raw materials at approximately 1500°C, and then
fast quenching in water or with cooled stainless steel rolls
to form a glass. There are hundreds of different
compositions formulated to provide specific coatings
properties, while using low cost and environmentally safe
materials [ Bar97] . Frit compositions can also influence wet
glaze rheology through surface charge development and
exchange of soluble ions with hydroxyl or hydronium ions
[ Yoo97, LaC97] .
The main advantages of frits are
1. Fritted raw materials melt at a lower temperature and
reach equilibrium in less time than their corresponding
crystalline phases. This allows for compositions higher
in A1203 and Si02, which improve mechanical and high-
temperature stability while increasing the fusion
temperature. Vitreous silicates soften gradually
through a range of relatively low temperatures when

108
compared to crystalline phases. This occurs because
varying amounts of energy are required to detach network
atoms which are structurally nonequivalently distributed
[ Chi97] .
2. Ingredients are prereacted and thus require less time
and temperature to achieve a homogeneous blend and the
desired phases. Little or no decomposition gases are
evolved.
3. Incorporation of Zr02 in the frit rather than as a ZrSi04
mineral causes nucleation and crystallization of a
greater number of uniform, finely dispersed ZrSi04
particles, which improves opacity and whiteness.
4. Suspensions obtained with raw materials varying in
density and particle size are more prone to settling and
segregation.
5. Solubility and potential toxicity of elements such as
lead, barium, cadmium and zinc are reduced. Lower
solubility of Ca+2, B+3, F~, etc. results in more stable
glaze rheology.
Fast-fire frits and glazes differ from traditional slow-
fire formulas in that concentrations of alkalis (principally
Na20 and K20) and B203 are lower, while oxides such as CaO,

109
ZnO and SrO are higher [Amo97, Teu95] . This aids in delaying
the fusion temperature and thus keeps the coating permeable
for a longer period of time and facilitates body outgasing
without the creation of glaze bubbles.
Major categories of fast-fire frits tend to vary
mainly in alkali/silica ratio, use of SrO versus ZnO for
the secondary flux, and Zr02 content. The Zr02 normally
ranges from 0% to 8% by weight, depending on the desired
opacity. Frits which do not incorporate Zr02 contain
increased A120,, and CaO in order to maintain sufficient glass
strength.
There is an increasing interest in replacing ZnO with
SrO, even though SrO increases the cost. ZnO is classified
as a "313" regulated chemical by the EPA. Manufacturers must
monitor and report spills of materials containing ZnO to
federal and state governments, while no such regulations
exist for SrO.
Fast-fire glazes are generally classified in four groups
[ Toz86] :
1. Opaque gloss
2. Transparent gloss
3. Matte
4. Special (satin, rustic, etc.)

110
Table 2.11 lists typical empirical formulas for fast-fire
tile gloss and matte glazes.
The largest volume of whitewares sold worldwide is
coated with opaque gloss glaze. Opacity arises mainly from
zirconium crystalline phases (zirconia and zircon), added to
the batch as a ZrSi04 mineral or divitrified from the frit.
The resulting micro-heterogeneities (<10 pm) have a
significantly different index of refraction (2.05-2.40) from
the glassy matrix (~1.50—1.70) and thus reflect and scatter
light.
Transparent glazes and frits contain little or no Zr0o
and include more CaO and A1203 to make up for the lost
mechanical strength (i.e., Mohs hardness and abrasion
resistance). Otherwise, their compositions are similar.
High CaO and ZnO in these glazes can cause immiscible
liquid/liquid phase separation and opalescent opacification
that is not otherwise noticeable in zircon-opacified glazes.
Phase separation varies with the firing cycle and can impart
shade variations in colored glazes.
Matte glazes have many crystalline phases which create
local discontinuities in refractive index. This yields high
diffuse (opacity) and low specular (gloss) reflectances. The

Ill
Table 2.11 Typical Empirical Formulas, in Molar Equivalents,
for Fast-Fire Ceramic Tile Gloss and Matte Glazes [Amo94,
Apa94, Toz86] .
Opaque
Gloss
Transparent
Gloss
Calcium
Matte
Alumina
Matte
Si02
2.263
1.988
1.292
1.967
ZrC2*
0.293
0.054
0.070
0.277
0.199
0.095
0.233
0.113
ai2o3
0.191
0.195
0.102
0.412
Na20
0.090
0.071
0.139
0.170
K,0
0.088
0.094
0.092
0.056
CaO
0.346
0.434
0.769
0.515
MgO
0.069
0.110
0.000
0.130
SrO
0.134
0.128
0.000
0.000
ZnO
0.273
0.163
0.000
0.129
Notes:
*1) Frits normally contain Zr02 in the range of 0.000 to
0.175. Additional Zr02 is added to the glaze batch in
the mineral form ZrSiC>4.
2) Clay is added to glazes as a suspending agent;
therefore, gloss frits normally incorporate less A12C>3
than listed above.

112
three types of matte glazes are produced with high CaO, ZnO
or A1203 concentrations.
Calcium mattes derive CaO from the frit, whiting (CaCO,)
or wollastonite (CaSi03) . The Ca+2 ion in glass has a high
bond strength, and crystal formation is energetically
favorable [ Toz68] . If the concentration of ions (~> 15% by
weight CaO) and cooling rate are sufficient, wollastonite and
sometimes diopside (CaO•MgO•2Si02) crystals form [ Esc96] .
Zinc mattes are produced when lamellar hexagonal
willemite (Zn2Si04) crystals form during the cooling cycle.
Typically, at least 20% by weight of ZnO is necessary.
Willemite nucleation and growth require a low melt viscosity
and relatively long cooling cycle for sufficient ionic
migration to create the species and is thus difficult to
obtain with fast-fire cycles.
Alumina mattes are formulated with high mill additions
of A1203 particles, which cause the glaze to be somewhat
underfired due to an increased melting temperature and
viscosity. Underfired glazes appear matte but do not mature
and develop the optimum amount of glassy phase during firing.
A more detailed description of the kinetics of
dissolution, nucleation and growth in glass and their

113
potential influence on color development in fast-fire glazes
is discussed in Section 2.2.5.
2_JLA Ceramic Colorants
Ceramic colorants are predominately inorganic mixed
metal oxide pigments. This represents a small but very
important part of the inorganic pigment family, listed in
Figure 2.18. Inorganic mixed metal oxide pigments are the
most chemically inert and heat resistant of all colorants
[ Dry82] . They are used not only for ceramics coatings but
also for coloring paints and plastics when weather and heat
resistance is necessary. These manufactured minerals are
produced by reacting selected metal oxides at high
temperature to induce ionic interdiffusion and form the
desired crystalline matrix [ Jac96] .
A less common method for producing color in ceramics
coatings is to utilize single metal oxide raw materials or
additives. The oxides dissolve in the glass during heating
and are reprecipitated as metal ions. The color developed
is influenced by the metal ions valency state and nearby ions
in the material structure that can alter the electronic
energy levels of the metal ions [ Tay86] . Metal oxides of
transition metals such as Cr, Fe, Co, Mn, Ni, Cu, V and Ti

METAL
OXIDES
114
Fe :Oi
Spinels
Cd Pigm.
Metallics
Magnetite
TiOt
BaSOj
Cr:0<
Rutiles
Cr Yellows
Lusters
Hematite
Sb:0<
CaC03
ele.
Zircons
Mo Orange
Ultramarines
Ochre
ZnO
Kaolins
ele.
Fe Blues
Gold Pinks
Umber
ZnS
Silicas
ele.
etc.
etc.
etc.
Talcs
etc.
Figure 2.18. Inorganic pigment family. [ Bur7 9]

115
and the lanthanide group's Pr and Ce are normally employed.
Since development of their color is highly sensitive to
processing conditions and interactions with other materials
in the coating formulation, they are now rarely used in high
volume whitewares processing where color control is
essential.
Figure 2.19 diagrams the chroma of colorants commonly
used in ceramics coatings. The distance from the origin
indicates the purity of the pigment color. Table 2.12 lists
glaze pigments, their maximum use temperature and glaze
formulation requirements for optimum color.
The research of this dissertation focuses on the doped
zircon structure pigment group of the mixed metal oxide
family. It is the most widely used pigment system for
ceramic whitewares coatings.
2,2.4a Zircon Triaxial Pigments
Zircon (ZrSi04) pigment systems have been used in the
ceramics industry for over 30 years [ Blo94a] . Zircon's
tetraganol structure has the ability to accommodate a number
of impurity ions substitutionally, and its high chemical and
thermal stability make it ideal for use in ceramics coatings.

116
+b
o
Q ZrSiO^Pr
q Yellow
60 "
t Titania
Maple
Zircon
â– h Green
Mixture
40 T
O
Cadmium
Orange
q Victoria
Green
A Cr2Ü3
Green
â–¡ â–¡
4-
20*
Zirconia
Vanadia
Yellow
c9
O
O
CrFeZnAI Brown
j. CrFeZn
Brown
3
ZrSi04:Fe
Pink
Cadmium
Red
, ~ CrAI Pink
Black,Greys CrSn Pink
a* r-,-20 â–¡ 0
° □ n
Cobalt Chromite
+
Blue Green
â–¡ â–¡
k
k
ZrSi04:V
Blue
-b
20
40
+a 60
Cobalt Alumínate and
-20 + Cobalt Silicate Blue
V
Figure 2.19. CIE a* and b* chroma of ceramic pigments.
[ Epp87]

Table 2.12. Glaze Pigments and Their Requirements [ Tay86] .
Color
Pigment
Maximum
Firing
Tempera¬
ture (°C)
Glaze Base Characteristics for Optimum
Color Strength and Stability
Blue,
light
Zircon, vanadium*
1300
Suitable for use in all glazes
dark
Olivine, cobalt silicon
1400
Suitable for use in all glazes
light
Spinel, cobalt zinc aluminium
1300
Suitable for use in all glazes
Green,
light
Zircon, vanadium/praseodymium
1250
Suitable for use in all glazes
dark
Spinel, chromium cobalt (zinc)
1400
Suitable for use in most zinc-free glazes
light
Garnet, calcium chromium silicon
1200
Zinc-free, high calcium compositions
Orange
Baddeleyite (indium, yttrium)
vanadium
1400
Suitable for use in all glazes,
especially those firing at high
temperatures
Pyrochlore, lead antimony
1050
High lead content glazes
Rutile, titanium chromium
1100
Leadless glazes
Yellow
Zircon, praseodymium*
1250
Suitable for use in all glazes
Cassiterite, vanadium
1300
Suitable for use in all glazes
Cadmium sulphide
1050
Glaze must be specially designed to give
temperature stability to the pigment
Red
Cadmium sulphoselenide
1050
Glaze must be specially designed to give
temperature stability to the pigment.

Table 2.12--continued.
Color
Pigment
Maximum
Firing
Tempera¬
ture (°C)
Glaze Base Characteristics for Optimum
Color Strength and Stability
Pink
Zircon, iron*
Zircon, cadmium sulphoselenide
Sphene, chromium tin silicon
Spinel, chromium aluminum zinc
1300
1300
1200
1300
Suitable for use in all glazes
Suitable for use in all glazes
Zinc-free, high calcium compositions
High zinc and aluminium glaze with low
calcium content
Brown,
dark
Spinel, iron chromium zinc
iron chromium (nickel)
1300
1300
Very suitable for use in zinc-containing
glazes
Very suitable for use in zinc-containing
glazes
light
iron aluminium chromium
zinc
1300
Very suitable for use in zinc-containing
glazes
Black
Spinel, chromium cobalt iron
nickel
1300
Zinc-free compositions
Grey
Cassiterite, antimony vanadium
1300
Suitable for use in all glazes
* Vanadium, praseodymium and iron zircons are the most common triaxial pigments.
CO

119
The triaxial system is based on blending zircon-vanadium
blue, zircon-iron coral and zircon-praseodymium yellow
structures to obtain a wide variety of colors. They can be
either milled into the glaze batch or, more commonly, stirred
into a prepared glaze base. Stir-in pigments require narrow
particle size distributions and carefully controlled
surfactants for good dispersion. Most calcined ceramic
pigments are in the 1-12 pm mean particle diameter range,
with no particles greater than 44 pm.
Zircon pigment concentrations in glazes range from 0.1%
to 10% by weight, with most systems containing closer to 0.5%
to 2.0%. Zircon-based pigments furnish a lower color
strength than metal oxides but provide much more
opacification and high temperature stability. They are
insensitive to variations in kiln atmosphere, and their
solubility varies with the melt composition.
Zircon is isomorphous with a tetragonal structure 4/m
2/m 2/m (one fourfold and two twofold axes with perpendicular
symmetry planes) as shown in Figure 2.20. It consists of a
chain of alternating edge-sharing Si04 tetrahedra and Zr08
triangular dodecahedra extending parallel to the c-axis. The
prismatic habit of zircon compounds results from edge-sharing

120
e
Figure 2.20. Typical forms of zircon crystals. (a-c):
a{ 100} , m{ 110} , p{ 101} , x{ 211} ; and zircon lattice structure
(d, e) . [ Wyc65, Mas68]

121
dodecahedra, which join the chains. There are four molecules
per unit cell [ Gre82] .
The Zr03 polyhedra are two interlocked, tetragonally
distorted tetrahedra. One tetrahedron is compressed, the
other is stretched parallel to the c-axis. The Si04
tetrahedra are elongated parallel to the c-axis. The two
angles bisected by the c-axis are less than the 109.4° ideal
tetrahedral angle and the remaining four are greater. Both
tetrahedra Si04 and dodecahedra Zr08 site symmetry are D2d,
which is Schoenflies notation for a twofold principal axis
and two twofold axes normal to it, with diagonal symmetry
planes [ Klu74] .
Zircon-doped pigments are prepared by solid state
reactions which approach equilibrium with difficulty
[ Epp87] . The research of this dissertation focuses on the
zircon-vanadium blue pigment; therefore, its mechanisms for
color production will be discussed in the most detail in this
section. Phenomena involved with loss of color strength and
stability of Zr-V pigments during processing also apply to
the other two triaxials, which are based on the same zircon
structure.
Zircon-vanadium blue is ordinarily formed from batches
of 60-70% Zr02, 26-36% Si02, 3-5% V205 or NH4V03 and 0.5-5.0%

122
NaF by weight. The materials are heated to 750-900°C for one
to eight hours to complete the solid state reaction. It is
then washed and leached to remove soluble V205 because V+5
provides unwanted green color [ Par73j . Without a
mineralizer-catalyst such as NaF, the V+4 which provides the
blue color does not form on the zircon lattice. NaF aids in
the development of the solid state solution by lowering the
required sintering temperature and increasing the mobility of
Zr+4 and Si+4 for forming zircon. Without a catalyst, melting
occurs for Zr02 at 2700°C and Si02 at 1713°C, while ZrSi04
begins to decompose at 1720°C [ Wea79, Sha92] .
Some details regarding the origin of the blue color are
still unexplained. A valency of V+4 is present, but the
critical role of NaF and the exact location of V+4 in the
lattice remain obscure. With an ionic radius of 0.61 Á, V+4
could theoretically occupy Si+4 (0.40 Á) , Zr+4 (0.87 Á) or
intersticial sites [ Ric92, Sha92] . Demiray, Nath and Hummel
surmised that V+4 is solely responsible for the blue color and
occupies the dodecahedral Zr+4 site (Zr08). They also stated
that NaF increased the solubility and thermal stability of V+4
in zircon but did not affect coordination or valence [ Par73] .
Gregorio, Greenblatt, Pifer and Sturge published that the

123
optical spectra suggests that V+4 enters the tetrahedral Si+4
(Si04) rather than dodecahedral site [ Gre82] .
Vanadium metal has an outer electronic structure of
3d3 4s2. The cation V+4, which produces a blue color, has a
3d1 4s° configuration. When the coordination of anions about
a single d electron is fourfold or eightfold, the electron's
initial fivefold degeneracy (shown in Figure 2.7) is
disrupted. This phenomenon is described with ligand field
theory in Section 2.1.2d. The anion coordination about V+4 in
the zircon structure results in the original d orbital
electron energy level ground state (2D) splitting into a new
doublet ground state (2E) and an excited triplet state (2. ) ,
as presented in Figure 2.21. The figure is based on a choice
of axes where xz and yz planes contain the anion ligand
atoms. When crystal site symmetry is reduced to tetragonal,
the orbitals are further split into two ground and two
excited states. The G and K symbols shown in Figure 2.21 are
differently polarized transitions due to orbital coupling.
Mixing of split orbitals due to atomic vibrations can also
occur. It has been suggested that this phenomena makes it
difficult to evaluate the true d orbital energy levels in
vanadium doped zircon [ Gre82] .

124
D
2d
cr TT
2E(xz,yz)
28<(x2-y2)
2
At(Z2)
2B2(xy)
Figure 2.21. Splitting of the d orbital in V+4 by tetrahedral
(T¿) and tetragonal (D2d) crystal fields. (Possible electron
transitions are shown with dotted and solid lines) [Adapted
from Gre82]

125
The resulting V+4 electronic structure absorbs incident
light in the range of approximately 600 nm-700 nm. These
wavelengths correspond to the energy of light absorbed, which
is equal to the difference in energy levels between the split
orbitals which allow electron transitions from the ground to
excited state. Vanadium is colorless when incorporated in a
glass structure because its coordination does not allow for
split orbitals and electronic transitions which correspond to
the energy associated with the visible spectrum.
Zirccn-ircn coral is the high temperature reaction
product of 50-75% ZrC2, 15-50% Si02, 2-10% NaF and 0.25-25%
Fe203 by weight. The Fe+3 loses an electron and converts to
Fe+4 (3d4) and substitutes for Zr+4 in zircon. This results in
approximately 40% ZrSi04 with 0.25% Fe+4 in the lattice. The
wavelength of incident light absorbed by the structure falls
in the range of 400 nm-570 nm [ Par73] .
Zircon-praseodymium yellow is manufactured by reacting
35-80% Zr02, 10-55% Si02, 0.25-3% NaF, 0.25-8% KNaO and 0.5-
10% Pr6On by weight. There is discord in theories regarding
the solid state reactions that occur in forming the Zr-Pr
structure. Decker [ Dec901 suggested that Pr6On is reduced to
Pr203 and then Prfi converts to Pr+4 by losing an electron.
The ?rT4 substitutes for Zr+4 in zircon, and some unreacted

126
PrgOl;L is left dispersed in the material. Unfilled f orbitals
(4Í1) allow for split energy levels, and resulting selective
absorption occurs in the range of approximately 400 nm-470 nm
wavelength of light.
Light scattering from opacifiers such as zircon counters
pigment absorption and thereby lowers color strength. It is
difficult to overcome the whitening effect of unpigmented
zircon except by adding large quantities of colorant. An
optimum system would incorporate zircon pigments sized to
scatter light in wavelength regions other than where its
dopant transition metal absorbs light. In this way,
increases in chroma through a decrease in reflectance in the
pigment's absorbing wavelength area can be obtained [ Blo94b] .
For example, for zircon-vanadium blue pigments which absorb
light in the 600 nm-700 nm range, a zircon particle size of
0.75 pm or 750 nm would scatter light without affecting
chroma. Current pigment manufacturing processes do not allow
for such tight particle size control. In addition, extremely
fine pigments can cause problems with agglomeration in the
wet glaze base and increased solubility at high temperature.
Optimizing color strength through improved particle size
distributions is an area for future research.

127
Although zircon doped pigments are the most stable
ceramic colorants up to approximately 1200°C, significant
color variations can occur. Resulting color is a function of
the stability of the zircon structure at high temperature and
the optical properties of the coating matrix phases. The
color of the substrate upon which the coating is applied is
also important for translucent glossy coatings.
2.2.4b Kubelka-Munk Analysis of Colorant Lavers
Pigment particles produce almost perfect diffusion of
light flux in every direction by Fresnel reflection or light
scattering. Light proceeding parallel to the boundary of the
colorant layer has just as much light diffused to the left as
the right and thus may be canceled out for mathematical
interpretations [ Jud65] . Consider two diffuse light fluxes,
one traveling downward (i) and the other upward (j), through
an elementary colorant layer of dx which starts at zero from
the unilluminated side, as shown in Figure 2.22. The
thickness of dx is small compared to the total layer
thickness (X), but large compared to pigment particle
diameters. The downward flux of light decreases due to
absorption into pigments by Kidx and reverses direction
through scattering by Sidx, where K is the absorption

128
Top of colorant layer
1
Elementary layer of thickness, dan 1
>
t
Distance, x.
between backing
Colorant layer
and an elementary —
of thickness, AT —
layer, dx
f
..Backing of reflectance
^///////zz/////////////////
Figure 2.22. Schematic of basis for Kubelka-Munk analysis of
colorant layers [Jud65].

129
coefficient and S is the scattering coefficient. Upward
reflected light (j) is also reduced by absorption and
scattering. The changes in light flux caused by an
elementary colorant layer can be written [ Jud65]
dj = -(S+K)jdx + Sidx (2.61)
-di = -(S+K)idx + Sjdx (2.62)
where dj and -di are the total changes in upward and downward
proceeding fluxes, respectively.
Exponential solutions to differential equations (2.61)
and (2.62) in terms of reflectance were derived by Kubelka
and Munk in 1931. They are equations (2.31) and (2.32) in
Section 2.1.2c of this dissertation. From (2.32), the ratio
of absorption to scattering coefficients is given by
K/S = (1-RJ2 / 2 R„ (2.63)
Important requirements of equations (2.31), (2.32) and (2.63)
are 1) each colorant layer must be evaluated one wavelength
unit at a time (normally in 0.5 nm increments), since K and S
are a function of wavelength, and 2) complete hiding power
or opacity of the substrate backing is assumed.
Beer's law (equation (2.41) in Section 2.1.2d) does not
hold true for complex-subtractive colorant mixing and instead
the Kubeika-Munk equations are applied. Absorption and
scattering coefficients of materials are used to predict

130
colorant concentrations required to produce a color or to
evaluate the hiding power of a coating. Equation (2.63) has
been used by Selling [ Sel47] for formulating dyes and by
Duncan [ Dun49] for paint formulations. Murdock, Wise, Eppler
[ Mur90] and Blonski [ Blo93, Blo94b] evaluated ceramic
pigments in glaze using Kubelka-Munk theory.
Application of equation (2.63) to pigment colored glazes
presumes the composite system is the concentration of
weighted sums of the individual components [ Blo94b] . An
absorption factor (K/S) [ Epp97a] can be used to measure the
color strength of the coating, which is a function of the
pigment concentration (Cpigment) :
K/S
(1-R)2
2R
C K _J_
^pigment ^pigment ^ glaze
c s + s
pigment pigment glaze
(2.64)
The requirements of (2.64) for evaluating pigmented ceramic
coatings are
1. Spectrophotometric measurements of reflectance (R) are
substituted for reflectivity (R„) in (2.63) if an
increase in coating thickness does not change the
reflectance values.
2. Reflectance is diffuse; the specular component must be
excluded.

131
In equation (2.64), (1-R) converts measured diffuse
reflectance into absorption, since light that is not
reflected by an opaque object must be absorbed. The factor
2R allows that more than all of the incident light cannot be
absorbed.
^glaze
Since as C • nt —>0; K/S — , K/S calculated with
^ glaze
the minimum reflectance (R . ) value measured across the
visible spectrum is where absorption of the glaze matrix is
minimal and absorption due to the pigment is maximum. This
provides the best indication of color strength due to pigment
absorption. A higher K/S at the minimum reflectance
indicates greater pigment strength.
Calculated absorption factors (K/S) of various pigments
and pigment/glaze systems are compared to weigh color
strength. Plots of K/S versus pigment concentration are
often collated and exhibit a linear relationship over most of
the range [ Epp97bj . The vertical distances between log K/S
versus wavelength curves are proportional to pigment
concentration, or layer thickness if the coating does not
completely hide the substrate.

132
2.2.5 Frit Influence on Color Development
The appearance of ceramic glazes is greatly affected by
the structure and composition of the glassy matrix phase and
dispersed crystalline or immiscible liquid phases which cause
opacification and color. Light scattering in partially
crystallized glasses can result in a range of optical
properties from clear to completely opaque and white.
Depending upon the starting frit composition and
corresponding high-temperature properties, crystalline phases
such as zircon, z.irconia (monoclinic or tetragonal) ,
wollastonite and willemite have been found to precipitate in
slow-fire glazes during firing [ Eld94, Esc96, Amo94, Gru78,
Apa94]. Dissolution rates of zircon pigments are also a
function of the glass composition and properties. In
particular, nucleation, growth and dissolution of crystals is
largely controlled by the glassy phase viscosity-time-
temperature conditions during processing.
2.2.5a Crystallization Mechanisms
In a glass melt at high temperature, unit lattices
formed from attractive forces continually disintegrate due to
the thermal motion of the atoms. Thermal agitation and the
rate of rupture decrease as the melt is cooled. Atoms crowd

133
closer together, bonds are strengthened and crystal formation
becomes more readily attainable. Crystallization at a
sufficiently low temperature depends on three main factors:
1) number of nuclei formed, 2) crystal growth rate, 3) melt
viscosity [ Par73] .
Nucleation begins when two atoms with sufficiently low
velocities collide and come together at rest. If there is
not sufficient energy to agitate the resulting mass, it
becomes an easier target than a single atom. The pair is
struck by other atoms and a nucleus begins to form. The free
energy change (AG) required for homogeneous nucleation is
(assuming spherical geometry) [ Sac96]
2 4 3 (2.65)
AG = (4nr“)y + (— nr )AGV
The first term on the right side of (2.65) is the free energy
required to form new interfaces, as a function of interfacial
energy (y) and nucleus radius (r). This term always has a
positive value. The second term is the free energy for the
bulk transition as a function of the free energy change per
unit volume of the new phase formed (AGV). This term can be
positive or negative. Overall, AG must be negative for
nucleation to become thermodynamically favorable. The
critical nuclei radius (r“ ) where AGv has a sufficiently

134
negative value to overcome the energy required to form new
interfaces is determined from the 3 (AG)/3r = 0 maxima and is
given by
r*
(2.66)
At embryo sizes greater than r*, the chemical potential of the
embryo is less than that of the bulk phase, and it becomes a
growing crystalline nucleus. An approximate equation for AGv
for condensed phases is [ Dor94]
AGV = AH f (T-TJ / VTn (2.67)
where AHf is the heat of fusion, Tm is the melting temperature
and V is the molar volume of the matrix phase.
The activation energy required for nuclei formation
(AG*) is
16 ny3 (2.68)
AG* = -
3 ag;
The rate of nucleus formation (I) is related to AG* by
-AG* 3 (2.69)
I = Kexp ( ) (no./cm /s)
kT
where k is Boltzmann's constant and K is a coefficient
assumed independent of temperature. By combining (2.67),
(2.68) and (2.69)
— 167tyj VT,
3AH2f (
1
1 = Kexp
kT
/
(2.70)

135
Equation (2.70) indicates that because of the (T-Tm)2 factor,
without a transport limitation there is a sharp increase in
the nucleation rate at some critical temperature below Tm.
The transport factor, or viscous flow, constrains the maximum
nucleation rate possible with decreasing temperature. Thus,
K is sometimes estimated as 103G cm--1' sec-1 poise for oxide
glass formers [ Chi97] .
The nucleation rate can be further related to the glass
matrix viscosity (T|) by [ Dor94]
leg IT) = K1/T(T-Tn)2 (2.71)
where is a temperature independent factor. Equation (2.71)
points out again that viscosity and nucleation rate are
inversely proportional.
Glass systems are normally not completely pure, and thus
heterogeneous rather than homogeneous nucleation occurs. The
activation energy is lower for heterogeneous nucleation
because the amount of required interfacial energy is reduced
by the presence of nucleation catalysts. The catalysts can
be surfaces, seed particles, contaminant particles, second
phases, bubbles or defects. A higher nucleation rate results
when interfacial energy is lowered due to the segregation of
dissolved species at interfaces [ Dor94] .

136
In equations (2.65)- (2.71), heterogeneous nucleation can
be accounted for by substituting an effective interfacial
energy (ye) fory [ Dor94] :
ye = Z f iY (2.72)
i
where fi is a correction factor for nucleation catalysts of
class i. In most cases, there is more than one catalyst
present that speeds up nucleation at different undercoolings.
Nuclei formed are typically 5-lOxlCT3 microns in diameter.
Figure 2.23 is a schematic of nucleation and
crystallization in a glaze system as a function of
undercooling temperature. Nucleation rates are low until the
undercooling is sufficiently large, then rates increase
rapidly with undercooling. Nucleation rates decrease at
lower temperatures because the increase in viscosity slows
diffusion transport.
The shapes of and spacing between nucleation and growth
rate curves vary for different glass and glaze systems. The
narrower the temperature range between peak nucleation and
crystal growth races, the higher the number and size of the
crystals formed, providing that viscosity permits transport.
Opposing factors for nucleation and growtn occur where the
curves overlap. In this region as temperature decreases, the

Viscosity
Nucleation rate
Crystal growth rate
137
Figure 2.23. Relationship between viscosity and temperature
favoring nucleation and growth in glazes. [Adapted from
Par73]

138
driving force for nucleation increases, but the diffusion
transport rate lowers enough to inhibit growth. Both
processes also need time to occur. If cooling is too rapid,
nucleation and crystallization may not result.
Crystallization from solution is a function of impurity
and nuclei types and concentrations, the number of phases
separating, time-temperature processing, matrix phase
viscosity and diffusion coefficients (D).
The rate of growth of a spherical particle of radius r
is [ Dor94]
dr D
—— = a —
dt r
(2.73)
where a is a coefficient dependent only on concentrations.
Thus, upon integrating separable equation (2.73), it becomes
evident that the crystal radius grows proportional to the
square roots of both time and the diffusion coefficient.
The diffusion coefficient is related to the gradient of
concentration c in the x direction [ Ban86] :
3c 3~c (2.74)
dt 3x'
and has an Arrhenius type relationship to temperature:
D
Du exp (-Q/RT)
(2.75)

139
where Do is a material factor with the same units as D, Q is
the activation energy for the transport process and R is the
gas constant.
When the matrix phase is near its equilibrium
composition during the later stages of growth, the crystal
radius is proportional to the cube root of time (t):
3 3 _ 8YCePVmt (2’76)
r r° 9RT
where rQ is the mean crystal radius at t=0, Ce is the crystal
concentration and Vm is the molar volume of particles.
The thermodynamic driving force for crystallization is
related to heat flow and rearrangement at the melt-crystal
interface. It is measured as the difference in Gibbs free
energy between the liquid and crystal. This is also referred
to as the molar free energy of crystallization of an
undercooled liquid [ Chi97], (AGm) given by
AGm = AHf (T —TJ/Tm = AHf-TmASm (2.77)
where AHf is the heat of fusion at the melting temperature T„.
At the melting point, AGm is zero. The high viscosities of
glass melts and corresponding small growth rates enable the
interface temperature (T) to be taken as the glass bath or
furnace temperature [ Kin76] , which are easier to measure.
Equation (2.77) is useful for small undercoolings, but fails

140
at very large undercoolings since AHf and ASm are temperature
dependent.
For large undercoolings, it has been presumed [ Chi97]
that the heat capacity difference between the liquid and
crystal remains constant with undercooling, and AGm can be
better estimated with Hoffman's equation:
AHf (T-TJ T (2.78)
AG™ = : 1”—
Tm
In many multi-oxide glass systems, viscosity tends to
increase rapidly with decreasing temperature and
undercoolings are large. For this case, Hoffman's equation
provides the best approximations. (All of the equations in
this section are left in terms of the standard relation
(2.77), but equation (2.78) can be substituted when
appropriate.)
The nature of the matrix/crystal interface has a strong
influence on the kinetics of crystallization. Models used to
predict the crystallization process are based on different
assumptions regarding interface geometry.
In oxide glasses the velocity of crystallization u is
often estimated with
AGm
U
3k )ni
(2.79)

141
where X is the thickness of the transition layer between
liquid and crystal. Equation (2.79) demonstrates the
reliance of crystallization on temperature, viscosity and AHf.
Often, crystallization velocity versus temperature is derived
using (2.79), but inaccuracies in calculated values may
result from uncertainties in geometrical factors.
One factor that has been used to estimate the
suitability of applying (2.79) was formulated by Jackson
[ Dor94] :
a = AHf/RT (2.80)
e m
where R is the gas constant.
If tt < 2.0, rough interfaces form and (2.79) is not
accurate. Values of a > 4.0 also result in discrepancies
when using (2.79) for reasons not clear [ Dor94] .
The growth rate for a typical glass is zero at the
melting point, increases with undercooling, rises to a
maximum, then decreases as the viscosity increases. This is
demonstrated with Avrami's equation [ Rah95] , which also
indicates that the volume fraction of crystals b with a given
thermal history is an exponential function of time (t)
[ Rah95] :
b
?xp
— 7tlu t
Ttlu't'
(2.81)

142
Crystalline phase formation can be observed as
exothermic transformations on DTA or DSC curves, while fusion
and melting are endothermic. Activation energy for crystal
growth (Q) has been determined using DTA measurements and
[ Pop77] :
In
d (AT)
dt
+ KAT
-Q
+ const.
RT
(2.82)
where Cp is the heat capacity of the sample and sample
holder, K is the heat transfer coefficient and AT is the
temperature difference between the sample and reference at
time t. The heat transfer coefficient can be calculated with
AK = KA (2.83)
where AH is the total heat evolved during the reaction and A
is the area under the DTA exothermic peak.
For DSC studies with a constant heating rate, the
activation energy can be derived from the temperature at the
peak maximum point (T ) [ Dor94] :
In
dT
TI/
F dt
= In (Q/R) - In v + Q/RT.
(2.84)
where dT/dt is the heating or cooling rate, v is a constant
and R is the gas constant. When In [t~/(dT/dt)1 is plotted vs.
1/Tp for various heating rates, the slope of the curve is Q.

143
The beginning particle size of a glass frit also has an
effect on crystallization. An increase in mean particle
diameter raises the starting temperature of crystallization
and lowers nucleaticn and crystallization rates. This is
depicted on DSC or DTA with a peak exotherm shift to a higher
temperature and a reduction in the maximum exothermic band
height. Larger particles lower the specific surface area
available for surface nuclei formation [ Amo94] .
If heating rates are sufficiently low, crystal species
will precipitate in seme glaze compositions during heating
and then partially or totally dissolve at the peak
temperature, as demonstrated in Figure 2.24. In the case of
ceramic tile processing, where cooling is much faster than
heating or critical cooling rates, crystals that do not
completely dissolve at the peak temperature are "frozen" in
their high-temperature state during cooling.
It is common for multi-oxide glasses to develop
anisotropic crystal morphologies and growth rates.
Pronounced asymmetry in the crystallization rates of multiple
species is observed at and in the vicinity of the melting
temperature. For this reason, existing models, including
those described in this section, may not be completely
accurate in describing crystallization kinetics [ Kin76] .

-1.5 -1.0 -0.5 0 +0.5 +1.0 +1.5
Crystal growth (mm/hr)
Figure 2.24. Crystal growth rate as a function of
temperature in Na20-Ca0-Al203-SiC>2 glass. [Adapted from
Kin76]

145
The equations described in this section help to identify
variables which control nucleation and growth during
crystallization. Models for accurately estimating
crystallization processes for all multi-oxide systems have
not been developed. Hypotheses of varying complexity have
been proposed, but there is still a great deal of dissention.
2.2.5b Zircon Crystallization and Dissolution
Zircon is an excellent opacifier for glazes due to its
high refractive index and ease in forming small crystals that
scatter light. Its structure (described in detail in Section
2.2.4d) has good resistance to chemical attack and is
relatively stable at high temperature due to its high atomic
bonding strength and kinetics of thermodynamic equilibrium.
The zircon structure is also widely used as a ceramic
colorant due to its thermal stability and ability to
accommodate rare-earth metallic ions. Even with its
relatively stable properties, dissolution and crystallization
of zircon has been found to occur at high temperature in some
glazes [ Blo94a, Cas97] . This can result in significant
variations in optical properties.

146
Zircon was found to crystallize (c) from Zr02 amd Si02
oxides in a vitreous (v) multi-oxide glaze frit during
heating, according to sequence [ Esc96] :
Zr(v) + 02 > Zr02(v)
(2.85
~ 800C
Zr02 (v) > ZrO, m(c)
(2.86
~ 850C
(2.87
ZrO. - m(c) > ZrO~-t(c)
where monoclinic (m) and tetragonal (t) zirconia continue to
precipitate and can be present up to as high as 1100°C
[ Amo94] .
Reactions (2.86) and (2.87) are controlled by diffusion
of Zr02 structural units to the crystal-glass interface.
Finally,
> 850C (2.88)
Zr02 (c) + Si02 (v) > ZrSiO,. (c)
which is a diffusion controlled, first order irreversible
reaction. The temperature of maximum crystallization of
zircon shifts upward with an increase in heating rate, since
the process is time dependent. Normally, crystallization of
all species in glazes subjected to rapid heating cycles
occurs at > 900°C [ Kin76, Amo94] .
Some frits crystallize zircon during the firing process,
providing opacity and whiteness. Crystallites formed are

147
normally <10pm in size and strongly orient along the crystal
c-axis lying in the plane of the glaze surface [ Blo93] .
Maximum whiteness occurs at a zircon mass fraction of
approximately 0.16 [ Esc96] . Crystallite size and amount
depend on heating time vs. temperature and the frit
composition and particle size. Phase diagrams can be used to
predict zircon crystallization for simple 2-3 oxide systems.
Nucleation and growth of zircon from frit predominates
heterogeneously on the frit particle surfaces. As the
specific surface area available for nucleation on particles
>40pm decreases with increasing size, so does the amount of
zircon precipitated in a given heating profile. Larger sizes
cause crystallization to occur at higher temperatures and at
slower rates. Deviations in frit mean particle size between
>10pm and <40pm, which is the normal range for industrial
frits, did not cause variations in crystal growth [ Esc96] .
Undissolved zircon seeds <1.0% by weight in the original frit
also had no effect on the results due to the predomination of
surface nucleation [ Amo94] .
Ceramic tiles processed in industry are subjected to
nonisothermal firing cycles. At temperatures where
nucleation and growth can occur, typical fast-fire cycle
heating rates of 40-50°C/min., peak temperatures of 1000-

148
1150°C and cooling rates of 100-300°C/min. are utilized.
Devitrification of two zircon opacified frits was found to
occur during the heating step and hold at the peak
temperature, but was not a function of the cooling cycle.
Ranges in cooling cycles tested which fell outside of the
limits normally employed to manufacture ceramic tile had no
effect on the amount or morphology of zircon present in the
fired glaze [ Esc96] .
Dissolution of zircon opacifier and pigment crystals in
the glaze melt leads to loss of opacification and color.
Dissolution rates increase with higher temperatures, smaller
particle sizes and lower viscosities. The solubility limit
of zircon in most silicate melts is 3-5% [ Kin76, Con97,
Par73] but is also dependent on composition and processing
conditions.
Zircon by itself dissociates into Zr02 and Si02 at 1720°C
± 20°C, as shown in the phase diagram in Figure 2.25 [ Lev69] .
Melting points of the individual oxides are 1713°C for SiO,
and 2700°C for Zr02. Zircon begins to dissolve in a glaze
usually below 1100°C, but this varies significantly with the
melt composition. No accurate models have been developed for
predicting zircon dissolution given the starting glaze or
glass frit oxide composition.

149
2800
2600
l^-27000
2400
\
l_l
I
i
LI
2200
2000
1800
1600
0
ZrOr
\
X
Zr02 SS
+
Liquid
(L)
1775°* 10°
Zr02 SS
+
ZrSiOA
i_L
ZrSiO +1
4 u 1675°* 5°
\ SiO?+L
» \ -4
■\ \ 1713°
>
\ v
W-
ZrSi04+ Si02
88%
20
40
60
80
100
Si0o
Figure 2.25. Binary phase diagram of Zr02 and Si02 system
(temperature in degrees celcius). [Adapted from Lev69]

150
Castilone et al. (1997) and Concepcion et al. (1977)
[ Cas97, Con97] found that in one unopacified glaze frit slow-
fired to 1100°C, 3% to 5% weight additions of zircon
completely dissolved and created Zr02 as a network former.
Continuously less zircon dissolved with concentrations
increasing from 5% to 13%. From 13% to 23% additions, all of
the zircon crystallized and none was present as Zr02 in the
final glass structure. They theorized that complete
dissolution of crystals for satisfying Zr02 requirements in
the glass does not occur at high zircon concentrations.
Volume is lost from each crystal but the original number of
sites remain. The sites function as nuclei seeds for
recrystallization during the slow cooling cycle.
Recrystallization during the cooling step in fast-fire
processes was found not to occur [ Esc96] .
The effect of particle size on the melting temperature
can be related with [ DeH93]
T (H) = T(H=0) - 2yVsH/AS (2.89)
where T(H) is the melting temperature of a system containing
solid particles with a mean curvature H, T(H=0) is the bulk
melting temperature (found on most phase diagrams), y is the
specific interfacral free energy (ergs/cm2) , Vs is the solid

151
particle molar volume and AS is the entropy of fusion.
Equation (2.88) assumes all particles are the same size and
spherical. Generally, as a particle size gets smaller, its
mean curvature increases, and from (2.89), the melting point
is lowered.
The solubility (c) of a nonspherical surface is [ Chi97]
CO
i
i
!—1
r- ^
c = cG exp
RT
r, r,
\ 1 - /J
where r1 and r2 are principal radii of curvature and c is the
equilibrium solubility. Again, as the radius of curvature
decreases and the interfacial free energy gets larger, the
solubility of the solid in the liquid phase increases.
Interfacial free energy is a function of the chemical
potential of the solid and liquid phases, temperature and the
specific interfacial excess of moles per unit area at the
interface.
Blonski [ Blo94a] found that the weight % of zircon
dissolved in a slow-fire ceramic tile glaze increased
exponentially with a decrease in mean particle diameter. He
determined that for zircon pigments with an aspect ratio >1,
dissolution was a function of the narrowest cross-sectional
dimension.

152
The sizes of zircon pigments used in industry, usually
l-12pm in mean diameter, provide optimum dispersability in
liquid glazes. The sizes of zircon opacifiers utilized
varies widely, depending on opacification requirements,
firing schedules, glaze compositions and cost.
2.2.5c Liquid-Liquid Phase Separation
Special glaze or glass compositions are formulated so
the melt separates into two different liquid phases. More
than one amorphous phase remains in the cooled glass and may
cause opalescent opacification.
A thermodynamic description of liquid-liquid phase
separation is given with AGmix vs. composition plots at
different temperatures. If the plot is concave upward, a
single phase has a lower free energy than any mechanical
mixture of two phases with the same average composition and
is thus most stable. But a miscibility gap occurs in phases
that have a positive excess Gibbs free energy at sufficiently
low temperatures. In this case, a single phase is not most
energetically favorable, and the plot is concave downward.
Spontaneous unmixing of an initially uniform phase, known as
uphill diffusion, reduces the difference in chemical
potential of the components and lowers the Gibbs free energy

153
of the system. This thermodynamic process, referred to as
spinodal decomposition, is responsible for liquid-liquid
phase separation [ DeH93] .
Phase separation in silicate glasses results from a
combination of silica network-forming limitations and
competition between cations to surround themselves with the
lowest energy oxygen configuration [ Kin76] . Nonglass¬
forming cations with a strong oxygen bond strength can raise
the energy of the glass system by breaking up the network. A
more energetically favorable system is achieved when two
separate phases form; one nigh-silica network and another
favoring the lowest-energy modifier phase. The ionic
potential of a cation is used to measure its tendency to
cause phase separación in silicates.
Phase separation in ceramic glazes is typically
encountered in slow-fire compositions high in B203. The
resulting two-framework continuous structure consists of one
low temperature, boron rich, high alkali B03 phase and a B04
phase.
In alumino-silicate glasses with high B-,0-, and low alkali
content, BT' assumes BO^ tetrahedral and some B0? planar
coordination. An addition of alkali and alkaline earth oxide
modifiers ra.'.ses the amount of B04 formed by increasing

154
connectivity, with local change compensation by the alkali.
As the alkali oxide concentration is increased beyond
approximately 16 mole %, a loss in simple tetrahedral
coordination occurs. Less tetrahedral sites are available
for B+3 and Al+3, and they compete for tetrahedral grouping,
with A104 favored [ Paw96] . Planar B03 content increases with
higher alkali and B203. The amorphous phases associated with
BO, and B04 equilibrium tend to separate from the liquid melt.
Consistent optical properties of ceramic glazes are
easier to achieve if phase separation is not a factor.
Opacification and color are easier to control with inert
particulates suspended in the glassy matrix. The following
steps are taken in order to avoid phase separation in fast-
fire glazes [ Toz86, Par73, Paw96, Tay86, Chi97] :
1. B203 contents are held < 10.0 wt %.
2. Seger's rule #2 is followed (Section 2.22), where
the ratio of alkalis to other oxides in the RO group
is < 1:1. The overall alkali content is kept at < 10.0
wt %.
3. Care is taken to avoid water incorporation in frits
during smelting. Introduction of OH” groups tends to
disrupt the glass network and convert B04 to B03. Water
can be minimized by using unhydrated batch raw materials

155
and smelting with methane fuel which tends to liberate
less water during combustion than natural gas.
4. AI2O3 (> 2.0 wt %) and Zr02 additions increase the glass
viscosity and have a marked effect on slowing diffusion-
driven processes such as immiscibility.
5. Phase separation is suppressed with rapid cooling from
the melt.
2.2.5d Viscosity Relationships
Crystallization and melting conditions are markedly
affected by viscosity. Glass viscosity varies enormously
with composition and is a strong function of temperature. As
demonstrated with equations (2.70), (2.71), (2.79) and Figure
2.23, in previous sections of this dissertation, nucleation
rates are a function of l/r| and crystallization rates are
proportion to T/r|. The glass-forming tendency is low and
crystallization rates are high in glasses that exhibit a low
viscosity increase with decreasing temperature near the
liquidus point [ Sim33] .
Viscosity is related to the internal friction of atoms
or groups. Stronger, more complex bonds and low disruption
of glass-forming chains results in higher viscosity. Low
viscosity enhances diffusion, and related process such as

156
nucleation and growth, by increasing atomic mobility and
chemical changes. The addition of oxides to silica always
drops its viscosity. As shown in Figure 2.26, the lowering
is least with alumina and alkaline earth oxides and greatest
with alkali oxides, ZnO and B203, which tend to break up the
silicate lattice [ Dor94] .
At high temperature, glasses exhibit Arrhenius behavior
[ Dor94] :
71 = T|0 exp (Q/RT) (2.91)
where Q is the activation energy for viscous flow and 7p is a
temperature independent coefficient.
Investigators found that glasses follow Arrhenian
behavior at very high temperatures and below the glass
transition temperature (T ) , but the true viscosity is greater
than predicted by (2.91) at intermediate temperatures above
[ Sim93J .
The Vogel-Fulcher-Tamman equation (VFT) has been applied
to calculate r| while accounting for the non-Arrhenian stage:
or
t| = Tlo eXP
A
T-T
O
log 71 = A + B/ (T-T.)
where Tq, A and E are constants.
(2.92)
(2.93)

157
Figure 2.26. Viscosities of some commercial silicate
glasses. [Dor94]

158
Figures obtained from the VFT equation are more accurate
than equation (2.91) for a wide temperature range, but
calculated values begin to diverge from actual viscosities
near T .
Ceramic whitewares manufacturers and researchers in
Europe and the U.S. normally calculate the viscosities of
glass frits and glazes using the VFT equation 2.93. The
constants TQ, A and B are derived with mathematical models
which require parameters that are measured with dilatometry
and heating microscopy techniques [ Bur96, Con97, Sch62,
Pag97] .
The heating microscope measures dimensional changes of a
sample as it is heated. The technique, in principle, is
based on the relationship between dimensional changes of a
sample (AL/L0) and its surface tension (y) and viscosity under
gravitational force (g) during heating. At a fixed
temperature, the sintering relationships are
AL 3y
a = =
Lo 4rir
and for nonisothermal conditions are
(2.94
da (3y) / (4r|r) (2.95)
at - (dT) / (at)
where r is the principal radius of curvature of the material.
As viscosity drops with increasing temperature, the sample's

159
surface tension tends to pull it into a spherical shape and
the force of gravity pulls it downward. Considering that
viscous flow is not in equilibrium during firing, the dynamic
conditions are described with
2y 2y (2.96)
— - pgh - — = ^(dD/dx)
b r
where b is a constant p is sample density, g is gravitational
force, h is the height of the sample and 3d/3x is the rate of
change of kinematic viscosity with thickness.
VFT equation (2.93) constants T , A and B are calculated
for glazes and frits from models which assume the following
characteristic points [ Bur96, Con97, Sch62] :
T| = 1013 poise at the dilatometric Tg
T) = 1010-25 poise at the dilatometric softening
point, Ts
r| = 104.55 poise at the half-sphere point
determined by a heating microscope, I1/2
The models are
T0 =
13Tg— 4.55T1/2+ (10.25TS—13Tg) (T1/2-Tg) / (Ts-Tgi
8-45 — 2.75 (T1/2-Tg) / (Ts - Tg)
A =
10.25 Ts - 13 Tg + 2.75 T.
T - T.
B = (T.-T0) (13 - A)
(2.97)
(2.98)
(2.99)
where all temperatures are in degrees Celsius.

160
These models yield useful viscosity data for
crystallization and dissolution studies involving fast-fire
ceramic glazes and frits because
1. The characteristic viscosity data at T , Ts and T1/2 are
well known for these specific materials.
2. Crystallization and dissolution in fast-fire systems
occurs at temperatures exceeding 900°C [Amo94, Toz86,
Apa94, Con97] . The T , where accuracy of the VFT
equation is reduced, is far below this temperature at
T < 700°C.
g
3. Glasses with high Ts, such as silicate fast-fire frits,
tend to follow Arrhenian-type behavior over most of the
melt-forming temperature range [ Dor94] , which improves
VFT equation accuracy.
Lehmann, Engell and Hellbrugge found a good correlation
between the dimensional changes in a glaze during firing and
the measured viscosity of bulk glaze in a crucible. James
and Norris identified that glaze exhibited Newtonian flow as
an ideal fluid at high temperature, even when loaded with 5%
inert pigment particles. Glazes behave in accordance with
traditional glass viscosity theories and viscosity-
temperature curves [ Par73] .

CHAPTER 3
EXPERIMENTAL PROCEDURES
3.1 Materials and Methods
A total of thirty-two batches were prepared from eight
glass frits loaded with four different zircon-vanadium
pigment concentrations of 0%, 0.5%, 2.0% and 5.0%. The
batches were blended with water and a suspending agent and
applied to white opaque ceramic tile substrates using a wet
spray method. The green coatings were fired to 1000°C,
1050°C and 1100°C peak temperatures utilizing a typical
industrial "fast-fire" ceramic tile firing curve. A total of
96 different fired samples were produced for examination.
3.1,1 Glass Frits and Zr-V Pigment
Compositions of the eight glass frits investigated are
listed in Table 3.1. They incorporate the most cost
effective and environmentally safe oxides for producing
frits. Formulations do not include BaO or PbO which are now
designated as hazardous materials by the EPA. The range of
oxide contents formulated exceeds the range normally utilized
161

Table 3.1. Frits Investigated.
Weight Percent
Oxide Compositions:
A
B
C
D
E
F
(
3
H
Si02
55.
.00
50
.00
55,
. 00
50
.00
55.
.00
50.
.00
55.
.00
50.
,00
b2o3
6.
. 00
6
.00
6,
.00
6
.00
5.
. 00
5.
.00
5.
.00
5.
, 00
Na20
2.
.00
4
.00
2 ,
.00
4 .
. 00
2 .
. 00
4 .
. 00
2.
.00
4 .
,00
K20
3.
. 00
6
.00
3.
.00
6.
.00
3.
.00
6.
. 00
3.
.00
6.
.00
CaO
8.
, 00
8,
.00
8 .
.00
8 ,
. 00
13.
. 00
13.
.00
13.
. 00
13.
.00
A1203
4 .
,00
4 .
.00
4 .
. 00
4 ,
.00
8 .
.00
8.
.00
8.
. 00
8 .
.00
Zr02
8.
.00
8.
.00
8 .
. 00
8.
.00
0.
.00
0.
. 00
0.
. 00
0.
,00
MgO
2.
,00
2.
.00
2.
.00
2.
.00
2.
.00
2.
. 00
2.
.00
2.
.00
SrO
0.
00
0.
.00
12.
, 00
12.
.00
0.
.00
0.
.00
12.
.00
12.
.00
ZnO
12.
00
12 .
.00
0.
, 00
0.
.00
12.
.00
12.
.00
0.
.00
0.
.00
100.
00
100.
.00
100.
,00
100.
.00
100.
.00
100.
. 00
100.
. 00
100.
.00
162

Table 3.l--continued.
Molar Equivalents
(Seger):
A
B
C
D
E
F
G
H
Si02
2.267
1.778
2.459
1.907
1.857
1.494
1.984
1.583
b2o3
0.214
0.184
0.232
0.198
0.146
0.129
0.156
0.123
Na20
0.08
0.138
0.087
0.148
0.065
0.116
0.07
0.123
K20
0.079
0.136
0.086
0.146
0.065
0.114
0.069
0.121
CaO
0.353
0.305
0.383
0.327
0.47
0.416
0.502
0.441
A1203
0.097
0.084
0.105
0.09
0.159
0.141
0.17
0.149
Zr02
0.161
0.139
0.175
0.149
0
0
0
0
MgO
0.123
0.106
0.133
0.114
0.101
0.089
0.108
0.094
SrO
0
0
0.311
0.266
0
0
0.251
0.22
ZnO
0.365
0.315
0
0
0.299
0.265
0
0
Molecular weight
247.7
213.7
268.7
229.3
202.9
179.6
216.8
190.2
Si: A1
23.37
21.17
23.42
21.19
11.68
10.60
11.67
10.6
RO: R02
0.41
0.52
0.38
0.49
0.54
0.67
0.50
0.63
Alkalis .’Other RO
0.19
0.38
0.21
0.42
0.15
0.30
0.16
0.32
Z R2o3
0.31
0.27
0.34
0.29
0.30
0.27
0.33
0.29
I R02
2.43
1.92
2.63
2.06
1.86
1.49
1.98
1.58
Calculated Expansion
Coeff. (Hall)
(50-450°C; X 10-6/C
7.3
8.9
7.8
9.4
7.7
9.3
8.2
9.7
163

164
for glossy fast-fire glazes, conforms to Seger's empirical
rules outlined in Section 2.2.2 and was designed to avoid
phase separation by following the steps outlined in Section
2.2.5c. Frits for producing glossy coatings were evaluated
because the vast majority of whiteware glazes are glossy.
Major variations in compositions between the frits involved
Zr02 content (0% and 8%), replacement of ZnO with SrO, and
alkali:Si02 ratio. The alkalis (K20 and Na20), SrO and ZnO
were of special interest because of their extensive use as
fluxes and their marked influence on modifying the SiO, glass
network. In frits containing no Zr02, A1203 and CaO were
increased to follow industrial practice.
The frits were produced by Ferro Corporation in
Cleveland, Ohio, by laboratory smelting raw materials at
approximately 1400°C, water quenching to avoid unwanted
crystallization, then ball milling to achieve the desired
particle size. Compositions were confirmed using AAS and XRF
analyses outlined in Section 3.2.1.
A zircon-vanadium (Zr-V) blue pigment 41715A dispersible
stain from Ceredec Corporation was the colorant tested. This
is the most commonly used Zr-V pigment in the whitewares
industry. Its mechanisms for color production are detailed
in Sections 2.1.2d and 2.2.4a. Of the three triaxial

165
pigments, blue was chosen because it has intermediate high
temperature stability when compared to Zr-Pr yellow (lowest
stability) and Zr-Fe coral (highest stability) [ Dec93] .
Also, color changes due to precipitated crystal species are
more distinguishable when using blue colorants because most
devitrified phases tend to scatter light of wavelengths
near blue pigment's peak absorption in the red to yellow
region.
.3., 1 , 2 Coatings Preparation and Application
Coatings formulations consisted by dry weight of 0%,
0.5%, 2.0% or 5.0% Zr-V, 2.5% Bentonite B from Milwhite Inc.
and frit. The materials were blended with water and applied
over a 2"x6" opaque ceramic tile body substrate using a wet
spray method.
Bentonite (68.6% Si02, 18.4% A1203, 2.4% Fe203, 1.9% MgO,
1.4% CaO, 1.1% Na20, 6.2% LOI) was necessary for suspension
and avoidance of settling prior to spraying. Levels of Zr-V
pigment concentrations comprised the range applied in
industrial glazes, where 0.5% to 2.0% zircon pigment loading
is most common. The substrate was ceramic wall tile body
which contained a 0.25 mm thick layer of Florida Tile's white
opaque engobe primei. The substrate was prefired before

166
coatings were applied in order to avoid body volatilization
influence on the results.
The thickness of a ceramic coating influences its
opacity (equations 2.31 and 2.32 in Section 2.1.2c). An
opaque substrate is required for implementation of Kubelka-
Munk equations (Section 2.2.4b) for calculation of color
strength due to the pigment. For these reasons, careful
control was maintained to ensure that a constant volume of
coatings solids was applied to each substrate in order to
negate this as a color influencing variable.
A volume of 2.70 cm3 of glaze solids on a 2"x6"
substrate represents average industrial specifications.
Water was added to provide a 1.75 specific gravity for each
wet glaze, and then the application weight which provides
2.70 cm3 volume of solids was calculated with
Vs M
W = —1
P
and
M-MS
P =
S-MS
where
W = coating application weight (grams)
Vs = volume of solids =2.70 cm3
(3.1)
(3.2)
P = weight % solids

167
M = density of solids (g/cm3)
S = specific gravity of coating = 1.75
The procedure used for determining M is outlined in Section
3.2.2 and the results are discussed in Section 4.1.
3.1.3 Firing Curves
Tiles were fired in a relatively small (20 m length)
industrial pilot roller kiln at Florida Tile Industries. The
high volume of air and gas flow in the kiln ensures uniform
temperature distribution and close correspondence between
thermocouple readings and actual tile temperatures. Figure
3.1 displays the time-temperature profiles employed,
consisting of three different curves with peak temperatures
of 1000, 1050 and 1100 degrees Celsius. The shapes of the
curves (i.e., heating and cooling rates, hold at the peak
temperature, etc.) equate to typical industrial fast-fire
tile processing. A previous study of fast-fire tile
processes concluded that the heating rate above the glaze
softening temperature (approximately > 700°C) and the hold
time at the peak temperature influenced glaze development,
but the cooling rate had no effect [ Amo94] . Figure 3.1 shows
a heating rate above 700°C of 43°C/min. and a 4-minute hold
time at the peak temperature.

168
O
o
k_
3
O
CL
E
o
h-
10 15 20 25 30
Time (min.)
35
40 45
Figure 3.1. Time-temperature profiles used to fire the
tiles.

169
Following the initial firing of samples, preparation and
firing of tiles with frits A and E were repeated to verify
test method consistency. No visual differences were observed
between initial and repeated tests.
3.2 Materials Characterization and Analytical
Techniques
3.2.1 AAS and XRF
Smelted frit compositions were confirmed using a Perkin
Elmer model 2100 atomic absorption spectrometer (AAS) for all
oxides except B203, ZnO, SrO and Zr02. The accuracy of the
device is approximately +0.5% for Si02 and A1203, +0.1% for
alkaline earth oxides and +0.05% for alkalis. A Rigaku
3370 wavelength dispersive x-ray fluorescence (XRF)
spectrometer confirmed the weight percent of the other oxides
to within 0.5% accuracy by implementing an Agel computer
program.
3.2.2 Frit Density Determination
Weights of 50, 100 and 150 grams of dried frit were
added to 300 cc of water and blended thoroughly. The
specific gravity of each mixture was measured and input along
with the weight percent solids into equation 3.2. The

170
equation was solved for M, and the average M for the three
weights was taken to be the frit true density. The bentonite
density was found using the same method. The results were
input into equation 3.1 in order to determine the required
coating application thickness.
3.2.3 Laser Diffraction Particle Size Analysis
A Coulter LS 100 Fraunhofer laser diffraction device was
used to measure frit and pigment particle size distributions.
Frits with very close particle size distributions were
desired for the experiments.
3.2.4 Spectrophotometry and Color Calculations
A Macbeth White-Eye 3000 spectrophotometer was employed
to generate reflectance versus wavelength data from 360 nm to
740 nm. The general method applied was previously outlined
in Section 2.1.4a. The light source was a pulsed xenon flash
lamp which produces light with almost exactly the same
wavelength distribution as daylight. When compared to
tungsten sources, xenon provides more accurate measurements,
and no heating of the sample occurs. Illuminance near 103
cd/m2 is furnished, which relates to cone vision described in
Section 2.1.3.

171
The spectrophotometer was configured to exclude the
specular reflection component in order to eliminate
variations in gloss (from differences in surface roughness
due to poor coatings applications) as a variable in color
calculations. Specular reflectance was also excluded so that
Rmin could be applied in equation 2.64 to calculate color
strength.
The diffraction grating, photodetectors and
microprocessor were arranged to output results in terms of a
D65 noon daylight standard source and a 2-degree angular
subtense.
Spectral reflectance curves were converted into CIE L*,
a* and b* values so correlations to human vision could be
achieved. For each tile sample, measured reflectance,
standard observer weighting functions and light energy values
in terms of wavelength were integrated from 400 nm to 700 nm
using equations 2.47-2.49 in order to calculate tristimulus
values X, Y and Z. These values as well as Xn and Zn factors
shown in Table 2.4 for a 2-degree observer were used to
calculate L*, a* and b* from equations 2.53-2.55.
Equation 2.56 was applied to calculate AE* for
quantifying color changes resulting from firing tiles to
different peax temperatures. A smaller AE* indicates that a

172
color is less sensitive to variations in firing temperature
and thus is more stable. A AE* of 1.0 is "just noticeable"
and ANSI specifications for ceramic tile glazes list a
tolerable range of AE* <3.0. L* , a* and b* values at 1050°C
and 1100°C were used to calculate AE* because this is the
peak temperature range typically utilized by ceramic tile
manufacturers.
Absorption factors (K/S) were calculated using equation
2.64 and were employed to quantify color strength. Values of
K/S were calculated using R-values at the wavelength where
maximum absorption of visible light by the pigment occurred.
3.2.5 Gloss Measurements
Gloss or specular reflectance was measured separately
from color, using a Hunter Lab ProGloss PRO-3 gloss meter.
The data were used to supplement the XRD analyses in
identifying devitrification and to provide a better
description of the visual quality of the fired coatings.
Light was directed onto the samples at 60° from
perpendicular, which is standard for glazes and corresponds
to an index of refraction of n = 1.7 according to Brewster's
law (equation 2.27, Section 2.2.1b).

173
3.2.6 Heating Microscopy
Melt viscosity and flow of coatings batched with 2.0%
Zr-V pigment were studied using a Misura 2.0 Heating
Microscope (HSM). Dry batches were pressed into small samples
approximately 1x1x5 mm in dimension. The samples were heated
in the microscope's furnace at a rate of 43°C/min. up to
1150°C peak temperature in order to simulate the experimental
firing curves. Sample images were acquired during heating
with a CCD TV camera and lens digital system, with a
resolution of 1 pm. Photographs of the samples at various
temperatures were generated and their geometries were
compared and related to flow behavior. Parameters such as
sample height, width, contact angle and their temperature
derivatives are automatically measured by the HSM software to
as low as 0.1% dimensional change.
Half-sphere temperatures (T1/2) established using heating
microscopy, and dilatometric T and T_ data were entered into
g
equations 2.97-2.99 and then 2.93 in order to generate log r)
versus temperature plots based on the models described in
Section 2.2.5d.

174
3.2.7 Dilatometrv
Dry coatings batches were pressed into 25-50 mm long
rods and heated with a Netzsch dilatometer at a rate of
43°C/minute to just above the glass softening temperature.
Both T and T. were identified and coefficients of thermal
g s
expansion were calculated.
3^ 2,-8 X-Ray Diffraction (XRD)
X-ray diffraction scans were performed on all unfired
and fired coatings containing 2.0% Zr-V pigment. This
colorant loading level was selected for the analyses because
a range of 0.5-2.0% zircon pigment is most commonly used in
industry, but the 2.0% level would better conceal the
substrate and thus improve the XRD quantitative analysis
accuracy.
The XRD scans and JCPDS cards were used to qualitatively
identify all crystalline phases that developed during firing.
A quantitative analysis of zircon content in the unfired and
fired coatings served to measure any pigment dissolution or
zircon crystallization in frits containing Zr02. These
results were equated to the color strength and stability of
the coatings.

175
A Siemens D500 Diffractometer scanned samples at a 20
range of 10-60° with a step size of 0.02° 20 and a counting
time of 2 seconds/step using CuKa radiation. In order to
achieve maximum diffraction intensity and to ensure X-rays do
not detect phases under a coating, the thickness of a coating
(t) should be [ Ale74]
3.2 p sin 0
t > (3.3:
HP'
where p/p is the mass absorption coefficient, p' is the
coating density and 0 is the XRD scan angle. Calculations
with (3.3) using the mass absorption coefficients (CuKa) of
the elements in the coatings at the minimum application
thickness tested revealed that coating densities > 1.8 g/crrr
are sufficient for avoiding transmission through to the
engobe substrate. Section 4.1 provides the results of the
frit density determinations.
Measurement of zircon content first involved scanning
unfired coatings with known quantities of zircon from 50-57°
20 and computing the integrated areas under the 53.5° 20
reflection peaks. A 20 of 53.5° for zircon corresponds to
[ 312] Miller index reflection. A zircon weight % versus
integrated intensity baseline curve was established from the
unfired standards using regression methods. This

176
relationship was applied to estimate the quantity of zircon
in fired coatings based on their measured integrated [ 312]
intensity.
Previous studies found that zircon forming at the glaze
surface has a strong orientation along the c-axis lying in
the plane of the surface. This causes substantial
exaggeration of [ hOO] and [ hkO] lines. Therefore, mixed
index lines such as [ 312] are less susceptible to surface
orientation effects and more accurate for measuring zircon
content in ceramic coatings [ Blo93, Blo94a, Cas97] .
3.2.9 Scanning Electron Microscopy (SEM) and Energy
Dispersive X-Rav Spectroscopy (EDS)
A JEOL JSM-6400 SEM was used to generate micrographs of
crystalline phases present in the fired coatings, at 1000X
and 6000X magnification. Crystalline precipitate sizes,
distributions and morphologies resulting from different frits
and firing temperatures were compared and related to color
development.
Crystalline species that were difficult to identify with
XRD due to strong preferred orientation were analyzed with a
Tracor System II EDS under SEM view. A semi-quantitative
analysis of elements present in unidentified crystals

177
assisted in distinguishing their phase. Samples evaluated
with EDS were coated with carbon rather than gold-palladium
because carbons peaks overlap less with the peaks of other
elements in the frits.
3.3 Statistical Methods for Deriving Equations
Rigorous statistical analyses of the results data were
conducted for the purpose of quantifying variable
relationships. Several linear and nonlinear equations were
developed from the experiments:
1. L* (perceived lightness) versus b* (perceived
yellowness-blueness).
2. L*, a* (perceived redness-greenness) and b* versus K/S
(pigment absorption factor) (3 equations).
3. K/S versus frit composition for Zr-V loadings of 0.5%,
2%, 5% and 0-5% (4 equations).
4. AE* (color change from 1050°C to 1100°C peak
temperature) versus frit composition for Zr-V loadings
of 0.5%, 2%, 5% and 0-5% (4 equations).
5. Weight percent zircon versus zircon [ 312] XRD integrated
peak intensity.

178
6. K/S, L* , a* and b* versus weight percent zircon
precipitated from frit loaded with 2% Zr-V pigment (4
equations).
7. Log T| versus temperature and frit composition with 2%
Zr-V loading.
Equations 1-7 were created using experimental design and
analysis computer software [ Bow87] . Independent variables
with calculated absolute t-values >1.9 were considered to
have significant effects (95% confidence level) on the
dependent variable and were included in the equation. A t-
value is given by [ Bow90]
where bi is the regression coefficient for the ith variable
and S(bi) is the standard deviation of the ith coefficient.
Equation coefficients were generated with a least
squares regression method [ Dow83] . The procedure involves
minimizing the sum of squares of the residuals by
differentiating with respect to each unknown coefficient,
setting the equations equal to zero and solving for the
coefficients using matrix methods. Linear, nonlinear and
interactive relationships were identified using this method.

179
Other statistics were also applied to confirm equation
validity. Coefficients of multiple correlation (R2) were
computed to indicate the total variability of the dependent
variable explained by an equation.
The R2 is given by [ Bow90]
(n-p-1) (S ) (3.5)
R = 1 L—
(n - 1) (STOT)
where n is the number of experiments, p is the number of
terms in the equation, ST0T is the standard deviation of
dependent variable actual values and S, . is the standard
deviation of the residuals. The expected accuracy of a
statistically derived equation is +2S, . , with a 95%
y x
confidence level.'
Statistics that were evaluated in order to detect
outlier experiments, unidentified time effects and model
overfitting included externally studentized (E.S.) and
standardized (Std.) residuals, Cook values and Durbin Watson
(DW) statistics. Absolute values of E.S. or Std. residuals
>3.0 or >(1.5X the next largest residual) are considered
outliers. For DW < 1.0, unidentified time-related effects
exist, while DW > 2.5 suggests an overfitted equation.
Comparisons of Cook values were made to determine the impact
of each individual experiment on the overall equation.

CHAPTER 4
RESULTS
4.1 Frit and Pigment Properties
AAS and XRF analyses confirmed the compositions listed
in Table 3.1. Trace quantities of Fe203 (<0.08%) were also
found in each of the frits.
Densities, mean particle diameters and coatings
application weights calculated based on the frit densities
are listed in Appendix C. Frit mean particle diameters
ranged from 20 to 28 microns with mean/median ratios of 1.50-
1.75. These values conform to industrial standards. The
pigment particle size distribution was typical, with a mean
of 8.9 microns and a ±2 standard deviations range of 0.4 to
12.29pm. Frit densities ranged from 2.70 to 2.91 g/cm3, which
were well above the minimum of 1.80 g/cm3 (Section 3.2.8)
required to avoid x-ray transmission through to the
substrate. This assumes the densities are not significantly
lowered during fast-fire ceramic tile processing.
180

181
4.2 Color of Fired Coatinas
Appendix D tabulates the frit composition, Zr-V loading
and peak firing temperature versus L*, a*, b*, AE* , K/S,
gloss and glaze fit (i.e., crazing) observations for all 96
experiments. None of the optical properties were found to
correlate to differences in the coatings application weights
which were required to supply a constant volume of solids on
the substrate.
Normally, the appearance of a colored opaque ceramic
tile glaze is uniform throughout the entire surface while
translucent coatings are darker at the tile edges. Greenware
tile edges have a higher pressed density than the center of
the piece. The resulting lower pore volume at the edges
causes water to be absorbed from the applied wet coating at a
slower rate. The longer drying time allows for fine pigment
particles to segregate to the surface above coarse frit
particles, due to the pigment's slower settling rate. This
segregation at the edges results in a denser color layer and
a darker fired appearance only if there is not enough
dispersed pigment and/or crystalline phases present to
produce a completely opaque coating.
Coatings with frits A-H fired to 1000°C and 1050°C, and
frits A, B, C and K fired to 1100°C appeared to be opaque.

182
Thus, their colors are a function of incident light reflected
by the pigment crystal structure and other species that
precipitated in the coatings during firing, minus light
absorbed by the vanadium metal atoms on the pigment lattice.
Effects due to transmission of light through to the engobe
substrate were not evident.
Coatings with frits D, E, F and G fired to 1100°C
appeared to be slightly translucent. Thus, their colors are
mainly a function of reflection by the Zr-V pigment and other
crystalline phases in the glasses, minus absorption by the
vanadium. In addition, a small amount of reflection from the
opaque engobe substrate background may have influenced the
appearance of these coatings. The engobe was very light and
slightly yellow (Figure 4.1), so it would tend to lessen the
blue color from the Zr-V.
The crystallization and pigment dissolution which caused
differences in the optical properties of the coatings are
detailed in Section 4.5 and Chapter 5.
4.2.1 Spectral Reflectance Curves
Figure 4.1 shows the spectral reflectance versus
wavelength profiles for the unfired batch materials and the
engobe substrate. Maximum absorption by the Zr-V pigment

Reflectance
183
-♦-Average of Frit
Powders A-H.
-♦-Average of
Unfired Coatings
A-H.
-♦-Tile Engobe
Substrate
-B- Bentonite
Powder.
Zr-V Pigment
Powder.
Wavelength (nm)
Figure 4.1. Spectral reflectance of unfired raw materials
and the engobe substrate backing (coatings contain 2% Zr-V,
2.5% bentonite and 95.5% frit by weight).

184
occurs at 640 nm. The Zr-V reflectance value of 22 at 640 nm
may be artificially high because powders were measured in
clear plastic bags that may scatter a small quantity of
light. Maximum reflectance from Zr-V within the visible
spectrum occurred from 440-480 nm. Thus, Zr-V absorbs yellow
to red light and reflects blue to green.
From equation 2.1, Zr-V characteristic peak absorption
or Rmin at 640 nm equates to 1.94 electron volts of energy.
This is the difference in energy between split d orbitals in
V+4 which allows electron transitions from the ground to
excited state. Dissolution of the Zr-V pigment into the
glass structure during firing would change the coordination
about V+4, resulting in a loss of split orbitals and the
energy gap that causes absorption of visible light. Hence, a
reduction in the absorption of 640 nm light and a
corresponding loss in blue color would occur. Therefore, K/S
calculated with equation 2.64 using reflectance values at 640
nm could be applied as an indicator of pigment color strength
and dissolution.
In Figure 4.1, the reflectance of unfired coatings is a
weighted combination of the frit, bentonite and Zr-V curves.
The unfired coatings were opaque; therefore, their
reflectance curves were not influenced by the engobe backing.

185
Frit and bentonite increase the amount of yellow light
reflected and thus lower the quantity of 640 nm light
absorbed by the pigment. Spectral curves of all unfired
coatings were compared, and one trial exhibited a
significantly lower reflectance of yellow light, indicating a
batching error where the bentonite was missing. This batch
was repeated and yielded a reasonable reflectance curve.
Figures 4.2 and 4.3 display changes in reflectance
curves of fired coatings due to pigment loading. Increases
in pigment content caused the curves to shift towards the Zr-
V pigment powder profile in Figure 4.1. Higher Zr-V resulted
in a maximum reflectance shift towards lower wavelengths
which approach 460 nm (blue light) and reduced reflectance
values above 460 nm. Comparisons of Figure 4.2 and 4.3
demonstrate the extreme variation in optical properties
resulting from different frits. With 2.0% Zr-V content, frit
C fired to 1000°C had the highest absorption at 640 nm (R =
15.0) while frit H fired to 1100°C yielded the lowest (R =
51.6). Spectral properties of the unglazed engobe substrate
were found to be stable over the firing range investigated
and thus did not cause color shifts due to changes in peak
temperature. Appendix E reveals spectral reflectance curves
for all 96 fired coatings.

Reflectance
Figure 4.2. Spectral reflectance of coatings batched wi
frit C, fired to 1000'C.

Reflectance
187
Wavelength (nm)
0% Zr-V 0.5% Zr-V 2.0% Zr-V 5.0% Zr-V
Figure 4.3. Spectral reflectance of coatings batched with
frit H, fired to 1100'C.

188
Figures 4.4, 4.5 and 4.6 illustrate peak firing
temperature effects on reflectance from coatings batched with
2.0% Zr-V. At 1000°C, frit A produced the lowest absorption
at 640 nm and frit C the highest, while the overall shapes of
all the curves are alike. From 1000°C to 1100°C several
changes occurred. Peak reflectances from E, F, G and H, the
frits containing no Zr02, shifted to higher wavelengths. From
1000°C to 1100°C, maximum reflectance raised from 460 nm to
approximately 490 nm for E and G, and 500 nm for F and H.
Reflectance at 640 nm increased slightly with frits A, B, C,
D and to a much greater degree with frits E, F, G and H. The
data show a strong decrease in the amount of yellow light
absorbed from fired coatings batched with frits containing no
Zr02 as the peak firing temperature was raised from 1000°C to
1100°C.
4.2.2 Pigment Absorption Factors (K/S)
Absorption factors (K/S) calculated using R640nm values
are plotted versus weight percent Zr-V in Figure 4.7.
Although Figure 4.7 confirms a previous study which found a
linear correlation between K/S and pigment concentration
[ Epp97b] , the strong influence of matrix conditions on the
slope of the curve is evident. The relationship between K/S

Reflectance
189
Figure 4.4. Spectral reflectance of coatings batched with
2.0% Zr-V, fired to 1000*C.

Reflectance
190
-4-Frit A
Frit B
-x- Frit H
—a— Frit F
Frit E
Frit G
FritC
—)— Frit D
Figure 4.5. Spectral reflectance of coatings batched with
2.0% Zr-V, fired to 1050'C.

Reflectance
191
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0 ♦ h-
360 400 440
480 520 560 600
Wavelength (nm)
—i
640 680
720
-x-FritH
—A—Frit F
-a-Frit A
—Frit B
—•—Frit G
—x— Frit E
FritC
—+- Frit D
Figure 4.6. Spectral reflectance of coatings batched with
2.0% Zr-V, fired to 1100"C.

192
Figure 4.7. Pigment absorption factors versus weight percent
Zr-V batched in coatings fired to 1000‘C, 1050'C and 1100‘C
peak temperature. Labeled are data points and linear
regression lines that represent frit/peak temperature
combinations which yielded the highest (C, 1000*C) and lowest
(H, 1100’C) K/S values.

193
and Zr-V content remains nearly linear for each frit
composition/peak firing temperature combination, but the
slopes of the curves vary greatly with changes in frit
composition and temperature. The temperature and frit
composition effects on K/S are quantified with equations
derived in Section 4.4.1 of this dissertation.
4.2.3 Color in L*, a* and b* Values
Figures 4.8 and 4.9 compare color values which relate to
human vision, for coatings fired to 1050°C and 1100°C,
respectively. Good correlation between trends at different
pigment loadings confirms test method repeatability. The a*-
value (redness-greenness) plot at 5.0% Zr-V was the only
ambiguous data. Otherwise, color progressed darker (lower
L*), greener (lower a*) and bluer (lower b*) with increasing
Zr-V pigment concentration.
At 1050°C, frits A, B, F and H produced the lightest,
least blue color while C and D are the darkest and bluest.
At 1100°C, F and H overtake A and B as the lightest colors
and C and D remain the darkest and bluest. Also at 1100°C,
frits E, F, G and H overtake A and B for generating the least
blue. The blue color loss in frits without ZrO~, (E-H) with

100.0
ro
>
*
80.0
60.0
40.0
10.0
5.0
0.0
-5.0
| -10.0
£ -15.0
¿ -20.0
-25.0
-30.0
-35.0
-40.0
0% Zr-V -m- 0.5% Zr-V 2.0% Zr-V 5.0% Zr-V
8%Zr02
12% ZnO
5% Alkali
8% Zr02
12% ZnO
10% Alkali
8% Zr02
12% SrO
5% Alkali
8% Zf02
12% SrO
10% Alkali
No Zr02
12% ZnO
5% Alkali
No Zr02
12% ZnO
10% Alkali
N0Z1O2
12% SrO
5% Alkali
No Zr02
12% SrO
10% Alkali
Figure 4.8. Color values of coatings batched with frits
(A-H) and Zr-V pigment, and fired to 1050'C.

195
o
_3
ni
>
h
A;
Bi
C;
Bl
El
El
Ql
til
8% Zr02
12% ZnO
5% Alkali
8% Zr02
12% ZnO
10% Alkali
8% Zr02
12% SrO
5% Alkali
8% Zr02
12% SrO
10% Alcali
No Zr02
12% ZnO
5% Alkali
Nc _r02
12% ZnO
10% Alkali
No Zr02
12% SrO
5% Alkali
No Zr02
12% SrO
10% Alkali
0% Zr-V
0.5% Zr-V
2.0% Zr-V
5.0% Zr-V
Figure 4.9. Color values of coatings batched with frits
(A-H) and Zr-V pigment, and fired to 1100'C.

196
increasing peak temperature correlates to the shifts in their
spectral curves, noted in Section 4.2.1, Figures 4.5-4.6.
From 2% to 5% Zr-V, the observed reverse from increasing
green appearance (-a*) with pigment loading to a lessening of
green (shown in Figures 4.8 and 4.9) can be explained by
comparing the coatings spectral reflection curves, standard
observer weighting functions in Figure 2.13, and equations
2.47-2.49 and 2.53-2.55. The a*-value magnitude is
nonlinearly related to the ratio of x : y weighting
functions. As Zr-V concentration in the coatings is raised,
peak reflectance shifts towards lower wavelengths near 460 nm
and reflection of light greater than 460 nm drops. In
comparison, the x magnitude drops from 600 nm to 500 nm where
a minima is reached, then abruptly increases with decreasing
wavelengths to 450 nm. In this wavelength interval, the y
magnitude continues to drop and the curves cross at about 475
d x d y
nm, where > . Therefore, as Zr-V concentrations from
dX dX
2% to 5% shift a greater portion of reflectance towards
wavelengths between 450 nm and 500 nm, at one point the
increasing x curve begins to dramatically overtake the y
curve, resulting in a redder appearance. This explains the
abrupt reverse in a*-value trend with increasing pigment
levels. Care should be taken to avoid formulations which

197
create this phenomenon; otherwise, slight batching errors
could result in large deviations in the color perceived.
Figures 4.10, 4.11, 4.12 and 4.13 present color values
for coatings fired to different peak temperatures with Zr-V
concentrations of 0%, 0.5%, 2.0% and 5.0%, respectively.
Figures 4.11-4.13 use the same y-axis color scales, but
Figure 4.10 required a blown up scale because unpigmented
coatings resulted in less color variations.
In Figure 4.10 where no pigment was present, all
coatings progressed greener at higher temperatures although
the a* scale shown represents a very small noticeable
difference (AE*=2.0). With increasing temperature,
noticeable lightening of frits without Zr02, E, F, G and H,
well as a lowering of yellow (b*) in all frits except E, F
and H was observed.
At 0.5% Zr-V (Figure 4.11), the most significant
variations in color occur in the b*-value. From 1000°C to
1100°C peak temperature, E, F, G and H (frits without Zr02)
became less blue (higher b*), while the blue color of A-D
(frits with Zr02) was stable. Frits F and H lost all blue
color at 1100°C and appear slightly yellow. The yellowing
shifts the color towards that of unpigmented coatings.
as
Figure 4.12 shows the same trends for 2.0% Zr-V as

198
10.0
ál
B;
Qi
—
Dl
E:
El
Qi
til
8% Zr02
12% ZnO
5% Alkali
8% Zr02
12% ZnO
10% Alkali
8% Zr02
12% SrO
5% Alkali
8% Zr02
12% SrO
10% Alkali
No Zr02
12% ZnO
5% Alkali
No Zr02
12% ZnO
10% Alkali
No Zr02
12% SrO
5% Alkali
No Zr02
12% SrO
10% Alkali
-*-1000C -m- 1050C -*-1100C
Figure 4.10. Color values of coatings batched with frits
(A-H) and no Zr-V, and fired to 1000’C, 1050’C or 1100*C.

199
©
3
re
>
10.
5.
0.
-5.
-10.
-15.
-20.
-25.
-30.
-35.
-40.
0 --
0 --
0 --
0 --
0 --
0 --
0 --
0 --
0 --
A:
Bl
Qi
Bi
E;
E:
Qi
til
8% Zr02
12% ZnO
5% Alkali
8% Zr02
12% ZnO
10% Alkali
8% Zrf)2
12% SrO
5% Alkali
8% Zr02
12% SrO
10% Alkali
No Zr02
12% ZnO
5% Alkali
No Zr02
12% ZnO
10% Alkali
No Zr02
12% SrO
5% Alkali
No Zr02
12% SrO
10% Alkali
1000C
1050C
1100C
Figure 4.11. Color values of coatings batched with frits
(A-H) and 0.5% Zr-V, and fired to 1000'C, 1050’C or 1100’C.

200
10.0
5.0
0.0
-30.0
-35.0
^0.0
A;
Bl
Ql
D:
El
E:
Ql
til
8%Zr02
12% ZnO
5% Alkali
8% Zr02
12% ZnO
10% Alkali
8% Zr02
12% SrO
5% Alkali
8%Zr02
12% SrO
10% Alkali
No Zr02
12% ZnO
5% Alkali
No Zr02
12% ZnO
10% Alkali
No Zr02
12% SrO
5% Alkali
No Zrt32
12% SrO
10% Alkali
-•-1000C 1050C -A-1100C
Figure 4.12. Color values of coatings batched with frits
(A-H) and 2.0% Zr-V and fired to 1000'C, 1050'C or 1100'C.

201
Figure 4.13. Color values of coatings batched with frits
(A-H) and 5.0% Zr-V, and fired to 1000'C, 1050*C or 1100'C.

202
demonstrated with Figure 4.11, but to a greater degree.
Color intensity in frits without Zr02 (E-H) was significantly
reduced with increasing temperature. Of those frits,
compositions with 10% alkalis (F and H) had lower color
stability over the range of firing temperatures than frits
which contained 5% alkalis (E and G).
Figure 4.13 exhibits color values at each peak
temperature with coatings containing 5.0% Zr-V. The L* and
b* changes with temperature for frits E-H are in the same
direction but with less magnitude than coatings with 0.5% and
2.0% Zr-V. This indicates that the higher pigment
concentration provides less color sensitivity to the glass
matrix or pigment dissolution changes due to temperature.
However, the effect of temperature on perceived greenness is
shown to be greater than with lower pigment concentrations.
The increased sensitivity of a* to temperature is due to the
shifting of a great portion of the spectral curve past the x
weighting function minima near 500 nm. From 1000°C to 1100°C
peak temperature, color progressed greener in frits E, F, G
and H with 5.0% Zr-V.
Figure 4.14 relates the relationship between observed
lightness and blueness for all 96 coatings trials. The Zr-V
pigment creates color which becomes exponentially darker as

203
b*-Value
Figure 4.14. Relationship between lightness (L*) and
blueness (-b*) of coatings batched with Zr-V pigment and
fired to 1000'C, 1050*C and 1100’C peak temperature. Note
b*>0 in coatings without Zr-V.

204
blueness increases. A second order polynomial relationship
well conforms (R2=.96) to the experimental data. Lower L* and
b* values resulting from higher Zr-V concentrations yielded a
better fit to the equation. Points on the graph where b* < -
16.4 depict pigment concentrations of 2.0% or 5.0%. Positive
b* values represent coatings which accommodate no Zr-V. The
figure indicates that at a low Zr-V contents, the
relationship between L* and b* is less predictable. The
discussion in Section 5.1.2 regarding crystallization and Zr-
V dissolution will provide insights into the causes of this
observation. Figure 4.14 also substantiates the use of
either L* or b* for evaluating visual blue color saturation
with Zr-V pigments.
All three color parameters were found to be logarithmic
functions of the Zr-V pigment absorption factor (Figure
4.15). Values of R2 for the relationships are highest for L*
and b* (R2 = .93). The redness-greenness of Zr-V
systems is less predictable from R640nm (R2 = .87). With the
equations in Figure 4.15, statistical models (in Section 4.4
of this dissertation) for predicting K/S can be converted
into numbers that are proportional to human vision. The
absorption factor was chosen as a primary variable for
quantifying color strength due to its direct correlation with

205
Figure 4.15. Zr-V pigment absorption factor relationships
with L*, a* and b* color values of fired coatings.

206
pigment concentration. Thus, Zr-V dissolution and/or
reflection of light from crystalline precipitates which
counters the characteristic 640 nm absorption peak are easier
to evaluate with K/S values. However, as Figure 4.15
illustrates, changes in K/S are most visible for K/S < 1.0
and least observable for K/S ^ 2.5. Any Zr-V contents <2.0%
yielded K/S <2.4 for all frits and processing conditions.
Thus, changes in K/S and corresponding pigment concentration
in frit systems which incorporate Zr-V pigments are most
noticeable for Zr-V < 2.0%.
4.2.4 Color Stability
Color variations resulting from an increase in peak
firing temperature from 1050°C to 1100°C are represented with
AE* values in Figure 4.16. Since previous data showed that
K/S never increased with temperature, AE* indicates the
magnitude of color loss which correlates to human vision.
The AE* from frits containing Zr02 (A-D) at all pigment
concentrations, and unpigmented frits with no ZrCc (E-H, 0%
Zr-V) fell below the ANSI requirement of AE* < 3.0 for
compatible ceramic tile glazes. However, manufacturers
prefer to conform to AE* <1.0 for matching surface
appearances. Only frits A and B, which unlike any of the

Color Change (Delta E*) from 1050C to 1100C Peak Temperature
207
Figure 4.16. Color changes (AE*) due to a variation in peak
firing temperature from 1050*C to 1100‘C for coatings batched
with frits (A-H) and Zr-V pigment.

208
other frits contained both Zr02 (8%) and ZnO (12%), met this
standard. Pigmented frits E-H exhibited considerable color
changes due to the 50°C increase in peak firing temperature.
These data concur with the figures in Sections 4.2.1 and
4.2.3 in identifying that frits with no Zr02 yield the most
temperature sensitive color. Frit composition and
temperature influences on AE* are quantified with statistical
models in Section 4.4.2.
4.2.5 Specular Gloss
Except for coatings with frit B, there was very little
influence of pigment concentration on gloss, as shown in
Figure 4.17. Also excluding B, frits containing Zr09 had more
stable gloss over the 1000°C to 1100°C peak temperature
range, while frits E, F and G exhibited increased gloss at
higher temperatures. Frit H yielded by far the lowest gloss
in all cases. Overall except for B, frits with ZrO, produced
a higher gloss. The gloss results were used in later
sections of this dissertation to collaborate with other
evidence of crystallization in some coatings.

209
(ft
V)
o
O
_ro
3
O
0)
a
V)
V)
V)
o
JS
3
O
O
a
(0
100
80
Fired to1050C
O 60 --
40
20 --
â–  0% Zr-V
â– 0.5% Zr-V
â– 2.0% Zr-V
—X— 5.0% Zr-V
A:
A
B;
4
Qi
B;
1
E;
p
E:
Qi
til
8% Zr02
12% ZnO
5% Alkali
8% Zr02
12% ZnO
10% Alkali
8% Zr02
12% SrO
5% Alkali
8% Zr02
12% SrO
10% Alkali
No Zr02
12% ZnO
5% Alkali
No Zx02
12% ZnO
10% Alkali
No Zr02
12% SrO
5% Alkali
No Zr02
12% SrO
10% Alkali
Figure 4.17. Specular gloss of fired coatings at a 60' angle
of incidence.

210
4.3 Viscosity of Coatinas Purina Heating
Results in viscosity measurements for coatings batched
with 2.0% Zr-V are detailed in this section and are related
to crystallization and pigment dissolution data in Section
4.5. In Section 5, these data are further linked to color
properties obtained with all three pigment loadings, based on
previous correlations found between color and Zr-V
concentration. The goal was to determine if the viscosity of
a coating at high temperature can be applied as a single
indicator of color stability and strength for industrial
quality control purposes.
4.3.1 Heating Microscope Images
Several of the images acquired using heating microscopy
techniques are pictured in Figures 4.18 and 4.19. The
dimensional changes versus temperature shown reveal distinct
patterns:
1. Raising the alkalis from 5 to 10% by weight at the
expense of silica (frits B, D, F and H) caused the
greatest increase in molten flow at all temperatures.
2. Overall, frits with Zr02 (A-D) flowed less at low
temperature up to approximately 1000°C and required
higher temperatures to reach the "sphere" point (contact

211
900°C : 1000°C:
A:
8% Zr02
12% ZnO
5% Alkali
B1
8% Zr02
12% ZnO
10% Alkali
Qi
8% Zr02
12% SrO
5% Alkali
Di
8% Zr02
12% SrO
10% Alkali
E:
No Zr02
12% ZnO
5% Alkali
E;
No Zr02
12% ZnO
10% Alkali
Qi
No Zr02
12% SrO
5% Alkali
Hi
No Zr02
12% SrO
10% Alkali
1100°C :
Figure 4.18. Heating microscope images of coatings batched
with frits (A-H) and 2.0% Zr-V, at 900’C, 1000’C and 1100‘C.

212
Stage 1: Stage 2: Sphere! 1/2 Spheral Stage.il
Ai
8% Zr02
12% ZnO
5% Alkali
Bi
8% Zr02
12% ZnO
10% Alkali
&
8% Zr02
12% SrO
5% Alkali
Di
8% Zr02
12% SrO
10% Alkali
El
No Zr02
12% ZnO
5% Alkali
El
No Zr02
12% ZnO
10% Alkali
Si
No Zr02
12% SrO
5% Alkali
Hi
No Zr02
12% SrO
10% Alkali
Figure 4.19. Heating microscope images of characteristic
stages of flow of coatings batched with frits (A-H) and 2.0%
Zr-V. (The temperature, % of original sample height and
contact angle are listed under each image.)

213
angle = 57-72°, % original height/contact angle =
.80-1.14).
3. Frits containing both Zr02 and high alkali content
exhibited the greatest flow at high temperatures near
1100°C.
The temperature at the 1/2 sphere stage (T1/2) was
combined with dilatometric data to generate log ri versus
temperature curves for each coating, as described in Section
2.2.5d.
4.3.2 Dilatometric T7 and
Dilatometric curves, including identification of glass
transition (T ) and softening (Ts) temperatures, and
calculated coefficients of thermal expansion (listed as
C.o.E. under each graph) are displayed in Figures 4.20 and
4.21 for all 2.0% Zr-V coatings. The crazing observed on
tile samples made with frits D, F and H could be attributed
to the high thermal expansion of the frits (91.8, 96.9 and
105.1 X 10“7 K-i, respectively, from 50-450°C) when compared
to the body/engobe substrate (65-67 X 10-7 K-1) . None of the
other frits crazed.

O 50 100 150 200 250 500 350 400 450 500 550 600 650 700 750 000 650 MO
Temperature pCj
Date: 13/02/98 Comment: HP frit (1050*c7_45 min.) Florida Tile
Lab.: Ferro I.C.C. RtD
Sample: FAB 30.55 mm C.o.E. ( 20/400): 66.1 10-7 K-l
Zero: ZERO-10 ( 50/400): 69.5 10-7 K-l
Platon: Quartz(100/400): 76.0 10-7 K-l
0 50 100 150 200 250 500 350 400 450 500 550 600 650 700 750 600 650 900
Temperature PC|
Date:
26/01/98
Comment: HP frit
(1050‘C
45 Bln.) Florida Tila
Lab.:
Ferro I.C.C. RtD
Sample:
FBB 49.85 mm
C.o.E. ( 20/400):
76.6
10-7 K-l
Zero:
ZERO-10
( 50/400):
81.9
10-J K-l
Platon:
QUARZ
(100/400):
87.0
10-7 K-l
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 600 650 900
Temperature rC|
Date:
Lab.:
16/02/98
Ferro I.C.C. RtD
Comment: MP frit
<1050*C, 45 Bln.) Florida Tila
Sample:
FCA 27.15 ma
C.o.E. ( 20/400):
70.7 10-7 K-l
Zero:
ZERO-10
( 50/400):
75.4 10-7 K-l
Platon:
QUARZ
(100/400):
79.0 10-7 K-l
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 600 650 900
Temperature [*C3
Date: 16/02/98 Comment: HP frit (1050’C, 45 min.) Florida TÜe
Lab.: rerro Z.C.C. RtD
Sample: FDA 49.80 m C.o.E. ( 20/400): 95.4 10-7 K-l
Zero: XERO-10 ( 30/400): 91.9 10-7 K-l
Piston: QUARTZ(100/400) : 97.0 10-7 K-l
Figure 4.20. Thermal dilatometric analyses of coatings A-D batched with 2.0% Zr-V.
(Heating rate of 43*C/min.)
214

SO 100 ISO 200 290 300 350 400 460 500 550 600 OSO 700 730 000 OSO 000
Date:
Lab.:
Sample
Zero:
Platon:
16/02/98
Perro I.C
FEA
ZERO-10
QUARZ
C. R&D
34.60 mm
Coma
C.
lent:
o.E.
MP frit (1050'C
( 20/400): 73.1
( 50/400): 77.8
(100/400): 81.0
45
10-
10-
10-
â– in.
K-l
K-l
K-l
) Florida Tile
Date:
Lab.:
Sample:
Zero:
Platon:
17/02/98
Ferro I.C.C. RfiD
FFA 49.85 aua
ZERO-10
QUARZ
Comment:
C.O.E.
MP frit (1050*C,
( 20/400): 91.2
( 50/400): 96.9
(100/400): 101.0
45 min.
10-7 K-l
10-7 K-l
10-7 K-l
Florida
Tile
•• •**
. -
. -.
~~
*
rtc-ar
Frtttm
CHTo
Mrt
le
rTf»
dikci
V
T»
me
!Q
7
JL
250
\J
/-u
A
~r
V-
1 \
T
>
r
7
v—
200
T
• a /
T
H rif 1
1
-1
r ru '
r~P
3^
A.
1
1
¿j
jL\
!
S
i\
s'
r
i â– 
|
1
S
O
^1.
T0 (5
Xáü
)
3^
1 too-
too-
.
V
80-
/
80-
0-
0
84
10
0 18
0 20
0 28
O 30
0 38
O 40
T«
O 48
npera
0 80
turef
0 88
q
O 60
0 6Í
W 70
0 78
>0 60
10 68
•0 8C
0
0
84
to
0 16
0 20
O 28
<0 30
0 38
0 40
Tei
0 45
Tipers
0 60
turep
0 88
6 60
0 68
O 70
0 78
<0 60
0 68
0 00
10
Date:
Lab.:
Sample:
Zero:.
Platon:
17/02/98
Ferro I.C.C. RfcD
FGA 34.25 mm
ZERO-10
QUARZ
Comment:
C.O.E.
MP frit (1050*C,
( 20/400): 75.0
( 50/400): 80.7
(100/400): 86.0
45
10-7
10-7
10-7
â– in.
K-l
K-l
K-l
Florida
Tile
Date:
Lab:
Sample:
Zero:
Platon:
18/02/98
Ferro I.C.C. R«D
FHA 31.90 mm
ZERO-10
QUARZ
Comment:
C.O.E.
MP frit (1050 C,
( 20/400): 97.6
( 50/400): 105.1
(100/400): 111.0
45 min.)
10-7 K-l
10-7 K-l
10-7 K-l
Florida
Tile
Figure 4.21. Thermal dilatometric analyses of coatings E-H batched with 2.0% Zr-V.
(Heating rate of 43'C/min.)
215

216
The Tg and Ts data disclose the following:
1. Alkali content in the frits had the largest influence on
T . The average Ts with frits containing 5% alkalis was
721°C. When the alkalis were raised to 10% at the
expense of Si02, Ts dropped to 688°C.
2. Changes in SrO versus ZnO caused the greatest shift in
Tg. The average Tg in frits containing SrO was 627°C,
and lowered to 596°C when replacing SrO with ZnO.
3. Substituting SrO for ZnO in the glass structure produced
a significant increase in the ratio of Tg:Ts. Less of a
temperature rise after Tg was required to reach Ts with
SrO present.
No other effects of the oxides on T and T were
g s
distinguished.
4.3.3 Viscosity vs. Temperature
Data from Sections 4.3.1 and 4.3.2, as well as equations
2.93 and 2.97-2.99 in Section 2.2.5d, were used to generate
the log T) vs. temperature curves in Figures 4.22 and 4.23.
The temperature range of interest begins near the average
softening point of the frits (~700°C) and ends at the highest
peak firing temperature tested. This is the temperature
range where crystallization and dissolution can occur.

Log Viscosity (poise)
217
Temperature (C)
Figure 4.22. Log T| versus temperature of coatings batched
with 2.0% Zr-V. Frits A-D incorporate 8% ZrC>2.

Log Viscosity (poise)
218
Temperature (C)
Figure 4.23. Log T| versus temperature of coatings batched
with 2.0% Zr-V. Frits E-H contain no ZrC>2.

219
In both figures, a drop in viscosity across the whole
temperature range is shown for parallel compositions where
only the alkalis/Si02 ratio is increased (i.e., A—>B, C—»D,
E—»F, G—>H). The increase in nonbridging alkali ions in the
glass has a marked effect on weakening the structure and
lowering viscosity. An increase in the slope of the curves
near the softening point of 700°C for frits where SrO
replaced ZnO is also evident. The influence of other oxides
in the frit on viscosity is not obvious from the plots, but a
rigorous statistical analysis of the data in Section 4.4.3
identifies significant variables. In the Discussion, Section
5.1.3, viscosity results will be related in detail to other
phenomena which affected the optical properties of the
coatings.
4.4 Derived Statistical Models
Simultaneous variations of oxide levels between the
frits necessitated the implementation of tedious statistical
methods in order to quantify the oxide's true effects on the
dependent variables. Several equations were derived using
the statistical techniques briefed in Section 3.3. Color
strength (K/S), color stability (inverse of AE* ) and melt
viscosity relationships with coating composition and

220
temperature have been quantified. The equations are useful
for identifyinq the weights of influence of the significant
independent variables on color and melt viscosity, and for
interpolating predictions within the experimental parameters
of this dissertation. However, predictions based on
extrapolations beyond the range of conditions tested during
this research (i.e., firing curves, coating thicknesses,
substrate composition, etc.) will be much less accurate.
Scientific interpretations of the physical basis for the
results of this section are given in Chapter 5, Discussion.
4.4.1 K/S vs. Coating Composition and Temperature
The following mathematical models were developed using
experimental results from each individual pigment loading
(4.1-4.3) and for compositions with Zr-V as an independent
variable (4.4):
ariable
Coefficient
t-Value
K/Sq.5% zr-v = + Y-intercept
1.136
0.191
2.162
5.227
- ZrC>2
+ SrO
+ SrO X Zr02
- SrO X T
21.046 X 10-4
1.84
8.87
-1.88
-5.82
6.65
-3.41
-1.42 (4.1)
- Al203/alkalis 0.654
+ (Al203/alkalis) 2 0.451
T
71.179 X 10-5

221
Variable
Coefficient
t-Value
K/S2.o% zr-v = + Y-intercept
1.912
—
- Zr02
1.449
-1.90 (4
+ SrO
24.224
5.08
+ SrO X Zr02
30.396
9.08
- SrO X T
2.272 X 10
-2 -5.02
- Al2Ü3/alkalis
3.588
-5.63
+ (Al203/alkalis)2 2.585
6.72
Variable
Coefficient
t-Value
K/S5.0% zr-v = + Y-intercept
8.825
—
- Zr02
12.725
-3.21 (4
+ SrO X Zr02
130.854
7.54
- SrO X T
34.558 X 10-4 -1.70
- Al203/alkalis
17.861
-5.39
+ (Al203/alkalis)2 12.277
6.14
Variable
Coefficient
t-Value
K/So-5% zr-v = + Y-intercept
1.254
+ Zr-V/T
799.062
14.76 (4
- Zr02
3.000
-1.47
+ SrO X Zr02
36.632
5.84
- Al203/alkalis
5.480
-2.94
+ (Al203/alkalis)2 3.792
3.38
In equations 4.1-4.4, units for Zr02, SrO, A1203 and
alkalis (Na?0 + KVO) are molar
equivalents of
the oxides in
the frits, Zr-V is the weight ]
percent pigment
added to the
batch and T is the peak firing
temperature in
degrees
Celsius. Although there was a
correlation between molar
equivalents of CaO and A1203 in
. the frits, A1203 provided fo
a better equation fit and was thus included in the models.
Equation 4.2 for 2.0% Zr-V was developed first, using

222
variables whose absolute t-values were £1.9. Then the other
models were derived with the same variables if their absolute
t-values were £1.4. Calculated R2 values of 0.95, 0.95 and
0.88 for equations 4.1-4.3, respectively, verify thorough
explanation of K/S variation in terms of the independent
variables in the models. Equation 4.4, which includes all
Zr-V levels, was less accurate (R2 = .75) than the other
models. Evaluation of residuals (E.S. and Std.) and Durbin
Watson statistics (all 2.0-2.5) revealed no outlier
experiments and validated the models. The precision of an
equation is estimated with the standard deviation of the
residuals (Syy). Values of Sy x for equations 4.1-4.4 are
0.030, 0.169, 0.878 and 0.987, respectively.
The magnitudes and signs of coefficients in all four
equations corroborate the following conclusions:
1. Higher SrO in the frit tends to raise the color strength
(K/S) unless the pigment concentration is high (5%) . In
this case, SrO boosts K/S only in frits containing ZrO,.
2. Increased Zr02 lowers K/S, unless SrO is present (in
place of ZnO). An interaction between SrO and Zr02
boosts K/S at all pigment loadings.
3. The curvilinear (X2-X) effect of the Al203/alkalis ratio
reveals that (a) at low levels, increasing Al203/alkalis

223
tends tc slightly decrease K/S, but (b) at high levels,
K/S is significantly raised as Al203/alkalis is
increased.
4. At a low pigment concentration (0.5%), higher
temperatures cause K/S to drop in all of the frits. At
2% and 5% Zr-V, only frits which include SrO experience
a decrease in K/S with increasing temperature.
Figure 4.24 relates the good fit between actual data,
and equation 4.2 predicted results for all 24 trials with
2.0% Zr-V. Note K/S is predicted to remain constant over ail
three temperature points for frits which include ZnO (A, B, E
and F) rather than SrO (C, D, G and H), indicating superior
color stability with ZnO.
The statistically adjusted effects of individual
variables from equation 4.2 (2.0% Zr-V) are displayed in
Figure 4.25. Data for the plots were generated by inputting
the actual range of composition tested of the x-axis variable
into equation 4.2, while holding values of all other
variables in the equation at their average level. Thus,
Figure 4.25 depicts mathematical interpolations based on the
statistical model. An increase in Zr02 (from 0 to 0.311 molar
equivalents) is shown tc raise K/S in the presence of SrO (at
average of 0.131 equivalents), but decrease K/S when ZnO (at

Pigment Absorption Factor (K/S)
224
Coatings (A-H) Fired to 1000C, 1050C and 1100C
Figure 4.24. Pigment absorption factor actual and equation
(4.2) predicted results for fired coatings batched with 2.0%
Zr-V.

225
Zr02 Molar Equivalents
AI203 / Alkalis Molar Equivalents
Figure 4.25. Individual effects of frit oxides on K/S based
on statistical model (4.2) for 2.0% Zr-V and a peak firing
temperature of 1050°C.

226
average of 0.156 equivalents) replaces SrO. The middle graph
illustrates that higher SrO (with ZnO still present at .156
equivalents) raises K/S to a much greater degree when ZrO., (at
average of 0.078 equivalents) is in the frit. The bottom
plot portrays the curvilinear relationship between K/S and
Al203/alkalis ratio.
Frits manufactured without Zr02 are normally designed
with a higher Al203/alkalis because more A1203 is added to
maintain glass strength and durability. With Zr02 in the
glass, increasing Al203/alkalis (<0.5 molar equivalents)
caused K/S to lower slightly. As Al203/alkalis levels were
increased (>0.80) when no Zr02 was present, a significant
increase in K/S occurred.
The weights of influence of independent variables on
K/S, derived with models 4.1-4.3, are disclosed in Figure
4.26. A variable's percent total influence (%I) was
calculated by
%I * 100 l4-5)
where
K/Si = the maximum range of K/S obtained by
inputting all values tested of independent
variable (i) into the statistical equation,

227
while holding other variables at their
average level.
X K/S^.h = summation of all the K/S- ranges
calculated.
Figure 4.26 does not consider interactive effects but
provides a picture of how Zr-V loading changes the systems.
The graphs reveal that with increasing pigment concentration,
the impact of Al203/alkalis increases, while Zr02 and
temperature become less of a factor. SrO influence remains
similar at all three Zr-V contents.
Cause/effect relationships between the frit oxides and
color results are not easily distinguishable without
statistically adjusted data. This is because more than one
color influencing oxide was varied simultaneously between the
experiments. Thus, regression methods were employed to
separate the true effects of eacn individual oxide and
generate Figures 4.25 and 4.26.
Figures 4.27-4.30 reveal nonadjusted, actual results for
each experiment with 2.0% Zr-V loading. Although less
distinct, the plots show trends which correlate to the
statistical models.

Figure 4.26. Variables weight of influence on K/S, based on
statistical models 4.1 (0.5% Zr-V) , 4.2 (2% Zr-V) and 4.3 (5
Zr-V) .
0.5% Zr-V

229
AI203 / Alkalis Molar Equivalents
Figure A.21. Pigment absorption factor trends with frit
A1203:alkalis ratio of coatings batched with 2.0% Zr-V and
fired to 1000°C, 1050°C or 1100°C. For each group of three
data points at a specific A^C^/alkali, the highest point
represents a coating fired to 1000°C peak temperature; the
lowest point was fired to 1100°C. Lines shown intersect
average K/S values for each group of three data points.

Figure 4.28. Pigment absorption factor versus frit ZnO and SrO molar equivalents of fired
coatings batched with 2.0% Zr-V and fired to 1000°C, 1050°C or 1100°C peak temperature.
N)
CO
o

231
Figure 4.29. Influence of frit ZrC>2, in the presence of SrO
or ZnO, on K/S of fired coatings batched with 2.0% Zr-V and
fired to 1000°C, 1050°C or 1100°C. All data points for
1000°C, 1050°C and 1100°C peak firing temperatures are
plotted. Lines shown intersect average K/S values for each
ZrC>2 and SrO or ZnO combination.

232
0.00
O
o
o
T- -0.20
O
o
o
o
T—
E
o
-0.40
(O
2
•“ -0.60
Q)
O)
C
ra
.c
ü
-0.80
-1.00
â–  8% Zr02
â–  0% Zr02
Figure 4.30. Variations in pigment absorption factor (K/S)
due to changes in peak firing temperature, for coatings
batched with 2.0% Zr-V and frits containing 8% ZrC>2 versus 0%
ZrC>2.
Frits also include the following weight percentages:
A and E: 5% alkalis, 55% SÍO2, 12% ZnO;
D and F: 10% alkalis, 50% SÍO2, 12% ZnO;
C and G: 5% alkalis, 55% SÍO2, 12% SrO;
D and H: 10% alkalis, 50% SÍO2, 12% SrO.

233
Figure 4.27 confirms that when Zr02 and SrO were
present, the highest K/S was achieved and was not dependent
upon Ai203/alkalis. Coatings with frits containing ZnO and
Zr02 exhibited a sharp drop in K/S with increasing
Al203/alkali and produced a weaker color than the SrO + Zr02
combination. Coatings without Zr02 in the frit showed
intermediate color strength, which increased with a higher
Al203/alkalis.
Figure 4.27 also demonstrates color loss with increasing
temperature. For each group of three data points at a
specific Al203/alkali, the strongest color (highest K/S)
represents a coating fired to 1000°C while the lowest K/S
represents a peak temperature of 1100°C. The vertical spread
in each group of data points, which represents color
variation with temperature, is greatest for the coatings
without Zr02 and least when both Zr02 and ZnO are present.
A comparison of K/S values resulting from frits with SrO
and ZnO at 2.0% Zr-V is made in Figure 4.28. An overall
pattern of increasing pigment absorption and color strength
with a reduction in ZnO or higher SrO is evident. Color
degradation from 1000°C to 1100°C was dramatically reduced
with frits containing high ZnO and no SrO. Frits F and H
generated data that slightly deviated from the overall

234
trend. They had in common the lowest Al203/alkali ratio of
the frits with no Zr02 and produced coatings with the lowest
gloss due to high devitrification/crystallization in the
glass.
Figure 4.29 relates the interaction between Zr02 and SrO
or ZnO. For frits that incorporated SrO, K/S increased with
Zr02 content. Absorption of light by the pigment was not
significantly influenced with higher Zr02 when the frit
contained ZnO rather than SrO. The range of K/S with
temperature is again shown to be lower with ZnO. Overall K/S
variation due to peak firing temperature is lower for both
ZnO and SrO frits when Zr02 is present.
In Figure 4.30, frits with equal weight percentages
of silica, alkalis and SrO or ZnO are grouped together on
the x-axis (A and E, B and F, C and G, D and H). In each
case, K/S changes less with peak temperature in frits with
Zr02. The largest drops in K/S stability when Zr02 was
removed (increased vertical distance between two bars)
occurred when shifting to higher alkalis (A-E to B-F and
C-G to D-H) and when SrO replaced ZnO (A-E to C-G and B-F to
D-H) .

235
4.4.2 AE* vs. Coatina Composition
Values of AE* were calculated as the color difference
between coatings fired to 1050°C° versus 1100°C peak
temperature. Since K/S never increased with temperature, AE*
denotes the magnitude of color loss which correlates to human
vision, and the degree of instability of the system. The
following statistical models were developed using
experimental results from each individual pigment loading
(4.6-4.9) and for all compositions with Zr-V as an
independent variable (4.10):
Variable
Coefficient
t-Value
Y-intercept
6.49
—
ZrC>2
47.07
-10.99
SrO X Zr02
49.77
3.79
Al2C>3/alkalis
5.80
1.48
(A1203/alkalis)2
6.10
-2.59
Variable
Coefficient
t-Value
Y-intercept
4.98
—
Zr02
76.04
-24.83
SrO
15.70
9.51
SrO X Zr02
49.29
-3.67
Al203/alkalis
26.84
10.48
(AI2O3/alkalis)2
22.51
-14.59

236
Variable
Coefficient
t-Value
AE* 5.0% zr-V ~
Y-intercept
3.90
—
-
ZrÜ2
46.31
-5.92
(4
+
SrO
8.43
2.42
+
Al203/alkalis
14.01
1.66
—
(Al203/alkalis)2
11.63
-2.29
Variable
Coefficient
t-Value
lnAE*o-5% Zr-V =
Y-intercept
1.89
—
-
Zr02
16.57
-12.99
(4
+
SrO
1.48
1.94
+
SrO X Zr02
20.88
3.52
-
(AI2O3/alkalis)2
0.59
-3.95
+
Zr-V
0.37
2.76
—
(Zr-V)2
6.13 X 10-2
-2.45
Oxides in equations 4.6-4.9 are
in molar equivalents.
As
4.9)
previously noted for K/S equations, although CaO and A1203
fritted molar equivalents correlate, A1203 in the models
provide for better accuracy. Equation 4.7 for 2.0% Zr-V was
developed first using variables whose absolute t-values were
^2.0, then other models were derived using the same
variables, with absolute t ^ 1.4. Calculated R2 values of
0.98, 0.99, 0.94 and 0.92 for equations 4.6-4.9,
respectively, validate the models. Values of S, „ for
equations 4.6-4.9 are 0.60, 0.39, 1.29 and 0.37,
respectively.
Equation 4.9 for all Zr-V levels required a logarithmic
transformation of AE* in order to establish a correlation

237
with the independent variables. The logarithmic dependency
found between K/S and L*, a* and b* in Figure 4.15, Section
4.23, and the linear relationship between K/S and pigment
concentration (reviewed in Sections 4.2.2 and 2.2.4b) explain
this relationship. Small changes in perceived color with
pigment absorption are nearly linear around 0.5%, 2.0% and
5.0% Zr-V areas. Large swings in L* , a*, b* and
corresponding AE*, resulting from significant changes in
pigment content or glass matrix compositional changes that
influence K/S, require a logarithmic function to describe the
whole range.
Formation of equation 4.9 also required the omission of
two outlier data; frit H fired to 1100°C at 2.0% and 5.0% Zr-
V. Calculated Durbin Watson and residual statistics fell
within the acceptable range. Model 4.9 attests that pigment
loading has an influence on color stability. The curvilinear
relationship between InAE* and Zr-V content indicates that an
intermediate pigment level is least visually stable with
temperature. This conforms with Figure 4.16, where AE*
results were overall greater for 2.0% Zr-V than 0.5% or 5.0%.
Although AE* 1050_1100°c and K/S1050 results did not
correlate well to each other (linear correlation coefficient
r=-.33), the same independent variables were found to

238
significantly affect both, albeit with different magnitude
and direction in some cases. Evaluation of the coefficients
in 4.6-4.9 reveals
1. Higher Zr02 in the frit improves color stability (lower
AE* ) .
2. Increased SrO lowers stability (raises AE*).
3. The (X-X2) effect of Al203/alkalis dictates that at low
levels, increasing Al203/alkalis tends to slightly raise
AE*, but at high levels, stability is improved.
Figure 4.31 compares measured and equation predicted AE*
for all eight fired coatings batched within 2.0% Zr-V.
Displayed is the good fit of the model, and the tendency for
increased color sensitivity to temperature when Zr02 is
removed (E, F, G, H). Color variation is further increased
in E-H when the alkalai level is raised (F and H) and when
SrO replaces ZnO (G and H) .
Individual effects of oxide compositions on AE* at 2.0%
Zr-V, calculated by inputting the range of composition tested
into 4.7 and holding the other oxides at their average level,
are displayed in Figure 4.32. Note the y-axis scale of 0 to
18 represents a dramatically noticeable color change. A AE*
^ 1.0 is noticeable with the human eye. Linear trends of
improved color stability with higher Zr02 (0 to .175 molar

239
18
2 16
a
E
4)
I-
J*
re
O
a.
O 12
14
O
©
u>
E
o
«
1X1
re
o
D
03
03
C
re
£
O
i_
o
o
o
10
2
â–  Measured Values
â–¡ Equation
Predictions
jmti
B
H
Frit
Figure 4.31. Delta E* actual and equation (4.7) predicted
results for fired coatings batched with 2.0% Zr-V.

240
E
o
c
•* -X
re
uj £
2 o
0) o
a?
©T_
O) Q
re O
â– c o
O «5
o
SrO Molar Equivalents
3 ®
AI203 / Alkalis Molar Equivalents
Figure 4.32. Individual effects of frit oxides on Delta E* ,
based on statistical model 4.7 for 2.0% Zr-V.

241
equivalents) and lower SrO (.311 to 0 molar equivalents) are
shown. The middle graph illustrates that SrO additions
increase AE*, but the overall color stability is better if
Zr02 is also present. The bottom plot shows increases in
Al203/alkalis at high levels (>0.80 molar equivalents)
provided great improvements in color stability.
The weights of influence of the oxides on AE* for each
pigment loading, based on 4.6-4.8 equations coefficients, are
disclosed in Figure 4.33. Percentages were calculated using
equation 4.5 and the procedure described in Section 4.4.1.
The bar graphs reveal that Zr02 content in the frit has the
greatest overall influence on color stability, followed by
Al203/alkalis and SrO. As the pigment concentration
increased, SrO had a slightly larger influence on color
stability.
Experimental, nonstatistically adjusted data which
relate the oxide influence on color stability are disclosed
in Figure 4.34 There are some discernable patterns shown
which appear to substantiate the statistical interpretations.
Systems containing Zr02 are the most stable. Replacement of
SrO with ZnO also improves stability, and to a greater degree
when no Zr02 is present. Higher Al203/alkalis results in a
significant increase in stability when no Zr02 is in the frit

Cl)
tr
3
CD
\
a
(/)
iQ
CD
C
a
h
•
0)
00
o
p
cn
C/l
U)
•
rt
U)
o
CD
•
oV>
rt
H-
IS1
c/>
<
rt
a>
1
H*
(-!
<
O
H-
CD
(D
•
h-»
tr
I '
3
(B
o
(/>
a
CD
£
M
CD
(J)
H-
(Q
tr
•
r+
CT>
O
-—.
Hi
o
H-
cn
o\o
H,
h—1
tsi
C
M
0)
1
P
<
o
—
a>
o
3
O
n>
i—1
N>
rt
•
CU
O
o\o
tn
*
(SI '
H
I
<
Percent of Total Influence on
Color Change (Delta E‘)
Percent of Total Influence on
Color Change (Delta E*)
Percent of Total Influence on
Color Change (Delta E*)

243
AI203 / Alkalis Molar Equivalents
Figure 4.34. Color changes due to variations in peak firing
temperature and frit AI2O3:alkali ratio of coatings batched
with 2.0% Zr-V.

244
but has an insignificant influence at low levels when Zr02 is
present.
4.4.3 Log Viscosity vs. Coating Composition and Temperature
For coatings batched with 2.0% Zr-V, their viscosity (rj)
during heating from 700°C to 1100°C was thoroughly explained
by the model:
Variable
Coefficient
t-Value
- Y-intercept
7.049
—
+ Zr02
5.278
9.03
- SrC
1.531
1
Cn
00
+ Al2C>3/alkalis
1.736
12.73
+ 1/T
10,914.600
53.03
where oxides are in molar equivalents in the frit and
temperature (T) is in degrees Celsius. Equation 4.10 fits
the experimental data very well (R2 = 0.98) and no outlier
experiments were detected. The standard deviation of the
residuals (S ) was 0.29. Thus, the accuracy of equation
4.10 for predicting log r| from 700°C to 1100°C was estimated
to be within +0.58 log poise, with a 95% confidence level.
Equation 4.10 is of the form of the Vogel-Fulcher-Tamman
(VFT) equation 2.93 in Section 2.2.5d. The A constant in
2.93 can be taken as the temperature-dependent oxide

245
composition effect of the first four factors in 4.10. The B
constant in 2.93 is the coefficient for 1/T in 4.10.
The same frit oxides found to significantly modify AE*
and K/S also controlled viscosity, but with different
magnitudes of influence. Temperature, of course, had the
largest effect with its inverse relationship to log r\. The
second greatest overall shift in viscosity resulted from the
range in frit Al203/alkali equivalents tested. Unlike
relationships with K/S and AE* , the Al203/alkalis molar
equivalents in the frit did not produce a curvilinear effect
on viscosity. Log r\ linearly increased with A1203 (from 0.084
to 0.170 molar equivalents) and decreased with alkali content
(from 0.130 to 0.294 equivalents). At an average temperature
of 900°C and with other oxides in 4.10 held at their average
molar equivalent level (Zr02 = 0.078, ZnO = 0.156, SrO =
0.131), the range of Al203/alkalis tested produced a 1.74
change in log T). Using the same procedure, Zr02 had the
second largest oxide effect, where increasing Zr02 through the
range experimented (0 to 0.175 equivalents) caused log r\ to
raise by 0.92. SrO content had the least influence of the
significant independent variables, where increasing SrO from
0 to 0.311 equivalents resulted in an estimated lowering of
log T) by 0.47.

246
In the Discussion, Section 5.2.2, of this dissertation,
the preceding viscosity data are related to color strength
and stability results. The potential for applying viscosity
data to estimate zircon pigment color stability in new frit
compositions is considered.
1^_5 Evolved Crystalline Species
X-ray scans of the unfired frits without pigment or
bentonite exhibited no crystalline phase peaks.
4.5.1 XRD, SEM and EDS Evaluations
Analyses of crystalline phases present in fired coatings
batched with 2.0% Zr-V are detailed in this section. These
data are further linked to color properties in the
Discussion, Section 5.1.2.
4.5.1a Frits with ZrO?
XRD patterns from fired coatings batched with frits A,
B, C and D, all of which contained Zr02, are displayed in
Figures 4.35, 4.37, 4.39 and 4.40, respectively.
Corresponding SEM micrographs are pictured in Figures 4.36,
4.38 and 4.41 (for C and D).

0.00 Cps 2200.00
2-Thete - Scale
Figure 4.35. XRD patterns from coatings batched with frit A and 2.0% Zr-V, and fired to
1000°C, 1050°C and 1100°C. Frit A includes 8% Zr02, 12% ZnO, 0% SrO, 5% alkalis.
247

248
Figure 4.36. SEM micrographs of a coating batched with frit
A and 2.0% Zr-V, and fired to 1100°C. Magnification is (a) X
1,000 and (b) X 6,000. Shown are large Zr-V particles
surrounded by fine zircon precipitates.

0-00 cps 721.58
2-Thata - Scale
Figure 4.37. XRD patterns from coatings batched with frit B and 2.0% Zr-V, and fired to
1000°C, 1050°C and 1100°C. Frit B includes 8% Zr02, 12% ZnO, 0% SrO, 10% alkalis.

250
Figure 4.38. SEM micrographs of a coating batched with frit
B and 2.0% Zr-V, and fired to 1100°C. Magnification is (a) X
1,000 and (b) X 6,000. Shown are large Zr-V particles, fine
zircon precipitates and large zircon fibers.

0.00 Cps 2500.00
E-Theta - Scale
1I00*C Z J
z
Z = Zircon (ZrSi04)
z
.. f . . . Ü
1050°C . i
^— . . fl
- ionn°r .
—■— 1 1 1—» " i '1-' 1—>—1 • < 1—■—
Figure 4.39. XRD patterns from coatings batched with frit C and 2.0% Zr-V, and fired to
1000°C, 1050°C and 1100°C. Frit C includes 8% Zr02, 0% ZnO, 12% SrO, 5% alkalis.

2-Thata - Scale
Figure 4.40. XRD patterns from coatings batched with frit D and 2.0% Zr-V, and fired to
1000°C, 1050°C and 1100°C. Frit D includes 8% Zr02/ 0% ZnO, 12% SrO, 10% alkalis.
252

(b)
D 1 1 00 V
•
t
•
% * 0
4k
■'* «
••
^ •
•
___ 1 O u rr,
t
1 5KU
1 U r in
XI ,000
13mm
Figure 4.41. SEM micrographs of coatings batched with 2.0
Zr-V, (a) frit C and (b) frit D, and fired to 1100°C.
(magnification X 1,000). Particles shown are Zr-V pigment

254
Patterns from all four coatings align well with the
zircon profile shown in Figure 4.42, where strong peaks exist
at reflection planes and 20 of [ 101] 20.20°, [ 200] 26.98°,
[ 112] 35.63° and [ 312] 53.48°. Zircon was the only phase
detected in A,C and D, while additional phases were observed
in B. A previous investigation [ Blo93] noted that zircon
tends to strongly orient along the c-axis in the plane of a
coating's surface. With zircon's tetragonal structure (a =
6.604 A, c = 5.979 Á), orientation along the c-axis results
in enhanced [ hOO] lines. Zircon in B, C and D causes
exaggerated [ 200] lines, which tend to rise with peak
temperature. An increase in temperature lowers the coating
viscosity and allows for faster transport of zircon from the
bulk melt to the surface.
The zircon profile from frit A (Figure 4.35) has a
stronger [ 112] line than B, C or D. SEM micrographs for A in
Figure 4.36 reveal numerous small (0.2-0.5pm) zircon
precipitates in the glass matrix surrounding a large
particle, whose morphology matches that of the Zr-V pigment
pictured in Figure 4.43. The small round precipitate clumps
are responsible for the [ 112] intensity relative to the
exaggerated [ 200] from the pigment. Thus in frit A, both

255
Figure 4.42. X-ray diffraction profile for zircon, ZrSi04.
(JCPDS card file data from Ferro Corporation, Cleveland, OH)

Figure 4.43. SEM micrograph of zircon-vanadium (Zr-V)
pigment, Ceredec 41715A, X 1,000 magnification.

257
zircon precipitating from Zr02 in the frit and zircon from the
Zr-V pigment structure were identified.
Enlarged XRD patterns from frit B in Figure 4.37 show
the presence of not only zircon but also Zr02-tetragonal (a =
5.12 Á, c = 5.25 A) peaks at [111] 30.2° 20, [002] 34.5° 20
and [ 200] 35.3° 20, again with exaggerated [ hOO] .
Intensities of the Zr02-t lines drop significantly with an
increase in firing temperature from 1000°C to 1100°C. Very
small peaks of Zr02-monoclinic (baddeleyite) also appear at
1000°C at exaggerated [ 100] 17.45° 20 and [ -111] 28.19° 20.
The Zr02-monoclinic completely disappears at 1100°C. Thus,
both zirconia phases crystallized below 1000°C and dissolved
from 1000°C to 1100°C. SEM focus on the frit B coating fired
to 1100°C (Figure 4.38) revealed Zr-V pigment, smaller
particles which resemble the zircon precipitates in frit A,
as well as large fibers of zircon. Overall, SEM showed there
was much less fine zircon precipitated from B when compared
to A. In addition to strong XRD peaks at [ 200] (pigment) and
[112] (spherical zircon precipitates), the fibrous zircon
resulted in strong [400] (55.6° 20) intensities.
Small compositional differences between frits A and B
produced very different crystallization tendencies. Mainly
the Si02:alkaiis ratio varied, where A consisted of 55% Si02

258
and 5% K20+Na20, and B contained 50% Si02 and 10% K20+Na20.
The 5% increase in alkalis and 5% reduction of silica may
deter zircon precipitation and induce formation of lower
temperature Zr02 phases observed in B for the following
reasons:
1. Less Si02 is available for formation of Zr02*Si0,.
2. An increase in nonbridging alkali ions in the glass
results in lower energy requirements for breaking up the
network. Figures 4.20 and 4.22 disclose that raising
the alkalis from 5% to 10% dropped the softening
temperature from 711.3°C to 684.2°C. Log r\ was also
reduced by a factor of 0.75 to 1.0 across the
temperature range. More rapid diffusion transport due
to a drop in log T| to 6.8-7.6 at the temperatures where
Zr02 phases have been found to begin precipitating (800-
850°C) results in significant Zr02 crystallization.
Enough Zr02 is precipitated at low temperature to where
there is still a small amount of undissolved crystals
remaining at 1100°C.
3. More energy is required to form zircon from crystalline
Zr02 and vitreous silica than when both oxides are in
the vitreous state. Crystallization data for B and
Figure 4.17 indicate that variations in gloss in tiles

259
produced with frit B could be attributed to the
precipitation and dissolution of Zr02 phases.
4. The molar A1203:alkalis ratio was lower in B, thus less
Al+2 was available to compete with Zr+4 for glass
network sites. More Zr+4 was required to maintain
network connectivity in B and less was freed up to form
zircon.
Figures 4.39, 4.40 and 4.41 confirm that Zr-V pigment is
the only crystalline phase in frits C and D at all peak
firing temperatures. Frits C and D vary from A and B only in
that ZnO was replaced with SrO. Crystallization of zirconium
phases did not occur in the presence of SrO, thus SrO raised
the AG of formation. Sections 4.3.2 and 4.4.3 demonstrated
that substituting SrO for ZnO in the glasses results in
higher glass transition temperatures and lower melt
viscosities (equation 4.10). Lower viscosities would tend to
favor rather than inhibit crystallization, so this does not
provide an explanation for the lack of crystallization which
resulted in increased color strength with SrO. The roles of
SrO versus ZnO in zircon precipitation, dissolution and color
development will be further discussed in Section 5.1.1c.

260
4.5.1b Frits without ZrO?
XRD patterns from fired coatings batched with frits E,
F, G and H are displayed in Figures 4.44, 4.46, 4.48 and 4.50
respectively. Corresponding SEM micrographs are pictured in
Figures 4.45, 4.47, 4.49 and 4.51.
Each of the coatings E-H at all three peak firing
temperatures still contained some zircon pigment, which is
evident by [ 200] lines which are enhanced due to surface
orientation. All of the zircon lines dropped with increased
peak temperature, indicating pigment dissolution in the frits
without Zr02. Zircon precipitación and dissolution are
detailed with a quantitative analysis in Section 4.5.2.
Dissimilar to coatings A-D, precipitation of non¬
zirconium species was pronounced in E-H, and the amount of
crystal phase increased with higher frit alkali levels. SEM
and XRD also indicated that unlike A-D where crystallization
was nearly complete by 1000°C, the total amount of
precipitate increased in E-H from 1000°C° to 1100°C.
Enhanced crystallization in the frits without Zr02 could not
be explained with changes in viscosity or ionic potentials of
the cations. Overall, no significant differences between
viscosity curves for A-D (figure 4.22) versus E-H (Figure

0.00 Cpa 800.00
2-Theta - Scale
Figure 4.44. XRD patterns from coatings batched with frit E and 2.0% Zr-V, and fired to
1000°C, 1050°C and 1100°C. Frit E includes 0% Zr02, 12% ZnO, 0% SrO and 5% alkalis.
261

262
E 1 050
•
. - *
â– â– â–  â– 
»
*
»
1
*
• '
• •
. 1 evm
F
. 15KU XI,€00
13mm
Figure 4.45. SEM micrograph of a coating batched with frit E
and 2.0% Zr-V, and fired to 1050°C. Magnification is X
1,000. Shown are large Zr-V particles surrounded by
dispersed diopside.

0.00 Opa 1B00.00
2-Theta - Scale
1
1100°C
, 1
II
H
, f-
t *
HD DII H
1 1 1 1
Z = Zircon (ZrSi04)
D = Diopside (CaMgSi206)
II = Ilardystonite (Ca2ZnSi207)
I
.H "
- 1050°C
L_J
1000°C
r—~r
4—
' —-■ái ■ ¡
35 40 45 50 GO
Figure 4.46. XRD patterns from coatings batched with frit F and 2.0% Zr-V, and fired to
1000°C, 1050°C and 1100°C. Frit F includes 0% Zr02, 12% ZnO, 0% SrO and 10% alkalis.

264
(b)
i â–  ' '
Jk£ 1
A- iN¿
> <
< — • 1
15KU X6 ^ 0 0 0 | Gmm
Figure 4.47. SEM micrographs of a coating batched with frit
F and 2.0% Zr-V, and fired to 1100°C. Magnification is (a) X
1,000 and (b) 6,000. All particles shown are hardystonite.

2-Thata - Scale
Figure 4.48. XRD patterns from coatings batched with frit G and 2.0% Zr-V, and fired to
1000°C, 1050°C and 1100°C. Frit G includes 0% ZrO?. 0% ZnO, 12% SrO and 5% alkalis.
265

266
G 1 1 00
1 0 H m ,
15KU XI,000 14mm
9
Figure 4.49. SEM micrograph of a coating batched with frit G
and 2.0% Zr-V, fired to 1100°C (magnification X 1,000).
Shown is crystallized SrCa2SÍ30g.

.00 Cps 1500.00
2-Theta - Scale
Figure 4.50. XRD patterns from coatings batched with frit H and 2.0% Zr-V, and fired to
1000°C, 1050°C and 1100°C. Frit H includes 0% Zr02/ 0% ZnO, 12% SrO and 10% alkalis.

268
Figure 4.51. SEM micrographs of a coating batched with frit
H and 2.0% Zr-V, and fired to 1100°C. Magnification is (a) X
1,000 and (b) X 6,000. Shown is crystallized SrCa2Si309.

269
4.23) are discernable. Although Zr+4 has a relatively high
ionic potential of 4.6, summations of the moles of each
cation multiplied by its ionic potential (from Table 2.5)
reveal that the overall bond strengths of E-H (range from
12.54 for H to 13.45 for E) are not significantly different
from A-D (range from 12.69 for D to 13.60 for A). The
increase in A1203 (ionic potential =5.3) in E-H makes up for
the loss of Zr02
The strong progression of crystallization in E-H caused
significant variations in the optical properties of the
coatings. This is evident by the large shifts in spectral
curves (Figures 4.4-4.6), L*, a* and b* values (Figures 4.8-
4.13) and AE* (Figure 4.16) over the range of peak
temperatures tested. Overall, lower gloss was also exhibited
when no Zr02 was present in the frit (Figure 4.17).
Diopside, CaMg (Si03)2 with a monoclinic structure, was
identified in E at 1000°C° and 1050°C, where several XRD
lines in Figure 4.44 match the standard pattern. Diopside
peaks lowered from 1000°C to 1050°C, then disappeared at
1100°C. A SEM micrograph of the surface of E fired to 1050°C
(Figure 4.45) shows large pigment particles surrounded by a
very faint pattern of crystals. The faint crystals appear
granular in morphology with particle sizes near 2 to 4 pm.

270
An EDS of the faint crystals confirmed the presence of
diopside by detecting mainly the elements Ca, Mg, Si and 0.
No faint crystals were viewed under SEM for E fired to
1100°C; only the pigment was present.
Hardystonite, Ca2ZnSi207 with a tetragonal structure, was
the predominant crystalline phase in F. Figure 4.46 shows
that hardystonite peak intensities at [001] 17.67°C 20 and
[ 002] 35.70° 29 rise dramatically from 1000°C° to 1100°C peak
temperature. An increase in the number of hardystonite
crystals with basal [ 00k] planes strongly aligned at the
surface would account for the exaggerated peaks. Small peaks
for diopside are also present at 1000°C but not at 1050°C or
1100°C. Apparently diopside dissolves with temperatures
exceeding 1000°C but hardystonite continues to form.
Crystals (2-6 pm) of hardystonite are displayed in the SEM
micrographs in Figure 4.47.
An increase in alkalis from 5% in E to 10% in F at the
expense of silica resulted in (a) more precipitate in the
glass, and (b) Zn+2 replacing Mg+2 as the cation bonding with
Ca+2, Si+4 and O"2 in the crystal structure. Possible
explanations for these observations are
1. From E to F, the glass softening temperature dropped
significantly from 736°C to 653°C (Figure 4.21). The

271
increase in alkalis also reduced the melt viscosity by
a log T) of 1.8 at 700°C and 0.2 at 1100°C. This
increased the slope of the log T| versus temperature
curve. Enhanced diffusion transport due to the large
viscosity decrease at lower temperatures where
nucleation and growth can occur would tend to raise the
number and size of crystals formed. Small drops in
viscosity at high temperatures may aid in inhibiting
dissolution.
2. Additions of higher radius ions (Na+1 = 0.97 Á, K+1 =
1.33 Á) open up the structure and allow for easier
movement of Zn+2 (radius = 0.74 Á) to re-arrange and
replace smaller Mg (radius = 0.66 Á) during crystal
formation.
3. Because less Si02 was available as a glass former and
the addition of nonbridging ions weakens the structure,
Mg+2 is retained in the glass over Zn+2 because its ionic
potential (2.6) is higher than zinc's (2.4).
4. Since both frits contained only 2.0% MgO versus 12.0%
ZnO, the reduced viscosity resulting from higher alkali
content would enable faster completion of formation of
diopside with ail of the MgO present. Lower temperature
dissolution of diopside could occur, while hardystonite

272
could form throughout the cycle due to the large amount
of ZnO present.
Greater fluctuations in the optical properties of F when
compared to E (Figures 4.4-4.6, 4.8-4.13, 4.16) can be
attributed to the crystallization differences. Figure 4.17
also shows the large variation in gloss exhibited with F but
not E.
For frit G fired to 1000°C and 1050°C, very small peaks
for wailastonite, CaSiO,, with a triclinic structure, are
observed at 20 values of 23.2°, 25.3° and 26.9° (overlaps
with zircon) in Figure 4.48. The peaks disappear at 1100°C.
By far the predominant crystalline species in G, and the only
precipitate detected in H, is represented by strong
reflections at 20 values of 17.95° and 36.38° in Figures 4.48
and 4.50. Although a JCPDS card search initially revealed no
matches, the following steps were taken to identify the phase
present [ B. Blonski, personal communication, engineer at
Ferro Corporation, Cleveland Ohio, May 1998] :
1. XRD scans indicated a line phase rather than a solid
solution because the patterns did not shift from
different firing temperatures or across the surface.
The data were input into Jade Text File Browser,
Materials Data Inc., computer software, which calculates

273
d-spacings for different structures. The d-spacings
observed for G and H matched well with a hexagonal
system (a = b * c, a = (3=90°, y= 120°) with a = 3.99 Á, c
= 9.87 A and a cell volume of 136.36 Á3. Snyder and
DeWolf figures of merit based on a statistical
comparison of observed versus calculated d-spacings were
M(10) = 37.3 and F(10) = 16.5, respectively. Figures of
merit ^10 are considered a good match, thus there was
significant indexing between actual and calculated
patterns.
2. A JCPDS card search of hexagonal silicates showed that
the c-lattice constant found was within normal range,
but the a-constant gives a smaller cell volume than
expected.
3. A standard procedure of rotating the coordinate system
by 45° was performed. The (hOl) cell direction was
moved to (hkl). This yielded lattice constants of a =
6.91 A and c = 9.87 A, which were in a reasonable range
and also provided d-spacings that closely matched the
XRD scans.
4. An EDS under SEM detected primarily Sr, Ca and Si in
approximate atom ratios of 1:1-2:3-5, respectively. The
proportion of Si was probably overestimated due to its

274
high concentration in the background matrix. Similar
hexagonal structures found in the JCPDS card file were
BaZrSi309 (a = 6.755 Á, c = 9.98 Á) and ZrK2Si3Og (a =
6.93 Á, c = 10.21 Á). The present structure was
compared to these phases in order to surmise the correct
stochiometry of Ca.
5. The estimated cell volume of the present phase is
between the two comparable structures. Atomic radii of
Ca (0.99 Á) and Sr (1.12 Á) also fall between Zr (0.79
A) and Ba (1.34 Á) or K (1.33 Á). The Sr and Ca radii
are relatively close and interchangeable in the silicate
structure. Using Pauling's rules [ Kin7 6] , coordinations
of anions about the cations are Ba, K and Sr = 8, Ca =
6-8 and Zr = 6. Thus, based on comparisons of cell
volumes and atomic radii, substitutions of Sr and Ca for
K or Ba and Zr are all possible. A stochiometry with
two moles of calcium results in SrCa2Si309. The
cyclosilicate (Si3Og-6) subclass of silicate minerals
consists of three closed rings of tetrahedra, each
sharing two oxygens. It provides 4-fold tetrahedral
coordination between Si+4 and O-2 and 2.25 Á wide
interstices between each 3-member ring. The Si308-4
configuration, which results in SrCaSi3Og, is not

275
energetically as favorable and is not a structural
classification for silicate linkages [ Mas68] .
Given these background data, the phase in G and H was
identified as SrCa2Si,Oq. This has a similar stochiometry to
hardystonite (ZnCa2Si,07) found in F which contained ZnO
rather than SrO. The SrCa7Si309 is a new line phase that is
not currently in the x-ray JCPDS data file.
Peak intensities for SrCa2Si309 in both G and H increase
substantially with peak temperatures. Especially noticeable
in Figures 4.48 and 4.50 are the [ 002] 17.95° 20 and [ 004]
36.28 20 reflections, which are exaggerated due to surface
orientación. SEM micrographs in Figures 4.49 and 4.51
display very large (10-30 pm) crystals of SrCa2Si309 in
coatings fired to 1100°C peak temperature. Coating H is
shown to have more precipitated material at the surface. In
accordance with prior evaluations, higher alkalis in H (10%
K20-rNa2Q) resulted in more crystal formation and greater
fluctuations in color with temperature than observed with G
(5% K20+Na20) .
The following points summarize the results of Sections
4.5.1a and 4.5.1b:

276
1. Frits with Zr02
• Zircon and a small amount of Zr02 were the
only phases that precipitated during
firing. Crystallization was complete by
1000°C.
• Replacement of ZnO with SrO in the frit
prevented zircon crystallization, regardless
of the alkali level tested.
• Increasing the alkalis (K20+Na20) in frits
with ZnC (without SrO) reduced the quantity of
zircon precipitate and caused its morphology
to be large and fibrous rather than small and
spherical.
2. Frits without Zr02
• Crystallization and zircon pigment
dissolution were pronounced and caused large
shifts in color from 1000°C to 1100°C peak
temperature.
• The amount of crystallization and dissolution
increased from 1000°C to 1100°C.
• In frits with ZnO, Ca?ZnSi207 tended to
precipitate, while SrCa2Si,09 crystals formed
in frits with SrO.

277
• Higher alkalis (K20+Na20) and/or replacement
of ZnO with SrO in the frits increased
doth crystallization and pigment
dissolution.
Proposed mechanisms and their effects on optical
properties will be detailed in Chapter 5.
4.5.2 Zircon Quantitative Analysis
Each fired coating batched with 2.0% Zr-V was rescanned
with XRD at 20 values near 53.48°. Integrated intensities of
zircon [ 312] reflections were determined as described in
Section 3.2.8. Several unfired coatings with known amounts
of zircon were also scanned in order to establish the
relationship between integrated [ 312] intensity and weight
percent zircon in Figure 4.52.
The quantity of zircon in each fired coating, based on
XRD rescanning results and Figure 4.52, is plotted in Figure
4.53. The only source of zircon in E, F, G and H is the Zr-V
pigment, while A, B, C and D could also precipitate
additional zircon from the frit.

Estimated Weight Percent Zircon
278
Integrated X-Ray Peak Intensity of Zircon [312] Reflection
Figure 4.52. Relationship between XRD [ 312] integrated
intensity and weight percent zircon in unfired coatings.

279
14
13
12
11
10
i 9
o
1-
Ñ 8
•*->
c
o
a 7
o
O.
£ 6
O)
'a>
$ 5
4
3
2
1
0
Figure 4.53. Weight percent zircon in fired coatings batched
with 2.0% Zr-V.
& Qi Qi Di Ei El Si H;
8% Zr02
8% Zr02
8% Zr02
8% Zr02
No Zr02
No Zr02
No Zr02
No Zr02
12% ZnO
12% ZnO
12% SrO
12% SrO
12% ZnO
12% ZnO
12% SrO
12% SrO
5% Alkali
10% Alkali
5% Alkali
10% Alkali
5% Alkali
10% Alkali
5% Alkali
10% Alkali

280
4.5.2a Frits with ZrO? and ZnO
Amounts of zircon much greater than the 2.0% pigment
addition were found in A and B.
Zircon in A was found to be between 12.0% and 13.5%.
All of the Zr02 in the frit would be required to precipitate
13.4% zircon. It is unlikely that pigment loss occurred in A
because zircon crystallization was energetically favorable
and the AE* from 1050°C-1100°C (Figure 4.16) was extremely
low ( Zircon content in frit B increased from 8.0% at 1000°C
to 9.7% at 1050°C and 1100°C. This is the temperature range
where the x-ray scans showed that the small amount of
zirconia crystals disappeared and were probably converted to
zircon. This may account for the drop in gloss disclosed in
Figure 4.17. Since zircon precipitation was energetically
favorable and AE* was <1.0, no pigment loss was evident in B.
4.5.2b Frits with ZrCh and SrO
Zircon pigment was the only crystalline phase found in C
and D with XRD and SEM. Frit C yielded Zr-V near 2.0% at all
firings with no trend downward with increasing temperature.
A slight decrease in zircon content from 1050°C (1.8%) to
1100°C (1.6%) was measured with D. When previous data are

281
compared, C and D (with SrO) experienced larger AE* values
(Figure 4.17 and model equation 4.7), shifts in K/S with
temperature (Figure 4.28) and energetically unfavorable
zircon precipitation when compared to A and B (with ZnO).
This provides evidence of slight pigment dissolution in C and
D, to within the approximate zircon test measurement accuracy
of ±0.2%.
4.5.2c Frits without ZrO-?
Significant dissolution of the Zr-V pigment occurred
with increasing temperature in all of the frits without ZrO-.
Nearly all of the pigment is present at 1000°C, but at 1100°C
Zr-V levels for E, F, G and H drop to 1.4, 0.1, 1.0 and 0.3%,
respectively. Greater dissolution occurred when alkalis
increased from 5% (E and G) to 10% (F and H). No significant
difference in pigment loss is evident between frits with ZnO
(E and F) or SrO (G and H). Thus, the energetics of the
frits without Zr02 were favorable for (a) dissolving Zr-V
pigment to add Zr02 to the glass structure and (b) removing
Ca, Zn, Sr and Si to form crystalline species.

CHAPTER 5
DISCUSSION
5.1 Color Strength and Stability Dependency
The previous chapters identified that color stability at
high temperature in glass frit/zircon-vanadium pigment
systems is optimum when both Zr02 and ZnO are present in the
frit and a relatively high pigment concentration is utilized.
In frits without Zr02, increases in the Al203/alkalis ratio
were found to be beneficial. In all cases, replacement of
ZnO with SrO lowered stability.
The highest color strength was achieved with frits
containing Zr02 and SrO, but no ZnO. The weakest colors
resulted in frits with no Zr02 and low Al203/alkalis. In
frits without Zr02, increases in the Al203/alkalis ratio
improved color strength and interchanging SrO and ZnO had no
significant effect.
In frits without Zr02 (E-H), where crystallization
increased with firing temperature, colors not only progressed
lighter and less saturated (lower chroma) at higher
282

283
temperatures but also shifted in hue. This was evident by
changes in the L*:a*:b* ratios.
Frits containing both Zr02 and ZnO not only yielded
excellent color stability but also produced a glossy, highly
opacified coating due to the crystallization of fine zircon.
The results also revealed that the most stable "transparent"
glossy frit, in which no crystallization occurs during
firing, incorporates both Zr02 and SrO, but no ZnO. This
combination of oxides provided enough Zr02 in the glass
structure to preserve the zircon pigment and avoid
precipitation of nonzircon species, while SrO replacement for
ZnO prevented crystallization of opacifying zircon. The
optimum transparent frit formula does not conform to the
current industry practice of always excluding Zr02 from these
types of frits.
Chapter 5 proposes explanations for these noted color
phenomena. Section 5.2 resolves the validity of applying
coating viscosity data, based on heating microscopy and
dilatometry measurements, for industrial color quality
control purposes.

284
5.1.1 Coating Composition
5.1.1a Zr-V Loading
Color strength due to pigment absorption in terms of K/S
was found to increase linearly with pigment content (Figure
4.7), while visual color intensity in terms of L*, a* and b*
values varied logarithmically (Figure 4.15). Higher Zr-V
concentrations exhibited less perceivable color changes per
weight percent pigment addition. Thus, industrial batching
errors are less visible in fired coatings with saturated
colors where Zr-V ¿2.0%. In these coatings, more than
enough vanadium atoms are present for contacting nearly all
of the incident light and achieving maximum absorption of 640
nm radiation.
Precipitation of typical mineral structures in the glass
matrix surrounding the pigment tends to increase reflection
at 640 nm and lower reflection at 400 nm, as shown for
bentonite in Figure 4.1. The resulting spectral curve
redistribution is related in Figure 5.1. Larger shifts in
the shapes of the curves occur for 0.5% and 2.0% Zr-V when
compared to 5.0%. The 5.0% saturated color tends to be least
sensitive to changes in glass matrix conditions due to

285
0.5% Zr-V
- - * - • 2.0% Zr-V
5.0% Zr-V
Figure 5.1. Changes in reflectance distributions at 400 nm
and 640 nm wavelengths due to increases in peak firing
temperature for coatings batched with 0.5%, 2.0% and 5.0%
Zr-V.

286
crystallization in E-H. Thus, higher pigment concentrations
are recommended for frits which do not incorporate Zr02.
Visible color changes due to variations in firing
temperature were greatest for an intermediate pigment loading
of 2.0% (Figure 4.16, equation 4.9) in frits where pronounced
crystallization occurred (E-H). In frits with Zr02 which
exhibited little or no crystallization or pigment loss from
1000°C to 1100°C, pigment loading was not a significant
variable for color stability. An intermediate Zr-V content
exhibited the greatest visible color change due to a shift in
peak firing temperature possibly because
1. Very light colors have (a) less pigment and color to
lose during processing, and (b) there is a logarithmic
relationship between the amount of light reflected and
our ability to see it (Figure 2.14), thus, the human eye
is less sensitive to changes in light colors.
2. Dark, saturated colors contain more pigment particles
than required for maximum absorption of incident light;
therefore, pigment dissolution is less noticeable.
5.1.1b ZrCn
Of all the frit oxides evaluated, Zr02 had the largest
influence on improving the color stability of the coatings at

287
any Zr-V content (Figures 4.16, 4.33 and equations 4.6-4.9).
The statistical analysis found an inverse linear relationship
between AE* 1050_1100 and Zr02 molar equivalents. In frits
without Zr02, large precipitates formed during firing and
crystallization increased at higher temperatures, causing a
reduction in color strength and stability.
Besides Si+4 and B+3 which are strictly glass formers,
zirconium has a greater ionic potential and single bond
strength with oxygen than any other cation in the frits
except Al+3. Glasses with high Zr02 contents tend to have a
strong, high melting, viscous and dense structure. Heating
microscope images (Figure 4.18), softening temperature (Ts)
data (Figures 4.20 and 4.21) and viscosity profiles (Figures
4.22 and 4.23) disclose that increases in molten flow and
decreases in Ts and viscosity due to increases in alkali
content were of much less magnitude when Zr02 was present in
the frit. A higher 0:Si ratio, due to a reduction in glass
forming Si+4 and more nonbridging alkali ions, was partially
compensated for by Zr+4 bridging the excess oxygen in the
structure. Frit B (10% alkalis) precipitated less zircon
than frit A (5% alkalis) because more Zr+4 was retained in the
glass structure to bond with the extra O'2 from the higher
alkali content. In frits C and D, where SrO replaced ZnO,

288
all of the Zr+4 remained in the glass structure at both alkali
levels tested, for reasons that will be discussed in the next
section.
The effect of Zr02 on AE* is related to the glass
density as shown in Figure 5.2. Frits with Zr02 are denser
and yield improved color stability, while frits without ZrO-
are shown to be more color sensitive to temperature and
alkali level. By comparison, the density of Zr02 is 5.5
g/cmJ, while A1203 and CaO which replaced Zr02 in frits E-H
are 3.8 and 3.3, respectively. Higher density glasses
contain less permeable structures, smaller rates of removal
of latent heat of fusion and lower diffusion coefficients and
mass transport rates, resulting in inhibition of
devitrification and dissolution reactions.
Fritted Zr02 lowered Zr-V color strength (K/S) when ZnO
was present (Figure 4.25, equations 4.1-4.4) and raised K/S
when SrO replaced ZnO. Zircon precipitation which occurred
in frits with ZnO and Zr02 increased the overall reflectance
from the coating and lightened its color. Zircon formation
was complete by 1000°C and resulted in good color stability
from 1000°C to 1100°C. When SrO replaced ZnO in frits with
Zr02, the solubility of Zr02 increased to the point where no
zircon crystallized from the frit. Since no zircon

289
Frit Density (g/cm3)
o 5% Alkalis ♦ 10% Alkalis
2.95
Figure 5.2. Influence of frit density on the color stability
of coatings batched with 2.0% Zr-V pigment.

290
precipitate was present in the coating to reflect light in
the wavelength region where pigment absorption occurs, a
strong blue color was achieved.
In all cases, regardless of whether ZnO or SrO was in
the frit, much less, if any, Zr-V pigment dissolution
occurred when Zr02 was present. In the frits tested, Zr02 >
0.139 molar equivalents was sufficient for satisfying most of
the zircon solubility requirements and maintaining minimum
energy configurations without the necessity of dissolving a
significant amount of zircon pigment to acquire more ZrCu for
the glass structure.
5.1.1c SrO vs. ZnO
Further evidence that SrO replacement of ZnO raises ZrO~
and zircon solubility (and AG of formation of zircon) is
given in Figure 5.3. A graphical method was employed to
estimate how much of the zircon present in fired coatings A-D
was pigment rather than zircon precipitated from Zr02 in the
frit. (The quantitative x-ray analysis of zircon content in
Section 4.5.2c identified how much pigment was in E-H, where
the frits contained no Zr02 and Zr-V was the only form of
zircon in the coatings.) Since K/S-»0 as Zr-V->0, positive x-
axis intercepts estimate quantities of dissolved pigment.

Frit A
in
2
H 1 1 1 1-
1.5 2 2.5 3 3.5
Weight % Zr-V Pigment
FritC
♦ 1000C ■ 1050C A 1100C
••• -Linear(1000C) Linear(1050C) Linear(110OC)
Figure 5.3. Pigment absorption factors (K/S) for coatings batched with Zr-V pigment and
frits containing 8% Zr02, and fired to 1000°C, 1050°C or 1100°C peak temperature.

292
Intercepts for frits A and B are at zero, while frit C varies
from 0.1 to 0.3 and D is 0.4 to 0.5. Figure 5.3 indicates
that a small amount of pigment dissolution occurs in frits
with Zr02 when SrO is present but not when ZnO replaces SrO.
The dissolution, however, is still much less than found in
frits without Zr02 (E-H). The interaction variable -(SrO X
T), derived in statistical models 4.1-4.4 and illustrated
with Figure 4.28, was found to significantly affect K/S.
This provides additional proof that SrO increases the
dissolution of Zr-V pigment with temperature.
Both SrO and ZnO modify Si04 network interconnectivity
by increasing the 0:Si ratio, resulting in lower Tg and
viscosity. Similarly, Sr+2 and Zn+2 may act as glass
modifiers and occupy intersticial sites to provide local
charge neutrality near nonbridging oxygen. However, unlike
Sr4-2, Zn1-2 can also contribute in part to the network
structure 3S an intermediate [ Chi97] . The atomic radii (Zn+2
= C.74 A; Zr+4 = 0.79 A) and oxide densities (ZnO = 5.6 g/cm3;
Zr02 = 5.5 g/cm3) of zinc and zirconium are comparable. The
radius of Sr+Z is higher (1.12 Á) and the density of SrO is
lower (4.7 g/cm3). Therefore, favorable substitution of Zn+2
as an intermediate in the glass structure for some Zr+4 is
likely to lower the activation energy and AG of formation of

293
zircon from Zr02 in the frit. Because Sr+2 cannot function as
an intermediate and maintain some network interconnectivity,
Zr+4 remains in the structure when SrO replaces ZnO.
Thus, if maximum color stability or opacity due to
zircon precipitation is desired, frits containing both ZnO
and Zr02 are recommended. If color strength or a relatively
stable transparent frit is the highest priority, frits
containing Zr02 and SrO but no ZnO are best.
In frits containing no Zr02, the statistical models
substantiate the same decrease in color stability (higher
AE*) when SrO replaces ZnO, but no effect on K/S was found.
The overall color strength cannot be raised with SrO through
a reduction in zircon precipitation because no Zr02 is present
in the frit to form zircon. Pigment dissolution, however,
and the corresponding high color variation which temperature,
still increase when SrO replaces ZnO in frits without Zr0o.
5.1.Id AloO^/Alkalis
The data and statistical models revealed that the A120, :
alkalis ratio had a larger influence on AE* and K/S than
other oxides in frits without Zr02. When no Zr02 was present,
higher ratios resulted in improved color strength and
stability at high temperature. Conversely, increases in

294
Al203/alkalis slightly lowered color strength and stability in
frits which incorporated Zr02. Also, the greatest increases
in melt viscosity occurred in frits with or without Zr02 when
the Al203/alkalis was raised.
Although Al+3 has a large ionic potential (5.6) and
increases the softening point, viscosity (and correspondingly
lowers mass transport and diffusion rates) and density of a
glass, in the present study it was not as effective as Zr+4 in
inhibiting pigment dissolution and crystallization above
1000°C. Frits with Zr02 yielded much more stable optical
properties than frits where Zr02 was replaced with A1203 and
CaO. When ZrM is missing (frits E-H) , Al+3 takes over as the
main glass intermediate but is not as effective in
maintaining the structure because it connects with less
oxygen and requires adjacent alkalis for charge neutrality.
Substitutions for Zr+4 or Si+4 require Al+3+Na+1 or Al+3+K+1,
where Al+3 occupies the centers of A104 tetrahedra and
converts nonbridging to bridging oxygen. An "equivalency"
point is reached when Al203/alkalis = 1.0. At ratios other
than the equivalency point, excess alkalis or Al+3 behave as a
modifier. Network connectivity with small radius Al+3 (0.51
A) combined with large radius Na+1 (0.97 Á) or K+1 (1.33 Á)
occupation of nearby intersticial sites creates a larger

295
strain on the structure than when Zr+4 (0.79 Á) is
incorporated in the glass network. The increased strain
results in lower thermodynamic stability and a greater
tendency for devitrification.
In frits where no Zr02 is present, higher Al203/alkalis
ratios reduce crystallization and pigment dissolution by
raising the melt viscosity, which inhibits diffusion
transport. This provides improved color strength and
stability at high temperature. Increases in A1203 help to tie
up excess oxygen from the alkalis.
In frits with ZrC2, the A1203:alkalis ratio has a
different effect on the fired color than found in frits
without Zr02. Additions of A1203 allow Al+3 to compete with
Zr+4 for network sites, and some Zr+4 is freed up to form
zircon crystals. The color strength and stability are
slightly lowered due to the increased quantity and variation
of zircon crystallization. Although higher alkalis in these
frits lower the viscosity, less zircon precipitation results
and stronger colors are produced. Zirconium gets tied up in
lower temperature Zr02 phases which precipitate and dissolve
in the lower viscosity medium. Thus, the statistically
derived curvilinear relationships between Al203/alkalis and

296
both K/S and AE*, as shown in Figures 4.25 and 4.32 and
discussed in Chapter 4, are elucidated.
5.1.2 Crystalline Species
Crystallization during firing significantly influenced
the color strength, stability, gloss and opacity of the
coatings. In frits with Zr02, zircon was the only crystalline
phase that formed. Most of the quantitative work of the
present study was performed on these frits because of their
predominant, use in industry.
Frits without Zr09 tended to precipitate crystals with
cations of Ca, Si and either Sr or Zn when present. While
precipitation of zircon in frits altered reflectance values
but not the wavelengths where peak reflectance and absorption
occurred, crystallization of nonzircon species changed the
peak reflectance wavelength (Figures 4.4-4.6) and
significantly altered the shapes of the spectral curves.
5.1.2a Zircon
The small radius of zirconium provides a high diffusion
rate for forming crystals. Figures 4.36 and 4.38 revealed
numerous 0.2-0.5 pm particles and 2-5 pm fibers of zircon

297
precipitated from Zr02 and Si02 in the frit, surrounding
pigment particles 8-10 pm in size.
The particle size and index of refraction of a material
determine its ability to scatter light. Zircon's dense
tetragonal structure yields a high index of refraction (n =
2.05) when compared to the coating matrix (n -1.55). Maximum
scattering from zircon occurs with particle sizes of 0.60 to
0.75 pm [ Blo94b]. The 0.2-0.5 pm zircon precipitate
particles in frits A and B are slightly smaller than the
optimum size for scattering the most visible light. However,
sizes closer to 0.4-0.5 pm (400 nm-500 nm) increase the
amount of blue light scattered and lessen the quantity of
yellow light reflected because near 450 nm corresponds to the
wavelength of blue light. This enhances the blue color
produced by the Zr-V pigment, which absorbs yellow light
without greatly scattering other visible wavelengths due to
its large size.
The small 0.2-0.5 pm zircon precipitates also produced
high gloss, where crystallization of larger particles in
frits F, G and H (2.0-30.0 pm) caused a significant increase
in diffuse reflectance and lowering of specular reflectance
(Figure 4.17). The gloss of frit B was lower than A because
some large fibers of zircon (2-5 pm) had precipitated along

298
with the smaller particles. The increased alkali content in
frit B resulted in a lower viscosity medium which enhanced
crystal growth.
Figure 5.4 illustrates how zircon precipitates and blue
pigment modified a coating's color. Linear relationships
represent fired coatings where the only zircon present is
pigment (^2.0%) and nonlinear plots are for coatings
containing 2.0% Zr-V plus crystallized zircon. The two
curves converge at 2.0% zircon where only pigment is present.
Data from all three peak firing temperatures are graphed.
Increases in the quantity of zircon precipitate in the
coatings from 2.0% to 13.0% resulted in greater losses in
lightness (AL* = 16) than greenness (Aa* =8) or blueness
(Ab* = 9). The precipitates caused more scattering of all
visible wavelengths and less absorption of 640 nm light,
which reduced the overall color strength. However, the blue
color was partially preserved due to the enhanced scattering
of 400-500 nm light from the 0.2-0.5 pm particles. The
result is a very light but distinct, high-chroma blue. Frits
A and B produced the highest blue (-b*) values of light
colors with L* values > 75.0 due to this phenomenon. An
analysis of the spectral reflectance data also shows that, by
comparison, frit A yielded the highest reflectance (=75) at

299
L*-Value
a*-Value
-30.0 -25.0 -20.0 -15.0 -10.0 -5.0 0.0
b*-Value
♦ Precipitated Zircon + 2% Zr-V Pigment
® Only Zr-V Pigment Present
Figure 5.4. Visual lightness (L*), greenness (-a*) and
blueness (-b*) as a function of weight percent zircon in the
fired coatings batched with 2.0% Zr-V.

300
460 nm blue wavelengths, when 640 nm reflectance values were
near 40. Support of this hypothesis is given in Figure 4.14
of previous Section 4.2.3. Data points for light colors with
higher than predicted L*-values above the L* versus b* curve
represent frits A and B where zircon precipitated. All of
the other coatings appeared to be darker, yellower and
greener than A and B.
The relationship between color and weight percent zircon
precipitate was found to be logarithmic:
L* = 8.23
In
(Z)
+ 57.72
(5.1)
a* = 4.4 5
In
(Z)
- 20.0
(5.2)
b* =4.94
In
(Z)
- 29.8
(5.3)
where Z is the weight percent of zircon precipitate.
These relationships correspond to the phenomenon noted
in Background, Section 2.1.3, where perceived lightness has
been found to be a logarithmic function of the actual light
level reflected from an object [ Hun87] . At higher zircon
contents and reflectances, a less noticeable color difference
is perceivaole per unit change in reflectance. Equations
5.1-5.3 also clearly reveal that a* and b* change less than
L* with increasing zircon precipitate content. Thus, a
method has been quantified for producing very opaque, glossy,
light and "clean" blue color by maintaining a low b* as L*

301
goes up. Similarly, color from other pigments can be
enhanced through controlled crystallization of zircon to a
particle size that tends to scatter more light of wavelengths
corresponding to the desired color and less light where
absorption by the pigment occurs. Sections 4.5 and 5.1.1
revealed that zircon crystallization could be controlled by
varying the SrO:ZnO and A1203:alkali ratios.
Pigment dissolution when no zircon precipitate is
present, as denoted with the linear plots in 5.4, yields
nearly the same effect on L* and b*. Adding or subtracting
Zr-V at quantities ^2.0% does not produce unique color. As
noted in Section 5.1.1, minimum zircon pigment loss and
correspondingly greater color stability is achieved with
frits formulated to precipitate zircon.
The relationship between light absorption by the
pigment, in terms of K/S values, and weight percent zircon is
displayed in Figure 5.5. The curves again converge at 2.0%
zircon where only pigment is present. Zircon affected K/S by
K/S = 3.975 (Z)-°-8283 (5.4 )
Comparisons of equations 5.1-5.3 and Figure 5.4, with
equation 5.4 and Figure 5.5, confirm that the 0.2-0.5 pm
zircon precipitate crystals influenced lightness (L*) to a
greater degree than the absorption of light near 640 nm which

302
Figure 5.5. Pigment absorption factor (K/S) as a function of
Zr-V pigment and zircon contents in coatings batched with
2.0% Zr-V.

303
causes blueness (-b*). The asymptotic curve in Figure 5.5
reveals that raising the amount of zircon from 5% to 13.5%
dropped K/S by only 0.3 Correspondingly, blue color lowered
by only Ab* = 3.0 while lightness changes were much more
visible, where AL* = 6.0. Thus, precipitation of 0.2-0.5 pm
zircon from 5% to 13.5% by weight does not significantly
alter the blue color strength if the Zr-V pigment is
maintained.
The wider spread of data points on color and absorption
plots (Figures 5.4 and 5.5) where only pigment is present is
indicative of the simultaneous Zr-V dissolution and
pronounced crystallization of large nonzircon species in
frits E-H. In frits where no zircon precipitated, both
pigment dissolution and extensive crystallization were more
prevalent and caused significant color variations (Figure
5.6). Precipitation of zircon was complete before 1000°C,
and the amount of zircon in systems where its crystallization
was energetically favorable stays nearly constant from 1000°C
to 1100°C. The energetic preference for crystallizing zircon
in these glasses inherently preserves zircon pigments and,
consequently, yields good color stability.

304
Weight Percent Zircon at 1100C
Figure 5.6. Changes in color stability denoted by K/S and
Delta E*, as a function of weight percent zircon in coatings
batched with 2.0% Zr-V.

305
5.1.2b Diopside
A small amount of diopside, CaMg(Si03)2 with a monoclinic
structure was identified in coatings batched with frit E and
fired to 1000°C and 1050°C (Figure 4.45), but the diopside
was completely dissolved at 1100°C. This progression,
combined with Zr-V pigment dissolution, altered the colors of
these coatings.
Diopside's density (p = 3.3-3.6 g/cm3) index of
refraction (n = 1.66-1.76, depending on crystal orientation)
and corresponding ability to scatter light are less than
zircon's. Its particle sizes (2-4 pm) in the fired coating
were too large to cause preferential scattering of visible
light wavelengths. Strong x-ray peaks at a 20 of 29.85° from
[-221] reflection result from the diopside structure. Figure
5.7 weighs the impact of pigment dissolution and diopside
crystallization signified by [ -221] reflection, on color
changes. Simultaneous dissolution of diopside and pigment
occurred with increasing temperature. The color changes are
mainly due to pigment dissolution rather than the small
amount of diopside formed, as denoted by the low x-ray peak
intensities. Because frit E contained ZnO rather than SrO
and a high Al203:alkali ratio (1.22), it had the most stable
glass structure of the frits without Zr02. These conditions

306
Figure 5.7. Color lightness (L*) and blueness (-b*)
progression with diopside crystallization and pigment
dissolution in coatings batched with frit E and 2.0% Zr-V.

307
yielded the least crystallization and Zr-V dissolution, and
best color stability and strength when no Zr02 was present in
the frit.
5.1.2c Hardvstonite
A dramatic increase in the magnitude of crystallization
and pigment dissolution occurred when the Al203/alkali molar
ratio was decreased from 1.22 in frit E to 0.61 in frit F.
When the ratio falls below the equivalency point of 1.0,
excess alkalis serve as modifiers and lower viscosity,
enhance devitrification and increase zircon pigment
solubility. The drop in gloss exhibited in Figure 4.17 and a
reduction in color stability (AE* =3.9 for E, AE* = 13.2 for
F at 2.0% Zr-V) were consequences.
Crystallization of hardystonite, Ca2ZnSi307 with a
tetragonal structure, increased with peak temperature in
coatings batched with frit F (Figure 4.47). Its density (p =
3.4 g/cm3), relatively low index of refraction (no values were
found in the literature, but n is proportional to p) and
large precipitate size (2-8 pm) produced less scattering of
visible light than zircon. Progression of hardystonite
crystallization with peak temperature is denoted by the
strong x-ray [ 001] reflection peak intensities shown in

308
Figure 5.8. Over the range of hardystonite precipitated, Zr-
V pigment weight dropped from 1.5% to near 0%. Unlike the
case where fine crystals of zircon preserved some of the blue
color in frits A and B, simultaneous precipitation of
hardystonite and loss of Zr-V in F result in a larger
reduction in blue (Ab* = 14.3) than increase in lightness
(AL* = 5). Thus, a much greater lowering of chroma with
increasing lightness occurred.
5.1.2d Strontium Calcium Silicate
The magnitude of crystallization and pigment dissolution
in fries without ZrO;. is further increased when SrO replaces
ZnO, as was the case in frits G and H. A reduction in melt
viscosity and greater zircon pigment solubility (for reasons
detailed in Section 5.1.1c) accompanying the substitution
resulted in lower color stability (AE* = 8.3 for G and 16.3
for K at 2.0% Zr-V) and gloss.
A previously unidentified crystalline phase, SrCa2Si309,
crystallized in coatings batched with frits G and H (Figure
4.49 and 4.51). The large, 10-30 pm precipitates with a
hexagonal structure, relatively low density (approximately
3.0-3.3 g/ern2 estimated based on comparisons with similar
structures) and small index of refraction produce less

309
Figure 5.8. Color lightness (L*) and blueness (—b*)
progression with hardystonite crystallization and pigment
dissolution in coatings batched with frit F and 2.0% Zr-V.

310
visible light scattering than zircon. The amount of
crystalline phase formed in the glass increased with peak
temperature while the quantity of zircon pigment decreased.
Figure 5.9 illustrates the color changes due to SrCa2Si309
crystallization, denoted by x-ray [ 002] reflection peak
intensity, versus pigment loss. The plots show greater Zr-V
dissolution in frit H where the Al203/alkali ratio (0.61), and
thus the melt viscosity, are lower. Higher XRD peak
intensities result from frit G (Al203/alkalis = 1.22), perhaps
due to preferred orientation at the surface of the coating.
The gloss readings (Figure 4.17) and SEM micrographs (Figures
4.49 and 4.51) indicate that more SrCa2Si305 is present in
fired coatings batched with frit H.
Ultimately, color is weakened more by pigment
dissolution than the crystallization of large, low reflecting
precipitates. Thermodynamic instability at high firing
temperatures in the relatively weak glass structures without
Zr02 is a catalyst for both processes. Conversely, inhibition
of diffusion transport processes in molten glass with
increased Al203/alkalis helps to preserve color by
simultaneously limiting pigment loss and the precipitation of
new species.

311
Figure 5.9. Color lightness (L*) and blueness (-b*)
progression with SrCa2Si309 crystallization and pigment
dissolution in coatings batched with 2.0% Zr-V.

312
5.2 Melt Viscosity
Even if a given state is thermodynamically favorable, it
may never occur unless adequate kinetics conditions exist.
Diffusion transport kinetics in glasses are governed by
viscosity. As the viscosity and density of a glass decrease
with rising temperature, rates of removal of latent heat of
fusion and mass transport increase, and diffusion
coefficients get larger. Consequently, the tendency for
crystallization and dissolution reactions increases.
5.2.1 Influence on Crystallization and Zr-V Dissolution
At a given temperature, differences in viscosities
between the coatings were primarily attributed to variations
in molar Al203/alkali ratios in the frits (Figures 4.22 and
4.23, statistical equation 4.10). Modifying Na+ and K+ alkali
cations enhance flow by disrupting bond linkages and breaking
up the glass network, while high ionic potential Al+3
counteracts this effect. To a lesser degree, coating
viscosity also correlated directly to Zr02 content and
inversely with SrO.
Nucleation proceeds at a lower temperature and higher
viscosity than crystallization. In frits with Zr02 and ZnO
where zircon crystallization was most favorable (frits A and

313
B), a greater number of finely dispersed spherical
precipitates resulted from increased viscosity conditions
(frit A, Figure 4.36), where the Al203/alkalis was higher. In
high viscosity mediums with low diffusion rates, internal
energy due to interfaces is reduced with spherical particles.
Larger precipitates with a fibrous morphology and a smaller
number of particles crystallized in the lower viscosity
medium (frit B, Figure 4.38) where the growth rate was
higher. The demonstrated size and morphology control of
zircon precipitation through viscosity regulation can be used
to alter color, gloss and opacity in glass coatings.
In frits with ZrC2 where zircon did not crystallize (C
and D) , higher slopes in the log r| versus temperature curves
are noted near the softening points of approximately 700°C.
Less crystallization tends to result when there is a high
viscosity at the softening point (Ts) and a viscosity which
increases rapidly below Ts [ Kin76] . Hence, there is less
opportunity for nuclei formation, and the subsequent volume
of crystals precipitated at high temperature is lower.
Replacement of ZnO with SrO in these frits prompted this
outcome.
In frits without Zr02 (E-H), the size and volume of ZnO-
based (Figure 4.47) or SrO-based (Figures 4.49 and 4.51)

314
precipitates increased with lowering viscosity. This caused
severe loss of color and gloss in the coatings. Although
replacement of ZnO with SrO in these systems also raised the
slope on the log r| curves near T , no corresponding reduction
in crystallization was observed. Precipitation of zircon,
where most of the nucleation and growth occur below 1000°C,
is more influenced by the values and rates of changes of
viscosity near T .
Comparisons of frits with and without Zr02 suggest that
Zr-V pigment dissolution in both systems cannot be directly
linked to the same viscosity characteristics. For example,
frits B (8% Zr02) and F (0% Zr02) had similar viscosity
curves, but all of the pigment was still present in B after
firing to 1100°C where less than 0.1% was measured in F.
However, if the frits without Zr02 are evaluated separately,
an inverse correlation between viscosity and both Zr-V
dissolution and crystallization becomes evident. The melt
viscosity of frit systems has less of an influence on
dissolving zircon pigments when enough Zr02 is present in the
frit to cause zircon to be the only energetically favorable
crystalline phase. When sufficient Zr02 is present in the
glass for achieving network connectivity and stability, there

315
is no driving force for dissolving the zircon pigment in
order to free up Zr+4 for the structure.
5.2.2 Value as a Predictor of K/S and AE*
Plots of melt viscosities versus K/S and AE* of fired
coatings batched with 2.0% Zr-V are given in Figure 5.10.
X-axis values were derived for each coating by inputting
their frit oxide molar equivalent values into equation 4.10
and then integrating the remaining function of temperature
from 700°C to 1100°C. The integrated area provides an
overall representation of the magnitude of viscosity during
the firing cycle. Excellent correlations exist for frits
containing no Zr02, where an increased viscosity throughout
the firing cycle results in better color strength and
stability. No correlations for frits with Zr02 are evident.
Figure 5.10 demonstrates the usefulness of comparing
areas under viscosity curves for predicting potential color
strength and stability achieved with zircon pigments in frits
without Zr02. In this case, where crystallization and pigment
dissolution proceed through the whole firing range above T
S t
the overall melt viscosity is a useful indicator. Higher
viscosities inhibit both processes in these types of frits.
However, in Zr02 containing frits, crystallization proceeds

316
Figure 5.10. Integrated log viscosity from 700°C to 1100°C
versus fired color strength (K/S) and stability (Delta E* ) in
coatings batched with 2.0% Zr-V.

317
rapidly to equilibrium soon above Tg. Thus, viscosity over
the whole temperature range is not a significant factor in
crystallization and corresponding color alterations.
For frits with Zr02, the slope of the log viscosity
curve near the softening point (T ) was found to correlate
positively with overall color strength (K/S) but inversely
with stability. Figure 5.11 reveals that this relationship
can be applied to predict K/S (R2 = .86) and AE* (R2 = .84).
Low R2 values for frits without Zr02 verify no significant
correlations.
For frits with Zr02, color strength is greatly increased
with a higher slope due to the corresponding lower
crystallization of zircon that tends to lighten color. A
very slight, hardly noticeable decrease in stability (AE*
from 1.0 to 2.0) resulting from a higher slope is due to the
increase in Zr-V solubility as a consequence of higher SrO
levels. If high color strength is the main objective,
tighter control of firing temperatures can be used to
compensate for the slight loss of stability.
In summary. Figures 5.10 and 5.11 demonstrate that color
strength and stability in glass frit/zircon pigment systems
can be controlled by regulating the melt viscosity.

318
Figure 5.11. Slope in the log viscosity versus temperature
near the softening point versus fired color strength (K/S)
and stability (Delta E*) in coatings batched with 2.0% Zr-V.

319
Increasing the magnitude of the viscosity curve by raising
the molar ratio of Al203/alkalis is effective for improving
color strength and stability in frits without Zr02. In frits
with Zr02, raising the slope of the log viscosity curve near
the softening point by increasing SrO at the expense of ZnO
can be applied to increase color strength with a slight
lowering of stability. Modest improvements in color
stability in frits with Zr02 can be accomplished by reducing
the slope, but at the expense of significant color strength
losses.

CHAPTER 6
SUMMARY AND CONCLUSIONS
The present research investigated the development of
blue color in glass frit/zircon-vanadium pigment systems for
ceramic whitewares coatings. The primary objectives were to
(a) determine if cost effective, environmentally safe glass
frits could be formulated to improve the fired color strength
and high-temperature stability of ceramic coatings and
(b)relate the frit composition, pigment content, firing
temperature, melt viscosity, crystallization and pigment
dissolution to the fired color.
"Fast-fire" ceramic tile manufacturing accounts for a
major portion of the whitewares industry and was the chosen
processing method for producing fired coatings. Other
whitewares industries, including manufacturers of
sanitaryware and dinnerware, are currently attempting to
convert to the same fast-fire roller kiln firing technology.
Silicate glass frits were of interest for this study
because they are the main ingredients in fast-fire glazes.
Frit is also the most reactive part of the formula and the
320

321
most corrosive to ceramic colorants. Experimental frit
compositions were designed to include low cost, nonhazardous
oxides, and conform with Seger's rules. The range of oxide
contents tested encompassed and exceeded the range normally
employed for glossy ceramic tile glazes. Special emphasis
was placed on comparing compositions with Zr02 (opacified)
versus no Zr02 (unopacified), SrO versus ZnO and low versus
high alkalis (K20, Na20) . The SrO and ZnO were of special
interest because of their extensive use as secondary fluxes
and their marked influence on modifying the Si02 glass
network. Other oxides in the frit formulas were A120-,, CaO,
MgO and B203. The B203 contents were kept low in order to
avoid liquid-liquid phase separation. A zircon-vanadium
(Zr-V) blue pigment was tested because it belongs to the most
common and stable group of colorants used in ceramic glazes,
the zircon triaxial pigments.
A coating's fired color strength was quantified by its
pigment absorption factor (K/S) , calculated with the
reflectance value at the Zr-V pigment peak absorption
wavelength of 640 nm. Color stability was expressed as the
inverse of AE*. The AE* was calculated as the magnitude of
the position vector in L* a* b* color space, between the
colors of two coatings fired to 1050°C and 1100°C peak

322
temperature, respectively. The L* a* and b* values,
calculated based on spectral reflectance curves and standard
observer tristimulus weighting functions, correlate to
lightness (L*), redness-greenness (a*) and yellowness-
blueness (b* ) color perception by the human eye. Equations
were developed from the results that relate pigment
absorption of 640 nm light to the color visualized.
Significant improvements in coating color strength and
stability through adjustments in frit oxide compositions were
demonstrated. Over the range of compositions tested, Zr02,
SrO, ZnC, A1203 and alkali contents had varying degrees of
influence over a coating's melt viscosity, crystallization,
pigment dissolution and corresponding fired color. Equations
were derived for predicting color strength and stability, and
melt viscosity based on the frit oxide composition, Zr-V
loading and firing temperatures. Thus, the results provide
information useful to industry where material costs versus
desired color properties can be weighed and prioritized.
Optimum color stability was achieved with frits which
incorporated both Zr02 and ZnO, but no SrO. These frits also
yielded good opacity due to zircon precipitation from Zr02 and
Si02 in the frit. The quantity of zircon that crystallized
increased with higher ZnO and lower alkali molar equivalents.

323
The greatest color strength resulted from frits
containing Zr02 and SrO, but no ZnO. The replacement of ZnO
with SrO prevented zircon crystallization during firing, and
thus these glasses can be classified as transparent frits.
These frits exhibited better color stability than
conventional transparent frits which incorporate no Zr02. In
frits without Zr09, increases in the Al203/alkalis improved
both color strength and stability.
Crystalline species that developed during firing were
related to the color perceived. Unique light blue color was
produced from a combination of Zr-V pigment and finely
crystallized 0.2-0.5 pm zircon particles in the coating.
Zircon precipitates 0.4-0.5 pm in size tend to enhance the
scattering of 400-500 nm blue light while increasing the
overall lightness. A new, previously unidentified
crystalline species (SrCa2Si309) precipitated in two frit
compositions which contained no Zr02 or ZnO.
This research also established that melt viscosity data,
estimated with models that utilize heating microscopy and
dilatometry measurements, can be applied as an industrial
quality control tool for predicting a frit's potential for
producing strong and stable color. The area under a log
viscosity versus temperature curve, from Ts to the peak

324
temperature, was proportional to color strength and stability
in frits without Zr02. In these frits, crystallization and
pigment dissolution occurred throughout the firing cycle
above T . The slope of the curve near Ts was proportional to
color strength and inversely proportional (with a small
effect) to color stability in frits with Zr02, where
crystallization was complete below 1000°C.
Significant results of this research are summarized in
the following sections.
6.1 Zr-V Pigment and Color Values
• Maximum absorption of visible light by the zircon-
vanadium pigment (Zr-V) occurs at 640 nm wavelength (1.94
eV), in the red-yellow region of the spectrum. This is the
difference in energy between split d-orbitals in V+4 on the
zircon lattice.
• Maximum reflectance from Zr-V occurs with 460 nm
(2.69 eV) blue-green light.
• There is a linear relationship between the pigment
absorption factor (K/S) calculated at 640 nm and Zr-V
content. However, the slope of the line varies significantly
with frit oxide composition and firing temperature.

325
• With increasing Zr-V content, the perceived color of
a ceramic coating progresses darker (lower L* ) and bluer
(lower b*). However, the trend of increasing green (lower
a*) with pigment content from 0% to 2% Zr-V reverses from 2%
to 5%, where the greenish hue appears to shift direction and
progress towards red (higher a*). A shift in maximum
reflectance to wavelengths below the x standard weighting
function minima correlates to the abrupt, nonlinear shift in
human perception.
• Variations in crystallization and pigment dissolution
have a lesser effect on the perceived color when high pigment
loadings ^5% are utilized.
• Pigment content increases from 0 to 5% have no
significant influence on fired gloss.
• At a constant pigment loading, most of the color
variation, due to changes in frit composition or firing
temperature, occurs in L* and b*.
• For Zr-V pigments, a second order polynomial function
well describes (R2 = .96) the positive correlation between L*
and b*.
• Values for perceived color, L*, a* and b*, are all
logarithmic functions of K/S at 640 nm, where a higher K/S
results in lower color values (higher chroma).

326
6.2 Frit Oxide Composition
• The strongest blue color, or highest K/S, is produced
with frits containing high Zr02 and SrO, and no ZnO. The
weakest color results in frits with no Zr02 and low A1203/
alkalis.
• In frits with Zr02, color strength (K/S) is increased
to the greatest degree when SrO replaces ZnO. This inhibits
precipitation of zircon from the frit which ultimately
lightens color. A relatively stable transparent frit results
from glass compositions with Zr02 and SrO but no ZnO. The
statistical models predict a linear relationship between SrO
and K/S when ZnO is also present.
• In frits without Zr02, SrO versus ZnO causes no
significant difference in K/S.
• If SrO is present in the frit without ZnO, an
increase in Zr02 molar equivalents in the frit linearly raises
K/S by reducing the solubility of the zircon pigment. In
frits with ZnO where zircon precipitation is favorable,
higher Zr02 content linearly lowers K/S.
• Raising the Al203/alkalis molar equivalents slightly
lowers color strength in frits with Zr02 which tend to
precipitate zircon. This occurs because Al+3 replaces some
Zr+4 in the glass network, which releases more Zr+4 to form

327
zircon and lighten the color. When no Zr02 is present, higher
Al203/alkalis ratios significantly increase K/S and color
stability at high temperature by inhibiting crystallization
and Zr-V dissolution due to an increased melt viscosity.
• The effects of Zr02 and temperature on color strength
are progressively reduced with increasing Zr-V content. The
same quantity of pigment dissolution is less noticeable in
saturated colors.
• The least noticeable color changes in fired coatings
due to variations in peak firing temperature are achieved
with high Zr02 and ZnO, and no SrO (AE* was ^ 0.9). The
lowest color stability results from frits without Zr02 and low
in Al203/alkalis (AE* ranged from 7.0-16.3).
• Color stability is increased to the greatest degree
by raising the Zr02 content, which inhibits both
crystallization near the peak firing temperature and zircon
pigment dissolution. There is an inverse linear relationship
between Zr02 molar equivalents and AE*.
• Raising the SrO molar equivalents in frits with or
without Zr02 linearly reduces high temperature color stability
due to an increase in zircon pigment solubility.
• In frits with Zr02, color stability lowers very
slightly with higher Al203/alkalis. When no Zr02 is present,

328
significant improvements in color stability are accomplished
by raising the Al203/alkalis, which increases viscosity and
thus inhibits crystallization and Zr-V dissolution.
6.3 Viscosity. Crystallization and Zr-V Dissolution
• Data obtained with dilatometry and heating microscopy
analyses of frits can be correlated to melt viscosity,
crystallization and color development with zircon pigments.
• The melt viscosity of frits can be described with a
model of the form of the VFT equation, where the A constant
can be taken as the temperature independent oxide composition
effect.
• An increase in the molar equivalents ratio of Alo03/
alkalis has the largest influence of the frit oxides on
raising the glass softening point and melt viscosity.
• To a lesser degree, higher Zr02 and lower SrO results
in increased viscosity.
• Replacing ZnO with SrO has the largest effect on
increasing Tg and the ratio Tg:Ts.
• In frits with Zr02 and ZnO, a greater number of fine
(0.2-0.5 pm) spherical particles of zircon precipitate when
the viscosity is increased by raising the Al203/alkalis ratio.

329
A more elongated, fibrous morphology predominates from lower
viscosity conditions.
• Precipitation of zircon from Zr02 in the frit is
inhibited when the slope of the log T) versus temperature
curve near the softening point is increased by replacing ZnO
with SrO.
• In frits without Zr02, the size and volume of
precipitate increases with lowering viscosity, resulting in
weaker color and lower gloss.
• In frits without Zr02, there is no correlation between
crystallization and the slope of the log r) curve near the
softening point.
• In frits without Zr02, a lower viscosity causes more
pigment to dissolve, resulting in a weaker color. Viscosity
is not a dominant factor in Zr-V dissolution in frits
containing Zr02.
• A higher integrated area under the log r| versus
temperature curve from Ts to the peak firing temperature
correlates well with increased color strength (K/S) and
stability (lower AE*) in frits without Zr02. A greater
overall viscosity inhibits crystallization of nonzirconium
species in these frits. In frits containing ZrO-, where

330
zircon precipitation is complete by 1000°C, no such
correlation exists.
• A higher slope in the log r\ versus temperature curve
near the softening point correlates well with increased color
strength and a slight decrease in stability (AE* ^2.0) in
frits with Zr02 because zircon nucleation and growth is
inhibited. No such correlation exists for frits without Zr02,
where crystallization occurs through the range of firing
temperatures.
• In frits with Zr02, zircon is the only energetically
favorable precipitate present at normal peak temperatures of
1050°C and 1100°C.
• Zircon crystals near 0.4-0.5 pm in size in the glass
with Zr-V allow for blue color with a very high lightness
(L*), due to enhanced scattering of blue light.
• Values for the color perceived; L*, a* and b* are all
logarithmic functions of the weight percent zircon
precipitated in the glass. Higher zircon causes more
lightening (higher L*) than losses of green (higher a*) and
blue (higher b*).
• The pigment absorption factor (K/S) increases
exponentially with a decrease in the amount of zircon
precipitate.

331
• Perceptible color alterations due to variations in
firing temperatures are reduced in systems that crystallize
zircon from Zr02 in the frit. Crystallization is complete
before 1000°C and the amount of zircon left in coatings fired
to 1000°C, 1050°C and 1100°C remains nearly the same.
• In frits where zircon does not crystallize, Zr-V
pigment dissolution tends to occur, with the greatest loss
resulting in frits without Zr02 and a low Al203/alkalis ratio.
In frits with Zr02, zircon crystallization does not occur when
SrO replaces ZnO.
• L*, a* and b* increase and K/S decreases linearly
with the weight percentage of Zr-V pigment dissolved (from 0
to 2%) in the glass.
• In the frits tested without Zr02, the main phases that
crystallized were diopside and hardystonite when ZnO was
present and SrCa2Si309 when SrO replaced ZnO. The SrCa^Si309
is a new line phase unidentified on the standard x-ray JCPDS
card file.
• The quantity of hardystonite or SrCa2Si3Og precipitate
increases with temperature from 1000°C to 1100°C, resulting
in poor color strength and stability. Crystallization of the
two species causes a shift in the maximum reflectance from
460 nm at 1000°C to near 500 nm at 1100°C. This results in

332
higher L* and b* values and greatly perceptible changes in
color (high AE*). A greater amount of either material
precipitates when the A1203/ alkali content is low.
In conclusion, the research of this dissertation
quantified methods for improving the color strength and
stability of glass frits colored with zircon-vanadium pigment
for whitewares coatings applications. Glossy opaque and
transparent frit compositions which yield excellent color
strength and stability over a range of firing temperatures
were formulated. In addition, a technique for producing
uniquely light but high chroma color with zircon pigments
through control of zircon precipitate particle size was
illustrated. The results also apply qualitatively to other
ceramic pigments which use the same zircon structure to
incorporate colorant metal ions.
The color, crystallization and dissolution data
presented, as well as the mathematical models derived during
this study, will be of interest to materials scientists and
industries that manufacture ceramics coatings, glass frits
and ceramic colorants. The applications established for
using melt viscosity data as an industrial color quality
control tool will also be of use to the ceramics industry.

CHAPTER 7
FUTURE WORK
The present dissertation uncovered several potential
research topics which are of interest for both basic science
and industrial purposes. Possible future investigations
include
• Expand the mathematical models established in this
study for predicting K/S, AE*, L* , a*, b* and log r\ by adding
as independent variables
(a) the other two zircon triaxial pigments; zircon-
praseodymium yellow and zircon-iron coral,
(b) complete time versus temperature firing profiles,
rather than the peak temperature of a specific
cycle, and
(c) a wider range of glass oxide compositions.
• Derive statistical models to describe the influences
of frit oxide composition and crystallization on gloss.
• Develop matte whitewares coatings (gloss < 20.0) at
60° incidence) with improved color strength and stability.
Investigate whether batches containing fritted Zr02 and high
333

334
mineral additions of relatively coarse A1203 are optimum.
Based on the findings of this research, fritted Zr02 would
tend to inhibit crystallization and pigment dissolution that
cause color variation while A1203 minerals may provide for an
"underfired" matte crystalline texture.
• Control the quantity, size and morphology of zircon
precipitating from Zr02 in the glass frit by varying ratios of
SrO:ZnO and A1203:alkalis. Explore the range of colors and
opacification that can be achieved with zircon pigments
through selective scattering of light by the zircon
precipitates.
• Study the influence of zircon pigment particle size
on the color of a fired coating. Investigate if 450 nm Zr-V
blue, 580 nm Zr-Pr yellow and 650 nm Zr-Fe coral pigment
sizes provide optimum color strength and opacity by enhancing
the reflection of the desired color of light. Develop
surfactants that allow for easy dispersion of such fine
particles in an industrial coating base. Devise methods for
avoiding dissolution of the small zircon particles.
• Compare the effects of batch additions of zircon
minerals versus fritted Zr02 on the melt viscosity,
crystallization and color development of ceramic coatings.
Determine what maximum zircon mineral particle size preserves

335
the color by dissolving into the melt before the zircon
pigment is attacked.
• Determine whether other intermediate or glass former
oxides tend to stimulate or inhibit zircon precipitation from
Zr02 in the frit. In the present study, ZnO enhanced zircon
crystallization and SrO prevented it. For example,
investigate BaO, which has a density (4.7-5.5 g/cm3) , cation
radius (1.34 Á) and cation charge (+2) that are comparable to
SrO (4.7 g/cm3, 1.12 Á, +2 charge).
• Perform quantitative thermodynamic and kinetic
studies of crystallization and zircon pigment dissolution in
frits without Zr02. Determine the AG of formation of each
crystalline species.
• Synthesize the new SrCa2Si309 crystalline phase and
thoroughly study its structure and properties. Add the new
information to the existing x-ray JCPDS card file.
• Compare microwave versus conventional heating methods
for firing colored ceramic coatings. Evaluate if melting
behavior, crystallization, zircon pigment dissolution and the
corresponding color stability and strength are altered due to
possible changes in diffusion transport processes.
• Synthesize zircon pigments using alternative methods
such as microwave processing. Determine if a greater number

336
of dopant color ions can be incorporated on the zircon
lattice than currently achieved with conventional techniques.

APPENDIX A
UNITS FOR DESCRIBING LIGHT AND COLOR [ Nas83]
Al.l Units Related to Energy
Color
Wavelength
(A,) in
Nanometers
(nm)
Frequency
t) (Hz)
(xlO14)
Wave
Number v
(cm-1)
* Energy
in eV
* Energy in
cal mol-1
(xlO3)
Infrared,
far
30,000
0.10
333
0.041
0.95
Infrared,
near
1,000
3.00
10,000
1.24
28.6
Red
650
4.62
15,400
1.91
44.0
Orange
600
5.00
16,700
2.06
47.7
Yellow
580
5.17
17,240
2.14
49.3
Green
500
6.00
20,000
2.48
57.2
Blue
450
6.66
22,200
2.75
63.4
Violet
400
7.50
25,000
3.10
71.4
Ultra¬
violet,
long wave
366
8.19
27,300
3.39
78.0
Ultra¬
violet,
short wave
254
11.82
39,400
4.89
112.6
337

338
Wave Number (cm :) = u = 1/X
Frequency (Hz) = v = h x c
(Al)
(A2)
where c = the speed of light = 2.998 x 1010 cm/sec.
*Energy (eV) = E = h x 1.2399 x 10 4
^Energy (cal mol-1) = E = U x 2.8573
* Wavelength (nm) x Energy (eV) = 1239.9
(A3)
(A4)
(A5)
Al.2 Photometric Units
Candela (cd): SI unit for luminous intensity.
One square meter of blackbody at 2042K emits 600,000 cd.
One candela produces one lumen of flux per steradian of
solid angle measured from the source.
Lumen (lm): unit of luminous flux or light intensity emitted
by a point source of 1 cd through a unit solid angle
(steradian).
Luminance (L): unit of brightness in (cd m~2) .
The sun has L = 1.6 x 109 cd m-2 and a fluorescent lamp
approximately 10, 000 cd m-2.
Illuminance in lux (lx): One lm/m2.
One lux = 10.76 foot candles.
* Units apply to a single electron.

APPENDIX B
THE 15 CAUSES OF COLOR [ Nas83]
Transitions Involving Ligand Field Effects
1. Transition Metal Compounds: Many pigments, some
fluorescence, lasers, phosphors and turquoise.
2. Transition Metal Impurities: Ruby, emerald, red iron
ore, some fluorescence and lasers.
Transitions Involving Energy Bands
3. Metals: Copper, silver, gold, iron, brass, "ruby"
glass
4. Pure Semiconductors: Silicon, galena, cinnabar,
diamond
5. Doped or Activated Semiconductors: Blue and yellow
diamond, light-emitting diodes, some lasers and
phosphors
6. Color Centers: Amethyst, smoky quartz, desert
"amethyst" glass, some fluorescence and lasers
Transitions between Molecular Orbitals
7. Organic Compounds: Most dyes, most biological
colorations, some fluorescence and lasers
8. Charge Transfer: Blue sapphire, magnetite, lapis
lazuli, many pigments, graphite
339

Vibrations and Simple Excitations
340
9.Incandescence: Flames, lamps, carbon arc, limelight
10. Gas Excitations: Vapor lamps, lightning, auroras, some
lasers.
11. Vibrations and Rotations: Water, ice, iodine, blue gas
flame.
Geometrical and Physical Optics
12. Dispersive Refraction, Polarization, etc.: Rainbow,
halos, sun dogs, green flash of sun, "fire" in
gemstones.
13. Scattering: Blue sky, red sunset, blue moon, moonstone,
Raman scattering, blue eyes and some other biological
colors.
14. Interference: Oil slick on water, soap bubbles, coating
on camera lenses, some biological colors.
15. Diffraction: Aureole, glory, diffraction gratings,
opal, some biological colors, most liquid crystals.

APPENDIX C
DENSITY, PARTICLE SIZE AND APPLICATION WEIGHT DATA
Cl.l Density and Particle Size Based on Volume %
FRIT:
ABCDEFGH
41715A
Zr-V
Pigment
Density 2.87 2.91 2.85 2.85 2.78 2.77 2.70 2.75 —
(g/cm3)
Mean
Particle
Diameter 23.4 19.7 19.4 24.4 23.9 22.9 27.5 24.4 8.9
(pm)
Cl.2 Coatinas Application Weights (grams) at 1.75 Sp.Gr.
FRIT:
A
3
C
D
E
F
G
H
0%
Zr-V
11.7
12.0
11.6
11.6
11.2
11.2
10.7
11.0
0.5%
Zr-V
11.8
12.0
11.7
11.7
11.2
11.2
10.7
11.0
2.0%
Zr-V
11.8
12.1
11.7
11.7
11.3
11.3
10.8
11.2
5.0%
Zr-V
12.1
12.3
12.0
12.0
11.5
11.5
11.0
11.3
341

APPENDIX D
DATA FROM COATINGS BATCHED WITH FRIT, 2.5% BENTONITE AND
Zr-V PIGMENT, AND FIRED USING A 45-MINUTE
CERAMIC TILE CYCLE

DATA FROM COATINGS BATCHED WITH FRIT, 2.5% BENTONITE AND Zr-V PIGMENT, AND FIRED USING A 45 MINUTE CERAMIC TILE CYCLE.
31Q2
B2Q3
* Frit Composition (Weight %):
Na2Q K2Q £aQ AI2Q3
ZrQ2
MgQ
SrO
ZnO
** Wt% Zr-V
Pigment
Peak Firing
Temp. (Cl
L!
C.I.E.
a!
b!
Delta E*
1050-1100C
60-Degree
Angle Gloss
K/S at
640nm
*** Glaze
Ell
1
55
6
2
3
8
4
8
2
0
12
0
1000
936
-0.2
2.3
—
86.6
0.01
OK.
2
55
6
2
3
8
4
8
2
0
12
0.5
1000
87.3
-6.3
-6.9
—
82.9
0.13
OK.
3
55
6
2
3
8
4
8
2
0
12
2
1000
78.8
-9.1
-17 6
—
87
0.47
OK.
4
55
6
2
3
8
4
8
2
0
12
5
1000
72.1
-10.4
-23
—
81.6
0.93
OK.
S
55
6
2
3
8
4
8
2
0
12
0
1050
93.7
-0.6
1.6
—
88.1
0.01
OK.
6
55
6
2
3
8
4
8
2
0
12
0.5
1050
87.5
-6.1
-7
—
86.6
0.13
OK.
7
55
6
2
3
8
4
8
2
0
12
2
1050
78.8
-9
-17.8
—
88.7
0.47
OK.
8
55
6
2
3
8
4
8
2
0
12
5
1050
71.5
-10.4
-24
—
83.6
0.99
OK.
9
55
6
2
3
8
4
8
2
0
12
0
1100
93.6
-0.7
1.1
0.5
91.2
0.02
OK.
10
55
6
2
3
8
4
8
2
0
12
0.5
1100
87.7
-5.8
-6.6
0.5
88
0.12
OK.
11
55
6
2
3
8
4
8
2
0
12
2
1100
78.2
-9
-17.6
0.6
90
0.49
OK.
12
55
6
2
3
8
4
8
2
0
12
5
1100
71.2
-10.4
-23.6
0.4
88.1
1.01
OK.
13
50
6
4
6
8
4
8
2
0
12
0
1000
91.2
0.1
5.2
—
98.1
0.02
OK.
14
50
6
4
6
8
4
8
2
0
12
0.5
1000
83.3
-8.4
-7.6
—
92.4
0.24
OK.
15
50
6
4
6
8
4
8
2
0
12
2
1000
72.7
-11.7
-20.7
—
92.3
0.88
OK.
16
50
6
4
6
8
4
8
2
0
12
5
1000
62.8
-11.1
-29.8
—
88.2
2.13
OK.
17
50
6
4
6
8
4
8
2
0
12
0
1050
91.6
-0.5
3.6
—
82.3
0.02
OK.
18
50
6
4
6
8
4
8
2
0
12
0.5
1050
84.9
-7.9
-7.5
—
40.6
0 19
OK.
19
50
6
4
6
8
4
8
2
0
12
2
1050
74.1
-11.2
-20.2
—
54.8
0.77
OK.
20
50
6
4
6
8
4
8
2
0
12
5
1050
63.7
-11
-29
—
74.8
1.94
OK.
21
50
6
4
6
8
4
8
2
0
12
0
1100
92.4
-0.7
3.2
0.9
27.7
0.02
OK.
22
50
6
4
6
8
4
8
2
0
12
0.5
1100
85.2
-7.7
-7
0.7
33.6
0.18
OK.
23
50
6
4
6
8
4
8
2
0
12
2
1100
73.5
-11.5
-20.1
0.7
68 9
0.81
OK.
24
50
6
4
6
8
4
8
2
0
12
5
1100
63.6
-11.5
-28.2
0.9
76.9
1.96
OK.
25
55
6
2
3
8
4
8
2
12
0
0
1000
89.1
0.5
8.8
—
93.1
0.03
OK.
26
55
6
2
3
8
4
8
2
12
0
0.5
1000
77.7
-11.5
-7.3
—
89.3
0.45
OK.
27
55
6
2
3
8
4
8
2
12
0
2
1000
63.3
-17.3
-26.1
—
88.3
2.41
OK.
28
55
6
2
3
8
4
8
2
12
0
5
1000
51.6
-13
-38.1
—
86.9
6 84
OK.
29
55
6
2
3
8
4
8
2
12
0
0
1050
89.4
-0.4
7
—
87.8
0.03
O.K.
30
55
6
2
3
8
4
8
2
12
0
0.5
1050
79.9
-10.5
•6.9
—
91.7
0.35
OK.
31
55
6
2
3
8
4
8
2
12
0
2
1050
64.3
-16.8
•26.6
—
87.7
2 22
O.K.
32
55
6
2
3
8
4
8
2
12
0
5
1050
52.8
-12.8
-36.8
—
87.7
5.79
O.K.
33
55
6
2
3
8
4
8
2
12
0
0
1100
89.8
-0.9
5.1
1.9
85.9
0.03
O.K.
34
55
6
2
3
8
4
8
2
12
0
0.5
1100
79.6
-10.7
-8.7
1.9
94
0.38
O.K.
35
55
6
2
3
8
4
8
2
12
0
2
1100
65.7
-15.6
-25.6
2.1
92
1.90
O.K.
36
55
6
2
3
8
4
8
2
12
0
5
1100
54.6
-12.5
-34.6
2.9
90
4.60
O.K.

37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
62
53
54
55
56
57
68
59
60
61
62
63
64
65
66
67
68
69
70
71
72
S1Q2
B?03
* Frit Composition (Weight %):
Na20 K2Q CaQ AI2Q3
ZrQ2
MgO
SrO
ZnQ
** Wt% Zr-V
Pigment
Peak Firing
Temp. (C)
L!
C.I.E.
a!
b!
Delta E*
1050-1100C
60-Degree
Annie Glosa
K/Sat
545nm
Glaze
EJ1
50
6
4
6
8
4
8
2
12
0
0
1000
89.1
04
8.2
—
93.4
003
Crazed
50
6
4
6
8
4
8
2
12
0
0.5
1000
77.8
-11.3
-7.9
—
92.1
0.45
Crazed
50
6
4
6
8
4
8
2
12
0
2
1000
64.1
-16 8
-26.6
—
91.2
2.26
Crazed
50
6
4
6
8
4
8
2
12
0
5
1000
51.8
-13.4
-39.5
—
95
7.36
Crazed
50
6
4
6
8
4
8
2
12
0
0
1050
88 8
-06
7.4
—
91.9
003
Crazed
50
6
4
6
8
4
8
2
12
0
0.5
1050
78.2
-11
-7.7
—
89
042
Crazed
50
6
4
6
8
4
8
2
12
0
2
1050
64.1
-16.8
-26.3
—
91.8
2.24
Crazed
50
6
4
6
8
4
8
2
12
0
5
1050
50.4
-12.7
-39.4
—
91.8
7.96
Crazed
50
6
4
6
8
4
8
2
12
0
0
1100
88.3
-1.1
6.6
1.1
90.1
004
Crazed
50
6
4
6
8
4
8
2
12
0
0.5
1100
79.9
-9.5
-5.5
3.1
94.6
0.32
Crazed
50
6
4
6
8
4
8
2
12
0
2
1100
64.7
-16.4
-24.6
1.8
85.2
2.03
Crazed
50
6
4
6
8
4
8
2
12
0
5
1100
50.1
-12.7
-37.2
2.2
85
7.36
Crazed
55
5
2
3
13
8
0
2
0
12
0
1000
90.6
0.6
7.1
—
85.5
0.02
OK.
55
5
2
3
13
8
0
2
0
12
0.5
1000
81.8
-9.3
-6.5
—
83.1
028
OK.
55
5
2
3
13
8
0
2
0
12
2
1000
69.1
-14 7
-22.6
—
84.5
1.34
OK.
55
5
2
3
13
8
0
2
0
12
5
1000
57 8
-12 9
-33.1
—
84.1
3.56
OK.
55
5
2
3
13
8
0
2
0
12
0
1050
69.4
-0.3
7.5
—
87.8
003
OK.
55
5
2
3
13
8
0
2
0
12
0.5
1050
80.9
-9.5
-5
—
88.5
0 29
OK.
55
5
2
3
13
8
0
2
0
12
2
1050
68.5
-16.4
-21.5
—
85.5
1.46
OK.
55
5
2
3
13
8
0
2
0
12
5
1050
52.3
-13.8
-36.8
—
87.9
6 39
OK.
55
5
2
3
13
8
0
2
0
12
0
1100
88
-0.9
7.5
1.6
95.8
0.04
OK.
55
5
2
3
13
8
0
2
0
12
0.5
1100
82.2
-7.4
-1.1
4.6
92.9
0.21
OK.
55
5
2
3
13
8
0
2
0
12
2
1100
69.2
-15.7
-17.8
3.9
93
1.25
OK.
55
5
2
3
13
8
0
2
0
12
5
1100
53.8
-16.5
-33.3
4.7
93.2
5.79
OK.
50
5
4
6
13
8
0
2
0
12
0
1000
92
0.3
5.1
—
54.3
0 02
Crazed
50
5
4
6
13
8
0
2
0
12
0.5
1000
83.9
-7.8
-6.2
—
45.9
0.21
Crazed
50
5
4
6
13
8
0
2
0
12
2
1000
72.7
-12.3
-20.5
—
46.2
0.89
Crazed
50
5
4
6
13
8
0
2
0
12
5
1000
62.6
-11.1
-30.2
—
42
2.15
Crazed
50
5
4
6
13
8
0
2
0
12
0
1050
90.2
-0.3
5.7
—
55.5
0.03
Crazed
50
5
4
6
13
8
0
2
0
12
0.5
1050
84
-7.2
-3.2
—
55.4
0.18
Crazed
50
5
4
6
13
8
0
2
0
12
2
1050
72.7
-14.1
-17.8
—
59
0.91
Crazed
50
5
4
6
13
8
0
2
0
12
5
1050
59
-13.7
-31.6
—
59
3.19
Crazed
50
5
4
6
13
8
0
2
0
12
0
1100
87.2
-0.8
7
3.3
77.5
0.05
Crazed
50
5
4
6
13
8
0
2
0
12
0.5
1100
85.7
-3.5
3.9
8.2
86.4
0.09
Crazed
50
5
4
6
13
8
0
2
0
12
2
1100
78
-10.5
-6.2
13.2
93.7
0.41
Crazed
50
5
4
6
13
8
0
2
0
12
5
1100
61.3
-16.7
-25.7
7
01.8
2.67
Crazed
uy
4^

73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
S1Q2
B2Q3
* Frit Composition (Weight %):
Ns2Q K2Q CaQ AI2Q3
ZrQZ
MgQ
SrO
ZnO
** WL% Zr-V
Elgmenl
Peak Firing
Temp. (C)
L!
C.I.E.
a!
h!
Delta E*
105Q:110QC
60-Degree
AoaleJjloss
K/Sat
64Qnm
•** Glaze
Ell
55
5
2
3
13
8
0
2
12
0
0
1000
89 6
06
8 7
—
75.6
0.02
OK.
55
5
2
3
13
8
0
2
12
0
0 5
1000
80.1
-9.9
-5.4
—
60.2
032
OK.
55
5
2
3
13
8
0
2
12
0
2
1000
654
-15.9
-24.8
—
57.5
1.88
OK.
55
5
2
3
13
8
0
2
12
0
5
1000
55 9
-12.7
-34.5
—
54.9
4.15
OK.
55
5
2
3
13
8
0
2
12
0
0
1050
89.1
-0.4
7.5
—
79.1
0.03
OK.
55
5
2
3
13
8
0
2
12
0
0.5
1050
81.4
-8.7
-3.9
—
69.7
026
OK.
55
5
2
3
13
8
0
2
12
0
2
1050
67
-16.2
-23.4
—
65.1
1 67
OK.
55
5
2
3
13
8
0
2
12
0
5
1050
54.5
-14
-35.2
—
74.1
5.07
OK.
55
5
2
3
13
8
0
2
12
0
0
1100
88.4
-0.9
6 9
1.1
79.3
0.04
OK.
55
5
2
3
13
8
0
2
12
0
0 5
1100
83.2
-6.7
-0.5
4.3
81.4
0 18
OK.
55
5
2
3
13
8
0
2
12
0
2
1100
71.1
-14.9
-16.4
8.3
757
1.02
OK.
55
5
2
3
13
8
0
2
12
0
5
1100
56.3
-16.1
-31.4
4.7
78.3
4.37
OK.
50
5
4
6
13
8
0
2
12
0
0
1000
91.8
0.1
6.1
—
15.4
002
Crazed
50
5
4
6
13
8
0
2
12
0
0.5
1000
83
-8.7
-6.3
—
26.5
0.24
Crazed
50
5
4
6
13
8
0
2
12
0
2
1000
70.6
-13.6
-22.2
—
29.9
1.13
Crazed
50
5
4
6
13
8
0
2
12
0
5
1000
62.2
-10.8
-30.2
—
32 5
2.15
Crazed
50
5
4
6
13
8
0
2
12
0
0
1050
90.9
-0.5
5.8
—
19.5
0.02
Crazed
50
5
4
6
13
8
0
2
12
0
0.5
1050
85.6
-6.3
-2.1
—
26.3
0.14
Crazed
50
5
4
6
13
8
0
2
12
0
2
1050
74
-12.9
-17.4
—
27
0 78
Crazed
50
5
4
6
13
8
0
2
12
0
5
1050
65.5
-11.4
-25.9
—
26
1.56
Crazed
50
5
4
6
13
8
0
2
12
0
0
1100
90
-0.9
6
1
27
0.03
Crazed
50
5
4
6
13
8
0
2
12
0
0.5
1100
88.7
-2.7
3.4
7.3
25.4
0.05
Crazed
50
5
4
6
13
8
0
2
12
0
2
1100
82.8
-8.2
-4.5
16 3
30 7
0.23
Crazed
50
5
4
6
13
8
0
2
12
0
5
1100
72.1
-11.9
-17
11.1
27.6
0.84
Crazed
* Coatings were applied at 2.70 cm3 volume per 2" X 6" tile.
( average wet application weight of 11.5 grams at a 1.75 specific gravity)
** Pigment was Ceredec 41715A turquoise-blue dispersible stain.
*** Thermal expansion of the wall tile body/engobe substrate was 65 - 67 X 10e-7/C.
345

APPENDIX E
FRIT SPECTRAL REFLECTANCE DATA AND CURVES
AT EACH TEMPERATURE AND PIGMENT LOADING

Table E.l. Engobe and Frits A and B Reflectance Data
(+ percentages are wt.% Zr-V pigment added to frit)
Wavelength
1000C PEAK TEMPERATURE:
ImnJ
Enaobe
A.
A + 0.5%
A ♦ 2.0%
A + 5.0%
B.
B ♦ 0.5%
B + 2.0%
B + 5.0%
360
48.5
58.8
55.3
47.4
37.9
46.9
43.1
33.6
23.3
380
56.7
69.7
66.3
59.3
49.7
57.9
55.2
46.1
35.2
400
64.2
76.4
73.6
67.7
58.7
65.5
63.9
56.2
45.8
420
69.3
79.2
77.0
72.0
63.9
69.3
68.2
61.8
52.6
440
72.0
80.3
78.5
74.0
66.9
71.3
70.5
65.0
56.7
460
73.7
81.7
80.0
75.7
69.0
73.2
72.5
67.3
59.3
480
74.8
82.2
80.2
75.2
68.1
74 3
73.1
67.2
58.4
500
76.7
83.1
79.8
72.2
64.0
76.0
73.1
644
53.1
520
78.2
83.6
77.5
65.9
56.1
77.2
70.8
57.4
43.6
540
79.6
84.0
73.9
58.6
47.8
78.3
66.8
49.1
34 4
560
81 4
84.5
69.7
51.8
40.4
79.5
61.9
41.3
27.0
580
83.1
84.7
65.8
46.4
34.8
80.4
57.3
35.6
21.9
600
844
85.1
62.9
42.6
31.0
81.2
53.8
31.8
18.8
620
85.1
85.3
61.0
40.4
28.8
81.7
51.8
29.6
17.0
640
85.5
85.3
60.2
39.4
27.9
82.0
51.0
28.8
16.4
660
85,9
85.4
60.4
39.7
28.0
82.2
51.6
29.2
16.7
680
86.4
857
61.7
41.2
29.4
82.5
53.6
30.9
18.1
700
86.9
85.9
64.0
43.9
32.0
82.8
56.9
34 0
20.5
720
87.4
86.1
66.4
46.9
34.7
83.1
60.2
37.3
23.2
740
87.9
86.3
67.4
48.1
35.8
83.2
61.6
38.7
24.4
1050C PEAK TEMPERATURE:
Enaobe
A.
A + 0.5%
A + 2,0%
A + 5,0%
B.
B + 0.5%
B + 2,0%
B + 5,0%
360
444
59.6
55.8
47.6
38.1
50.5
47.7
36.3
24.5
380
52.6
70.9
67.3
59.5
49.8
61 9
59.9
49.0
36.5
400
61.0
78.0
74.7
67.9
58.9
69.3
68.1
59.0
47.1
420
67.1
80.8
77.9
72.1
64 0
-72.-1
72.1
64.4
53.7
440
70 3
81.8
79.2
74.1
66 8
74.2
74.1
67.3
57.5
460
72.0
83.0
80.5
75.8
68.8
75.9
75.8
69 5
60.0
480
73.2
83.4
80.5
75.2
67.8
76.8
76 4
69.3
59.1
500
75.3
84.1
80.1
72.3
63.4
78.2
76.2
66.4
54.0
520
76.9
84.4
77.8
65.9
55.3
79.1
73.9
59.5
44.7
540
78.1
84.6
74.3
58.5
46 8
79.8
69.9
51.3
356
560
79 5
84.8
70.2
51.7
39.3
80.3
65.0
43.6
28 2
580
80.7
84.7
66.3
46.3
33.7
80.6
60.4
37.8
23.1
600
81.7
84.8
63.4
42.6
30.0
81.0
57.1
34.0
19.9
620
82.4
84.8
61.5
40.4
27.8
81.2
55.0
31.8
18.1
640
82.8
84.8
60.7
39.4
26.9
81.4
54.3
31.0
17.5
660
80.8
84.8
61.0
39.6
27.1
81.5
54.8
31.4
17.8
680
83.7
84.9
62.4
41.1
28.6
81.8
56.8
33.1
19.2
700
84.2
85.0
64.7
43.8
31.2
81.9
60.0
36.2
21.7
720
84.8
85.2
67.2
467
34.1
82.2
63.2
39.6
24.5
740
85.2
85.3
68.2
47.9
35.3
82.3
64.5
41.0
25.6
1100C PEAK TEMPERATURE:
Engobe
A.
A+ 0,5%
A + 2,0%
A + 5,0%
B
B + 0.5%
B + 2,0%
B + 5.0%
360
43.1
60.2
56.4
45.9
37.8
53.2
47.5
34.6
23.9
380
51.3
71.6
67.9
57.8
49 4
64.8
59.7
47.5
35.7
400
59.5
78.6
75.1
66.3
58 4
72.1
68.0
57.5
46 1
420
65.8
81.3
78.2
70.5
63.2
75.3
72.0
63.0
52.6
440
69.0
82.1
79.3
72.6
65.9
76.6
73.8
65.9
56.4
460
70.6
83.2
80.4
74.4
677
78.3
75.5
68.0
58.9
480
71.8
83.5
80.5
73.8
66.8
79.2
76.1
67.9
58.2
500
74.1
84 2
80.0
70.9
62.6
80.5
76.1
65.2
53.5
520
75.6
84.4 77.9
64.7
54.7
81 4
74.0
58.5
44 6
540
76.7
84.4 74.7
57.5
46.4
81.9
70.3
50.3
35.6
560
77.7
84.5 | 70.8
50.8
39.0
82.3
65.6
42.7
28.1
580
78.5
84.2 67.0
45.5
33.4
82.4
61.2
36.9
23.0
600
79.3
84.1 64.2
41 8
29.7
82.5
57.9
33.0
19.8
620
79.8
84.1 I 62.4
39.7
276
82.7
55.9
30.9
18.1
640
80.2
84.0 61.6
38.7
26 7
82.8
55 1
30.1
17.4
660
80.6
83.9 | 61.8
39.0
26.9
82.9
55.7
30.5
17.7
680
81.1
84.0 63.1
40.5
28.4
83 1
57.7
32.3
19.2
700
81.5
84.0 6 5.3
43.2
31.1
83.2
60.8
35.4
21.7
720
82.0
84.2 67.7
46.2
34.0
834
63.8
38.8
24.6
740
82.4
84.2 ' 68.6
47.4
35.1
83.5
65.1
40.3
25.8
347

Table E.2. Frits C and D Reflectance Data
(+ percentages are wt.% Zr-V pigment added to frit)
Wavelength
1000C PEAK TEMPERATURE:
fnm)
c
C ♦ 0.5%
C + 2.0%
C + 5.0%
D
D + 0.5%
D + 2.0%
D + 5.0%
360
27.0
20.1
13.5
9.8
28.4
21.2
14.5
9.3
380
40.2
33.6
25.7
19.5
41.6
35.0
27.6
19.8
400
51.7
46.3
38.7
31.0
53.2
47.9
41.2
32.4
420
58.8
54.4
47.7
39.8
59.9
55.7
50.1
41.7
440
62.5
58.7
53.0
45.7
63.2
59.8
55.1
47.5
460
65.1
61.7
56.4
49.4
65.7
62.5
58.2
50.9
480
67.0
63.3
57.0
48.8
67.4
63.9
58.5
50.1
500
69.5
64.0
54.4
43.0
69.8
64.4
55.7
44.1
520
71.5
61.8
46.5
32.0
71.7
62.1
47.7
32.9
540
73.2
57.4
36.7
21.8
73.3
57.5
37.7
22.2
560
75.2
51.9
27.7
14.5
75.2
52.0
28.7
14.5
580
76.9
46.8
21.4
10.2
76.7
46.8
22.2
10.0
600
78.2
43.0
17.6
7.9
77.9
43.0
18.3
7.5
620
79.0
40.7
15.6
6.8
78.6
40.7
16.2
6.3
640
79.5
40.1
15.0
6.4
79.0
40.1
15.7
6.0
660
799
41.1
15.7
6.7
79.4
41.1
16.4
6.4
680
80.5
44.0
17.8
7.8
79.9
44 0
18.7
7.6
700
80.9
484
21.6
9.9
80.4
48.4
22.6
9.9
720
81.4
52.9
25.8
12.4
80.8
52.7
27.0
12.6
740
81.7
54.7
27.7
13.6
81.1
545
28.9
13.8
1050C PEAK TEMPERATURE:
£
C + 0.5%
C + 2.0%
C + 5.0%
Q
D + 0.5%
D + 2.0%
D ♦ 5.0%
360
29.3
24.4
16.1
11.5
27.8
22.0
14.6
8.9
380
43.2
38.6
28.9
21.6
41.4
35.7
27.6
18.9
400
548
51.1
41.9
32.9
53.3
48.5
41.0
30.8
420
61.8
58.8
50.5
41.2
60.3
56.4
49.9
39.7
440
65.2
62.6
55.4
466
63.6
60.3
54.7
45.3
460
87.5
65.1
58.4
49.8
65.9
62.9
57.7
48.5
480
69.2
66.4
58 8
49.2
67.8
642
58.0
47.7
500
71.5
67.1
56.0
43.5
70.2
84.7
55.3
41.8
520
73.2
651
47.9
32.9
72.0
62.6
47.5
30.9
540
744
61.0
38.0
23.0
73.4
58.2
37.7
20.7
560
75.6
55.7
28.9
15.8
74.6
52.7
28.7
13.5
580
76.5
50.7
22.5
11.4
75.3
47.6
22.3
9.3
600
77.3
47.0
185
9.0
76.1
43.8
18.4
7.0
620
77.9
44.8
16.5
7.8
76.6
41.7
16.4
5.9
640
78.2
44.1
15.9
7.4
76.9
41.1
15.8
5.6
660
78.5
45.1
16.5
7.7
77.2
42.1
16.5
5.9
680
79.0
47.9
18.7
8.9
77.8
44.9
18.7
7.0
700
79.4
52.0
224
11.0
78.2
49.0
22.6
9.1
720
79.9
56.0
26.5
13.4
78.6
53.0
26.8
11.6
740
80.1
57.7
28.4
14.6
78.9
54.7
28.7
12.7
z
1100C PEAK TEMPERATURE:
£
C + 0.5%
£ + 3.0%
C + 5.0%
D
D + 0.5%
D + 2.0%
D ♦ 5.0%
360
35.7
284
19.6
13.5
29.3
24.5
15.5
9.1
380
49.9
429
32.7
23.9
43.1
38.5
28.5
18.5
400
60.2
54 5
45.0
34.8
54.3
50.5
41.3
29.5
420
65.8
61.2
52.7
42.6
60.6
57.7
49.6
37.7
440
68.3
64.3
57.0
47.4
63.4
61.1
54.1
42.8
460
70.2
666
59.7
50.4
65.7
63.6
57.0
46.0
480
71.7
67.6
60.0
49.8
67.4
64.9
57.5
45.4
500
73.5
67.8
57.0
44.5
69.7
65.8
55.0
40.1
520
74.8
65.3
49.2
34.5
71 4
84.2
47.8
30.0
540
75.7
60.7
39.6
25.0
72.5
60.7
38.5
20.5
560
76.4
55.0
30.8
17.8
73.4
56.1
29.8
13.6
580
76.7
49.6
24.5
13.3
73.7
51.5
23.5
9.6
600
77.1
45.7
20.6
10.7
74.1
48.2
19.6
7.4
620
77.3
43.4
18.5
9.4
74.5
46.2
17.6
6.3
640
77.5
42.7
17.8
9.0
74.7
45.7
17.0
6.0
660
77.6
43.5
18.4
9.3
75.0
46.6
17.7
6.3
680
78.0
46.1
20.5
10.5
75.5
49.2
19.9
7.4
700
78.3
50.1
24.0
12.7
75.9
52.8
23.7
9.4
720
78.6
53.9
27.9
15.3
76.3
56.4
27.8
11.8
740
78.8
55.6
29.7
16.4
76.5
57.9
29.6
13.0

349
Table E.3. Frits E and F Reflectance Data
(+ percentages are wt.% Zr-V pigment added to frit)
Wavelength i
1000C PEAK TEMPERATURE:
(nml
E
E + 0.5%
E + 2.0%
E + 5.0%
E
F + 0.5%
F + 2.0% F ♦ 5.0%
360
35.1
31 4
23.3
16.1
51.3
42.7
32.0
22.5
380
48.1
44.6
35.8
26.7
60.3
54.0
44.3
34.2
400 58.6
55.8
47.6
37.6
67.5
62.9
54.9
45.2
420
64.8
62.8
55.5
45.8
71.2
67.7
61.1
52.4
440
677
66.2
60.0
51.2
73.0
70.0
64.5
56.7
460
69.7
68.4
62.8
54.5
74.8
71.9
66.9
59.4
480
71.1
69.5
63.2
54.0
75.8
72.7
66.9
58.5
500
73.2
69.9
60.7
48.7
77.4
72.9
64.4
53.3
520
75.0
68.1
53.5
38.5
78.7
71.1
57.6
43.6
540
76.5
64.3
44.4
28.6
79.8
67.6
49.2
34.1
560
78.2
59.4
35.8
20.9
81.1
63.2
41.2
26.5
580
79.7
54.7
29.5
16.1
82.1
59.0
35.3
21.5
600
80.9
51.1
25.4
13.2
83.0
55.7
31.4
18.5
620
81.7
48.9
23.2
11.7
83.5 | 53.7
29.3
16.8
640
82.2
48.2
22.4
11.1
83.8
53.1
28.6
16.3
660
82.6
49.1
23.0
11.5
84.1
53.9
29.1
16.6
680
83.1
51.7
25.1
12.7
84.5
56.2
31.1
18.0
700
83.6
55.7
28.7
15.1
84.8
59.8
34.5
20.6
720
84.1
59.7
32.7
17.8
85.2
63.3
38.2
23.4
740
84.5
61.4
34.5
19.0
85.4
64 9
39.9
246
*
1050C PEAK TEMPERATURE:
E
E + 0.5%
E + 2-0%
E + 5.0%
E
F + 0.5%
F + 2.0%
F ♦ 5.0%
360
25.2
22.5
17.2
9.9
40.6
33.5
25.0
15.7
380
39.2
36.3
29.9
19.4
51.4
46.1
37.9
26.8
400
52.2
49.4
43.0
30.6
60.7
56.9
49.9
38.6
420
60.7
58.2
52.4
39.7
66.0
63.3
57.4
46.8
440
64 8
62.6
57.6
45.6
68.6
66.5
61.5
51.7
460
67.1
65.1
60.5
49.2
70.5
68.7
64.0
54.7
480
68.8
66.5
61.2
48 8
71.9
69.9
64.6
54.4
500
71.3
67.5
59.5
43.4
73.8
70.8
63.3
49.9
520
73.1
66.0
52.9
32.8
75.2
69.9
57.8
40.3
540
74.4
62.5
44.0
22.7
76.3
67.4
50.0
30.2
560
75.8
57.9
35.2
15.2
77.3
63.8
41.9
22.2
580
76.8
53.4
28.5
10.8
78.0
60.2
35.5
17.1
600
77.8
50.0
24.2
8.4
78.7
57.5
31.3
14.1
620
78.4
47.9
21.9
7.2
79.2
55.8
29.0
12.6
640
78.8
47.3
21.2
6.8
79.4
55.2
28.2
12.1
660
79.2
48.2
22.0
7.1
79.7
56.0
29.0
12.5
680
79.7
50.8
24.4
8.2
80.1
58.2
31.4
14.0
700
80.1
54.7
28.6
10.4
80.3
61.3
35.5
16.6
720
80.6
58.5
33.1
13.0
80.6
64.3
39.8
19.6
740
81.0
60.2
35.2
14.2
80.7
65.6
41.7
21.0
1100C PEAK TEMPERATURE:
E
E + 0,5%
E 7 2.0%
E + 5.0%
E
F ♦ 0.5%
F ♦ 2.0%
F + 5.0%
360
23.0
20.7
14.5
8,4
25.5
24.3
19.9
11.6
380
36.5
342
26.9
18.0
38.6
37.9
33.5
23.2
400
49.3
47.2
40.0
29.6
50.4
50.3
46.1
35.6
420
57.9
56.1
49.4
38.8
57.7
58.0
54.2
44.3
440
61.9
60.5
54.6
44.6
61.0
61.6
58.3
494
460
64.3
63.1
57.7
48.1
63.2
64.0
60.9
52.4
480
66.2
64.8
58.8
48 3
64.9
65.7
62.4
53.0
500
68.7
66.3
57 8
44 2
67.3
67.7
63.2
50.5
520
70.5
65.8
52.5
34.8
69.0
68.7
61.4
43.0
540
71.7
63.7
44.7
24.9
70.1
68.7
57.5
33.8
560
72.8
60.7
36.8
17.0
71.1
68.0
52.5
25.6
580
73.5
57.4
30.5
12.0
71.7
67.0
47.8
19.8
600
74.1
54.9
26.4
9.2
72.3
66.2
44.3
16.3
620
74.6
53.4
24.2
7.9
72.7
65.7
42.3
14.5
640
74.9
53.0
23.5
7.4
73.0
65.6
41.7
13.9
660
75.2
53.8
24.3
7.9
73.3
66.1
42.5
14.5
680 75.7
55.9
26.7
9.3
73.7
67.3
44.8
16.4
700 760
58.9
30.7
11.9
73.9
68.7
48.4
19.8
720 763
61.7
35.0
15.0
74.2
70.0
51.8
23.4
740 1 764
62.9
36.9
16 5
74.3
70.5
53.2
25.1

Table E.4. Frits G and H Reflectance Data
(+ percentages are wt.% Zr-V pigment added to frit)
Wavelength |
1000C PEAK TEMPERATURE:
InmJ
S
G + 0.5%
G + 2.0%
G + 5.0%
H
H + 0.5%
H ♦ 2.0%
H * 5.0%
360
29.3
25.1
18.3
14.1
42.5
37.6
27.5
21.9
380
42.0
38.1
30.3
24.4
54.1
49.6
40.4
33.4
400
53.0
49.8
42.4
35.5
63.4
59.6
51.9
44.5
420
60.1
574
50.6
43.8
68.7
65.3
58.8
51.7
440
63.7
61.3
55.4
49.2
71.2
68.2
62.5
56.0
460
66.2
64.0
58.5
52.6
73.3
70.4
65.0
58.6
480
68.0
65.5
59.1
52.1
74.7
71.4
65.2
57.8
500
70.5
66.4
56.5
46.7
76.7
71.8
62.6
52.7
520
72.5
64.8
48.8
36.3
78.1
69.9
55.4
43.0
540
74.2
61.1
39.2
26.3
79.4
66.1
46.4
33.4
560
76.3
56.4
30.5
18.7
80.8
81.4
37.9
25.9
580
78.0
51.8
24.3
14.1
82.1
56.9
31.6
21.1
600
79.4
48.3
20.5
11.5
83.0
53.5
27.7
18.3
620
80.2
46.2
18.5
10.2
83.4
51.4
25.6
16.8
640
80.7
45.6
17.9
9.8
83.6
50.7
24.9
16.3
660
81.1
46.6
18.5
10.1
83.7
51.6
25.6
16.7
680
81.6
494
20.6
11.3
83.9
54.1
27.8
18.1
700
82.1
53.7
24.3
13.6
84.0
57.8
31.5
20.6
720
82.6
57.8
28.3
16.2
84.2
81.4
35.5
23.4
740
82.9
59.5
30.2
17.4
84.2
62.8
37.3
24.7
1050C PEAK TEMPERATURE:
s
G + 0.5%
G ♦ 2.0%
G + 5.0%
U
H + 0.5%
H 2.0%
H ♦ 5.0%
360
27.1
24.4
17.7
11.6
40.3
38.4
29.8
24.4
380
40.6
37.9
30.6
21.8
51.9
50.2
42.7
35.9
400
52.6
50.1
43.3
33.5
61.4
60.0
53.6
46.5
420
60.4
58.1
52.0
42.3
67.1
65.7
60.1
53.3
440
64.1
62.1
56.8
477
69.7
68.5
63.6
57.2
460
66.4
64.6
59.8
51.0
71.8
70.6
66.0
59.8
480
68.2
66.0
60.3
50.6
73.3
71.8
66.5
59.5
500
70.7
67.2
58.1
45.6
75.3
72.9
65.0
55.6
520
72.5
66.0
50.9
35.3
76.8
72.3
59.4
47.1
540
73.8
62.9
41.6
25.1
77.8
70.2
51.7
38.0
560
75.1
58.8
32.8
17.4
78.8
67.2
43.9
30.5
580
76.1
54.7
26.3
12.6
79.5
64.1
37.8
25.4
600
77.0
51.6
22.2
10.0
80.1
61.6
33.7
22.4
620
77.5
49.8
20.1
8.7
80.5
60.0
31.5
20.8
640
77.8
49.3
19.4
8.3
80.6
59.5
30.8
20.3
660
78.1
50.2
20.1
8.7
80.8
60.2
31.5
20.7
680
78.6
52.7
22.5
10.0
81.0
82.1
33.8
22.3
700
79.0
56.4
26.4
12.4
81.1
64.8
37.6
25.0
720
79.3
59.8
30.7
15.2
81.3
67.3
41.5
28.0
740
79.5
61.3
32.6
16.5
81.3
68.3
43.1
29.4
1100C PEAK TEMPERATURE:
S
G + 0.5%
G + 2.0%
G + 5,0%
H
H + 0.5%
H * 2.0%
H ♦ 5.0%
360
27.1
23.9
17.8
10.4
36.7
38.8
34.1
28.0
380
40.9
37.8
31.0
20.9
49.1
50.7
46.9
40.0
400
52.6
50,1
43.7
32.8
59.1
60.1
56.9
50.1
420
59.9
57.9
52.2
41.5
65.0
65.5
62.7
56.1
440
63.3
61.6
56.7
46.8
67.6
88.0
65.6
59.3
460
65.6
64.2
59.6
50.1
89.8
70.1
67.8
61.7
480
674
65.8
60.7
50.3
71.4
71.6
69.1
62.3
500
69.7
67.4
59.8
46.5
73.6
73.4
69.9
60.7
520
71.4
67.1
54 9
37.4
75.1
74.3
68.6
55.3
540
72.6
65.4
47.7
27.6
76.1
74.5
65.5
48.0
560
73.5
62.6
39.9
19.5
77.0
74.1
61.3
40.9
580
74.1
59.6
33.6
14.3
77.4
73.3
57.2
35.5
600
74.7
57.3
29.4
11.4
77.9
72.6
54.1
32.1
620
75.1
55.9
27.2
9.9
78.1
72.2
52.2
30.2
640
75.3
55.5
26.5
9.4
78.2
72.0
51.6
29.6
660
75.6
56.3
27.3
9.9
78.3
72.3
52.2
30.1
680
76.0
58.3
29.8
11.5
78.6
73.1
54.3
32.0
700
76.2
61.0
33.9
14.3
78.7
74.0
57.2
35.1
720
76.5
63.5
38.2
17.5
78.9
74.8
60.1
38.4
740
76.6
64 5
40.0
19.1
78.9
75.1
61.2
39.8

FRIT A; 1000C PEAK TEMPERATURE:
Wavelength (nm)
—*-0% Zr-V —0.5% Zr-V -*-2.0% Zr-V -«-5.0% Zr-Vj
FRIT A; 1050C PEAK TEMPERATURE:
Wavelength (nm)
-0% Zr-V
-0.5% Zr-V
-2.0% Zr-V
- 5.0% Zr-VI
FRIT A; 1100C PEAK TEMPERATURE:
Wavelength (nm)
I 0% Zr-V -*-0.5% Zr-V -*-2.0% Zr-V -x- 5.0% Zr-V I
« l
Figure E.l. Frit A spectral reflectance curves.

FRIT B; 1000C PEAK TEMPERATURE:
Wavelength (nm)
—0% Zr-V -*-0.5% Zr-V -e-2.0% Zr-V 5.0% Zr-V
FRIT B; 1050C PEAK TEMPERATURE:
Wavelength (nm)
—0% Zr-V -*-0.5% Zr-V -e-2.0% Zr-V -*-5.0% Zr-V I
FRIT B; 1100C PEAK TEMPERATURE:
Wavelength (nm)
-*-0% Zr-V -*-0.5% Zr-V -a-2.0% Zr-V -*-5.0% Zr-V I
Figure E.2. Frit B spectral reflectance curves

FRIT C; 1000C PEAK TEMPERATURE:
Wavelength (nm)
->-0% Zr-V -«-0.5% Zr-V -*-2.0% Zr-V -««-5.0% Zr-V I
FRIT C; 1050C PEAK TEMPERATURE:
Wavelength (nm)
¡ 0% Zr-V 0.5% Zr-V 2.0% Zr-V 5.0% Zr-V I
FRIT C; 1100C PEAK TEMPERATURE:
Wavelength (nm)
! —*—0% Zr-V -*-0.5% Zr-V -*-2.0% Zr-V -«-5.0% Zr-V I
Figure E.3. Frit C spectral reflectance curves

FRJT D; 1000C PEAK TEMPERATURE:
Wavelength (nm)
0% Zr-V 0.5% Zr-V 2.0% Zr-V 5.0% Zr-V!
FRIT D; 1050C PEAK TEMPERATURE:
FRIT D; 1100C PEAK TEMPERATURE:
Wavelength (nm)
-»-0% Zr-V —0.5% Zr-V -+-2.0% Zr-V -«-5.0% Zr-VI
Figure E.4. Frit D spectral reflectance curves.

FRIT E; 1000C PEAK TEMPERATURE:
Wavelength (nm)
—0% Zr-V —0.5% Zr-V -*-2.0% Zr-V -*-5.0% Zr-V
FRIT E; 1050C PEAK TEMPERATURE:
Wavelength (nm)
| -»-0% Zr-V -«-0.5% Zr-V -*-2.0% Zr-V -*-5.0% Zr-V|
FRIT E; 1100C PEAK TEMPERATURE:
Wavelength (nm)
-*-0% Zr-V -*—0.5% Zr-V -*-2.0% Zr-V -*-5.0% Zr-V l
Figure E.5. Frit E spectral reflectance curves.

FRIT F; 1000C PEAK TEMPERATURE:
Wavelength (nm)
: 0% Zr-V -*-0.5% Zf-V 2.0% Zr-V -x- 5.0% Zf-V |
FRIT F; 1050C PEAK TEMPERATURE:
Wavelength (nm)
-»-0% Zr-V 0.5% Zr-V 2.0% Zr-V -k- 5.0% Zr-V j
FRIT F; 1100C PEAK TEMPERATURE:
Figure E.6. Frit F spectral reflectance curves.

FRIT G; 1000C PEAK TEMPERATURE:
Wavelength (nm)
-4- 0% Zr-V -9- 0.5% Zr-V 2.0% Zr-V 5.0% Zr-V ¡
FRIT G; 1050C PEAK TEMPERATURE:
Wavelength (nm)
j —-0% Zr-V -«-0.5% Zr-V -*—2.0% Zr-V -w-5.0% Zr-V
FRIT G; 1100C PEAK TEMPERATURE:
Wavelength (nm)
i -»—0% Zr-V —•—0.5% Zr-V -*-2.0% Zr-V -*«-5.0% Zr-V I
Figure E.7. Frit G spectral reflectance curves.

FRITH; 1000C PEAK TEMPERATURE:
Wavelength (nm)
j 0% Zr-V -*- 0.5% Zr-V -*- 2.0% Zr-V -x- 5.0% Zr-V ¡
FRITH; 1050C PEAK TEMPERATURE;
Wavelength (nm)
—0% Zr-V -*-0.5% Zr-V -*-2.0% Zr-V -*«-5.0% Zr-V |
FRITH; 1100C PEAK TEMPERATURE:
Wavelength (nm)
- 0% Zr-V
-0.5% Zr-V
-2.0% Zr-V
-5.0% Zr-V I
Figure E.8. Frit H spectral reflectance curves.

REFERENCES
Ale74 L. Alexander and H. Klug, X-rav Diffraction
Procedures. 2nd Ed., John Wiley and Sons, New York
(1974) .
Amo94 J. Amoros, British Ceramic Transactions. 93 (6), 224-
228 (1994).
Apa94 J. Aparici, A. Moreno, A. Escardino, S. Mestre,
Oualicer 94. 2, 35-45 (1994).
Ask94 D. Askeland, The Science and Engineering of Materials.
3rd Ed., PWS Publishing Co., Boston MA (1994).
Aza97 J. Aza, N. Baldini, E. Barszczewski, L. Bernhardt, E.
Gutman, J. Kramer and C. Leinweber, Annual Book of
ASTM Standards. 15.02 (15) (1997).
Ban86 N. Bansal and R. Doremus, Handbook of Glass
Properties. Academic Press, Orlando, FL (1986).
Bar96 L. Barbieri, C. Leonelli and T.Manfredini, Ceramic
Engineering and Science Proceedings. 17 (1), 11-22
(1996).
Bar97 T. Barson, Ceramic Engineering and Science
Proceedings. 18 (2), 28-36 (1997).
BÍ167 F. Billmeyer and M. Saltzman, Principles of Color
Technology. John Wiley and Sons, New York (1967).
Blo93 R. Blonski, Ceramic Engineering and Science
Proceedings. 14 (1-2), 176-189 (1993).
Blo94a R. Blonski, Ceramic Engineering and Science
Proceedings. 15 (1), 249-265 (1994).
359

360
Blo94b R. Blonski, Ceramic Engineering and Science
Proceedings. 15 (1), 266-280 (1994).
Boa83 M. Boas, Mathematical Methods in the Physical
Sciences. 2nd Ed., John Wiley and Sons, New York
(1983).
Bow87 R. Bowles and S. Haller, Multiple Correlation.
Computer software, SSI Management Consultants, Inc.,
Cleveland, OH (1987).
Bow90 R. Bowles and S. Haller, Problem Solving with
Experimental Design. Multiple Correlation. Multiple
Property Optimization and Partitioning Variability.
SSI Management Consultants, Inc., Cleveland, OH
(1990) .
Bra86 G. Brady and H. Clauser, Material Handbook. 12th Ed.,
McGraw Hill, New York, NY (1986).
Bur79 A. Burgyan, Interceram. 28 (1) 1979.
Bur96 B. Burzacchini, M. Paganelli and H. Christ, Ceramic
Engineering and Science Proceedings. 17 (1) , 60-66
(1996).
Byr94 C. Byrne, S. Kutney and R. Pipoly, Ceramic Engineering
and Science Proceedings. 15 (1), 206-223 (1994).
Cas97 R. Castilone and W. Carty, Ceramic Engineering and
Science Proceedings. 18 (2), 83-84 (1997).
Che83 H. Chen, Theory of Electromagnetic Waves. McGraw Hill,
Inc., Fairfax, VA (1983).
Chi97 Y. Chiang, D. Birnie and W. Kingery, Physical
Ceramics. John Wiley and Sons, New York, NY (1997).
Con97 C. Concepcion, J. Oteo, E. Ocana, J. Rubio and M.
Velaso, Ceramic Engineering and Science Proceedings.
18 (2), 96-113 (1997).
C. Decker, Ceramic Engineering and Science
Proceedings. 11 (3-4), 307-313 (1990).
Dec90

361
Dec93 C. Decker, Ceramic Engineering and Science
Proceedings. 14 (1-2) , 155-162 (1993).
DeH93 R. DeHoff, Thermodynamics in Materials Science.
McGraw-Hill, New York (1993).
Din96 R. Dinger and C. Villa, American Ceramic Society
Bulletin. 75 (4), 93-95 (1996).
Din96a R. Dinger and C. Villa, American Ceramic Society
Bulletin. 75 (5), 80-83 (1996).
Dod94 A. Dodd and D. Murfin, Dictionary of Ceramics. 3rd
Ed., University Press, Cambridge, London (1994).
Dor94 R. Dormeu, Glass Science. 2nd Ed., John Wiley and
sons, New York (1994).
Dow83 S. Dowdy and S. Weardon, Statistics for Research. John
Wiley and Sons, New York (1983).
Dry82 Dry Color Manufacturers Association Classification of
Mixed Metal Oxide Inorganic Colored Pigments. 2nd Ed.,
Published by DCMA, Arlington, VA (1982).
Dun49 R. Duncan, Journal of Oil and Color Chemists'
Association. 32. 296 (1949).
Ear94 D. Earl, M.S. Thesis, University of Florida (1994).
Eld94 S. Eldefrawi, Second International Ceramics Congress.
1, 196-205 (1994).
Enr96 J. Enrique, J. Amoros and A. Moreno, in The Science of
Whitewares. American Ceramic Society, Westerville, OH,
357-370 (1996).
Epp87 R. Eppler, Ceramic Bulletin. 66 (11), 1600 (1987).
Epp96 D. Eppler and R. Eppler, Ceramic Engineering and
Science Proceedings. 17 (1), 77-87 (1996).
Epp97a D. Eppler and R. Eppler, Ceramic Engineering and
Science Proceedings. 18 (2), 141-142 (1997).

362
Epp97b R. Eppler and D. Eppler, Ceramic Engineering and
Science Proceedings. 18 (2), 150-158 (1997).
Esc96 A. Escardino, Oualicer 96. 1, 169-184 (1996).
Fer96 "The U.S. Tile Industry," Pamphlet distributed by
Ferro Corporation, Cleveland, OH (1996).
Fle93 J. Fleming, Experimental Techniques in Glass Science.
(C. Simmons and 0. El-Bayoumi, eds.), American Ceramic
Society, Westerville, OH, 1-29 (1993).
Gre82 S. Gregorio, M. Greenblatt, J. Pifer and M. Sturge,
Journal of Chemistry and Phvsicis. 76 (6), 2931-2937
(1982) .
Gru78 0. Grum-Grzhimailo, Glass Ceramic. 35 (1), 41 (1978).
Hal86 D. Halliday and R. Resnick, Fundamental of Physics.
2nd Ed., John Wiley and Sons, New York (1986).
Hum93 R. Hummel, Electronic Properties of Materials. 2nd
Ed., Springer-Verlag, New York (1993).
Hun87 R. Hunter and R. Harold, The Measurement of
Appearance. 2nd Ed., John Wiley and Sons, New York
(1987) .
Jac96 P. Jackson, Science of Whitewares. American Ceramic
Society, Westerville, OH, 395 (1996).
Jam85 P. James, Journal of Noncrvstalline Solids. 73. 517
(1985).
Jud65 D. Judd and G. Wyszecki, Color in Business. Science
and Industry. 2nd Ed., John Wiley and Sons, New York
(1965).
Kim59 K. Kim and F. Hummel, Journal of the American Ceramic
Society. 46 (6), 286-291 (1959).
Kin40 B. King, Journal of the American Ceramic Society. 23
(8), 221-225 (1940).

363
Kin76 W. Kingery, H. Bowen and D. Uhlmann, Introduction to
Ceramics. 2nd Ed., John Wiley and sons, New York
(1976) .
Klu74 H. Klug and L. Alexander, X-Ray Diffraction
Procedures. 2nd Ed., John Wiley and Sons, New York
(1974).
LaC97 W. LaCourse and W. Mason, Ceramic Engineering and
Science Proceedings. 18 (2), 237-245 (1997).
Lev69 E. Levin, C. Robbins and H. McMurdie, Phase Diagrams
for Ceramists. 2nd Ed., American Ceramic Society,
Westerville, OH (1969).
Mac91a Macbeth Optimatch Color Matching System Reference
Guide. Macbeth Corp., Newburgh, New York (1991).
Mac91b Macbeth White-Eve 3000 Spectrophotometer User's
Reference Guide. Macbeth Corp., Newburgh, New York
(1991).
Mac96 Macbeth ProPalette Computer Color Matching System
Version 1.6. Computer Index, Macbeth Corp., Newburgh,
New York (1996).
MacA35 D. MacAdam, Journal of Optical Society of America. 25
361-367 (1935).
MacA42 D. MacAdam, Journal of Optical Society of America. 32
247-274 (1942).
Mal96 R. Malgren, Ceramic Industry. 146 (10), 31-36 (1996).
Mas68 B. Mason and L. Berry, Elements of Mineralogy. W. H.
Freeman Co., San Francisco (1968).
Mei97 K. Meinssen, Ceramic Engineering and Science
Proceedings. 18 (2), 308-319 (1997).
Mue83 D. Mueller, "Ceramic Raw Materials," Class Notes from
CES314, Alfred University, Alfred, NY (1983).

364
Mur89 J. Murray, American National Standard Specifications
for Ceramic Tile, A137.1. Tile Council of America,
Princeton, NJ (1989).
Mur90 S. Murdock, T. Wise and D. Eppler, American Ceramic
Society Bulletin. 69 (2), 228-230 (1990).
Nas83 K. Nassau, The Physics and Chemistry of Color. John
Wiley and sons, New York (1983).
Pag97 Mariano Paganelli, Ceramic World. Editore Tile Italia,
S.r.l., Sassuolo, MO, Italy, 148-151 (1997).
Par73 C. Parmelee and C. Harman, Ceramic Glazes. 3rd Ed.,
CBLS, Marietta, OH (1973).
Pat67 D. Patterson, Pigments. Elsevier Publishing Co., New
York (1967).
Paw96 E. Pawlicki and D. Fu, Ceramic Engineering and Science
Proceedings. 17 (1), 173-179 (1996).
Pet72 R. Petrucci, General Chemistry. MacMillan Company, New
York (1972).
Pop77 M. Pope and M Judd, Differential Thermal Analysis.
Heyden and Son Ltd., Bellmawr, NJ (1977).
Rah95 R. Rahaman, Ceramic Processing and Sintering. Marcel
Dekker, Inc., New York (1995).
Ree83 J. Reed, "Optical Properties of Ceramic Materials,"
Class Notes from CES305 Ceramic Lab., Alfred
University, Alfred, NY (1983).
Ree95 J. Reed, Principles of Ceramic Processing. 2nd Ed.,
John Wiley and Sons, New York (1995).
Ric92 D. Richerson, Modern Ceramic Engineering. 2nd Ed.,
Marcel Dekker, Inc., New York (1992).
Sac86 Sacmi Labs, From Technology Through Machinery to Kilns
for Sacmi Tile, 2, Sacmi R&D Lab Notes, Imola, Italy
(1986) .

365
Sac96 M. Sacks, EMA 4645 Ceramic Processing Class Notes,
University of Florida (1996).
San57 C. Sanders and G. Wyszecki, Journal of Optical Society
of America. 47. 398-404 (1957).
Sch62 H. Scholze, Glastech. Ber.. 39. 63-68 (1962).
Sel47 H. Selling, Journal of the Society of Dyers Colorists.
63, 619 (1947) .
Sez98 G. Sezzi, Tile Italia. 1, 46-50 (1998).
Sha92 J. Shackelford and W. Alexander, CRC Materials Science
and Engineering Handbook. CRC Press, Boca Raton, FL
(1992).
She97 L. Sheppard, Ceramic Industry. 147 (1), 41-174 (1997).
Sim93 J. Simmons, Experimental Techniques in Glass Science
(C. Simmons and 0. El-Bayoumi, Eds.), American Ceramic
Society, Westerville, OH, 383-427 (1993).
Tay86 J. Taylor and A. Bull, Ceramics Glaze Technology.
Pergamon Press, Elmsford, NY (1986).
Teu95 K. Teuchert, R. Thomas and A. Muller-Zell, Interceram.
44. (4) , 230-233 (1995) .
Tip80 P. Tipler, Physics, Worth Publishers, Inc., New York
(1980).
T'oz86 R. Tozzi, Glazes for Fast-Firina. Italian Ceramic
Society, Rimini, Italy (1986).
Vic97 W. Vickery, Ceramic Engineering and Science
Proceedings. 18 (2), 409 (1997).
Wea79 R. Weast and M. Astle, CRC Handbook of Chemistry and
Physics, 60th Ed., CRC Press, Boca Raton, FL (1979).
Wor82 W. Worrall, Ceramic Raw Materials. 2nd Ed., Pergamon
Press, Elmsford, NY (1982).

366
Wyc65 R. Wycoff, Crystal Structures. 3rd Ed., John Wiley and
Sons, New York (1965).
Yoo97 C. Toon, W. LaCourse and W. Mason, Ceramic Engineering
and Science Proceedings. 18 (2), 439-454 (1997).

BIOGRAPHICAL SKETCH
David Earl was born in Buffalo, New York, on December
28, 1961. He graduated with a B.S. degree in ceramic
engineering from the New York State College of Ceramics at
Alfred University in 1984. In 1985 he moved to Parkersburg,
West Virginia, and worked as a quality control engineer for
A.B. Chance Co., a manufacturer of porcelain high voltage
insulators. In May of 1986 he joined Florida Tile Industries
in Lakeland, Florida, as quality assurance supervisor and was
eventually promoted to the positions of corporate statistics
consultant and senior engineer. In 1994, David earned an
M.S. degree m materials science and engineering from the
University of Florida in Gainesville. In July of 1994, he
joined Huntington/Pacific Ceramics, Inc., a ceramic tile
manufacturer, as plant manager of the Mount Vernon, Texas,
facility. David returned to Florida Tile Industries in
August of 1995 and has since held the positions of plant
manager and director of research and development. He was
367

368
admitted to doctoral candidacy at the University of Florida
in December of 1996.
His professional affiliations include memberships in
the American Ceramic Society (ACerS), National Institute for
Ceramic Engineers (NICE), Phi Kappa Phi international
academic honor society, Ceramic Manufacturing Council (CMC),
Ceramic Educational Council (CEC) and Keramos. He is also a
Distinguished Mentor for ACerS.

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
1 C ¿X
David E. Slark, Chairman
Professor of Materials Science
and Engineering
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
E. Dow Whitney
Professor of Matél
and Engineering
Is Science
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Professor of Materials Science
and Engineering

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Dinesh 0. Shah
Charles A. Stokes Professor
of Chemical Engineering
This dissertation was submitted to the Graduate
Faculty of the College of Engineering and to the Graduate
School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
December 1998
Winfred M. Phillips
Dean, College of Engineering
M. J. Ohanian
Dean, Graduate School

LO
2780
J99L
16 i X
UNIVERSITY OF FLORIDA
3 1262 08555 1017



xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EXBMYKTNK_R7S9DV INGEST_TIME 2014-07-25T02:41:50Z PACKAGE AA00021358_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES